PAGENO="0001"
COMPETITIVE PROBLEMS IN TF~E
DRUG INDUSTRY
~1O&3?I,~
HEARINGS
BEFORE THE
SUBCOMMITTEE ON MONOPOLY
OF THE
SELECT COMMITThE ON SMALL BUSINESS
UNITED STATES SENATE
NINETY-FOURTH CONGRESS
FIRST SESSION
ON
PRESENT STAPUS OF COMPETITION IN THE
PHARMACEUTICAL INDUSTRY
PAET 28
JANtIARY ~1% JL~L~Y~ 9 AND iO~ 1975
ORAL HYPOGLYCEMIC DRUGS
[CONTINUEDJ
Printed for the use of the Select Committee on Small Business
U.S. GOVERNMENT PRINTING OFFICE
56-592 WASHINGTON: 197.5
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price $4.00
PAGENO="0002"
SELECT COMMITTEE ON SMALL BUSthE~S
[Created pursuant toS.'Res. 58, 81st Cong.]
GAYLORD NELSON, Wisconsin, Chairman
JOHN SPARKMAN, Alàbáma JACO~ K. ~AV~1~S, New York
THOMAS J. McINTYRE, New Hampshire ~T. GLENN BEALL, Jil., Maryland
SAM NUNN, Georg1a~ BILL I3BQCK,, Tefipessee',
J. BENNETT JOHNSTON, Louisiana LOWJ~LL P~WEI~KER, JiL, Connecticut
WILLIAM D. HATHAWAY, Maine DEWEY F. BARTLETT, Oklahoma
JAMES ABOUREZK, South Dakota PALL ~4XAZT Nevada
FLOYD K. HASKELL, Colorado BOB PACKWOOD, Oregon
DICK CLARK, Iowa
WALTER F. MONDALE, Minnesota
WILLIAM B. CHEaItA~KY, Staff Director
BENJAMIN Goapon, Staff Eçoito'niist
JTJDAII C~ SoMMan, Minority Co~assel
~ KLATTfr Rei~earc1v Assis'tant
SuucoMl\4ITTgli ON MONOPOLY
GAYLORD NELSON, Wisconsin, Chairman
THOMAS J. McINTYRE, New Hampshire DEWEY F. BARTLETT, Oklahoma
WILLIAM D. HATHAWAY, Maine ,. ~. ~LEI~ BE~LL, Ja., Maryland
.TAMES ABOTJREZK, SO~th Dakotk ` ~OB I~ACKWOOD, Oregon
FLOYD K. HASKELL, Colorado JACOB K. JAVITS,* New York
*Ex officio member. `
(II)
PAGENO="0003"
CONTENTS
Testimony of-
Chayet, Neil L., Counsel, accompanied by Robert F. Bradley, M.D.,
Chairman, Committee for the Care of the Diabetic, JOslin Clinic, Page
Boston, Mass 13277
Chester, Edward M, M.D., Professor of MedLciI~e,, Case Western
ReservC University, and Director, Athbulatory Medicine Teaching
Clinic, Cleveland Metropolitan General Hospital 13309
Felig, Philip, M.D., Professor and Vice Chairman, Department of
Internal Medicine Yale University School of Medicine -. - - 13318
Lamer, Joseph, M.J3., Ph. D., Professor and Chairman, Department
of Pharmacology, and Director of the Diabetes and E~idocrinoJogy
Center, University of, Virginia SChool of Medicine_ - - 13324
Meier, Paul, Ph. D., Professor of Statistics, University of Chicago,
Chicago, Ill 13267
Falumbo, ~ ~ M.D., Assistant Professor of Medicine, Mayo Medical
School :~:_, 13273
Ricketts, [~Ienry T., M.D., Professor of Medicine Emeritus, University
of Chicago Medical School 13264
Schmidt, Alexander M., M.D., Commissioner, Food a~nd Drug Ad-
ministration, accompanied by~ Rich~ird Merrill, Chief Counsel, Food
and Drug Administration; J Richard Crout, M.D., Director, bureau
of Drugs, Food and Drug Administration; James M. Biletad, M.D.~
Group Leader, Division of Metabolism and EndOcrine Drug Prod-
ucts, Bureau of Drugs, Food and Drug Administration; and Robert
Wetherell, Director, Office of Legislative Services, Food and Drug
Administration 13200
Sims, Ethan A. I-I., M.D., Professor of Medicine, College of Medicine,
University of Vermont, Burlington, Vt 13329
White, Colin~ M.D.~ Professor of Public Health, Department of
Epidemiology and Public Health, Yale University School of Medi-
cine, New Haven, Conn 13258
Zele~, Marvin, Ph. D,, Professor of Statistical Science and Director,
Statistical Laboratory, State University of New York at Buifalo__ 13271
APPENDIX
Exhibits provided ~or the hearing record by the Subcommittee on
Monopoly:
"Report of the Committee for the Assessment of Biometric Aspects of
Controlled Trials of Hypoglycemic Agents," article from the Journal
of the American Medical Asaociation j~337
"The Effects of Long Term Therapy With Oral Hypoglycemic Agents
on the Oral Glucose Tolerance Teat Dynamics ift Chemical Di-
abetes," presentation at the 33d Annual Meeting of the Americab
Diabetes Association on June 23, 1973 in Chicago, Ill' 1337k
"Assessing Survival in a Diabetic Population," by Paula Helene
Kanarek, a thesis submitted to the Faculty of the Harvard School
of Public Health, Boston, Mass.; January 1973, excerpts 13393
(~III)
PAGENO="0004"
Iv
"A.M.A. Aide Let Upjohn Use Letter To Sell Drug," by David ~`age
Burnham, article from The New York Times, July 8, 1975 13413
"Oral* Hypoglycemic Agents," excerpt from the Pharmacological
Basis of ~Therapeutics, fifth edition, Goodman and Gilman pages
1519-1524 - 13414
Letter dated August 21, 1975, to hearing clerk, Food and Drug
Administration, from Max Miller, M.D~, professor, Case Western
Reserve University 13421
Presentation of Sidney M. Wolfe, M.D., and Anita Johnson, Public
Citizen's Health Research Group, ttr the FDA hearings on proposed
labeling for diabetes drugs, August 20, 1975 13423
"Oral Hypoglycer~iic Agents Are Worthwhile," by Robert F. Btad1~th
1VLD., JOslin Cliriic, ~o~tdn 1~'1a~ - . -- 13427
"AMk (~ffk~1aI Let cofir~3an~ LT~e ~nte" b~ ~tuM~t Arié~ba~h, th'tiÔ1~
from thQ Washington Post, July 9~ 1q75 13439
Letter dated ]~ebru~ry I~, 19'lS, t~ ~èh~tot Gaylord ~ Ohai~
nt~th, Select C imfttée ott ~m~ll ~itsinè~, U.S. ~ front
Dr. Paul Meier, Pro~e~O~ o~ ~tatiétiO~, ThiIvè±sity of OMOa~o 13440
Letter dateti ~arch 11, 1075, to Sen~tor~GaylOtd Nelsott, Cha1rttia~
Select Cmnnuittec ott Sttiall B~niness, [1.~. S~n~te, frorri ~ei9b~i~t IL
Mc1)a~de, ~fr., PresidOnt arid Chief Operating Officer, IJSV hai~n~-~
cetitic~1 Carp 13448
Letter dated April 1, 1q75, td ~1ott. Alé~iañder M. SCliñ~tidt aD.,
Coinffilssidtief, Food and Drug Adxttinisttatiott, from SOnath~
Gaylord Nelson, Chairman, Select Committee on Small is',
tLs. Senate 13452
t~ettet dated April 10, 1975, to Senator Gaylord Nelson, Chai~,
Select Committee ott ~tr1all Bushië~ T.Y.~. Sefiate, frdi~ Robert C.
Wetherell, Jr., Director, Office of Legisititi* Servic~, F~dd anti
t~iig Admi5istrtttiOn~- 13453
Lettei~ dated Febrttary 14, 197&, to Sen~tor Ga~r1ord NelsOn, Chaif-~
man, Select Committee on Small ~usiries~, U.S. Senate, frOnt
Neil L, Chayet, Attorney at LaW, Cha~et and Somrenreicb, P. C.,
with ae~oippatiy1tt~ enciOsttre 13454
Letter dated May 21, 1~75, to Neil L. Cht~yet, Attot4e~ at Law
dha3tet and Sontteru~e1cb, ?. C., front Benjathin Gord~tt, ~ta~
Economist, Select Committee on Small Business, U.S. senate 13457
Lettêi~ dated ~ul~r 10, 1975, to Be~j~tnirt ~4ordon, Sta~ff EConOntist,
select Committee on Small Bushies~, U.s. Senate, from Nell L.
Cha~*et, Attorney at Law, Ohayet ati~ Sotttten~cicb, P. C., ~th
acéontpanyh~g enclosure 13457
Not1c~ of Public Hearing arid Proposed Labelift~ by the Fci~d a~d
Iiru~ Adntittisti~tiOfl, Depnrtment o( Health, Education, afid
Weif~tre, 21 CPIt Part 31(1, I~ockét 140. 75N-0062~ July 1 I$7&~.... 13462
Letter dated June 10, 1975, to Alexander M. Schmidt, M.b., Com-
missioner, Food and Drug Administration, from Sidney M. Wolfe,
M.D, Health Research Groui 13530
~`The PI3A and Hypoglycemic IDrugs," by John K. DavidsOtt, M.D.,
Ph.D.~ article from the Journal of the American Medical Ass~cia~-
t~on, Vol. 232 No. 8,, May ~6, 1q75, pages 853-~55 13535
~`~ovérnthent f~r!s-l~tYA Proposes Important ReVisiOn," artiCle
from Drug Therapy, Janue~y 1915 .~ 13538
"Oral Antidiabetfo Agents Have a Tmite4 Place itt Manttgeinetit
and May Be Harwfi4," b~ Albert I. ~inegrad~ Ee~ S. c~eriient~
~r., and Anthony D. Morri~Oi~ ijtiiversity Of P~tttxs~lva~ttia ~ehOoi
of Medicine - 13541
"Ii~1if~o Efl~ects of' Tolbtttamidè on tl~O Sympat.ho-Adrølild ~y~teth,"
by Chung-Yi ~EIsu, Gary Brooker, Mtchaçl ~. Peac~i, Thotnas 0.
Westf all, and Joseph Lai~ner, artiCle frOm Diabetes, Vol. 24, Stipple-
merit 2, June 15, 1975 13556
"A Study of the Chronic Effects of Tolbutamide in the Rhesus
Monkey," by J. Borensztajn, G. S. Getz,S. Glagov, A. Rubenstein,
C. H. Ts'ao, and R. W. Wissler, FDA No. 72-114, February 15,
1975 13560
:Letter dated May 30, 1975, to Senator Gaylord Nelson, Chairman,
Select Committee on Small Business, U.S. Senate, from Robert C.
Wetherell, Jr., Director, Office of Legislative Services, Food and
Drug Administration, with accompanying enclosures 13568
PAGENO="0005"
V
Prepared statements~ ~
Chayet, Neil L., Counsel, accompanied by Robert F. Bradley, M.D.,
Chairman, Committee for the Care of the Diabetic, Joslin Clinic, Page
Boston, Ma~s 13620
"Role of Diabetes in Congestive Heart Failure: The Framingham
Study," by W. B. Eannel, M.D., M. Ujortland, Ph.D., and
W. P. Castelli, M.D., article from The American Journal of
Cardiology, Vol. 34, pages 29-34, July 1974 ~. 13637
Chester, Edward M., M.D., Professor of Medicine, Case Western
Reserve University, and Director, Ambulatory Medicine Teaching
Clinic, Cleveland Metropolitan General Hospital 13643
Table 1.-Comparative usage of oral hypoglycemic agents,
Cleveland Metropolitan General Hospital, 1968-1975 13653
Table 11.-Yearly use of insulin stocked in pharmacy, Cleveland
Metropolitan General Hospital, based on accessible records,
1968-1975 13654
Fel~g, Philip M.D., Professor and Vice Chairman, Department of
Internal Medicine, Yale University School of Medicine 1~655
Lamer, Joseph, M.D., Ph.D., Professor and Chairman, Department
of Pharmacology, and Director of the Diabetes and Endocrinology
Center, University of Virginia School of Medicine 13659
Meier, Paul, Ph.D., Professor of Statistics, University of Chicago,
Chicago, Ill 1~665
Palumbo, P. J., M.D., Assistant Professor of Medicine, Mayo Medical
School 13667
Ricketts, Henry T., M.D., Professor of Medicine Emeritus, University
of Chicago Medical School 1~3670
Sims, Ethan A. H., M.D., Professor of Medicine, College of Medicine,
University of Vermont, Burlington, Vt 13676
White, Cohn, M.D., Professor of Public Health, Department of
Epidemiology and Public Health, Yale University School of
Medicine, New Haven, Conn 13689
Zelen, Marvin, Ph.D., Professor of Statistical Science and Director,
Statistical Laboratory, State University of New York at Buff alo - 13694
Exhibits provided by the Food and Drug Administration:
Prepared statement of Alexander M. Schmidt, M.D., Commissioner,
Food and Drug Administration 13697
Proposed labeling indications, contraindications and warning sections
for oral hypoglycemic drugs of the sulfonyJi~ire~ c~tegory~ 1371i~
HEARING DATES
January 31, 1975:
Morning session ~. 13~7'
July 9, 1975:
Morning session 13289
July 10, 1975:
Morning session 13309
PAGENO="0006"
PAGENO="0007"
COMPETITIVE PROBLEMS IN THE DRUG `INDUSTRY
(Present Status of Competition in the Pharniac~utica1
Industry)
FRIDAY, JANUARY 81, 1975
U.S. SENATE,
SUBCOMMITTEE `ON MONOPQLY OF THE
SELECT COMMITTEE ON SMALL BUSINEss,
Washington, D.O.
The subcommittee met, pui~suant to recess, at 10:07 a.m., in room
1114, Dirksen Senate Office Building, Senator Gaylord Nelson (chair-
man of t'he full committee) presiding.
Present: Senator Nelson.
Also present: Benjamin Gordon, staff economist, and Kay Klatt,
research assistant.
The CHAIRMAN. Today the Monopoly Subcommittee of the Senate
Small Business Committee is resuming its hearings on the oral hypo-
glycermic drugs initiated on September 18, 19, and 30 of last year~
Our witnesses today will discuss recent studies dealing with the
safety, efficacy and usefulness of this class of drugs. The list of
witnesses includes four members of the committee selected by the
internationally, renowned Biometric Society as well as a member of
the Mayo Clinic.
Additional witnesses to appear are Mr. Neil Chayet and Dr. Robert
Bradley, counsel and chairman respectively of the Committee on the
Care of the Diabetic.
Our first witness this morning is Dr. Cohn White, professor of
public health, `Yale University School of Medicine, New Haven, Oonn.
Dr. White, your statement will be printed in full in the record, ai~d
you may present it however you desire, and extemporize as much as
you wish.
Would you identify the organization you represent `for the record,
or first, for the reporter. Perhaps each of you, starting on my far
right, would identify yourself for the reporter, so if yon address
yourself to some question we shall be able to identify you.
Dr. MElEE. I am Paul Meier, professor of statistics at the Unive~.~
s~ty of Chicago. ` , ` .`
`Dr. RIçKETTS. Dr. Ricketts, Un.i~rsity of Chicago. `
Dr. WHITE. Cohn `White, professor of public health, Yale
`University. , ` , " ,
Dr. ZELEN. Marvin Zelen, professor' of statistical science, State
University of New York, Buffalo. " " ,
13257
PAGENO="0008"
13258 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Dr. PALUMBO. Pat Palumbo, Mayo Clinic. [The statements and ob~
servations made in this testimony are my own and do not necessarily
represent official policy of the Mayo Institution.]
The CHAIRMAN. Thank you, gentlemen, for taking the time from
your busy work to come here and testify today.
Now, if each of you, when you speak, would pull ~he mi~ro~hone
up closely and speak directly into it, we si'~all be able to hear you.
Go ahead, Dr. White.
STATEMENT 01? COLIN WHITE, M,~., PROFESSOR OP PUBLIC
HEALTH, DEPARTMENT `OP EPIDEMIOLOGY AND PUBLIC
HEALTH, YALE UNIVERSITY SCHOOL OP MEDICINE, NEW
HAVEN, CONNI
Dr. Wnrn~. Senator Nelson, I am the chairman of a committee
which w~s appointed by the Biometric Society and, funded by the
National Institute of Health to carry Out the following mission:
One, to make an in-depth assessment of the scientific quality of the
UGDP study and in particular of the biometric aspects of the
design, conduct, and analysis of the trial; two, to make a similar
assessment of other `controlled trials of oral hypoglycemic agents.
Tl'~e CHAIRMAN. Would you identify the Biometric Society in at
least a brief description so that th~ record will be clear on that.1
Dr. W~II~E. The Biometric Society is an international society o~
people who are interested in the application of statistical data to
biological problems.
The CIIAtRMAN. Who are' the rnemb~rs? That is, what is' the
eligibility of your membership.?
Dr. Wnrri~. Membership is attained by application and the' corn~
mittee decides the qualifications of those who wish to join. In general,
membership is governed by interest in the work of the society.
The CHAIRMAN. Are there any special required scientific qualifica~
tions?
Dr. WrnPR. I think an expression of interest is all that is neôessary.
The CHAIRMAN. Is this an international soQiety?
Dr. WHrv~. It is.
The CI~Ani1~EA~. And are yQU president of this society?
Dr. WliITE. No, I am not. I am chairman of the committee that
was appoh~ted by the society for this particular purpose.
The CHAIRM4N. And the society selected the members of the com~~
mittee of which you are chairman for the purpose of. evaluating the
UGDP study?
Dr.'WrnTE. Yes.
The CHArEtMAN. k~'o ahead, doctor.
Di~. WmTE. The ~oinmittee eo~isisted of six members: J~ohn
Gilbert, Harvard 1J~iversity; Paul Meler, University of Chicago;
Chris L. Rumke, Free Uni~rer~1ty, Amsterdam; :ROdO1fQ Sara~ci,
PIsa, Italy; Marvin Zelen, S~tate t~iiversity `of Ne* York at Buffalo;
Cohn White, Yale University.
1 See Biometric Society study, page 13337.
PAGENO="0009"
~0~PiTIV~ PBQ~L~5 IN r~I~ ~RUG INDU~ 1a259
Th~ full thmmittee nie~ o~ si~ oeeasions over a 2-year period and.
ha~ completed a report which will be publisJie~ on :February 10 i~
the Journal of the American Medical Association.
Th~ `work of the IJGDP i~ ~tillin progreiss' and 1 tbhik it: 1~ fair
to say that diabetologists in general await with interest the finding~
on the treatmei~t by insulin. There has ne~rer been ~` study of corn-
parable scope and thoroughness on the long-term effects of this
agent in suMects with maturity~onset diabetes. In the meanwhile,
however, controversy ha~ arisen about the data concerr~iug
tolbutamide.
The committee saw as its maii~ task the investigation of the
reported excess cardiovascular mortality in the stibjects recelv*
ing this drug. It is interesting to note that the UGi~P pre-
sented results on phenformin which are quite comp~rabie to `those on
tolbutamide: the death rate from gardiova~cul~r catlse~ was apprô~i-
mately the same in the two cases. rflle 1~idings on phenfo~mix~, if one
can ~udge from the absence of criticism, appear to hav~e been accepted
by medical scientists, even if they have not so far been tr~isl~ted
effectively into medical praèti'ce. Yet these fi~d~ugs~ also were made
by the UGDP using the methQds that have come under he~vy
criticism when applied to tolbutamide,
Because of the many factors which iufluence survivorship in a
chronic disease such as maturity-onset diabetes, oaref~l method~ of
investigation are needed, and, in particular, eontrol groups `~re:~ssen~
ti~1. Consequently we reviewed oaly such trials as we~ cont~oJ4ed.
It then became' clear that the major st~&dy to cou~ider, other than tha
TJGDP, was the study in Bedford EngIand~ organized by X~'r~
H. .Abby Keen and Dr. R. 3. Jarrett. It should be said at once, h~w-
,that the Bedford study, based on 125 patients in e~eh &f th~
two' treatment groups was not comparable in Size or in detail t~ the
UGDP in which approximately 200 patients were followed `on ench
of five treatments.
The work of the committee appointed by `the Biometric Soci~t~
fell into four sections:
One: Visits were made to ~he t~GDP coordinating center and' t~
two of the' cooperating `ciin1~al centers to study methods used in
th~ triaL Two' T}t~ methods and findings `of the UGDP stud~.t `were
discussed with several authors who had written' about them, and'
the `BecU~ord study was discussed with I~r. Keen au~ Dr. Jarr~tt.
Three: The publish~d criticisms `of the' UGDP were reviewed in
t~iI. Comparable criticisms of the Bedford stud~r do not e~tst~
though several of the major criticisms made about the UG~P woui~
apply a f~'rtiori to the Bedf~rd ~tudy~ Four :Ne~ analyseS w~re
made o~ the data from the UGDP and Bedfo~rd studies, the. data
being kindly made available by the UireMtors e~ne4.
Critics have poin~ted out t~at ii~ the I~GDP study the total mpr-
tsiaty w~ not si~ificanUy higher in the tolh~e4ami4e groi~ than ~in
the placebo gi~oup, even though there ~wa~ a signifi~cant `dige~~*w~i ii~
the cse~ of~ deaths frem ca iovaseula~r causes. We consii~er that this
criti~isw has some weightbut is not eonirincing. Criticisms that have
been commonly made but which, in our view, are not Co1~eCt~ are'
PAGENO="0010"
13260 OOMPETITIVE PROBLEMS IN THE DRUG `INDTJSTW~
One: The `finding of excess mortality in the tolbutamide group was
due to the data obtained from just a few clinics.' These are objections
we do not find valid. .
Two: The studies of Keen et al. and of Paasikivi contradict the
IJGDP.
Three: The baseline differences among the treatment groups ac~
count for the finding of the adverse effects from tolbutamide. On this
point I might remark that none of the criticis, to my knowledge, has
given serious consideration to the multiple logisticS method that wa~
used by the UGDP to take the effect of `baseline risk factors' intQ
account. Until they do this they have not carried out an adequate
review of the TJGDP analysis.
The CHAIRMAN. And your group did do that?
Dr. WHITE. Yes, we did.
Four: The findings on the effect of tolbutamide are flawed by the
failure to adapt dosage to individual need.
Five: The evidence wa~ not adequate to justify the' discontinuation~
of the oral drugs.
In our analysis of the TJGDP data we have used the same multiple
logistic model as was employed' by th~ `UGDP investigators, but have
taken additional variables into account to allow for the time each
subject was under study and for differences `between `clinics. We
confirm the principal finding from the `simpler `study of failure rates;
namely, that the cardiovascular death rate was higher in patients
receiving tolbutamide than' in those receiving placebo. This differ-~
ence remains after adjustment for the effect of baseline ~ariablës
and cardiovascular risk factors.
We have also made an analysis in which the extent of adherene~
to assigned treatment was taken into account. The highest death rate
was found' in the tolbutamide group who adhered 100 percent to their
treatment and who did not modify the dose.
In an analysis of the data from the Bedford trial we found no
difference in death rate between the placebo and the tolbutamide
group. As indicated above, we do not interpret this failure to find a
difference as a contradiction of the more thorough UGDI? study.
`The conclusion of the committee is that it, remains with the pro-
ponents of the oral agents to conduct scientificall~ adequate studies
to justify the continued `use of such agents.
The CHAIRMAN. Well, ~put in different words, are you saying that
it is the judgment of the Biometric Society that it was a statistically
valid sample, and a scientifically conducted study, and that the result~
of `the' study-are the conclusions valid? Is that what yon are saying?
Dr. WIUTE. Yes. We support the principal findings of the UGDP
study. We `do make some mii~ior critiei~ms in the report, bu~ we do, im
general, support the main finding~ ` ` ` ` ` `
The CHAIRMAN. `And the `main finding is~hat?
`Dr. Wnim. That there is an excess mortality in the group receiv-~
ing tolbutamide as compared with the group `on the pl~eebo. ` `
The OI~ATRMA~. Well, did' you find any evidence at all' that th~
oral ~hypoglycemic drugs retarded or pr~evented ` vascular cornplica~
tions of' diabetes? ` ` ` ` ` ` ` `
Dr. WmTE. That aspect of the study is one that we did not under-
take. We considered that our main responsibility was to look intG
PAGENO="0011"
COMPETITIVE PROBLEMS I~ TI~E DRUG INDUSTRY 13261
the question of mortality effects. There is evidence still to come in
on the long-term effects of the various treatments that were used.
Mr. GoRDoN. One question.
You say it remains with the proponents of the oral agents to
conduct scientifically adequate studies to justify the continued~ use of
such agents.
Now, when we had the UGDP people here before us, they stated
that they discontinued the use of these agents because they founçl it
was ethically untenable to keep on giving these drugs to people~ be-
cause they were satisfied that it was causing a lot of harm.
How do you feel about that?
Dr. WHITE. I think that if the group decided that it was ethically
untenable, that would settle the question then and there as far as
public policy were concerned. If they could persuade a responsible
group otherwise, then the only kind of evidence that would be accept-
able to us is evidence obtained from a controlled trial.
The CHAIRMAN. Any of you gentlemen may comment on this.
Is my memory correct that the UGDP study then concluded that
diet was a better way of managing the problem than `by tolbutam~de
and other oral hypoglycemics? Was that their general conclusion?
Dr. WHITE. Yes. I think that is a question on which Dr. Ricketts
would have a more valuable opinion than I have.
Dr. RIOKETTS. Well, I suppose it was done because tolbutamidev
apparently was no better than diet.
The CHAIRMAN. Than diet?
Dr. RICKETTS. Yes. And since it was a little dangerous' they would
say naturally after a certain number of deaths that they had better
stop.
The CHAIRMAN. The general conclusion of the study was that diet
was better than tolbutamide or oral hypoglycemics of any kind. Is
that correct?
Dr. ~IOKETTS. Well, I am not quite sure that is the .way to put jt.
I think I just said, and I guess I will hive to repeat it, that tolbuta~
mide was no better than~ diet, and if that is true, and if it looked as
if the tolbutamide was rather dangerous, then anybody *ould say, let
us stop tolbutamide and do what we can with the diet.
Does that answer your question?
The CHAIRMAN. Yes.
Dr. MEIER. There is one point I would like to `emphasize, and that
is that neither our committee, the supporters of the IJGDP, nor those
who think it was `an invalid study, believe that this is a simple ques-
tion. It is complex, and I do not think it is capable of ,a simple
answer of `the form that was' suggested, namely: "Here is' a `drug
that is `of no value. It is toxic. We ought to abandon it." ` ` "
There are special subgroups of patients who a~'e not successful'
with diet, are unable to take insulin', and I think most `of the com-
ments before this committee and elsewhere have pointed to su~h.
subgroups. For these, tolbutamide may `haveS `definite value. But the
question of whether it should be' used more widely, as it now is, re-
mains difficult also. It is the case that the IJGDP investigators'
themselves were not unanimous about the desirability of dropping
tolbutamide from the UGDP study. The discussion that `went on
there was very well described in a paper by one of the participants~
PAGENO="0012"
1~262 ~oMr~a'rrIv~ ~crn~s IN ~ ~ INPT,I~RY
Dr. Theodôr~ Schwa~tz,1 The teii~ion~ the pulling and ha~ili~ig, the
malor differences of opinio~i that led ~o t~iat final decision, mu~ all
be taken into account in trying to interpret the mearnn~ of that cieci~
sion. it wa~ jnadged likely that toibutamide was toxic; but the evi~
deuce wa~ not considered conclusive.
So I do not think we can say that there is a clear, flat conclusion
tlmt comes out of thi~, and I think reasonable people may ~ome to
son~ewh~t di~erent conclusions,
rrhe C~AIRMAN. Clear conclusions about what?
Dr. MEI]~. About *rhether tolbutamide should be aband~oned by
all physicians in the treatment of diabetes.
The CuAIu~rA~. That; really is not the issue, is it?
Dr. MuIER. I think the issue is what we ought to do, not whether
we have reached a firm conclusion. I do not think we have reached a
completely firm conclusion as my statement will show, I deplore the
fact that we are not in a position to reach a firmer conclusion than
we now have, but I would support the final statement of the Bio-
metric Society comitdtte&s r~iort~ which suggests that; a new study
might be eonducted~ I think it would be ethically legitimate to con~;
duct a new study. I myself think it i~ not ethically legitimate to con-
titiue to use the drug without a new study.
The CHAIRMAN. The issue is not whether you should prohibit it~
use under any circumstance on any patient in any situation., The
question is, as a general proposition, should you use it in those cases
where the patient situation can be managed by diet?
For example, Dr. John Davidson said that at the Grady Memorial
Bospital it was finally concluded after the study-~I think they had
some 6~5OO patients, which I believe was the largest group in th~
~o~nt~y-~-that they would take them of~ the drug and if my recollec-
tion is correct their patients were better managed on diet. He sa~id
it was tough medicine to swallow because they had lived with oral
hypoglycemics, thought they did well, studied the iYGDP~ study,
which, they concluded, was right.
rrhen, in a mOre precise ansi*er-'-I believe I am correct, and if I
am not~ I will correct the record~-~-he thought that maybe in a. very,
very small percentage of cases, I think he said it might be I percent
or less, an oral hypoglycemic would be indicated to be used, He did
~ot state, what that case was, so I do not know whether that waS an
~nsurance policy statement or not,
But in any event, is that not the question: `Not whether 7~u should
abolish these drugs, but whether ~n those cases where diet can inan~
~ge the problem, it should be used? And is it not the con~iu5ion of
the TJGDP study, as well as Dr. Davidson at Grady ~[emoHai IIos~
pital-and the doctor from Mayo will address hithself to this 4u1es~
tloti also~'-that there is a very,; very small percentage of cases in
which it i~ indicated, but that it iS widely used ~n cases where it is
yiOt indicated4
Is that a fair genetalization?
Dr. MEmn~ I think the question reall3r is whether the evidence is
of such over~helmM~ clarity `that the conclusion `reached by these
gentlemen Should be' a regulation imposed by law up~m the ntedioal
1The~prntd~o~troversy ~ .& Pei,sonal Pet~speet1~e (A.fluaia of Intei~na1 Med1cI*~.
PAGENO="0013"
COMPE4~1~t11~ `r o~i~s m~ ~ inu~ ~ I32~$
co~iuniby generaLly. If we `were solidly convinced that ~tolbutami~e
were poison, there would be no doubt about it, and if the hcueüt~H~k
ratio wa~ çlefini~ly proved to be unfavorable, then I think we wouLd
seek' regulation to prevent its general use.
What I am saying is that I do not think the eiriden~co is thait~
Qlear. I think that sqine inv~sti~ators have come to the kind of con~
elusion that yçu cribed, an~ not being a phys~ciau, I have no
jndep~R1d~nt op1n~lo~ about whether their e*perieuce is one that
could be generalized to all physicians. I think there is enough room
for doubt th~,t I ~puld be hesitant to seek r~guiation to determine
absolutely that the drug may not be used except in that 1 percent
of cases. I çlo think that there is enou~ evidence against it that
even though we might allow the community to use its judgment
with relatively little restriction, that it would only be appropriate
to do that as long as we are setting about immediately to settle the
remaining doubts.
I am sorry that the state of affairs does not lead me to a feeling
that we know all the answers. I think there are important answers
we do not yet know, and therefore I would be reluctant to go so ~ar
as to say that the use of tolbutamide should by law be restricted to
the 1 percent subgroup.
The CHAIRMAN. I do not think anyone is dealing in absolutes
here, and of course there are all kinds of medicines in the mark~t-
place which are widely used for `nonindicated cases. This, it seems to
me, from what I have heard from the experts is what we are talking
about here.
The conclusions reached at Grady Memorial Hospital was that
there was a very, very small number of cases ~n which the or~I
hypoglycemics were indicated, that the large percentage was better
managed by diet, and that their results' after mor~ than 3 yCars
showed that the patients were better than they ~vere before, and this
is what the TJGDP study indicates,
I assume you agreed that the study was statistically valid ~al-
though being a scientist I am sure you want to say that nobody caji
be absolutely sure, which is of course true. Nobody is ab~oIuteIy sure
about anythii~g, but you' do endorse the position of the Biometric
Society in their evaluation of the UGDP study, is that correct?
Dr. MRIRR. Let me' say that I wholeheartedly endorse the report
that the Biometric Society Committee put out,' and I will disc~i~
that further in my statement.
The CHAIRMAN. Dr.: Palumbo, did you want to comment?'
~Jr. PA~tTMBO~ May `I? ` ` `
As a physician and clinician who is involved in the treatment o~
diabetic patients, I think that we have to make a reasonable ~id'g-
merit on the basis of a randomized clinical trial such as the tTGDP
as to what we are going to do for the pati~nt who sits in front of
us; find the decision here is based upon. the first principle that each
physician is committed to; and that is-4f I may, use the Latii2
phi~ase, "primum non nocere," which "translated means, ~`d~o ii~
narm.
And therefore, it has tO be el~ar that our treatment i~ n~t ing
harm to the patient. Now, we may, under unusual cireuth~tsA4ces,
elect for a risk-benefit ratio, but I think for the majority of our
PAGENO="0014"
13264 ~oM?ETTTIVE PROBLEMS IN ~THE DRVG INDUSTRY
practice and the practice in this country that it should be that these
agents should be curtailed. The UGDP study's conclusions should be
accepted.
Mr. GORDON. Dr. Palumbo, did you people at the Mayo Clinic stop
using these drugs?
Dr. PALUMBO. We have stopped using them routinely. We were
never very "gung-ho" about them in the first place, but we had used
them prior to 1970. When the results of the University Group Dia-
betes Program came out, we accepted the conclusions and, adjusted
our practice accordingly.
Subsequent to that~ a couple of our members of the department of
statistics and epidemiology looked into the matter with a whole
group of people with Dr. Cornfield in the group, and they came up
with the conclusion in 1971 or 1972 that the studies were valid and
that the conclusions were justified despite all of the possible, you
know, flaws or criticisms you can point out with any prospective
study.
We had accepted these conclusions as valid in 1970 when the re-
sults were promulgated. We do not use the agents routinely, only
under the unusual circumstance if a patient says, I absolutely refuse
to take insulin, then we usually assign them to an oral hypoglycemic
agent. I still have this reservation that we are using these agents
solely to control blood sugar; and we are not absolutely convinced
that the control of blood sugar makes any difference anyway.
In fact, that is one of the findings from `the UGDP study that
perhaps blood glucose did not have any relationship to complica-
tions, and so you are introducing' an agent to control blood sugar
which of itself may be harmful to `the patient.,
I think there is no question that this agent has to be curtailed.
The CHAIRMAN. Curtailed, did you say?
Dr. PALtrMBO. Curtailed, c-u-r-t-a-i-l-e-d.
The CHAIRMAN. Thank you very much, Doctor.
We will proceed to the next witness, and as I stated a few moments
ago, feel free to comment on any question asked or any statement
made by other witnesses.
Our next witness is Dr. Henry Ricketts, University of Chicago
Medical School, Department of Medicine.
Dr. Ricketts, we are very pleased to have you here this morning.
You may present your statement however you desire.
STATEMENT 0]? HENRY T. RICKETTS, M.D., PROPESSOR OP MEDI.
CINE EMERITUS, UNIVERSITY OP CHICAGO MEDICAL SCHOOL
Dr. RICKETTS. Thank you very much.
I feel a little embarrassed to read the first paragraph, but I sup-
pose I ought to declare myself as to what I am.
I studied-well, first of all, I am emeritus professor at the Univer-
sity of Chicago Medical School. I have studied diabetes and cared
for patients with diabetes and conducted researches in this specialty
for 34 years. I have been president of the American Diabetes Asso-
ciation and' cofounder and president of the Chicago Diabetes
Association.
PAGENO="0015"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13265
I have served on the study section of endocrinology and metabol-
ism, Grants Division, National Institute of Arthritis and Metabolic
Diseases, and have served as a contributor and an associate editor
of the journal "Diabetes." I think that is probably enough.
My connection with the Committee of the Biometric Society was
that of a consultant diabetologist, and I attended most of the meet-
ings. I was struck by the thoroughness with which the members of
the committee made their investigation. I detected no bias for or
against the UGDP study. The committee listened to more who
criticized the study than. to those who were less opposed or favor-
able. The committee did not hesitate to ask the coordinating center
in Baltimore for raw data when a point was in doubt, and members
made trips to the center and to several participating clinics to check
methods, procedures, and results. No uncertainty `was too small to
leave unresolved.
I should remind you that the TJGDP was set up to determine
whether various treatments for diabetes would minimize the mainly
vascular complications that notoriously accompany that disease. It
is ironic that a full report dealing with complications has not yet
been published because, in the third and fourth years of the study,
an alarming preponderance of deaths had accumulated in the tolbuta-
mide group. The investi.gators, then, per force, had to turn their
attention to mortality and survival.
I was not a participant of the TJGDP study, but I followed it
closely. Despite some imperfections, I think that the results' aIld
conclusions of the TJGIDP have shown tolbutamide and phenformiti,
and probably their cousins, to be dangerous drugs, especially when
taken for extended periods of time. I stand by my opinion of 4
years ago, expressed with the help of a committee of the America~i
Diabetes Association in the editorial statement accompanying the
first report of the TJGDP. I quote: "The TJGDP mortality study
shows that death rates were essentially the same in the IYA:E~
group"-I suppose I have to explain that.
Mr. GORDON. Is that the insulin variable group?
Dr. RICKETTS. Yes. I will explain that later.
The TJGDP mortality study shows that the death rates were
essentiafly the same in the people who had various dosages of insulin
and which maintained more nearly normal fasting blood glucose
levels than in the more poorly controlled groups of the placebo and
the other groups.
This would appear to mean that efforts to establish good control of hypogly-
cemia in the kind of population studied had no effect on mortality.
The real lesson of the data is that if diet plus insulin does not reduce inor-
tality below that experienced with diet alone, it Is highly improbable that oral
hypoglycemic agents will do so.
There is indeed no doubt about the reality of the greater number of cardio-
vascular deaths observed in the TOLB group as compared with all other treat-
ment groups. Inquiry into the reasons for this has been both intensive and
extensive. Aside from the most proximate explanation, that tolbutamide may
have been directly and solely responsible, the possibility that the tolbutamicle
population, by chance and despite randomization, entered the study with more
or greater risk factors than the other populations bad to be scrupulously in-
vestigated.
Although this possibility has, in the opinion of the ADA Ad Hoc Editorial
and Advisor~ Committee, not been excluded, the weight of statistical analysis
PAGENO="0016"
132(~ Co TIY~ p~j~ ~
makes It prc~bai~te that the ~ces~ ro~a~cuku~ mortality Fa TO~B is at-
tft~Yl~ either to 1i~e dr*g ~t~elf o~ to t~n idere4 and unknown ~aetors~ Iii
U~e absence of evidence for the Fatter, suspicion would naturally attach to tol-
butamide.
The WQrthilty study is at least s~~ge~tlve enough to put a damper on What
appears to be thC indiscriminate use Of all oral bypoglyceni&c age~ets in the
treatment oi~ mild or moderate, adult-~mset dia1~etes. Mtboi~h tolbutamide, for
practical reasons; has been the only sultouylurea drug investigated by UGDE.
This is 4 years ago.
The chance that other compounds of this fami1~ may be shxitlarl~ lnvol~ed
ca~inot be dismissed despite differences in molecular structnre,
Jt wou~4 not be justifiable at this point, bowe~er, to proiiU$t the mamiJfaetlire
and use of su1fony1w~ea drugs, for they will i~robably continue to fill a need in
special circUmstances.
If these drugs are dangerous, what course should we take? `You
have just heard that their manufacture of the drug should not be
forbidden, and for reason. For exampJe, how do we treat a diabetic
patient who ought to be taking insulin but is living alone with a
broken, or amputated, or paralyzed arm that prevents him from
using a syringe and needle? One who is blind and cannot measure
his dose of insulin? One who is old and tremulous? One who is
mentally disturbed? And finally, one who refuses to take insulin,
In another vein, there are diabetics who are engaged in hazardous
occupations and ought not to take insulin for fear of reactions.
We ought to make allowance for these j~atients, even though the
oral agents are not very effective and, I believe in the long run, may
be harmful.
The CHAIRMAN. Does this list of exceptions include most or all of
the exceptions that you could think of?
Dr. RIcK~'rTs. Well, I think so, yes. I might think further, but
that is quite a number.
The CHAIRMAN. All right, please go ahead.
Dr. RIOEETTS. But if we continue to make these agents available,
as I think we must, how do we protect other diabetics who would
like to use them but should not?
Insulin comes to the patient with a package insert that carries a
great deal of information, incli~ding certain warnings. The oral
agents come to the patient in silence because they have been re-
garded as innocuous, needing no instructions except the doctor's
directions for dosage and timing. This must change.
But it is the physician who should lead the way, and I hope that;
the report of the Biometric Society will in time convert the many
~urremt unbciiever~. Meanwhil -~$d this might seem t~ be prepos-.
terous-it zthgl~t not be too radical to ask the FDA, under proper
authority, to transfer the oral. hypogiycemi~ agents to the circum-
scribed schedule II of dangerous drugs along with barbiturates,
amphetamines, ~nd certain narcotics.
Phy~iciaus might learn that the oral agents are not exactly safe,,
and the requirements of BNDD prescriptions, if for dubious need,.
might become a salutary nuisance. This arrangement, of course,
would have holes in it-and I can see some-but It might have the
effect of helping to reduce the use of a product that too many pa..~.
tients c~uid well do without. ` . . .
PAGENO="0017"
CoMr~IVE PR~E~f~ i~( ~E~i~' mtTia; ~D~STBY I~32~7
Th~ ~CIA~R~A~. Than~k yo~ very much, Doctor. Our next w~itaëss
is D~. Paul MciM~, ~Dcpartment ~f Statistics, Tjniversity o~ `Chiçngo.
Dr. 2~feier.
STATE1YIERT OP PAUL HEIER, PH. D., PROPESSO1~ O~I? STATISTICS,.
UNIVERSITY 01' CHICAGO, CHIcAGO, ILL.
Dr. MI~ER. Mr. Ohairman, I speak as a member of the Biom~tri~
Society Committee on biometric aspects o~f controlled cliaiical t'±ials
of hypoglycemic agents, which report is under discus~ion ~todny.
Professor White has outlined our problem and our findi~gs.
Professor Zelen will speak about some of the particular criticisms~
made of the UGDP report. I shall speak a little more generally
about the role that I see for clinical trials in guiding our decisions
about tnod~s of therapy.
It happens that in March of 1970 I testified before this cOmmittee
on the subject of risks of thromboembolism due to the use of ~ai
contraceptives. I spoke then of the deplorable lack of prospective
controlled clinical studies on the effects of oral contraceptives. I
discussed possible reasons for that lack. Let me quote a few lines
from that earlier testimony.
I said:
Frankly, the required research, although important, is not especially appeal-
ing to scientists. It is not fundamental and it is not exciting. It is diffidult,.
it is expensive, and It is fraught with the risk of attack from all sides. Who
would willingly prepare himself for such a study, make an application to be
weighed competitively with others on scientific merit, and risk the loss of
support halfway through the study when a review committee with diffei~ent
views or priorities comes to consider renewal of support, all this when he stabds
to gain so little in scientific recognition or otherwise?
Evidently, for whatever reasons, there is no sound' body of scientific stu4les
concerning these possible effects available today, a situation wbic~i I regard e s
scandalous. If we proceed in the future as we have In the ~last, we will contlime
to stumble from one t~ntative and inadequately supported conclusion to anotl~er,
always relying on data which come to band, and which were not designed ~or
the purpose. The planning of better studies is difficult, and the recruitment
of investigators willing to commit their efforts to these purposes may be mOre
difflcttlt still, I believe both are possible and essential to the public welfare.
At the time those words were written, I had no knowledge of
the UGDP, but they could scarcely have been more apt.'
Let me interpolate in my prepared statement my warm commenda~-
tion for the group of, pi~ysicians and statisticians who undertook the
`UGDP study. With whatever limitations, this i~ far ai~d, away tlae
best evidence we have to date on tolbutamide toxicity. it `is an ~xcel-
lent study. No o~ie study can answer all ~f ~the relevant questiol3s,
but that is scarcely the fault' of these investigators, and 1 am led to
modify my statement about the lack of excitement a~d, interest that
such studies could generate.
I think this group has shown us that there is new ground to `be.
broken through some of the work that they have done in the theo~y
of the conduct of controlled clinical trial~ and they `have also con-
tributed substantial new knowledge. to1 an importa~it medical
probleni.
~T return to my statement. - , -~
56-592--75----2
PAGENO="0018"
13268 COMPETITIVE PROBLEMS IN TIlE DRUG INDUSTRY
The CHAIRMAN. May I ask a question? You state that although
the UGDP study has its defects; it is an excellent study proving the
case against tolbutamide. Is there a comparable study that proves
the case for tolbutamide, really?
Dr. METER. No, there is not.
The CHAIRMAN. What did you mean by that, then? Is there no
case? You were so equivocable in what you were saying awhile back
and now I do not quite follow you. You endorse the IJGDP study,
but then you say the case against the drug has not been proved. Do
you mean absolutely proved 1,000 percent, or is it 99, or what? I
cannot follow your testimony at all.
Dr. METER. I understand your question, Senator.
A major point that I hope to leave with you is that in this area of
clinical research we will often feel obliged to stop a study before we
achieve a high degree of certainty. We wish it were otherwise. It
would be very nice if we could say for certain. "These are the facts.
Now everyone must fall into line and follow the facts." IJnder the
circumstances we find that we must make decisions in the face of
substantial uncertainty. Whereas I believe that the TJ&DP is the
best evidence that we have, I believe that the study was indeed ended.
before we could be certain. Take note that I am not trying to make an
especially cautious statement about a virtually proven fact. The
evidence of toxicity is substantial, but in itself by no means
conclusive.
The CHAIRMAN. Before you could be certain what?
Dr. MEIER. That the drug is toxic. Before we could be dead certain
of that they pulled it off the study~
The CHAIRMAN. Before you could be certain that the drug was
toxic.
Dr. MEIER. Yes, before we could be certain that it causes heart
attacks. The evidence pointed that way but before it was certain, in
my opinion, they quite properly withdrew tolbutamide on ethical
(grounds.
Senator, I wish I could say that a good study necessarily gives a
solid answer to a reasonable question. A good study, ethically done,
may leave us with considerable residual uncertainty. I am sorry if
that is confusing but I feel that that is the circumstance.
The CHAIRMAN. It is confusing. I suppose you are familiar with
the Kefauver amendments of 1962. In 1938, the Congress, because of
the sulfanilamide disaster, passed legislation that there should be
adequately controlled studies to prove the safety of a drug before it
is marketed. Then in the midst of the dispute over the Kefauver
proposals the thalidomide case arose and the Congress passed legis-
lation that there has to be adequately controlled studies to prove the
efficacy of the drug.
I think most scientists agree that this is sound. You should not
put drugs on the market that are not safe, safe by a scientific meas-
urement in a cost-benefit ratio. Any active compound, as everybody
knows, has side effects and may be serious.
So we are dealing with a situation here where the question is do
you put into the marketplace for broad usage or even a narrow
usage a drug for which the efficacy has not been proved by carefully
controlled scientific studies? There are no adequately controlled
PAGENO="0019"
COMPETITIVE PROBLEMS IN THE DBtTQ INDUSTRT 13269
studi~s that prove it. Yet there is a comprehensive 10-year study
that raises a very serious cloud over both. the safety and efficaQy of
this class of drugs. That is what we ai~e dealing with, is it not?
Dr. M1~izR. indeed, it does raise a very serious clOud but you seem
to be urging me to conclude that it is proved, and there is qmte a
diffei~eiice between a very serious cloud and proof..
The CHAIRMAN. I am nOt trying to do that at all. What~ I am try-
ing to urge you to appreciate~ at least is what the. law is, and th~t is
that you do not int~oduce active compounds. for use in ñiedical~ prac.
tice and use them broadly unless there is proof they do some good,
~ind particrdarly when there seems to be some serious indications
that they do harm. That is the issue here, is it not?
We used to put drugs into the marketplace prior tO 1938, and there
was no proof of ~afety and no proof of efficacy. And in the whole
history of the development of drugs down through the history of
mankind there is hardly half a dozen of them that survived as being
safe or efficacious. Most of the drugs people have taken for hundreds
of years had no efficacy at all. They might have been safe because
they did nothing. . .
But we are dealing with a question here of a study that indicates
there are serious side effects and a study that indicates that there
does not appear to be any possible usefulness except in limited ca~es.
That is the issue we are dealing with.
Dr. METER. I agree, and I think the difference we are arguing
about is the difference in how solid the evidence is. I would further
agree that we need to define policy in the face of uncertainty, that we
cannot wait for final proof.
The CHAIRMAN. Let me ask you this question. If you had the
UGDP study before the drug was marketed, do you think it would
be marketed under the law?
Dr. METER. I doubt it.
Shall I continue?
The~ CHAIRMAN. Yes. Go ahead.
Dr. METER. It is true that the UGDP had defects. It is true, alsO,
that it falls short of proving the case against tolbutamide. Noneth~..
less, as Professor Cornfield remarked in testimony here last Sep~
tember, the UGDP today provides the best available information
on the possible toxicity of tolbutamide.
As to defects, there are no studies which are entirely free of them,
and it was the judgment of our committee that this study was well
conceived ai~d executed, and that those defeats we could identify
did not give reason to doubt the findings.
As to it being inconclusive, that was inevitable in the nature of
the case. Once the investigators became convinced that there was
substantial evidence of toxicity, and not of corresponding benefits
th~y had no choice but to withdraw the drug.
I hus we are left with an ominous yet inconclusive result, and I
believe that this is a typical outcome which we may~ expect to see
repeated in many other instances. It may be, in such a case, that the
commun~ity of physicians will decide that, although not conclusive,
that the evidence is sufficient to abandbn the drug. Or, on the con-
trary, as in the UGDP case, they may conclude that the evidence
does not require them to give it up.
PAGENO="0020"
1327(~ COM~iTIV~E I~EtOELEM?S IN `rBE ~E~RtY~G i~m~nsm~
in The ~la4~ter ~oase, ho~w~ver, I ean ~ee ~no a1tei~nativ~ to The ~nitia~.
tion of ~a ~new cl~ieal t~rial, oxiadu~ted ~y ~ s~cian~ iancommiced by
the first one. I ~houk1 extpect, ~fl any ~event, that i~boi~ p~hysicians rntd
pa~tient~s ~h!oqild be made as ~nHy in~formed abaiwt fhe e~d~enee i~s is
feasible.
I go so far as to* hope that fbhe experience to date ~ith ~orai hypo.
glycemic drugs may convince ins That &inicai trials should he a
ooiit~uin~ oom~pone~t of d\rug .eurveilance for any dr~g, from the
first 1da~ of its release, and so long ~s substanitial doubt abo~t the
baia~ace of risks and bene~ts remains.
~)he Cu~AIR~Aw. I think everybody would a~iree that your last
sentence would state an ideal situation which we would ~ii hope
someday would be a~hi~ed.
`I~r. M~n. Seiaatqr, I would hope that day would be ea~y rather
than late. I sp*e sentiments like this 5 years ago before `this
committee. I de~eri~hed in some detail ways in which authority might
be given to the FDA, and methods by which the funds could be
allocated to such studies. I was pleased to see that in testimony in
September Dr. Prout argued along quite similar ~ines. I do not
think I see anything in the line ot legislation that would tend to
move u~ in that direction, and I would hope there may be some.
The IO~UIRMA'N, I do not ~think we need the legislation, but prob~
ably do need the money. But I think There is no doubt that it would
b~ ~very sound to start good clinical trials once `a drug is marketed,
because if there is not, we `would have to rely upon the reports of
physicians' observations around the country. It may take a long
time for individual physicians to accumulate enough data to associate
with some adverse e~ect because individual observations would have
to be reported thirou~h medical journals or to each other, and that
would take quite a while. Your recommendation is very sound and
I do not believe anyone would disagree with you on that.
Dr. METER. I would just like to point out that in this case it
depended upon an interested academic group, physicians and statisrn
ticia~n~, to decide that it ought to be done and to con'vince an NIH
study section' that it ought to be funded, `at ~quite a high price, in
N~IH terths. That seems to me `to `be an unacceptable way `to operate.
If such `a drng is to be marketed, the sales of that drug not simply
the taxpayers' money, should contribute to carrying out a study. I
think there are proper ways in which that obligation could be laid
upon the. manufacturers who are selling the product t~ see that the
funds are supplied, not `because they feel like it, but because they
must do so. And that is the kind of le~i~lation I would hope to see.
The O~IA~R~1AN. As you might recall if you read the testimony in
additi~n `to "the testimony you gave yourself 5 years ago, ~when I
r~sed the ~iues'tion ~ibout `studies `to detet~mine how many thierograms
of estrogen could be put into' an ora~l ~ontraoeptive and still ~be
e~1FeQtJive, the answe~ was, well, it would be very hard to gst vOlun~
teers `to ru~i that risk. I `do not `think that is the case. I think there'
would be plenty of v~lunteers ~ho are seriously concerned about
whether or not' they got pregnant now or 6 months later, who `would
be put j~to a, `test ~ see whether you could dramatically reduce the
ipiqr~grarns of ~estrogen in the~' oral contraceptive, and it seems quite'
PAGENO="0021"
COMREW[~IV~ J?I~O~LBMS IN THE ]~EUG W~UST~Y 1a2Th
n~tt~ that s~ fa~ M I 1rnow,~ i~~ ex.periment~ of th~t~ kind ~
ye1~ been made, ai~d ~e h~d~ 15Q microgi~u pills in ther mark~tp1&ee,
wi~ile~ Ein~tuid ~ve~ ~head~ wit1~ ~O, We tpok testimoi~y~ front au
English scientist.
But you are absolutely rights we have not done the kind of eon~
trolled studies we ought tOr do.
Thank youi very much for your testimony.
Our next witness is Dr. Marvin Zelen, Statistical Laboratory~
~UNYAB, Amherst,. N.Y.
STATEME~IT OP l~ARVIN ZE~EN, PH. D~ PROPES$O~ OP STAT*U~
CAL SCIE1~OB A1~D DIREC~OB.,~ STATISTICAL LAB ATOB~Y~ ~4~&
UNIVERSITY OP NEW YORK AT BUFFALO
`Dr. ZELEN. Senator Nelson, thank you for tkis oppàrt~unMy to
appear before this committee. My general eomment~ w~li be divided
into two parts.
The first topic I wish to comment on is how is it that able ~nd
respected clinicians can disagree with the interpr~tation of the
UGDP data ~ The tolibutamide cardiovascular death rate us more
than dottble compared to ather treatments~ Yet many clinicians ~ho
treat adult onset diabetes find it difticuit to accept ~nth a fig~re.
For many of them, this elevated cardiovascui~r death rate, does z~ot
~ppear to have been perceived in the clinic.
I would like to examine other factors which may lead to ele~tatea
cardiovascular mortality. According to the UGDP data,. the eardi~o~
vascular death rate for individuals above the age of ~3 is appro~i~
mately five times that of individuals. ~3 or younger; pe~ipie with
nrteriai calcification at time of diagnosis have fou±~ times the cardip..
vascular death rate compared to those' without arterial caiciflc~atio~i;
the initial glucose tolerance test, called GTT, as used by the UGDP
investigators, shows that those with a GTT above ~&, the median
value, have double the rate of cardiovascular deaths conipau'ed ~o
those who have a GTT below the median; men have a donbl~d
cardiov~ular death rate compared to women. Although the nun~.
bers quoted are rounded for simplicity, it is clear that in the c1in~
there are many factors simultaneously influencing cardiovascular
deaths~ Several of these have greater or eqnal effect on th~ cardiO~
vascular death r~t~ compared to the effect of tolbutamide~ As a resuit~
it would be difficult for a clinician to' perceive an elevated cardió.
vascular death rate associated with tolbutamicle. Such an effect wmi4
be almost completely obscured by these other important factors.
Onl:y if there i~ careful and st~uctttred recordkee~ping on a Iarg~
numI~r of patients would a changed cardiovascular death rate o~
two to three be detected. The analysis of such multifaceted chtta re~
quires more sophisticated data analytic methods than those in com~
mon usage by clinicians.
Next, I wish to comment on some features of the Biometrics
Society report. A criticism of the original. UGDP ~nalysis is that i1~
failed to explore the effects of several factors acting simultaneously
on the cardiovascular mortality. Our committee did in fact consider
this matter very carefully. We found that when one examines the
PAGENO="0022"
13272 COMPETITI\~E PR0BL~MS IN TH~ DRUG INDUSPR~
group of older women, age greater than 53, the tolbutamide cardio-
vascular death rate is almost five times that of the placebo group~
It is in this group of older women where the tolbutamide excess
cardiovascular mortality is most dramatically shown.
Finally, I wish to comment on the problem of planning and analyz-
ing clinical investigations in which patients are expected to be on
chronic medications for a period of many years. It is important in
planning these long-term studies to allow the clinician to change the
medication if it is in the best interests of the patient. This can result
in ai~ altered dose or even a change in the medication. The UGDP
protocol did allow the clinician this freedom. A protocol which does
not `allow this flexibility may not be in the best interests of the
patients under study.
The, CHAIRMAN. In evaluating this study, did you or did you not
conclude that the authority of the clinician to alter protocol, which I
assume some did, had any adverse effect, or did it prejudice the study'
in any way?
Dr. ZELRN. No~!'
In addition to modified or changed medications, patients may, on
occasion, not take their medication at all. In' the Biometrics Report,
these problems were examined in considerable detail. It is our con-
elusion that the greatest statistically significant difference between
tothutamide and placebo occurs in the group who have taken their
prescribed medication in exactly the manner specified in the protocol'
for the entirO period `of followup.
* To conclude, I wish to state that the interpretation of the data is
difficult due to the small number of deaths relative to the total num-
ber of patients. In our endeavors we have analyzed the data in many
other ways which have not been put in our final report. Our conclu-~
sion is' that the weight of eviden~e points to tolbutamide as being
responsible for the excess cardiovascular mortality.
If I may, Senator Nelson, I would like to comment on some general
aspects of clinical trials that have, surfaced during' our di~cussion.
Obtaining scientific evidence using the clinical trial method is the
most difficult way of obtaining scientific evidence and should be' used
only as a last resort. I speak from long experience. My `research
group, the statistical laboratory at the State University of New York
at Buffalo, is involved in over 60 clinical trials at the present time
in all areas of cancer treatment. It is very difficult, `time cOnsuming,
there is a great~'deal of aggravation arising from the vagaries of the
funding agencies~ ` . ` ` ` `
I think to' mount long-term studies, either of oral hypoglycemic
agents .or anything else, shOuld only be taken after much carelul
thought and after all other' ways of attempting to obtain such evi-
dence have been `thoroughly examined. Mounting these trials should~
not be done very casually. ` ` ` ` `
The CHAIRMAN. Thank you very much. `
Our next witné~ss is Dr. Palumbo, the' assistant professor of medi-
cine, Mayo Medic~'al School, Rochester, Minn. . * *
You may present your statement however you desir~ and e~ttem-
porize on it if `yo~i `desire. ` , ` ` ` ~`
PAGENO="0023"
COMPETITIVE PROBLEMS IN THE DEtG INDUSTRY 13273~
STATEMENT OP P. i PALUMBO, M.D., ASSISTANT PROFESSOR OP
MEDICINE, MAYO MEDICAL SCHOOL
Dr. PALUMBO. The comparison of treatment for a disorder can't
only be `evaluated through controlled, randomized, clinical trials..
Hints and leads from retrospective `studies can be extremely valu-
able in leading to a new hypothesis and may be the basis of justiflàa-
tion of a randomized trial. However, standing alone they car~not
form the basis of any firm conclusions concerning treatment, effects.
The preliminary analysis of our data of the incidence, prevalence~
and mortality of diabetes mellitus in Rochester, Mimi~, between 1945
and 1970 contains some hints that survivorship may be `lower in
diabetics on oral antidiabetic agents, and we grouped them all to-
gether: These are sulfonylureas and phenformin.
Mr. GORDON. About how many people were you following?
Dr. PALtTMBO. We were following over 1,000 [1,090 to be exac'ti
patients with diabetes over that 25-year period. There were only 138~
on oral agents out of that group.
Mr. GORDON. How did ,they fare?
Dr. PALUMI~O. Their survivorship was less, but however there are
group differences that have `to be taken into account, and therefore'
we cannot make any firm, conclusions, Our statisticians are very loath.
to leave themselves open to the, criticism that a retrospective study
can lead to firm conclusions [regarding treatment].
All we can say is it suggests or hints that the oral agents pltis
other factors may affect survivorship of the diabetic. As a clinician-k-
and I am deviating from my statement-as a clinician, I would e~-~
pect that the oral agent group would be similar to the diet group~
the same group, the sanie isehemic heart disease, the sathe hyper~-
tension, et cetera, and I would have expected them [patients on ora'
agents] to have the same survival curve as the patients on diet alone.;~
that is, the oral agent group should have been similar' to those on diet'
alone.
However, the' survivorship of those patients on oral agents whei~
compared with a group of the general population, similar in age anc~
sex ~or our midwestern area, the death rate or rather relative sur-
vivorship for the group of diabetic patients showed that the ora~
agent group was much lower.
The CHAIRMAN. Now, wait a minute. Ton. said the `death rate and~
survival. You cannot have it both ways.
Dr. PALUMBO. Their survivorship was lower.
The CHAIRMAN. The higher `incidence of death.
Dr. PALUMBO. That is right,' and in the fir~t 3 years there was a
difference in the death rate for cardiovascular mortality in the oral
agent group, or there was a higher death rate from cardiovascular
deaths. ` ,
The CHAIRMA~T. This was retrospective? .
Dr. PALtTMBO. This' was retrospective. The groups `are not corn-
parable. The, insulin `group `is yot~nger, has a higher blood sugar
and in our study has a higher p~rcent'age of stroke actually, whic,h,
should f~ror a poor survivorship. The oral agent ~nd diet group-
PAGENO="0024"
~3274 `cOMPETiTIV~ P~O~LEM~ LW T~E D~RU$ INDUSTRY
and remember w~ do riot 1ia~ a~ placebo to compare this with, so
that for the di~t-~orai agent groi~p,~ they are pretty comparable with
regard to ischemic heart disease. There is Tess strOke in the oral
agent group than the diet group at th~ time of diagnosis of diabetes,
but retinopathy was higher in the oral agent group, and blood p~s~
sure was 6 percent higher in the oral agent group.
These are group differences that have to be taken into account. A~U
we c~an say from our study is that it suggests that the oral~gen~s may
be one of the factors that may adversely affect survivorship in the
diabetic.
Such an observation-I am returning to my statement now-such
an observation would point to the need for controlled, rando~mzed
~chiniaal trials to study the possible adverse effect of variou& treat-
ments on survivorship in the diabetic.
The University Group Diabetes Program was a ra13dom~ze~ trial
study. to evaluate the influence of treatment on diabetic eoznplaca-
tions. A statistically significant, adverse effect on survivorsliip was
noted after patients had been on tolbutamide and phenformin for 5
or more years. These data have been reviewed by others~ and the
review was published in the journal, I believe, "Diabetes~" by Dr.
Cornfield [the journal was .JAMA, 1971] and also~bad been reviewed
by our own statisticians, and the conclusions have been fonnd to be
sound. I have to rely on their conclusions because I am a clinician
and not an epidemologist or stati~tieian.
Was there a question, Senator ~
The CItAIRMAN. You concluded that the UGDP stucI~y was sound?
Dr. PAWMB0. The conclusions are sound; that is correct. In my
~opinion as a diabetologist, another randomized trial study of trcat~
ment in diabetes is not ethically justified, as the data from the
University Group Diabetes Program clearly indicate, from my stand~
point, an adverse effect of the oral antidiahetic agents on survivor-
ship in the diabetic. The use of these oral agents, therefore~ should b~
curtailed.
The CHAIRMAN. How long after the UGDP study was published
did the Mayo Clinic conclude that they would not use the oral hypo~
glycemic agents except in special circumstances?
Dr. PALuMBO. There was a meeting of the American Diaheto~
Association, I believe-and maybe Dr. Ricketta can correct me-ia
.June 1969, was it, that published those results, or maybe it was the
following year.
Dr. RICKETTS. 1970, actually.
Dr. PALUMBO. It was 1970, and subsequent to that time we began
to inform all patients about the risk involved with the use of the
oral agents~ We took patients o~ the oral agents and tried them on diet
alone afte~ir iAfonning them of the possible risks imvo~ved~ We have
not followed those patients to see how they have done, except thal~ our
own clinical impression is, as ~r, 4d~n hiaa ah~ea4y rp~b~d~from
his committee, that they do iust as weil, and I did not feel that a lot
of these patients needed to be on oral antidiabetic agents.
When. the plasma glucose. or their response to tr~atxueut to diet
~has not been satisfactory, we have advised insulin therapy, because
we feel insulin at least does no harm. Even if it has been shown not
to do any good, at least it does not do harm. It does protect the pa-
PAGENO="0025"
co~1~mVE ~OBLEMS i~ ~ p~c~ ~ 13275
tient from the acute complication, of ketoaci~sis if they are prone to
that. We would never have used these oral agents bt the ketoacidosis'
prone patients anyway, but in any event if I were to err now, I
would err on letting tho blood sugar drift a little bit upward aM
not worry so much at keeping it at a certain partienlar level.
And tl~erefore, our observed policy has been t~ curtail the u~e of
these `agents. I do not use them routinely. I tai~e patients o~ when
they are referred to us. We warn them about the possible hazards, and
we transfer them to insulin therapy. We are wore ~ tertiary center
than a primary center. We are describing here in the study patients
alluded to, our own patients from Rochester, Minn., from a popula~
tion of about 50~0O0, so we do provide primary care for that popular
tion, but the majority of our patients that we see in th~ diabetes
clinics, which number about 8,000 to 10,000 patients a year would be
told exactly the same thing, that the oral agents may be deleterious
to their health and that we would recommem&d, if t alone does not
control their diabetes, that they are placed on insulin therapy.
Most of our patients have been willing to accept this when we
have shown them how to adinjimister the insulin~ There has been no
particular problem.
It certainly would be nice to administer an agent orally and take
care of the blood sugar, but unfortunately if the agent has b~en
shown to cause an increased mortality from cardiovascular death,
we would be reluctant to use this agent.
As I stated previously, I feel a physician should do no harm.
The CH~th~AN. Do I understand you are saying t~iat `this posi~
tion is a policy of the clinic?
Dr. PALTJMBO. Well, as a member of the diabetes committee of the
institution, it is our recommended policy. Obviously, I cannot speak
for the 400 or ~O0 physician.~ we have on stalL There way be sothe
who migl~t be, but I think we have disseminated the information
widely in conferences and through memoranda,
I believe the position is pretty clear that we have accepted the
findings of the University Group Diabetes Program, that patients all
must be informed `about the hazards of these drugs and that only
under nisusual circumstances would they be prescribed.
There would be very few patients that woul4 not see us in the
diabetes `clinic, so that it is impossible that a smafl group of pa-
tients. might be treated with oral agents. I rather doubt that, since
we maintain close contact with all of our colleagues and dis~minate
mtermation through conferences audi memoranda.
The Ci ~ Does'anyone on the panel wish to mahe an obser-
~at~on on any of the points that have been raised thus far in the
testimony or on any questions that have been asited.
Dr. Bieice~'rs~ Yes, Senator Nelson, just a rather small point.
It ~s well J~n~wn that a great many people, and this is particularly
women, are overweight. I mean to say diabetic people. W~ struggle
and p'~ach and do all we can to~get them to lose weight, and fi~~ll~r,
some of them do. It does not last awfully long but nevertheless the~r
do, And of co~øse if the ~1~~ity is controlled, the blood sugar goes
elow,n~, and `this is what we want, And thus it went fo~r a lo~tg, long
time until tolbutamide came in, and then what happened? ,Doctm~s
began to give tolbutamide and tell them it is good for them, and
PAGENO="0026"
13276 ö0MPETITIVE PI~BL~MS IN THE DBTJG INDIJSTRT
they began to take them. But what did they do? They took their
toitbutamide but now they did not think that they needed to diet.
And this is very sad. It is a poor outcome of this business we are
talking about.
The CHAIRMAN. Thank you.
Dr. Zelen, did you have a comment?
Dr. ZELEN. Yes. There have been some who suggest that another
UGDP-like trial be mounted.
The CHAIRMAN. I am sorry. I did' not get the first part.
Dr. ZELEN. There have been some individuals who suggest that an-
other UGDP-type trial be mounted. Judging from the experience
with this one, it is likely to take 6 to 10 years before any conclusions
will be reached.
Furthermore, with the recent change in patient consent, people
face the following situation. If an individual comes to a clinician
who is participating in such a trial, the physician, by law, has to in-
form the patient of the risks involved. The scenario would go some~
thing like this. The physician would state:
There are a large group of people in the cuuntry who believe that tolbuta-
mide may be dangerous. A study has been completed purporting to show this.
However, there is conflicting evidence to believe that the Interpretation may be
in doubt. Consequently we are going to try again.
Well; I think most people would not like to be part of such a
scheme, and it might be very difficult to enlist patient volunteers..
The CHAIRMAN. Well, was there anything in evaluating the TJGDP
study that would indicate some necessity for repeating the same
study?
Dr. ZELEN. In my opinion, no!
Mr. GORDON. May I ask a question at this point?
`Dr. MEJER. I would just like to clarify my own position. I have
not taken a position on whether there should be restrictions on
tolbutamide. What I have said is that if it is to continue to be
widely used, then 1 think it imperative that another study be
mounted. I hope I make `that clear. It is now being widely' used
long after the UGDP report was `published and discussed. If that
situation is to continue, then I would see no ethical choice for those
who use it but to mount another study.
Mr. GORDON. But `what are they going to do in the meantime? Are
they going to keep on using it widely while the 6- or 8- or 10-year
study goes on?
Dr. MEIER. That is a matter that `I presume the FDA is actively
studying right now. The report of the TJGDP appeared, received
commendation from the ADA and the AMA, but as a matter of fact;
the community continued to use the drug. Barring administrativ~
action, I presume they: would still continue to use the drug~ and I
~m saying that if there is no action to prevent that, then I think
there should be action to further study the matter.
Mr. GORtON. How about the newer drugs that have not been re-
leased yet?
Dr. MEIER. I would hope that they would be `studied. If they are
to be released, I would hope that proper studies would be initiated
immediately. ` ` ` `
PAGENO="0027"
COMPETITIVE PROBLEMS IN TUE DRUG INDUSTRY 18277
Mr. GolmoN. In the Keen study, in the placebo group 30 percent
were over 70 years old, and in the tolbutamide group there was 18
percent who were over 70.
According to the Biometric Society report, the difference is sta-
tistically significant at the 5-percent level. Do you know the com-
parable differencç~ in the UGDP? It was much smaller, was it not?
Dr. MEIER. Yes, the IJGDP was conducted with, very careful ran-
domization, as described in our report. The Keen study used a much
more informal kind of allocation scheme, and indeed, in respect to
age, the Keen study had a much wider discrepancy between the
groups than did the UGDP.
Mr. GORDON. So the baseline characteristics were more similar in
the TJGDP than they were in the Keen study.
Dr. MEIER. Yes.
Mr. GORDON. I just wanted to clear that up.
The CHAIRMAN. Is there any other observation any of you gentle-
men have on any aspect of this?
Well, the committee wants to thank you very much for taking the
time to come here and present the results of your study for the
record of this committee. We appreciate it very much. Thank you.
If you have anything supplementary that occurs to you that y~u
think will be useful for the record, the record will be opened for
another 2 weeks and you may submit it for printing in the record,
Our next witnesses will be Dr. Robert Bradley, chairman of the
Committee on the Care of the Diabetic, Joslin Clinic, Boston, Mass.,
and Mr. Neil Chayet, counsel.
The committee appreciates you gentlemen taking the time to ap-
pear before the committee. You may present your statement however
you desire. It will be printed in full in the record.1
* STATEMENT OP NEIL L. CIIAYET, COUNSEL, ACCOMPANIED BY
ROBERT F. BRADLEY, M.D., CHAIRMAN, COMMITTEE FOR THE
CARE OP THE DIABETIC, JOSLIN CLINIC, BOSTON, `MASS.
Mr. CHAYET. `Thank you very much, Senator.
My name is Neil L. Chayet. I am a member Of th~ law firm of
Chayet & Sonnenrei~h, and I appear as counsel for the Committee
on the Care of the Diabetic.
The CHAIRMAN. Counsel for whom?
Mr. CHAYET. The Committee on the Care of the Diabetic.
The CHAIRMAN. Are you also counsel for the Medical Tribune?
Mr. CHAYET. I would be glad, Senator, to submit it, a list of all my
clients,' if you care to subpena it. I still believe, however, in the con-
cept of attorney-client privilege and the confidentiality of that' rela-
tionship, which I kno* does not mean that much around here
anymore, but I still value it very highly.
If you care to sub'pena a list of my clienth, I would be glad to
provide it. We have many clients and represent many groups, indi-
viduals, publications; and others.
The CHAIRMAN. You are the first witness we have had who is
embarrassed `about .wh'om he represented. , "
1 prepared statement, page 13620.
PAGENO="0028"
13278 ~Q~TI'~IVE ~RO~4EMS ~N ~ PEUG ZNPU~TRY
Mr. Ci~~r. I am not embarrassed at all. As I said to y?~i, I am
willj~g to prpyide it nuder ~ proper subpei~. I believe in confi~
dentiaTity and that is my response to that question.
May I oontimie, sir?
The CH~ni~Aw, Go ahead.
Mr. CiI~Y~T. Thank you..
With me is Dr. Robert BraclJey, wh~ is director o Joshu Clime
and who is a1~o the eh~irman o~ the ~ommitt~e on the Car~ of the
Diabetic,
I have a written statement which I would like to submit and ask
that it be printed in full in the record.
The CII4nP~EAN, It will be printed in full in the record as though
read and you may present extemporaneously whatever you desire.
Mr. CHAYET. Thank you very much, sir.
This matter has continued~n~w for nearly 5 years and the Com-
mittee for the Care of the Diabetic as well as physicians and s~ieiitists~
throughout the country have been engaged fully in this controversy.
When my involvement began, ~t was solely as a lawyer for a client.
I now have another interest in this matter which I would like tc
disclose to the committee at this time.
Since I began handling this matter in 1971 my mother has been.
diagnosed as a diabetic, and she has become very severely ill and~
crippled by this disease; and so while I still function as an attorney
heçe, I also have a personal interest in this matter because of this
situation.
The Committee for the Care of the Diabetic is a group of leading
diabetologists from all over the United States which was formed
in November of 1970. rt initially sought to deal with the Govern-
ment administratively before seelciug legal counsel.
The CEAIRMAN. Are you saying this was the year that the Com-
mittee for the Care of the Diabetic was created?
Mr. CHAYET. Yes, sir. It was created shortly after the results of
the UGDP were first brought forth; I did not become counsel until
about a year later, It is clear from the record that there was exten-
sive correspondence between Dr. Bradley and the Committee and the
FDA in an attempt t~ settle this matter administrati~ely. These are
not litigious individuals. They chose the courts only as a last resort
beca1use there was simplir no place else to turn.
When I first reviewed this matter from a legal point of view it
concerned me that if a doctor continued to prescribe this medication
in the face of a package insert which indicated there was an in-
creased risk of cardiovascular disease, it appeared to the, having
looked at same recent cases, that such physician could well be sued
for malpractice. I have found several cases where the package insert
was introduced as expert testimony into evidence. That was my initial
concern from the legal point of view; there is, however, a far greater
concern, and that is the impact that this entire matter has had on
millions of patients throughout the country; it is the fear and the
panic that has be~en caused: by a combination of governmental action
and the press, and by the great confusion that has swirled around
this issue that has really done the damage.
And while the issue of potential malpractice actions is still pres-
ent, it is now much more a question of how patients react when they
PAGENO="0029"
dOM~T1flV~E ~Oi~t]~iM~ I~ ~It~ t~*UG ~ i*%7g
read press reports of deaths ~Ueged1y ~at~s~d by o~l hy~gl~th~iic
agents. how can they r~ta{ii itny conficl~n~e m thei± p~y~lan in light
of such reports ~ We are cOlt med with not only the pt'oteetiOn of
the physician from a niaipt'a~tice action, but of equal iutipôrtaiiee, the
protection of his patient from the actions that have ocei~rMd again
this very week; actiOn~ siffitiar to What oect~l±ed in 19~O~ ~h~II ~ln.a~
ture press re1~ase~ again he~aIded this a bia~d view of oontmv~y~
It is most unfôrttt~a1~O Senat;or.
The O1IAIRMA~. We1~l, I gtues~ ~~on have made it clear. Y~u a~re re~
ferring to what yoti belieire to be C~nftt~iolt and doubt~ *hlch have
resulted from stories respepting the TIGDP ~tttdy.
Is that What you are sayIng?
Mr~ CHAYET. Is there a~ny question about that Senator? For ~
ample, I will qttOte a TIPI report in the BOst~~ Globe, tuesday,
January 2, 1975. The report is pathetically ipa~ctirate "Ai1 intei~~
tional soientifie jury has supported the much debated ~ thai~ th~
oral diabetes drug used by 1.5 million Americans are probithly kiflifio~
10,000 to 15,000 ôl them pearly."
And the inaccttraey iS not the fault of the pres~ It is the fau~t ~
those who are giving the releases and the fault of those *h~ ~
written and released the editorial statement *hi~h acoomp~n1ed the
Biometric Report which reviewed the TJGDP sthdy~
The Biometric Study~ ilt many wayS is a vet7 ~cholat'iy e~udy, b~t
what was dane with It is most unfortunate. And that is Wher~ the
problem lies. Tt is a repeating pattern by thOse who seek to Stifle and
muzzle the couitroYei'Sy whiëli nobody can any longei4 deny.
The CtxAIRMA~. I have not seen all of those ~tOt1~s~ bitt the st~ries
I have seen were l?ased upon an editorial that is appearing, ap~aI~
ently, in support of the troi~~ study, ift the Journal of ~he Anierlèai~
Medical Society. Then the ~toties WetO written ~rOm that. N~*, I
have n~t seen what the joaftal said, but i~ the journal story wa~
exaggerated, that would be a matter of whoever te~ported th~ story~
I guess.
Mr. CHAYET. Well, that is the problem, Senator. Tt is ~ne aggera~
tioti on top of another. Yon have the l3iotn~tric Study, then, sOffi~One
wHte~ an editorial and refers to "possibly some 10,000 to 1~,000
deaths"-no statement of which appeared ii~ the Biometric Studjr~~~
and then the press reportS that an ititS national hhi~4ibboit j~iry
found 10,000 to i~,Oo0 deaths a year. And I think we are well aware
of the political proeesS to kfio* that this IS the way it goes; and we
in the legal and SOiehtiflO eommunities have to take steps to prevent
this frOnt ocdurthi~. Aftd thd~e StepS WetS nevet1~aken in this sittta~.
ti~h, aftd I regret that.
The O~AIu~A~. I uxidOrStthid What you ate Saying. I have been in
politics :Mr many, many years, and have been "dQnS in" maii~, hiany
times. Bitt I have not suggested that we abolish the freedom of the
press.
Mr. CnA~EI~. I guess when we get done In Senator, we at
have the riSk Of that because of Out pitblie ~QsIt1oii. But the millions
of people out there do iiOt take that risk, and that is why I am COn~
cerned about it. That is the poiitt. It is not the prOSS *h~ch sh~ttTd be
téstriCted bitt thoSe Who provide the erroneous Iiiforn~atithi tO the
preSS.
PAGENO="0030"
1.3280 cOMPETmVE PROBLEMS IN. THE DRUG INDUSTRY
As f Or the tIGDP study itself, I am, not at this point going to
discuss its scientific flaws in. detail. Although I am a lawyer, not a
scientist, I would `ethphasize, however, something which strikes me as'
very important; and it is. the final `paragraph of the U.GDP study,
which reads: "It should be noted that `any conclusion reached in `this
study pertains only `to the type of `patients studied"-~nd a very
particular .group of people were studi.ed'-"and only to the specific
hypoglycemic agents used. Extrapolation of findings obtained in the
IJGDP to other~ dosage schedules ~f the same drug"-and dosage
sche~14leS other than those used iii clinical practice were used in that
study-~---"or to othejr chemically related hypoglycemic agents not in-
cluded in this study, mu~t be made on a judgmental and nonstatistical
basis."
No~, those are `the words' of the study Itself. And yet,. in spite of
these words, w~ see an unfortunat~ extrapolation, contrary to the very
words of the UGDP study; and `I would only say, Senator, that ques-
tio~s of- , , ~, . , , .
The CHAIRMAN. May I say, just a~i~oment? I wa~ looking at the
final paragraph, and the final paragraph,. ~s it reads to i~ie, is: "In
cpnciusion,"-t'his is the-
Mr~. CrrA1'n~r. Excuse me, Senator; i~ is .th~ next-to-last paragraph
at page 814 of the stucly~ I am sorry.
Tl~e~ CHAI~AN. Well, `let me say this, so they juxtapose. "In con-
clusion, we consider, in the light of, the UGDP findings, it remain~.
with the proponents Qf.the oral hypoglycemic to ponduct scientifically
adequate ~tudies to justify the continued use of thich agents."
Mr. CIUYET. Unfortunately, yOu are reading from the wrong study,
Senator.: ` . , ` ` ,
The CHQI4RMAN. This i's the Biometric Soeie~y Study.
Mr. C~IAYET. I know. I am not talking about th~t, sir. What I
~a~d very clearly is that `I am talking about the TJGDP study. You
see, thia is how,it goes..
The CHAIRMAN. I misunderstood.
Mr. CiLj~ET~ This is how the con'fusion, esca'lat'es.
The, CHAIRMAN. I am sorry I misunderstood you. We will put ,the
two tog~ther, \sO that everybody can x~ead them. Go ahead with your
testimony. . . , ` ` ` `
Mr.~CEAYET. That will be fine,. Senator,
As~t said,, I am not. going to concentrate on the IJGDP study itself,
or even the specifics of the Biometric "study. I oniy want to `make one
point, and I *ould like to make it as clearly as I can. There is great
controversy in this. situation, and it is not going to go away. It does
not matter how many people are lined up on either ,side-and I am
perfccthy cognizant of the fact that `the `press releases describe the
bluç~ibbon jury of experts who are for ,the UGOP, and when any-
b'ody on th'e . other side is mentioned, they a±e re~orred ,to as a group
of practicing physicians. I realize these are subtleties, but they aic
subtleties, that have resulted from the fact that~the Government has
made a fundamental error in this situation; that is,' it has tried to
muzzle a controversy, which has `been put forth in good faith by very
eminent, very learned, and qualified people. `
Now, I know `that $8 million and 10 years is a long time, and `a lot
of money, a'nd criticism is difficult. `It is not to be given or taken
PAGENO="0031"
COMPETITIVE PROBLEMS. IN TIlE DIWG INDUSTRY 18281
lightly: But ~h~t `is the way it is; nothiflg is gOilig to make t~l's
study free frOm controversy. Nothing thus far has settled this con-
troversy, and Dr. Chalmers can write editorials entitled "Settling
the UGDP ~tudy" as long as he wants, and it is not going to resolve
the issue. I ~ould like to now move on and state exactly what the
Committee on the Care of the Diabetic is seeking to accomplish,
First of all, thay I state that, when I began this matter L sought to
restrict the TJGDP findings on the label. I thought the study was so
flawed, based on what I had learned from the physicians I represent
that it should not appear on the labeling. However, I later 1~led an
amended coinp'laiht in the/ Federtid court, because there is the possi-
bility that the, study `had some. merit Cv.en though it is flawed. We are
not' haying that the~e drugs' absQl'utely do not cause certain problems
because we do not know. But the UGDP study did not' give `us the
answer, and pretending it did does not help us at all.
`What we are `seeking' is a. label which reflects fair balauee~ which
reflects tl~ fact that there may be a problem with the drugs, which
indieates the stitdy' results and the controversy surrounding them.
There have, `been many "eminent people supporting both sides. The
evidence which has been ,presented clearly points up the existei~ce
of conflict and controversy which now must `be admitted by all. To~
fail' to~indicate such çontrpyersy on the label is most inappropriate.~
At this point, I would like to discuss the lawsuit, Senator, I belie~re
that the action is unprecedented. It is the first time, `to my knowledge,
that a group of physicians and patients-and I emphasize that the
plaintiffs include patients-in a class action representing' all physi-
cians and patients similarly situated-~-have sued the Government and
the mailufacturers to'prevent the Government from forcing the label
change along the biased lines `that it~ sought and the manufacturers
from' buckling nuder to `FDA~ pressures, `for a variety of reasons, . and
voluntarily changing th~ label. It is on the basis of that lawsuit that
we secured, in November of 1973, a preliminary'' injunction halting
the Government from ordering the label change and halting the com-
paiiies frQr~ vWuntarily altering it~ We' have not~ sought to h~ld up
labeling, and ~Ido not understand why it has taken `so' long for tl~e
FDA to move f0rs~y~rd,'with labeling. This case was. before the `court
of appeals on July 31, 1973. Why is it that 18 months later, we still
have no' revised, labeling? `One reason ma,y1be the' unrealistic expecta-
tion that the Biometric Study would settle the matter once and for,
all and then the. ~abe~ing could proc~td. ` `
Well, that has not happened. And. `the Committee for `the Care' o~
the Diabetic. will go back toc'c~urt~ and will, take evety `step it has
to `take, `to prevent a one~ided, bia~ed label' from emei~ging; and the
Biometric .Soci~ty Study does not a~lthr"our re~olve. I might add' that
we have had less than 2' days `to review the Biometric. Report. `Why
did we only have 2 days? `Why Was it not: `made available t'&rnen'l'ike
Dr. Bradley, and the other members of the Committee `on `the' Care
of the Diabetic? I asked the AMA why it was not; `and they' `~áid
they could not allow' this because the' Bioinetric Society `said `not to
release it to anyone; anyone except I)r. `Chalmers, `that ~is~
Why was the biometric Society so 5e~retivè abbut "this docnmeiit'?'
T would very much like you t~ find oflt~' Senator; the answer probably
is that NIH insisted pn secrecy. The result is that a document is going
PAGENO="0032"
1328~ ~dMrx~E1v~ ~LL\~tS i~ TH~ irnua INDTJeTRY
to be printed in ~ journal in 2 weeks, a document which is not given
to oth~r~ for re~'iew despite the known eontrover~y which exists. Then
the AMA ~a1l~ a preptiblication press conference~ and ftnnounees an
article that i~ to be printed in 2 ek~, and the restilt is a headlme
which appears in papers throughout the country which spea1~s of
15,000 people dying each year, which has nothing whatsoever to do
with any of the material to be pubUshed~ I think it is irre~ponsable,
and I ara e~treme1y disappointed that the AMA has seen fit to take
this n~iea~ure; I can only asôribe it to some naivete, they never should
have functioned in this manher.
The O~IA~RMAN. I hare not, by the way, see~1 either the editorial,
nor have I yet seen the report of the Biometric Society.
Mr. OiiA~'. You ought to read 1t~ It IS most interesting, particiF.
larly the editorial.
The CHAIIt~A~. Well, I have not seen it.
Mr. O~AY~t'. SenatorS I am sorry to go on for so long, but you said
something previously about the l~gal aspects involved in this matter,
and you ~efe'rred `to the Food, Drn~ and Oosmetio Act. I would like
to address myself to this aspect of thl~ very important ~4uestion.
W~ have depended on a regniatioti' of the FDA Itself, which reads
as follows:
The existence of a difference of opinion among e~perth ~nalffied b~ tr~i~thg
arid etpetietwe; as to the trrith of ri representation tha(Th or sug~este'd in the
labeling Is a i~Oct the falinre to re~a1 which ma~ reri~ler me labeling ml slead~
1ng~ if there is a material weight Of o~iniori contrary to such representations.
What thIs means is, if a manufacturer seek~ an NDA `and he knows
that there is controversy o~rer his product, he has a duty to come
forward, if there is a material weight of opinion' ~a~nst his drug,
and inforth the Ft~A. Wh~ `dOes not the aovcrnrnent h~e a similar
obli~çation' to inform physi~cians that there is a mate~ial weigltt of
opinioti contrary to its findings? Why has the Government sought to
repeal thi~ regulation When the ~a5e was returited ~to the agenc.y by
the ~Ourt of appeals? Does this principle e~i5t only for the manu~
facttirers? If there is material weight of opinion against a position
why is it stifled and not reftected-~u~t because it Is the Government
position? That i~ my question. I have not yet received an appropriate
answer.
The C~tAt~w~. Well, just let me ~ay~-if `we are addressing Onr~
Selves to the same thing-~4hat J carry no brief for everything the
FDA does, but, as a matter ol fact, they have in `a very massive way
done eñ~tly what y~u 5~y they have not done.' In ~ceordahce with
the 1~62 amendments to the ~efauve1t Act, ~the FDA contira~ted with
the National Academy of Sciences-National' Research cotiu~il wh'i~h
Set up panels on all kinds of drugs. These panels evaluated thotis~~
ands of `drugs and then made recommendations. And the FDA took
very positive action on a large number of drugs, perhaps `as many
as ~,000 of theth~
Mr. C~i~t. That is' true.
The cin~~. Informing the public abotit their `deficiencies.
Mr. CHATET. I `do not want to imply that the FDA dbés not do any~
thing pr~per ~t all. The FDA has done a great ~deal' of fine, very
valuab1~, very important work, and I never want to deprecate that
PAGENO="0033"
COMPETITIVE ,PROBLEMS I±~ ~HE DRUG INDUSTRY 13283
for a moment. But in t~his case,-an~I I do not know why, but-in this
case, the process ha~ broken d~w~ frretrievably. The. FDA is not liv-
ing up to its mandate in this case; it has not, and it' is not at' the
present time. And what its mandate is, it seems to me; is to pre~ent
information, but to do so fairly and impartially; and `it. is' not doing
so,
Let me explain what I mean by this. When one applies for. an
NDA, the substantial evidence test is ap~hed. The thrust of im~le-
mentation of the Food,, Drug and Cosmetic' Act has been to do `away
with clinical'opinion insomuch as possible, and reduce it to controlled,
statistical studies. That may be a very laudable goal in some cases,
but in other caseS, it can lead' to great difficulty, ~s in this case. There
is substantial evidence of efficacy of these drugs-as they do lower the
blood' sü~ar~nnd that is the efficacy which is claimed. The q~iest~on
is, what is the result of lowering blood sugar, and how dangerous are
these drugs. This is where one gets into difficulty.
There is `a question as to whether the TJGDP study can be con-
sidered substantial evidence, as it appears to be so flawed. I' think
that the `study is reflective of a great deal of effort and, in many~ ~v~ys,
it was a very complicated and very sophisticated study. It is indeed
unfbrtunate' that improper extrapolation has caused it to reach this
result.
Let us assume, then, that the IJGDP constitnte~ substantial eiri-
dence. It is important to realize that wq are de~tling with a warning
which is to appear on the label. If a drug is approved for efficacy
and is being m~rketèd and a problem d~elops. with the `drug,: I thi~ik
that problem ought to be reflected on the labeling, and in this ca~e,
it should be sq indicated as to the oral hypoglycemic `drugs.
So, what does one do? Let us say substantial evidence is not ~n
issue but, rather, someone comes up with a clinical opinion that there
is a problem here. That controversy should be adeq~uateiy refleet~d on
the labeL Where there is great controversy about findings such' as
TJ'GDP, we want to have the FDA indicate such controversy on tl~e
label, even if they are not exactly sure that it is a final or totally
correct' warning.
The fact that there is controversy about a particular finding, and
there is no final answer does not mean that you ignore the con-
troversy. Where you do a substantial evidence study that costs $8
million, and lasts 9 years, and it is wrong, you are confronted with
a major problem. You are in serious trouble. You cannot replicate
it, because it is going to take another 10 years, if you can find `the
$8 million to do it. So you have a problem-and I hope the Go~ern~
ment is not going to say that, just because we have a study that. cost
$8 million and lasted 9 years, therefore, that is substantial .evidence
and it cannot be wrong; because, as I say in my statement, there
have been mistakes made-very costly mistakes made-~iñ research, as
in all human endeavor,, and there will be again in the futur~e. For
this reason, I feel' that the FDA regulation itself `should be applied
in this particular `situation. We have already filed oqmments against
the suggestion by FDA that this regulation be repealed. And we will
go back to court, if we have to, to prevent this' regulation from being
repealed iii this particular case. W~ believe the UGDP study has
56-592-75---3
PAGENO="0034"
13284 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
such severe inherent problems that, unless the controversy is indi-
cated, it will be misleading to the public and to the physicians who
are responsible for treating this problem.
I mentioned this article in the Globe. Dr. Bradley will tell you
how many people have called the Josl'in Clinic in the last few days,
and have been so upset. I have heard similar statements from other
physicians, telling the how many of their patients become upset by
the press releases which have occurred. I do not blame the Bio-
metric Society. They said things such asthe following-I am reading
now from the Biometric Society: "There remains the question
whether tolbutarnide, although ineffective in a fixed-dose regimen,
might be an effective therapy as ordinarily used." That is right in
the Biometric Report, Senator. `In other words, if they used the drug
properly,, maybe it would' be~ effective therapy. Very interesting-
that is why I say, it is not these reports per se that are causing the
problem for the public. It is what' people are doing with these
reports.
The CHAIRMA~. I think it should be known you are reading ex-
cerpts.
The fact of the matter is that the members who did the study
agreed with the TJGDP study results. The fact is that one of the
greatest clinics in the world, the Mayo Clinic, has accepted the re-
port. It is accepted at Grady Memorial Hospital in Atlanta. Those
`were two of the witnesses.
*So sure, there are' scientific disputes, but this society which was
independently chosen of very distinguished people, have concluded
that it was a valid statistical study, but the record is clear on that,
anyway.
Mr. CHAYET. The record is not that clear, I would like in closing
to summarize my position for the record if I may, Senator.
The CHAIRMAN. Sure.
Mr. CHAYET. I think we need new labeling right away. I would
like to dispel any doubts at all that we are standing in the way of
new labeling. We need it, and I am sorry we do not have it up' `to
this `particular point, `and I hope we will `have it soon.
I also-
The ChAIRMAN. Excuse me for interrupting. The committee coun-
sel tells me that the FDA is going to testify very soon on the ques-
tion of labeling. I did not realize it was already scheduled.
Mr. CHAYET. Fine.
But I think we need this labeling very badly, by that I mean
labeling that will be fairly balanced and indicate the scientific con-
troversy. This raises the question of what do we do about the con-
troversy. Although I heard what the doctors said before about the
difficulty of launching another study lIke this, I really do not see
any way of avoiding it, unfortunately, at the present time, because
what we have at the moment does not answer the critical and crucial
questions raised. Regardless of an additional study, I think there
ought to be balanced labeling so that doctors will use this drug
properly and be aware of `the controversy while, as quickly as pos-
sible, we ought to have additional studies. I would `say, Senator,
there ought to be at least one study that is as lengthy and as fully
PAGENO="0035"
COMPE~1TIV~ PROBLEMS IN THE DRUG INDUSTRY 13~85
funded a~ this, and we ought to stop putting money into trying to
justify a flawed study. Instead we ought to try to correct it, cøme
up with a better study, and come up with the answer which retdly
is of such ithportance to millions of people throughout the countjy.
I thank you very much, Senator.
Dr. Bradley?
The CHAIRMAN. You are an articulate advocate, and I appreciate.
yOur testimony.
Mr. CHAThT. Thank you, sir.
The CHAIRMAN. Doctor, we are glad to have you back again.
Is there anything you would like to add?
Dr. BRADLEY. Yes. I 4would like to make a few comments, Senator
Nelson. I would like to start, if I may, with a slight anecodote, and
this just to change the pace a little bit. Yesterday one of my associ-
ates came up. to me proudly and said I want you to meet Mr. X.
He now is down to his ideal weight of 150 pounds. A little ovei~ a
year ago he weighed 275 pounds, and at that time was taking 80
units of insulin. Now he weighs down to his ideal weight, and his
glucose tolerance test is normal, so that this is an illusti~ation, I think,
of the benefits of weight loss and of diet. One of the first things I
want to affirm is that the idea of diet as treatment for diabetics
did not originate with Dr. Davidson, or with the UGDP or with
Dr. Ricketts. As a matter of fact, one of the reasons the Joslin
Clinic has been criticized over the years is that they have so rigidly
insisted upon diet as primary treatment, whether people are On
just diet alone, or whether they are taking insulin or pills. This
continues to be the case and has been so all along.
So if you will pardon my digression, I would like to come back
to the other issue if I may. I think Mr. Chayet quite rightly has
emphasized that the controversy remains, but I would like to ap-
proach it along three or four hues with you briefly, if I may.
First of all, the Biometric Study, I think most of us realized, would,
to a certain degree, be moderately supportive, . and I think those
words appear in the study of the TJGDP experiment. We did not
expect anything different. I would not have been here just because
I knew the BiometHe Study Report was going to appear, because it
does not answer the fundamental questions in this whole issue.
They did not address themselves specifically to two questions whicl1
I have raised all along, and now I realize I must send you a publica~
tion that has appeared within' the last year or so which I prepared
for "Controversies in Medicine," and I would be happy to send you
copies of that if you wish for the record.
The CHAIRMAN. Thank you.
Dr. BRADLEY. And without going into much detail, these relate to
unknowns, if you will, and they are what clinicians, those people
taking care of people with diabetes, are more aware of, perhaps a
little more humble about than others looking at it from a purely
scientific standpoint. There could be unknowns that we have no wa~
of knowing about, but there can be two specific unknowns that come
to mind, and these are critical, really, and they are as follows: First,
the level of coronary heart disease, the condition of the heart in
the diabetics at baseline was not known. I emphasized this in my
PAGENO="0036"
13286 cOMP~TIV~ PROffl~EMS IN. .TB~ DRV.~' ~I~T~Y
Ipr~vious testimony. The tools for evaluating the~e were crude. ~bere
* WaS some kTffiwledge about it, but not enough was known so that
one could really say these treatment grOups were the same.
Second, and. this is crucial to the whole thing, they did not' really
know how long diabetes had been present. NOW, diabetes is a risk
factor in all of the patients in all of the groups of the study, Most
clinicians know quite well nowaldays, and there is good statistical
evidence to back it up, that the major effect of diabetes ~n' vascular
disease relates to how long one has it-~-its duration. And unfortu-
nately, in ad~lts it is very difficult to know how long diabetes has
been present.
We have a little clue from the study. I think Dr.. Ze'len or one of
the other witnesses referred~ to the glucose tolerkm3e tests. The clue
is a very iiiterestin'g one in that when one looks at the glucose toler-
ance test in the group on. tolbutamide, there were more people in
the tolbutamide' groups who had higher fasting blood sugars, and
I think, more at every interval of the glucose tolerance tests, than
in any other of the treatment groups, certainly more than was the
case of .placebo~ . . .
This~ means, then, two . things. Subgroup `(~.)`: Their diabetes was
some*hat~ more severe as a group at baseline; and (2) they may have
had diabetes Thuger.
N~ow, we have no way of proving or disproving it~ We can present
evidence on both sides, but it raises the same.. kind of uncertainty
about the hardness of the TJGDP, and I think I gained~ this from
Dr. Meier~s testimony, a lack of hardness. Tile was not quite willing
~to say that this was a very hard kind of decision. I think this is
where we have to be very careful in accepting these results hook,
line, and sinker.
Now, we~ have a further hint that there may have been more risk
in that when you add up the numerators of the various risk factors
in the patients on tolbutamide, they were somewhat more.
The key issue here is, again, that this d~oes not prove that tolbuta-
mide and phenformin are not hazardous. They may still be. And our
problem is how do we resolve this. Sir, if I may, I would like to
make a few more suggestions that might help us all come to a
solution of the problem.
First: I' think the role of Government has unfortunately gotten to
be a. little bit too strong here. Mr. Chayet' has referred to this, and
it is alluded' to. in our statement. But I will focus specifically on the
FDA.
We have `been concerned~ all~ along that the Ffl~A~ has' not gone
ahead with labeling which was reasonabl~' balanced. That was all
we' `wanted, `somethii~g that acknowledged the fact that true con-
~troversy existed; that would not lay undue concern upon the mihds
of physicians and their patients, yet'enough so that they would be
caretul, but also allow us to get' about *our business' of educating
physicians and patients. This is what has fallen terribly behind and
is the reason these drugs are being used altogether too much today
in diabetic patients of this type throughout the couhtry.
* I will come back to that in a moment.
NOw, you asked a very important question, and that is about the
benefits. In my original statement in September I indicated that
PAGENO="0037"
COMPWI'ITIV~ ~RO~LEMS IN THE DRUG INDUSTRY 13287
these drugs are not oral hypoglycemic drugs. This is a semau~tic er~or
on the part of Dr. Palumbo, who may he here `to defend himself.
They really ai'e oral blood sugar lowering drugs~ They may pro-
duce hypoglycemia, but that i~ not their blood sugar lowermg agei~ts,
and they do this very clearly. No scientist in his right mmd wOuld
argue with this, that they do it very clearly in selected patiei~ts,
not all of the patients of the type that were studied in the TIGDP,
but in a majority of them. The problem is they do not continue to
do it.
And second: There may be a hazard from a cardiovascular stand-
point, so they are suspect in two areas.
Now, in terms of other kinds of benefits which the TJGDP study
was set up to evaluate, I do not believe they were set up to evaluate
cardiovascular mortality. They wanted to see, did these patients
have more diabetic neuropathy after a period of time? Did tl~ey
have retinopathy, involvement of the eyes? Did they have more
kidney involvement? Did they have more cataracts? Did they have
more infectious disease?
I think the end point was listed as vascular complications in gen-
eral, as a mixed bag, but they really were looking for those complica-
tions which are more specifically related to~diabetes.
At the moment we do not have data relative to such end points ~n
terms of effectiveness, long term effectiveness. The only data we have
is relative to the blood sugar. If these drugs do not lower the blood
sugar, no thoughtful diabetologist would continue to use them.
Now, certainly if he does not believe that lowering the blood sugar
might protect from these complications, he would not use these drugs.
So at the moment we must deal in terms of blood sugar lowerii~g
and maintaining it.
Now, what I would like to see happen, I would hope from these
hearings, is that we would get away from this controversy. Frankly,
I think it is ridiculous. I think the people who suffer from this p~-
tentially are patients. The controversy has held up our getting to-
gether and trying to approach the problem not by governmental fiat,
not by Dr. National Institutes of Health or by Dr. FDA, but by the
appropriate cooperative statements of objectives in terms of treat-
ment of patients. And if we cannot agree on objectives, maybe we
will have more controversy. But I think we should agree on objec-
tives that relate to using these drugs, if they are going to be used at
all, in such a way that the one effect we know about, namely, lower-
ing of blood sugar, is indeed guaranteed in patients. If we do not
do that, I think patients either will be stopped because of govern~
mental fiat, and a law forbidding them, or a policy, or else physician~
will go on using it in patients. And if there is a hazard, then Obr
viously they are being exposed to it.
I think we must realize there may be a hazard. There is no ques~
tion about it. And our committee has never tried to say that this wa~
not a possibility. All we wanted to do was to be sure that the situa-
tion was put in balance. That is all.
Thank you.
The CHAIRMAN. Well, thank you very much, Dr. Bradley. I ap-
preciate your taking the time from your busy schedule to come here
and testify today.
PAGENO="0038"
13288 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Well, thank you very much, gentlemen.
Mr. CHAYET. Thank you.
The CHAIRMAN. Did any of the members of the biometrics panel
wish to comment for the record on any of the statements made by
Dr. Bradley?
All right. Thank you very much.
[Whereupon, at 12:08 p.m., the subcommittee adjourned, subject to
the call of the Chair.]
PAGENO="0039"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
(Present Status of Competition in the Pharmaceutical
Industry)
WEDNESDAY, JULY 9, 1975
U.S. SENATE,
SUBCOMMITTEE ON MONOPOLY OF THE
SELECT COMMITTEE ON SMALL BUSINESS,
Washington, D.C.
The subcommittee met, pursuant to notice, at 10:15 a.m. in room
318, Russell Senate Office Building, Senator Gaylot~d Nelson (cha~ir-
man of the full committee) presiding.
Present: Senator Nelson.
Also present: Benjamin Gordon, staff economist, and Kay Klatt,
research assistant.
The CHAIRMAN. Today the Monopoly Subcommittee of the Sen~tte
Small Business Committee resumes its hearings on the safety, efficacy,
and use of oral blood-sugar-lowering drugs, which are taken by
diabetics.
Well-controlled studies showed that users of these drugs are 2½
to 3 times more apt to die from heart problems than thosc w~o
depend solely on diet or diet and insulin. According to expert testi-
mony, these drugs also have limited uses, and Dr. John Davidson,
the director of the largest university-based diabetes clinic in this
country, estimated that more than 99 percent of the people usii~g
these drugs should not be using them.
In early 1974 the FDA moved to change the labeling of the blo~d
sugar lowering drugs to reflect the latest scientific evidence about
their dangers and lack of efficacy, but the agency's efforts were
blocked by a court order. Nevertheless, recent additional studies
confirming the dangers of these drugs have made a change in tl~ie
labeling imperative, and the Commissioner of the Food and Drt~g
Administration is here today to discuss the new labeling of the or~d
blood~sugar-lowering agents as well as the recent human and animal
studies that confirm the need for prescribing physicians to be in-
formed in the clearest terms possible of the latest knowledge in tl~e
field.
The new labeling-as well as other aspects of these drugs-will be
discussed tomorrow by a number of medical experts.
Dr. Schmidt, you may present your testimony however you desire.
The statement will be printed in full in the record.'
See prepared statement, page 13697.
13289
PAGENO="0040"
13290 COMrETITIVE PROBLEMS IN THE DRtTG INDUSTRY
STATEMENT OF ALEXANDER M. SCHMIDT, M.D., COMMISSIONER,
FOOD AND DRUG ADMINISTRATION, ACCOMPANIED BY RICH-
ARD MERRILL, CHIEF COUNSEL, FOOD AND DRUG ADMINISTRA-
TION; L RICHARD CROUT, 1VLD., DIRECTOR, BU~U OF DRUGS,
FOOD AND DRUG ADMINISTflATION; JAMES M. ~BiLSTAD, M.D.,
GROUT LEADER, DIVISION OP METABOLISM A~TD ENDOCRINE
DRUG PRODUCTS, BUREAU OF, DRUGS, FOOD ANi~ DRUG ADMIN-
ISTRATION; AND ROBERT WETHERELL, DIRECTOR, OFFICE OF
LEGISLATIVE SERVICES, FOOD AND DRUG ADMINISTRATION
Dr. SCHMIDT. Thank you, Mr. Chairman.
Because the statement I have is relatively brief, I thought I wOuld
go through it. I am accompanied this morning by Dr. Richard Crout,
Director of the Bureau, of Drugs, on my right and your left, and
Mr. Richard Merrill, Chief Counsel of the Food and Drug Adminis-
tration, behind me. To my left is Mr. Robert Wetherell, Director of
our `Office of Legislative Services, and to my right, Dr. Bilstad,
our Group Leader of the Division of Metabolism and Endocrine
Drug Products. We are pleased to be here this morning to discuss
our current actions regarding the oral hypoglycemic drugs.
As you are well aware, labeling for this class of' drugs has been the
subject of extended public controversy and legal challenge for a num-
ber o~ years. The' Agency `has now published a proposed regulation
providing new labeling for this class of drugs. The ~proposa1 appeared
in the Federal Register on July 7, 1~97~ and asked for comment on
the labeling. It also announced a public hearing to be held ou
August 20, of this year to `afford interested persons a further oppor-
tunity to comment. ` `
Last September, I summarized before this subcommittee the, actions
of the FDA that followed the `report in 1970 of the results of the
utliversity group diabetes program study. TodayS 1 will review the
events `that have taken place since Fny previous testimony and will
discuss, in some detail, of course, aspects of the proposed labeling.
Mr. GORDON. May I interrupt you for just ~ second, Dr. Schmidt?
As I understand it, new labeling was originally proposed by the
FDA `in 1972. Is that correct?
Dr. SCHMtDT. That .is correct.
Mr. GoRDoN. So~ you have already had comments on that labeling.
You stated in your statement which appeared in the Federal Register,
that you did not expect any major new information. In fact, it is
on page 15 of the Federal Register insertion. You have the results
of Other studies including animal studies which support the UGDP
study.
Why do you, then, have to go through the same long procedures
again, that is, proposing changes, having 60 days for comments, hav-
ing administrative hearings, and so on? `Is that for legal purposes?
Dr. SCHMIDT. Well, we spent a considerable amount of time dis-
cussing and deciding on the best procedure to use in going ahead
with the labeling change and quite deliberately chose the form&. rule-
making procedure which in effect this is. And I think the reason the
rulemaking procedure is clearly the best way to go is that the goals
PAGENO="0041"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13~9i
that we have in this whole process include not only the revisior~ of
the labeling but `the dissefriinatioh' of information, the education of
the great number of people, physicians and others, who are inter-
ested in this subject. `
We are interested in educating everyone as to our firm' behef~ in
this area and the rulemaking procedure allows public discussion,
public debate, public comment, a process through which many peo~
ple can become informed and become edudat~d, and we thrnk~ the
benefits of this in' this area are obvious and all to the good.
Second, we would' wis1~ to avoid further litigation, if that is pos-:
sible, and One can perhaps avoid litigation by achieving' consensus
and one of the, best ways of achieving consensus is through public
debate and discussion.
We further think that the rulemaking procedure done formally
could strengthen any court case that might evolve. And' I could t~sk
Mr. Merrill to comment on that.
Mr. MEmaLL. `Senator Nelson, Mr. Gordon, there is a second reason
behind the way we are proceeding `now. Nothing would ple~se us
more th'an to avoid reinstitution of that lawsuit in the Court of
Appeals in the first circuit. But it is our belief that we should follQw
the regular rulemaking route prescribed by our procedures in the Ad-
ministrative Procedures Act.
We strengthen ourselves in any subsequent court challenge of this
labeling. It fortes the court to, in effect, conclude that we' were
demonstrably wrong. It puts the burden `on any challenge or to esti~b~.
lish that we were arbitrary and capricious~ on the basis of the evidence
and information that we assembled in this administnttive proceeding.
The CHAIRMAN. Go ahead. ` , ` `
Dr. SCHMIDT. Because' o~ the controversy' among ph~ysicians con-
cerning the TJGDP study on the oral hypoglycemia bibeling pze-
viously' proposed by us based on the use of the TiTGDP study, we
decided the publication of the propOse~ labeling in the Federal Begis-
ter should await c~ntpl~tion" ~fi' the ~etailed study of the 1JGDP
study by the Biontetriid Soci~y.
The report of the ~oóiety was published in the February 10, 1975,
issue of the Journal' of the~American' ` MediCal `Assoei~tiOr~, }n testi~
mony before the subcommittee last J'aiu~aty b~ the members of tl~e
society who conducted~ the re~iew,: a review of their conclusiofis was
provided for you. The Biometric Society committee asses~ed the
scientific quality of the TJGDP study, particularly the design, coi~-
duct, and' analysis of `the trial.
And, as well, the committee evaluated other controlled trials in-
volving oral' hypoglycemic agents.~ The committee discussed in detail
criticisms of the UGDP study and concluded that they found, "most
of the criticisms unpersua'sive~" ` ,
Specifically, the Biometric Society committee con'clud'ed that first,
the criticism'that patient selection was' inappropriate, was "largely
irrelevant" to the validity of the evidence for the toxicity of `the
oral agents~ `
Second, the critieism that total' thortality ifl the tolbutamide was
not significantly different from that in the placebo group had some
weight and "the `toxic' effeCt of th~ oral" hypoglycemics cannot be
affirmed with the certainty that would be present if total mortality
were significantly different."
PAGENO="0042"
13292 COMPETITIVE PROBLEMS IN THE I~RUG INDUSTRY
Third, excess mortality in tolbutamide-treated patients was not
confined to a few clinics, as critics have claimed.
Fourth, although there was a "puzzling anomaly" concerning the
distribution of sexes to the treatment groups within clinics, they
could find an assignable cause for this distribution and have no
reason to think that this study had been compromised by a breakdown
in the randomization of patients to the treatment groups.
The committee particularly analyzed the criticism that there were
important differences in baseline cardiovascular variables among the
groups and concluded that there was no evidence that the baseline
differences arising from the randomization contributed in any im-
portant way to the finding of adverse effect from tolbutamide.
Another conclusion was that the criticism that oral hypoglycemic
drugs were given in fixed dosage was not relevant to the question~
of whether the drugs were toxic.
The committee also noted that the fixed dose given was about equal
to average recommended dose. They further concluded that although
it would have been easier to interpret findings were there more data
on mortality, that is if the study had been carried out longer, they
did not criticize the UGDP investigators for having made the de-
cision when they did. The committee said:
Nevertheless, the result of that decision is to leave us with some residual
uncertainty about the meaning of the findings, a point that is well `understood
by the UGDP investigators themselves.
And last, the committee said that' other studies said to contradict
the findings of `the UGDP study do not in fact do so.
The CHAIRMAN. Dr. Schmidt, yesterday and today-yesterday in
the New York Times, today in the Washington Post-there is a story
referring to `a letter that was written early this year by Dr. James
Sammons, executive Vice president of the AMA to the Upjohn Co.
in which, as I read the story, he is critical of the UGDP study and
the evaluation by the Biometric Society of that study. Among other
things his letter states:' "A considerable body of expert scientific
opinion contradicts these published findings." Then the letter was
sent to the State medical societies and county medical societies, and
1,100 of detail men of Upjohn were furnished copies of the letter.
Obviously, it attacks the findings of the UGDP and as well the
evaluation of the Biometric Society of those findings, which appeared
in the Journal of the American Medical Association.
My question is, the UGDA study extended over 10 years; is that
correct?
Dr. SCHMIDT.' The study began in 1961, ~nd the evaluation of it
is still going on now.
The CHAIRMAN. It started in 1961. On page 4 of your prepared
statement you quote from the Biometric Society report that other
studies said to contradict the findings of the UGDP study do not
in fact do so.
Are you aware of any carefully designed scientific studies that
have been conducted that refute the findings of the UGDP?
Dr. SCHMIDT. No, sir, we are not.
The CHAIRMAN. So, as far as the Food and Drug Administration
is concerned, you are not aware of any scientific studies that contra-
dict the UGDP findings?
PAGENO="0043"
COMPl~TITIV~ PR0BLEM~ IN THE DRUG. INDUSTRY 13293
Dr. Sc~rn~xmT. N9 well-designed studies; no, sir.
The CH*IEMAN. Well, I will not ask you what you think the m~ti~
vation of Dr. Sammons i~ because I think everybody knows.
Go ahead, I will print in the record this article from the New
York Times as well as the Washington Post on this subject at the
appropriate place in the record.1
Dr. SCHMIDT. In addition to evaluating criticisms of the IJGDP
study, the Biometric Society conducted extensive new analyses of
the UGDP data, taking into account the effect of various baseline
variables and cardiovascular risk factors. These analyses coi~firmed
that cardiovascular, mortality was increased in the tolbutamide gro~lp.
This increase was statistically significant for the patient p~pi~la-
tion taken as a whole and in the subgroup of females, especially in
women over the age of 53, but not in the male subgroup. This dQes
not mean that the studies show that the drug carries less risk in
males. On this point, the committee concluded:
The data do not support the same conclusions for men, but one possible ~ea~
son is that the smaller number of patients in, the male group results~ in a lack
of sensitivity to detect differences of mo~erate magnitude.
An important finding was that the highest death rate, occurred in
the ~oup of patients who adhered most closely to the tolbutamide
regimen and did r~ot hai~e their dose modified. A1SQ,, when the analy~is
was cOnducted according to an approach called the survival modeIi~g
method, which takes into account the proportion of time each patient
received the assigned medication, women in the tQlbutamide group had
a statistically significant increase in both cardiovascular alid total
mortality.
The Biometric Society committee summarized its conclusions in
the final sections of its report as follows-and I need to point out
that all of page 5 on my copy is, in effect, taken from the conciusio~s
of the cOmmittee. And they said:
On the question of cardiovascular mortality due to tolbutamide and phen.
forthin, we consider that the UGDP trial has raised suspicions that cannot be
dismissed on the basis of other evidejice presently available.
It further went on:
We find most of the criticisms levelled against the UGDP findings on this
point unpersuasive. The possibility that deaths ma~V ha~ve been allocated to
cardiovascular causes preferentially In the groups receiving oral therapy exists,
and, in view of the "~ionsignlflcance" of differences in total mortality, some
reservations about the conclusion that the oral bypoglycemics are toxic mttst
remain. Nonetheless, we consider the evidence of harmfulness moderately
strong. The risk is clearly seen in the group of older women. Whether it affects
all subgrOups of patients cannot be decided on `the basis of the available data,
owing to the small number of deaths involved in these subgroups.,
In conclusion-
They Wbnt on:
We consider that' In the light of the IJGDP ~ndlngs, It remains with the
proponents of the or~il hypoglycemics to conduct scientifically adequate studies
to justify the continued use of such agents.
Mr. GORDON. You stated before to the chairman that they have nOt
come up with these scientific stt~dies.
1 See pages 13413 and~ 13439
PAGENO="0044"
13294' cOMPETITIVE PROBLEMS IN T1~E " ~DRIJG INDUSTRY
Dr. SC~IMIDT. That is coi~rect. The cornmitte~ c'oneluded~ that there
were no data that refuted the principal éonciusions of the UGDP
study, and we* agree with that.
Mr. GORDON. But setting aside, ~oi~ the thoment, the oa~rdiovascu-
lar deaths, have* the opponents of your proposed labeling supplied
si~bstaatial evidence-as required by law~*L~that the oral hypogly~emic
agents `have a beneficial effevt on the' long-term v~scul~r complica-
tions of, diabetes?. In other words, .1 am' talking ab&t~t efficacy in
treatincr diabetes.
Dr. ~`CflMIDT. Yes, I see; I have a little problem with your ques'-
tion because it implies that wI'iat is required by lai~r would be that
these~ drugs would have a beneficial e~ect on the bug-term vascular
complications of diabetes. And in fact, we have no' substantial evi-
dence on that point.
`Mr. GORDON. But, that is required. by law, is it not?
Dr. ScHMIDT. Well, no, because it depends upon the' cinims made
and if the claim fo~ these drugs was that `they infiuenae\ the long-
term mortality, the~ they would indeed need' substantial evidence.
But, if the claim is that they lower blOod sugar or relieve symptoms-
in other words, if they hai~e that effect and' b~e~e is su~stai~tini evi-
dence for that, then that is what is required by law for that W~iing.'
And we do have substantial evid~nce that these Qb~ugs~ h~wer blood
sugar and that they' relieve sympto~s~
Mr.' GouioN. Well, does it 5tate~ that the purpose is merely' to re-
lie've' symptom's aild ~hat is' all?'
`Dr. SCHMID~r~ Well, no,' but you se~, `what you are doing is two
things: oxie is you are pointing out the need for revised h~beling,
and' we firmly agreed with this. In times past, as `1 believe I said
list time I w~s here, it was believed b~r most physicians that lower-
ing the blood sugar in the dinbetic' ~oubd have a beneficial effect
upon the long-term mortality figure~ of diabetic patients. We analo-
gized this t~ the i~lea that lowering blood pressure w~uld~ prolong
the life of individuals with hypertension. And what' we are deter-
mining by some substantial `evidence is that `lowering blood pres-
sure in hypertensives does prolong lives in those `individuals that
have high blood pressure. ,
We are learning things aboi~t the lowering of the blood sugar in,
diabetics that surprise us. And so' we are in a different position now
than we were in the past when the labels may have been silent on
the issue of w]~ether or iiot lowering glucose prolongs the life of a
diabetic.
We' may have' applied this cause and effect relationship. What we
need to do now i.s to separate out now clearly the treatment of peo-
pie in order to relieve symptoms, which is very important.
I have taken care of many diabetics. And you can be sick if you
are a diabetic. You can feel terrible when you are' a d'iabetic~ And
in symptomatic diabetics, the normaiiza~ti~n of blood sugar, which
relieves symptoms as it does in some, is a very important thing.
But we have to separate that and substantial evidence for that and
that claim from the effect that that might or might not have on
longevity of individuals with diabetes.
Mr. GORDON. Now, even if the results of the' TTGDP study were
not conclusive-let's assume they are not conclusive-but ai e hkely,
PAGENO="0045"
COMPETITIVE PROBLEMS I~ THE DRUG INDUSTRY 13295
or even suggestive, would ~iot the absence of the beneficial effect on
the long-term complications of diabetes mean that the benefit-to-risk
ratio for these drugs is unfavorable.
Dr. SCHMmT. Well this comes back again to that same point of
separating out treatment of the symptomatic diabetic who cannot
take insulin, or who is not normalized by diet, that small grou~ of
people. Clearly, the benefit-risk ratio for that small group o,f mdi-
v~duals is such that we believe the drugs are safe and effective for
them and should be available for their treatment.
Mr. GORDON. Only for lowering blood sugar-is that right?
Dr. SCH1~IIDT, That's right.
Mr. GORDON. And for a limited period of time?
Dr. SCHMIDT. And for symptomatic patients. Now if you are talk-~
ing about asymptomatic patients, then my belief is that the drugs
simply should not be used.
Dr. GROUT. I would agree with Dr. Schmidt as a physician. On the
other hand, the asymptomatic patient is what the argument, the
true argument, is all about. So, I think when you see estimates, or
hear estimates of whether the oral drugs should be used in 1 percent,
or 10 percent or 20 percntor 50 percent of the people now taking them,
what you are hearing arc differences of medical opinion on whether or
not the lowering of the blood sugar in asymptomatic patients may stave
off long-term cardiovascular disease.
And I think an important point to realize is that we do not view
the IJGDP study as ~onclnsive on thut point. Nor did, the biometric
study review the study on that point. The point we feel cOhsiderably
more secure about is the evidence that the drugs may increase ca~dio-
vascular mortality. Whether the lowering of blood sugar staves, off
such mortality and is a compensating benefit for these drugs is an
unanswered question, and the labeling reflects that point.
The CHAIRMAN. Is it not also correct that the study concludes that
the purpose can be accomplished by diet better than the use of the
drug-except for. that rare small number you are making refereii~ce
to?
Dr. GROUT. I think a number of physicians, including ourseh~es,
would draw that interpretation from the study.
Your `question was, did the study. per se show that? And the
answer is, not precisely. But that would be the cOnclusion, pe~~le
would draw from the study and it is an importanb point., T3ecat~se
the question has been asked," if usage. of tl3ese drugs goes down, does
that mean that,, u~age of ~nsuhn will automatica1~1y go up? . And ~n
our opinion, `and I think, in the opinion of a number of physicians,
the answer to that is flC).
The best alternative therapy for the great majority of patients n~w
on these drugs is diet. We be'ieve that the changes in the practice of
medicme that ought to occur at this point in time will focu~ more
on the value of `diet than on replacement of oral' hypoglycemic'
drugs with insulin.
The CHAIRMAN. Dr. Davidson at Grady Memorial Hospital sajcl
in his testimony 1 that he thought perhaps the oral hypoglyoemics
1 See testimony of ~Tohn K. Davidson, M.D., Ph. D., director, Diabetes Unit, Emory School
of Medicine atid Grady Memorial Hospital, In bearings "Competitive Problems in the
Drug Industry," part 25, pages 10838-10854.
PAGENO="0046"
13296 CoMPETITIVE PROBLEMS IN TEE DRUG INDUSTRY
were indicated in about 1 percent of the cases.
I have a note here that Dr. Bradley thought 20 percent and Dr.
Weingrad 10 percent. I believe Dr. Davidson testified that they had
the largest university based diabetes clinic in the country. Is that
your memory?
Dr. CROUT. As far as I know, yes.
The CHAIRMAN. I think that was his testimony. You may recall
he testified that since they used the oral hypoglycemics for many,
many years, it was hard for them to conclude that they have been
wrong, but they concluded they were. And he testified that they had
taken everybody off the drugs. As I recall, he said they got better
results by managing his patients with diet alone than when they
were using the oral hypoglycemics.
Do you remember that testimony?'
Dr. CRouT. Yes.
The CiiAIRMAN. Are you aware of whether or not that is the ex-
perience in other clinics?
Dr. CROUT. I think that is a common opinion among many good
diabetologists. And I think that actually Dr. Bradley would harbor
the same opinion, I believe, and `you could confirm this with him, that
his figure of 20 percent represents his opinion of the number of dia-
betics who might need drugs if greater emphasis were given to diet.
So we do not contest those figures at all.
I would point out that nobody to my knowledge, including Dr.
Davidson, has ever published in the medical literature sharp studies
on this issue. Dr. Davidson's views are largely `in testimony before
this committee, not in the medical literature. Sq. real studies to define
1, percent, 10 percent, 20 percent are not available either.
Mr. GoRDoN. I t11~ink Dr. Davidsoir also publishd an article in the
journal of the American Medical Association recently, did he not?
Here it is: May 26, this year.'
Dr. CROUT. I think it was a comment. But to really present the
data that were presented before this committee in a full published
form has ,not occurred to my knowledge.
The CHAIRMAN. Go ahead Qoctor.
Dr. SCHMIDT. In addition to the Biometric Society report, other
information has recently become available. First, the IJGDP has
published recently their detailed report of, the results' of the phen-
formin study. In additiOn to reporting that cardiovascular mortality
and total mortality were greater in the phenformin-treated grotip
than in the other treatmeht groups, the report presented evidence
that pheuformin `therapy resulted in increased blood pressure and
heart rate, thus suggesting possible. mechanisms by which this drug
might influence cardiovascular mortality.
Mr. GoRDoN. Dr. Schmidt, are you saying here that the benefit-to-
risk ratio o1~ phenformin is even more unfavorable than for `the other
oral hypoglycemics?
Dr. SCHi'~çIDT. Yes, .that would be the, conclusion one would 4raw
from this study, yes.
Mr. GoRDoN. In a study published in a book called "Controversy
in Internal Medicine" by Winegraçl, Clements and Morrison, TJni-
versity of `Pennsylvania School of Medicine, Dr. Allan Winegrad,
PAGENO="0047"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13297
with whom, I am sure, you are acquainted as being an eminent
scientist and clinician, and others stated that phenformin has no role
in the treatment of diabetes mellitus. Do you agree with Dr. Wine-
grad?
Dr. SCHMIDT. Well, Dr. Crout and I have been discussing this for
some time. And with your permission, I will let him describe our
feelings on this.
Dr. CROUT. Let me wear two hats. The first hat is as a physician.
And I agree with him. The second hat, Director of the Burei~u of
Drugs, where we have to deal with the issue of action on that point.
Let me state the reason why the benefit to risk for phenformin is
less than for the other drugs. This drug may cause lactic acido~is, a
potentially fatal complication in patients who take it. So, there is a
clear added hazard, in addition to ~vhat it does to cardiovas~ular
mortality.
Now, the incidence of that lactic acidosis has. in the past been
thought to be quite low. As more and more information comes along,
it looks like it is higher than we had anticipated. And that issue
of lactic acidosis was brought before our Metabolic and Endocrine
Advisory Committee more than a year ago.
At that time they recommended that a warning be placed on the
drug, but that it stay on the market. We are gçing to take the issue
back to that committee and take up that issue again. But, it is a
separate issue, it is a separate issue from the labeling on cardio-
vascular mortality. On that, as far as we know, phenformin is the
same as the sulfonylureas. We are dealing here with two adverse
effects. And it is the sum of those two that I think is the import~nt
issue.
Mr. GORDON. As a medical scientist, could you tell us what the
medical justification is for having this drug on the market?
Dr. C~OUT. What my personal opinion is?
Mr. GoiwoN. Yes.
Dr. CROUT. I would support Dr. Winegrad. I personally would not
use the drug.
Mr. GORDON. Then you see no reason for this drug to be on the
market. Is that correct?
Dr. CROUT. You are asking me as a medical scientist?
Mr. GORDON. As a medical scientist.
Dr. GROUT. As a medical scientist, yes, I don't see any such reason.
Mr. GORDON. Well, then, why is it not being taken off the mar1~et?
Why aren't steps going to be taken to take it off if there is no n~edi-
cal, justification? After all, is that not the purpose?
Dr. CROUT. As I point out, we will take that back' to our advi~ory
committee. It is hardly a universal opinion. And neither I nor any-
body else in the Federal Government that I am awa~re of has the
power to simply exercise hi~ ,personal opinion in .the drug reg~ila-
tion business. , .
If I may wear that hat of the Director of the Bureau of Drttgs,
we will take that back to our advisory committee. We will atte~npt
to see what support there is for a po~ition that phenformin should
not be marketed. We will take appropriate action after that.
That is a separate issi~e from this labeling.
PAGENO="0048"
13298 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. MERRILL, Mr. Gordon, Dr. Grout makes a point that is too
important uiot to reemphasize. And we have been spending a lot of
time in the Food and Drug Administration trying to insist that drug
manufacturers back up their claims with adequate and well-controlled
clinical studies-scientific evidence and not on opinion. And for the
very same reason we are reluctant to rely upon our own instinctive
judgments. We would like the backing of the scientific community,
and strong scientific evidence. We think it is obtainable. But we want
to be sure.
The CHAIRMAN.' Go ahead Doctor.
Dr. SpHMIDT. At hearings before this subcommittee this past Janu-
ary, Di~. P. J. Palumbo reported that a retrospective study of dia-
betic patients treated at the Mayo Clinic suggests that survival was
lower in those patients treated with hypoglycemic agents compared
to those patients treated with either insulin or diet. Dr. Palumbo's
full study has not yet been published.1
Another study, a retrospective study of patients treated at the
Joslin Clinic reported in a doctoral thesis by P. Kanarek, can `be
interpreted as providing results that are consistent with those of
the TJGDP. Although we have seen this study, it has not yet been
subjected to a full review by statistical and epidemioio'gical experts.2
At this point we can say that certain subgrotips of insulin-
treated patients appear to have better survival rates than tolbuta-
mide-treated patients with comparable glucose abnormalities. Studies
of this type, however, always present problems in interpretation
because of doubts regarding comparability of treatment groups and
because treatments are not randomly allocated. Thus, although the
retrospective stu4ies of Palumbo and Kanarek may or may not, when
fully analyzed, `add sup~o'rt to the TJGDP findings, the prospective
UGDP study must be accorded far greater weight and is alone a
sufficient basis for our proposed actions.
DoctOrs Tan, Bradley, GIea~on and Soeldner have teported on
the long-term effects-4 years-of hypoglycemic agents' on the oral
glucose tolerance test ~nd blood lipids in chemical diabetics `at the
annual meeting of the American Diabetes Association, in 1978-
abstract in "Diabetes" 22 (~uppl. 1) 290, 1973. `The in~restigators'
abstract reported there was no significant difference in the improve-
ment in glucose tolerance between patients receiving oral hypo-
glycemi~ agents and patients receiving a placebo. `The full report of
this study has not yet been published, but it appears that the hires-
tigators studied glucose tolerance on the day following discontinua-
tion of the drugs. Their findings thus would indicate only that the
oral agents do not lead to improved glucose tolerance in the absence
of continued use of the drug.
- In another study, Dr. K W. Wissler, in an FDA-supported inves-
tigation, examined the chronic effects of tolbutamide in the rhesus
monkey. He found that coronary artery lesions were almost two
times more ferquen't and three times more severe in the tolbutamide-
treated animals than in the control animals. FDA recently received
1 HearIngs before the SubcQmlflittee on Monotoly~ Se~eet Committee on Small Business,
U.S. Senate, 94th Congress, 1st Session, on Safety, Efficacy, and tise of Antibiotjcs-
Clirulamycin and Lincomycln, January 28, 29, and July 8, 1975, Part 27.
2 study by P. Kanarek, page 13393.
PAGENO="0049"
COMPETITIVE PROBLEMS IN THE DRUG INWYSTRY 13299
the final report on this study, which has not yet been published in
the literature. FDA, at br. Wissler's request, is supporting further
study of the pathologic findings by seve~ai independent pathologists.
Dr. D. F. Wu, et aL, reported on the effects of tolbutamide on heart
function in dogs, with chemically induced thabetes, at the methng
of the American Federation for Clinical Research this past May.
The investigators reported that after 1 year of treatment With
tolbutamide the left ventricular function was reduced alid cardiac
morphology altered compared to control groups.
The animal studies do not necessarily bear directly on the ex~ess
cardiovascular mortality seen in tolbutamide-treated humans in the
TJGDP study, but they do suggest overall mechanisms by Which this
might have occurred.
Now, as you know, Mr. Chairman, it has been the vieW of the
FDA since 1970, that the findings of the TJGDP study should be
reflected in a warning in the labeling for oral hypoglycemic drugs
and in turn, in the use by physicians of these drugs. Let me e~n-
phasize that this view does not require that we conclude the study
provides absolute proof of hazard, The UGDP study is an adequate
and well-controlled study-by far the most exten~iite and best exa~n-
ination of the long-term effects of oral hypoglycemic agents yet ei7er
undertal~en.
The finding of an increased cardiovascular mortality in tolbuta-
mide and in phenformin-treated patients cannot be attributed to atiy
shortcomings of study design or execution. This finding, despite any
residual uncertainty that may remain, requires a clear warning to
physicians. Prudence dictates that a warning be issued whene~er
there is sufficient evidence to believe that a drug m~y be hazardq~is
d~ carry a risk, and that such a warning is necessary for the sa~fe and
effective use of the drug by physicians and patients. :Enough, time
h~s now pnssed for interested persons to. have studiet the ~i~nn~t~ic
Society report and the recent detailed UGDP report on ph~nformin.
The Agency has, therefore, published for, co~rnnerit a gu1,ati~n
proposing new labeling for the oral hypoglycemic labelh~g. ~ter-
ested' p~rsons may comment on the proposal by Septehibor ~, 1975,
and a public hearii~g will be held on August ~0, 1975. Final `iaheli~g
regulations will not be published until after all comments ar~d
materials have been considered.
Our proposed labeling has two sections of particular importance; a
boxed warning stating that there may be an increased risk of cardi~-
vascular death associated with the use of oral hypoglycemic drugs
and a new indications section that limits use of these drugs to p~-
tients whose symptoms or blood glucose abnorm~alities can~iot be
controlled by diet alone and who cannot take insulin for one of a
number of specified reasons.
N?w, I would like to discuss both of these sections in greater
detail. And they are both reproduced in full in an attachment to the
statement. The warning describes the UGDP study and its f1nding~.
It has been contended that certain studies said not to support th~
findings of the UG1T~P study should be mentioned in the warnin
section to provide the "fair balance." We have concluded, however, an
made clear in revised regualtions that if scientific data exist to sup-
port a warning, the warning must be presented in unambiguou~
56-592-75------4
PAGENO="0050"
13300 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
terms without disclaimers or qualifications that would undermine or
destroy its usefulness. There is, therefore, no mention in the proposed
warning of other studies involving the oral hypoglycemic drugs. The
mention of studies in which increased cardiovascular mortality was
not found would serve only to encumber the warning.
Mr. GORDON. Dr. Schmidt, you say that the warning must be
stated in vigorous terms without disclaimers. Why, then, does the
warning include-you do not have It in your statement, b~it it is in-
cluded in the Federal Register statement-"that, despite the con-
~troversy regarding the interpretation of these results, the findings of
the TJGDP study provide adequate scientific basis for this warning"?
Is it not true that the presence of controversy about the need for the
wording is irrelevant and is distracting? Is that not an encumbrance,
in a way?
Dr. SCHMIDT. No. We definitely do not feel it is an encumbrance,
and I ~personally feel that it is absolutely necessary to the creditibility
of warning that we state that the TJGDP study, despite this contro-
versy, provides sufficient evidence for the w~rning. I think that were
we to remain silent, or to ignore the fact that many physicians and
many experts have said, in efrect, forget the UGDP study-if we w~re
to remain silent on that, it would be so o~yious, cause so many ques-
tions, make people wonder why we were in a sense trying to finesse
the issue, that it is to me absolutely necessary in order to have a
credible statement that we say that very clearly. The IJGDP study
is sufficient to provide a sound basis ,for this warniflg, and the con-
troversy does not controvert the fact that the, study provides that
avenue.
Mr. GORDON. Well, I thought that it is not just the IJGDP shidy
that you are basing the label on. It is the TJGDP study, the B~p-
metric Society study; you referred to the Wissler stud~r, the Wu
studies, the Tan' studies.
Dr. SCHMIDT. No, I did just. now in the statement. The warning
is based on the UGDP study.
Mr. GORDON. Solely.?
Dr. SCHMIDT. Yes, sir. I can make a couple of other points. The
first is that the encumbrance of other studies is not present, `and I
think that the warning statement is `much better for another reason;
it relates to credibility. It makes the statement believable, and it
renders it, I think, much less. subject to attack and discredit. As I
said early on, one of our goals here is to achieve a professional con-
census about the use of these drugs. So I just strongly believe that
the warning, stated as it is, makes .a much better statement.,
Mr. MERRILL. May I inject one comment? Your question, Mr.
Gordon, is one that will be asked of us again. I have no doubt, that
the charge will be made that we are speaking out of both sides of our
mputh in these very two documents. In one it will be argued we say
there can be no encumbrance of the warning, and in the other we
seek to encumber i~t. That is not true. In the context of this warning,
the s~t~ternent that the administration of oral hypoglycemic drugs
pi~y be a~soqiated with inCre~sed cau~d~iovásc~ilar mortalities as corn-
pared to treatr~ient with' diet alone aád diet with insulin is tw~ full
paragraphs a*ay from the line you just quoted. In addition, the
PAGENO="0051"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13301
reference is to the controversy about the interpretation of the U~DP
study, and says nothing about the significance. of the warning. So I
do not have any difficulty at all reconciling the two statements.
Mr. GORDON. One other thing. You also proposed a boxed warn-
ing at the beginning of the warning section of the labeling. Is ~hat
not correct?
Dr. ScH~IIDT. Yes, in the proposal. That is where it i.s put.
Mr. GORDON. Why not put the boxed warning at the beginningS of
the labeling, as with chlorarnphenicol and lincomycin and 4hn-
damycin?
Dr. SCHMIDT. Speaking only for myself, when I went through it,
its location was logical. It was at the beginning of the warning sec-
tion, where many are, and the thought did not cross my mind that it
could be or should be or might be removed from the warning th~ng,
a~d stuck at the beginning. I think that it is an interesting sugge~-
tion, a valuable comment, and is one that we wilt evaluate, reevah~ate.
There is certainly no-it need not stay where it is.
The CHAIRMAN. With respect to the UGDP study, if my memory
serves me correctly, it was reviewed at the .re~juest of Dr. Marstan,
the then Director at NIH, by Dr. Chalmers, who at that time. was .at
NIH. It was reviewed by somebody else prior to calling upon the
Biometric Society to evaluate it, wa~ it not?
Dr. CROUT. It was reviewed by a number of people, and it was
reviewed by ~our own staff. It was rev~ewed by people, at the NIH.
It was reviewed also by critics. So, rather simultaneously, there were
several reviews that appeared in early- to mid~187I. It was re,vie~ed
by Dr. Seltzer. Dr. Schor and Professor Cornfield also had detailed
reviews of it.
The CHAIRMAN. Well, what were the conclusions of the other
groups that reviewed it?
Dr. CROTJT. The major conclusions of Dr. Seltzer and Dr. Sc]~tor
were that the study had flaws in its fundamental design and execu-
tion to the extent that it was worthless, and that was the begim~ing
of the controversy. Dr. Cornfield then felt that those criticisths,
while correct in certain respects, were insufficient to negate ~he
study. I must also say ~Dr. Alvin Feinstein also had a very critipal
review. So, the major critical reviews were those of Schor, Feinstein,
and Seltzer. The major supporting review has been the Biomet~ic
Society, and the review by Cornfield,
In our opinion, the review by the Biometric Society was in 1~ar
greater detail, far greater depth, than any of the others, and is again
the overriding review of the study.
The CHAIRMAN. The Biometric Society was a review for the per-
pose of evaluating the validity of the studies. Is that correct?
Dr. CROUT. That is correct~ and it is the only review conducted,
in a sense, by a third party. You see, the controversy. was begun by
Dr. Feinstein, Dr. Seltzer, and Dr. .Schor; and, in a sense, they
wëi~e parties to one~side of the controversy. Dr. Corn~1~ld was a con-
sultant to the study, . ~nd in defending it was, in a sense, on ~t~e
T.JGDP side. So the Biometric, Society review was, in ou~ opinion,
not only the most detailed and thorough, but it also wa~ by a thb~d
party. All the people were previously involved in the `controversy.
PAGENO="0052"
13302 CoMP~TITXVE PROBLEMS IN Tfl~ DRUG INDUSTRY
That is really why we waited so long for that report, and believe it
is so important,
The CnAntMAN. Go ahead.
Dr. ScrnumT. Our warning also points out that, although only one
sulfonylurea and one biguanidé were included in the TJGDP study,
it is prudent from a safety standpoint, in view of the similarities in
chemical structure and mode of action of drugs within each of these
two categories, to consider that the UGDP findings may apply to the
other drugs in each category. The warniug has thus been applied to
all sulfonylurea drugs, and the one biguanide drug marketed in this
country.
Finally, the warning sectiOn states that the clear implication of the
finding that tolbutamide and phenformin may carry a risk not asso-
ciated with insulin; the drug, the label will state: "Should be used
in preference to insulin only in patients with maturity onset diabetes
* whose symptoms or blood-glucose level cannot be controlled by diet
alone and only when the advantages in the individual patient justify
the potential risk (see Indications). The patient should be informed
of the advantages and potential risks of the (drug) and of alterna-
tive modes of therapy and should participate in the decision to use
this drug."
We have concluded that a patient population exists for which these
drugs, properly labeled, can be considered safe and effective. We
have also concluded, however, that this patient population is a
limited one.
The OI~[AIRMAN. This is part of the warning, is it?
Dr. ScIIMnxr. Yes, sir.
The CHAIRMAN. The label states that the patient should be in-
formed of the advantages and so on, and should participate in the
decision to use the drug. I am wondering what you are getting at
there. If in fact you have a patient, who cannot manage it for what-
ever reason,, psychologically, mentally, physically, or what have you;
could nOt use insulin or could, not b~ kept on a diet, and the doctor
~onclud~s you should reduo~ the blood sugar level, that is a case for
which, I would assume, there is general a~greement that an oral hypo-
glycemic would be indicated. Is that correct?
Dr. SCHMIWr. Yes, sir.
`The CHAIRMAN. Well, now, then, when you say they should par-
ticipate in the decision to use one of the drugs, are you then talking
about that very large number of people whose problem can be han-
d'l~d effectively by diet? Are you sayiiig to the physician that be~fore
you put them on an oral hypoglycemic. they must be notified it is the
behel of the physician that their problem can be handled by diet;
that it would be better to use the diet, and that the doctor should
attempt to persuade him to do ~o, and understand that the use of
th~ drug may have adverse effects, and that the patieiit would be
better off on a diet. Is that what you are talking about?
Dr. SCHMIDT. No. I think that this is really quite explicit. The
warning says that the drug should be used in preference to insulin
only in patients with maturity-onset diabetes, who cannot be con-
trolled by diet alone, and oniy ~vhen the advantages in the individual
patient justify the potential risk. Now, this narrows it down to that
PAGENO="0053"
COMP~Pfl~IVB PROBLEMS IN TI~E~ DRUG IN~IJSTRT 13&03
very small group of patie1~ts~ we h~ye talked~ about before? who for
whatever reason cannot be controlled by diet or cannot take insulin.
So first of all, we are talking about jUSt `this small g~oup~ and t~ien
second, we are say'ing that the patient i~iust be knowledge~b1e. Gpod
medical practice' dictates that the patient is knowledgeable about
the options of control by diet, control by insulin, and if the, choice
by the physician is to be the oral hypoglycemic agput, w~ are saym~
that the patient should be knowledgeable. about the risks of this drug'
to `which he may be subjected.
Now, I could use another analogy that may, help a little. bit. If one
has hypertension, he really ought to avoid salt. Now, I do not know
if you have ever tried a salt-fiee diet, but a salt-free diet or low-
salt diet is an extremely difficult, and in many respects uncomfort-
able sort of trial. When drugs came along that increased the renal
excretion of sodium, there was a very strong tendency on the part
of the patients to say, this solves my problem. I can now eat salt
again~ and have tasty food, and take the drug, and everything is fine.
I will be just as well off as I was when I was on a salt-free d~et.
Physicians. found it much easier to prescribe, a pill than it was to
fight-~and believe me, it is a constant struggle with patients ~o keep
them on a salt-free diet, or to control a diabetic by' diet. So,.. taking
the pill was really kind of a step toward the brave new world.
Now, when I used to. inform my patients. that, yes, they could io~er
their sodium. with this drug, and not be as ,strict about their di~t;
but when they took a drug, they were running these risks. Sometimes
for the first time, I got compliance on the part of my patients with
this sodium-free diet, the lo~ sodium diet. Now~, what I am talking
about is~ simply good medical practice tl~at would, be accepted as good
medical practice by, anyone. And we are say~ig, in this ~abelix~g,
that `patients, for the reasons I just illustrated with my analogy, m*st
be informed of this possibility of,increased risks;. a~u4 as pai~t o~ their
management, they mast know the options ,of insulin and qf th~ir
using. dietary coi$rol. *, ` ` `
The CHAIR~N~' Well,. then., if I iip4erstaud the whole paragraph.
quoted there, it is~ addressed to' a ~e~y narrow speotrun~, of patients.
Dr. Sc~xD~r. Quito ~o,, yes. . `
The CimAn~Aw. And, if I iñterp~t it, right, you are saying that if,
as a practical matter, the patient's blood sugar can be controlled 1y
diet, the doctor should not give him the drug.
Dr. Soim~mT. That is our opinion,' yes. flut, as~ I indicated,~ we do
believe that there is a small patient population for which these
drugs, properly labeled, can be considered safe and e~ëctive, ax~d
we have also concluded though that this patient population is quite
limited.
The CHAIrnXAN. I have forgotten the figure of the estimate of the
number of patients per year that were, receiving prescriptions fpr
oral hypoglycei~'ics? Was it 1.5 milli9n?.
Dr. SOUMIDT. About 11/2 million patients is a rou~h estimate.
The CHATRI~tAN. iou. testified a. few months ago that there has not
been any care±ui stuthes to show what. percentage of those receiving
oral hypoglycemics are receiving them for properly indicated. reasox~s.
PAGENO="0054"
13304 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
If the figure is 1 million and a half and if you take, even Dr. Brad-
ley's estimate of only 20 percent, Dr. Bradley is saying that 80 per-
cent of a million and a half people-about i,200,000-should not be
using them. In the view of Dr. Davidson, only 1 percent should he
taking them which~ means that `almost the whole million and a half
are exposed to needless risks; Only 150,000 diabetics are taking them
for proper indications. Is that correct?
Dr. SCHMIDT. It would be hard for me to believe that anymore
than, say, one out of five or two out of five, at the absolute outside,
of people who are now getting these drugs by these indications should
get them.
The CHAIRMAN. Well, if you go above one out of five you are
above Dr. Bradley, who has been one of the most vocal critics of the
UGDP study.
Dr. SC1Th~IDT. I have really no sound basis on which to pick a
number.
Dr. CROtIT. I seem to think you are emphasizing' the importance
of this relabeling and what it means to the practice of medicine. I
would also emphasize that' we are talking about the United States
and point out that these drugs are used worldwide. No country has
yet, to thy knowledge, put a warning of this type on the labeling,
and, yet, we do know that when the Food and Drug Administration
of the United States does something it tends to cascade worldwide.
These drugs are' used enormonsly in Europe. I am told, for exr
ample, that the number one selling prescription drug in `Germany is
not a tranquilizer as it is in the United States, but an oral hypo~
glycemic drug. So' this partièular action `will, I think, have world
impact., and we are sensitive to that. it is also whyit `is so terribly
important to the drug' industry because it is multinational.
Mr. GORDON. In an article in the Journal of the American Medical
Association, Dr. John Davidson stated':
There has been a striking increase in death rate and decrease in life expect-
ancy in maturity onset diabetics in America, Europe, Asia, Africa, and Aus-
tralia during th~ last 20 years. These changes have paralleled the increasingly
widespread neglected diet therapy and the almost unbridled enthusiasm among
many physicians and patients for the use of ~ulfonylureas and treatments of
choice. `They seem to parallel the rise, and the uses and increaCe in these drttgs
and increase in the death rate and the decrease in life~ expectancy among
diabetics.
Do you have any comments to make on that?
Dr. CROUT. We have not reviewed that situation, and I would
not want to engage in the sensationalization of putting thos~' two
things together. That statement may be true or not true. I do not
know.
The CHAIRMAN. Go ahead.
Dr. SCHMIDT. The limitation of the treatment population to pa-
tients on whom insulin cannot be used has been opposed' in the past
on the ground that it has interfered with the practice of medicine.'
We recognize that drug labeling has an impact upon the p1 actice
of medicine, and I think' it should' for this reason. The FoQd and
Drug Administration has an obligation to ensure that drug labeling
is as correct and accurate as possible. It must, moreover, meet the
statutory standard of describing the conditions and use under which
PAGENO="0055"
C0MPETITIVI~ PROBLEMS ~IN THE DRUG INDUSTRY 13305
the drug may be considered safe and effective. If a known hazard
and potential risk leads to the conclusion that a drug may be used
safely only' on certain patients, this limitation on use must be
expressed in labeling.
The indications section, in addition to describing the populatibu
in whom these drugs is indicated, points out that "in considering t~ie
use of (drug) in asymptomatic patients, it should be reco~niz~d
that whether or not controlling the blood glucose is effective in pre-
venting the long-term cardiovascular or neural complications of d~a-
betes is an unanswered scientific question." This emphasizes the
different benefit-risk considerations that obtain in the symptomatic
patient who, needs alternative treatment if insulin cannot be used,
and the asymptornatic patient, whose n'eed for alternative treatment
is debatable. I think we have discussed this point at some length
previously..
You asked that I comment bn `the promotion of these drugs. We
cannot conclude that advertising for these productsha~s been generally
violative. It has, however7 been based upon labeling that is in need
of modification. It is clear that promotional materials must change
radically to reflect the new warning and restricted indications. YQu
can be assured that we will be monitoring the advertising of the~e
products closely after the ne~ labeling becomes final to see that they
do, indeed, do so.
The CHAIRMAN. i~o you permit reminder `advertising, and how do
you handle them?
Dr. SCHMIDT. Well, we permit reminder ads, yes.
The CHAIRMAN. With no claim?
Mr. MERRILL. This drug, Senator,' carried a boxed warnin~. We
have in preparation a final order that is responsive to a notice~ of
proposed rulemaki~ig published last year that would prohibit ~the u~e'
of reminder ads for any drug that carries a boxed warning.
The CHAIRMAN. I see, so any advertising would include the `box.
Mr. MERRILL. It would include a brief summary-the full range `of
information. `
The CHAIRMAN. I see, go ahead.
Dr. SCHMIDT. It is important to realize that the use of these oral
hypoglycemics remains widespread despite the, UGDP study and
despite the rather limited ability of the drugs after a few years of
use, even to lower the blood sugar. Total prescriptions for this class,
according to the National Prescription Audit, have been stable b~-
tween 19 million and 21 million since 1967 (except for an apparent
dip in 1969.)
The CHAIRMAN. That is per year?
Dr. SCHMIDT. Yes.
The CHAIRMAN. What does that prescription mean in this context?
Dr. SCHMIDT. Well, any `one individual would during the year r~-
ceive more than one prescription.
The CHAIRMAN. Well, if 1½ million people `were getting it,' o~
course, it would not always necessarily be the same' person.
Dr. SCHMIDT. Well, if these figures are' accurate what that means
is the average person would receive over 15 prescriptions or one
prescription a month. I will not put my career on the line toward the'
PAGENO="0056"
13306 COMP'ETITI~VE~ PROBLEMS I~ T~IE DRUG i~puSmY
accuracy of those figures, though. Generally, physicians control when
they se~ patients~ by writing presc~iptions~ though, and I used, for
example, to he sure that I would see a patient at ~uch ~nd such a'
time by being sure that he ran out of medicine and had to cOme in,
so that it is not unusual for indications of this kind to have more
prescriptions than some others.
But the point of these relatively high figures is that there is a
great deal of common practice to overcome before use of the oral.
agents will proceed to what we would consider to be proper levels.
It is anticipated that publicity attendant upon publication of pro-
posed labeling by FDA and the announcement of the upcoming
public hearing, as well as publicity relating to today's. hearing, will
bring the new labeling to the attention of physician~s, as we begin
the long process of persuading them that the UGDP findings should
change the way they treat diabetics.
in addition we plan to issue a drug bulletin when the labeling for
these drugs is made final. We will monitor the use of these drugs and
will take additional measures as necessary to publicize the labeling.
This concludes my formal statement. I will be happy to respond
to additional questions.
The CHAIRMAN, Thank you, Doctor., I note your comment that
there is a great deal of ,coirimon practice to overcome before the
use of oral agents would recede to its proper levels. I think yOu face
a formidable task, considering that you have. Dr. SamrhOns of' The
AMA writing to all the State medical and county medical societies
in the country and then Upjohn Co., using his letter. We received a~
letter from 13r. Max Miller, who, as you know,, is director of the
UG'DP study. He sent along a copy of the letter that was sent to him
by a doctor r~porting on what the. Upjohn sale~man said to this
doctor about the study. Dr. Miller did not wish that the doctor's
name who wrote to him be disclosed, but he did not object to his own
name being used. , ,
This doctor wrote to Dr. Miller and said, "Deai~ Max,~, Here , i~ a
summary of what the Upjohn salesman said to me in, his visit yester-
day: (1) there is no' cause-and-effect relationship revealed in the
study between the use of Orin'ase and the incidents of coronary dis-
ease; (2). the statistics are so comp~1icated that only a student of sta-
tistics ban evaluate' them; (3) 2½ less coronaries in the study would,
not change the results; (4) 35 diabetologists do not accept the results.
of the, study; (5.) the director of the study in Cincinnati dOes not
accept the results; (6) two other men in the study do not accept the
results; (7) Dr. Kent of Cleveland does not accept the results of the
study; (8) Cincinnati added patients from it~ cardiac, unit to fill its
quota of diabetic patients; (9) most' of the coronaries came from
two centers; (10) there wasno follow-up on~ five patients in the study;
(.lj) the dose of .Oriila'se was fixed so, therefore, it was not a correct
dose for many patients; (12) the Joslin Clinic and other diabetes
clinics have reviewed their cases i~1 their clinics, and the result cannot
support the results of the university study; and (13) the FDA will
probably modify the ruling on Orinase. During the interyiew he. had
a copy"-that is the Upjohn representative-"of the Medical Tribune
in his hand which he referred to from time to time."
PAGENO="0057"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13~3O7
So, considering that Upjohn has 1,100 detail men and other efforts
by the industry, you have a long battle ahead of you to achieve a
standard of rational prescribing of oral hypoglycemics despite the
fact that there are no scientific studies refuting the careful studies
done by the UGDP or the Biometric Society's evaluation. As is cl~ar
from here, the Biometric Society refuted charges that are made by
the detail men. I would have this printed in full in the record.
Dr. SCHMIDT. I would sincerely hope that the medical profession,
itself, and particularly the diabetologists would respond to what to
me is a clear challenge in all of this and would be able to separate
out the principal issue and that is the following: For whom, for wijat
group of individuals, are these drugs suitable, given the IJGDP
study? And as I have indicated earlier in many, many conversations
I have had with the most vigorous opponents of the UGDP study,
there is an agreement that these drugs are grossly overus~d.
Once there is agreement to that, that identifies a very serious
problem, which is, in essence then, apart from the controversy of the
UGDP study. And if 80 percent of these drugs are misused, that
identifies a problem to which the medical profession itself must ~e-
spond, and I will be bitterly disappointed if it does not.
Mr. GORDON. Doctor, suppose you are disappointed and use does
not go down? That is a possibility. What do you think of the idea±-
I brought this up yesterday with respect to lincomycin and clinda-
mycin-about having corrective advertising with surveys by the FD~A.
and continuation of the corrective advertising, as in the FTC's
Hawaiian Punch case, until the use actually drops?
Dr. SCHMIDT. We will, as I indicated, monitor the use of these
drugs. We will certainly monitor the advertising, and we will see if,
indeed, these drugs do become in effect unsafe and this can be showb,
then I would probably have a long talk with Mr. Merrill, but I do
not know. It is hard for me to hypothesize.
Mr. GORDON. May we get periodic reports on the use of the~e
drugs as you get them?
Dr. SCHMIDT. Certainly. I would be happy to.
The CHAIRMAN. You get reports monthly?
Dr. SCHMIDT. Well, we follow certain surveys that are done, such
as the prescription audit survey that I mentioned and others, Thei~e
are some commercial sources of data and others that we follow t4t
can give at least an estimation of the use of the drugs. We can also
undertake surveys of our own.
Mr. MERRILL. Our information is not as good about drugs in this
class as it would about antibiotics, which are certified, of course, and
thus, we know how much is being produced. But we have access to
pretty good numbers.
The CHAIRMAN. Thank you very much, Dr. Schmidt, for your very
valuable testimony today. The hearings will open tomorrow mornin
at 10 o'clock in the same room. Senator Abourezk will preside as
have hearings starting on the energy legislation in the Finance
Committee that has come over from the House. Thank you.
{Whereupon, at 11 :35 a.m., the subcommittee recessed to reconvene
at 10 a.m., the next day.]
PAGENO="0058"
PAGENO="0059"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
(Present Status of Competition in the Pharmaceutical
Industry)
TEURSDAY, ~~ULY 10, 1975
U.S. SENATE,
SUBCOMMITTEE ON MoNOPoLY OF THE
SELECT COMMITTEE ON SMALL BUSINESS,
Washington, D.C~
The subcommittee met, pursuant to notice, at 10:05 a.m., in room
318, Russell Senate Office Building, Senator James Abourezk pi'e-
siding.
Present: Senator Abourezk.
Also present: Benjamin Gordon, staff economist, and Kay Klatt,
research assistant.
Senator ABOUREZK. The hearings will come to order.
Part of the panel is here this morning, Dr. Sims and Dr. Chester.
Is that right? Dr. Felig and Dr. Lamer will be here soon.
Unfortunately, I have to run over to the Senate Democratic Caucus
for a few minutes. I am going to open the hearings now, and Y?U
can begin your testimony. I shall ask Benjamin Gordon, our staff
economist to receive your testimony until I come back, so that we do
not interrupt the hearings. I shall he back as soon as I can.
So, if you are ready to begin your testimony, we are ready ~o
receive it.1
STATEMENT OP EDWARD M. CHESTER, M.D., PRGFESSOR OP MED~[-
CINE, CASE WESTERN RESERVE UNIVERSITY, AND DIRECT0~,
AMBULATORY MEDICINE TEACHING CLINIC,' CLEVELAND
METROPOLITAN GENERAL HOSPITAL
Dr. CHESTER. Mr. Chairman, I am pleased to respond to ~he inv~-
tation to testify concerning the use of oral hypoglycemic agents i~i
the treatment of diabetes mellitus.
I am a professor of medicine at~ Case Western Reserve University
and director of the ambulatory medicine teaching clinic at Cleve-
land Metropolitan General Hospital. A large segment of my time
is devoted to teaching 3- and 4-year medical students during their
clinical clerkships.
My research efforts have been directed toward an understanding
of the eye changes which are associated with diabetes meUitus. Eoi~
1 See prepared statement, page 13643.
13309
PAGENO="0060"
13310 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
the past 16 years I have been active in the direction of the diabetes
clinic at Cleveland Metropolitan General Hospital. Prior to 1959,
when I joined the full-time faculty of Case Western Reserve Urn-
versity at Cleveland Metropolitan General Hospital I had been
engaged in the practice of internal medicine for 18 years in a subur-
ban area of Cleveland. My practice dealt chiefly with patients who
suffered from cardiovascular disease and/or diabetes mellitus. Dur-
ing the years of* practice I served as `a part-time teacher at Case
Western Reserve University at Cleveland Metropolitan General Hos-
pital.
The diabetes clinic at Cleveland Metropolitan General Hospital
provides care to approximately 500 patients with diabetes mellitus
per year. This totals approximately 2,000 patient visits per year.
Approximately 80 percent of the patients have the maturity onset form
of the disease. Prior to 1973, the majority of this group of patients
were treated with oral hypoglycemic agents with a limited degree
of success. Although dietary instruction was provided for the patients,
there was little compliance.
When the results of the UGDP study were issued in 1970, we be-
came concerned with our use of the oral hypoglycemic agents. We
urged the physicians who cared for patients with diabetes in the
clinic and hospital to pay heed to the results of the above study and
to reevaluate their treatment of the maturity onset group of patients.
In an attempt to learn the extent of the use of oral hypoglycemic
agents and their cost; the amounts of these medications di~pensed
by our staff were recorded from 1968 to early 1972, and I would
refer you to table I. Review of these data diselosed an alarming in-
crease in the use of these agents from 1968 through 1970. Response
to the recommendations of the UGDP study was reflected by~a modest
decrease mthe use of the oral agents during 1971 and 1972. Because
we behevecF that the use of these agents was still excessive, the fol-
lowing letter was dispatched to the chairman, pharmacy committee
of the hospital on May 24, 1973.
The results of the University Group Diabetes Program, UGDP, Diabetes, 19,
Supplement 2, 747-830, 1970, allows one to develop the following conclusions
concerning the safety and effectiveness of the oral hypoglycemic drugs, specific-
ally the sulonylurea group, Tolbuta~ide and Chiorpropamide. (1') In the group
treated with Tolbutamide there was a significant increase In deaths from
cardiovascular disease, as compared with those treated with either insulin or
strict adherence to a calculated diet. (2) ribat Tolbtttamide was not as effec-
tive as either insulin or strict adherence to an isocaloric diet in the control of
levels of, blood, sugar.
The UGDP study subsequently reported comparable results with the use of
Phenformin, JA.M.A., 217 No. 777-784, 1971.
It is only fair to point out that there are skeptics who do not accept the
results of the above study.
I accept the results of the study and believe that the use of Sulfonylureas,
Tolbutamide and ChlorprOpamide, and the Biguanides, Phenoformin, should be
restricted because they appear to be hazardous to health and are far less effec-
tive and more expensive than Insulin.
I suggest that we implement a form of control which would restrict the use
of $ulfonylnrea drugs, Tolbutamide and Chlorpropamide, and Phenformin with
the following exceptions:
One, patients who cannot administer Insulin to themselves because of severe
visual impairment or other physical handicaps such as neurologic disorders
which impair use of arms and hands.
Two, patients who refuse to use insulin.
PAGENO="0061"
~O1~tPETITIVE PROBLE~ IN. ~ DRUG I~pus~ir 13311
In order to acc~mpIish. ~ contrQl, the department. of medicine Would PFO-
vide a list of phystciaz~s who could authorize the use o~ the drugs unaer dis-
cussion. Other se~vtces may wish to provide a ~ithilar mechanism.
`In 1972, $30,000 were expended' for T~lbutamid~, Pheirforniln, and Oblor-
jropamide. Substitution of insulin would be less costly.
The results of this educational reminder and form of contrdl~ro-
ditced the results noted in table I.
Table II indirectly indicates that many of the patients prev1o~isly
receiving oral agents were started on insulin therapy. I would like
to add that i,f we totaled the cost of the oral agents in 1970, it i~eaohed
$56,000. By 1974, this was reduäed 55 percent (sic) to $9,676. That in-
formation is not recorded on the table.
Continuous review of the use of the oral agents is' in progi~ess'with
the intent of f~irther decreasing their u'ae except un~et' th& circinn-
stances noted' in the letter of Ma~ 24, 1973. It is a~pp'arelIt `that re-
striction of the use of these `medications in a hospital can b~ a~cco~np-
lished by education of patients and physicians `atn4' `by proy4dih~g a
method for control.' The problem is unfOrtunately not A~ simple' for
a variety of reasons when one attempts to achidve similar `results
with patients who are under the care of privatu physicians. Among
these reasons are: One, that the conclusions of the `UGDP study are
not accepted by some diabetologists. " ,
Mr. GOlwoN. May I interrupt for just a momthiti
E'~en the critics of the UGDP study admit that `i~ `i~ tlie~b~st~ st~idy
conducted in this field. Other studies, including animal stadies, have
confirmed the validity of its'conclu~ions.
Why, then, have some physicians-even some with, prestigious
`names in the diabetes field-continued to attack these studies, even
though they acknowledge lack of effectiveness of these drngs?
Dr. CnEsn~R. M~. Gordon, this is extremely difficult to'understand.
I can envision that some of them believe that lowering the blood
sugar may prevent the vascular disease. However, there is no evi-
dence to support that contention.
Mr. GOItDON. Would any'other witnesses care to comment?
Dr. SIMS. Would you be willing, Dr. Chester, to add as a qualifying
phrase in thIs group of noninsulin dependent, predominantly over-
weight diabetics? ]~n other words, I am asking you, would you make
the same statement for the juvenile diabetics?
Dr. CHESTER. Would I make a different statement for the juver~ile
diabetics? No.
I think that there are no data to support the concept that the con-
trol of diabetes, as we measure it, namely levels of blood sugar and
quantities Of sugar in the urine, will' prevent the development of ~he
vascular diseases that we see in the' coronary arteries, the peripheral
vessels, and in ~the eye. `
M~. GoiwoN. Dr. Felig, would you care to cowmet~t?
Dr. Fi~pi,G. I' agree with Dr. Chester's `remarks.
I think' that the kinds of treatments available t~' us, be it insulin,
oral agenth, or diet, ar~ such that we do not fully restore the patie4t,'s
metabolism to normal, and I think that we can at least explain, why
we `might not see the improvement in' prevention' of ~va~seular diseas~.
I might comment on the point that you raised. as to wh~ people in
the fi~ce of evidence of lack of effectiveness still criticize the UGI~P.
PAGENO="0062"
13312 COMPETITIVE PROB~JEMS IN THE DRUG INDUSTRY
It should be recognized that some of the most vocal critics of the
UGDP have raised issues regarding experimental design, statistical
methodology, and some basic aspects of the approach to clinical trials
without necessarily tending to promote the use of the oral agents,
as such. But, this has then been misinterpreted as a validation of
the oral agents. Specifically, one of my own colleagues at Yale, Dr.
Albert Feinstein, has been a notable critic of the UGDP study. Dr.
Feinstein is an outstanding biostatician, clinical epidemiologist, as
well as internist. The basis of his criticism is not directed at trying to
promote the use of the oral agents; but he has looked upon this
particular study and has raised some statistical questions. Others
have unfortunately used his objections as evidence for the perpetua-
tion of a treatment which they themselves say is hardly effective.
Mr. GORDON. Dr. Feinstein, as I understand it, has never made that
clear, that he is not really pushing the drugs.
Dr. FELIG. Well, while this may not be clear from his public state-
ments, knowing him well and havi~ng worked closely with him, and
since we both have appeared on papels on this issue in New Haven, it
is quite clear that he does not particularly favor the utilization of these
agents but, he is merely raising questions of experimental design..
Dr. SIMS. If I could just add a word.
I do agree with Dr. Felig that,as ordinarily carried out, the partial
regulation of blood glucose has not had demonstrable effect in pre-
venting gross cardiovascular lesions in the group of largely oyer-
weight people in the UGDP, most of whom did not actually need
insulin,
I do believe, however, and Dr. Crout made this point yesterday,
that we do not have the evidence from the UGDP to justify extend-
ing a spirit of therapeutic nihilism with respect to blood sugar regu-
lation to. all types of lesions and in all types of diabetes.
I do feel that with Dr. Felig, that has ordinarily accomplished the
regulation of blood glucose, it is not altering the cardiovascular effects
from all the measurements we have up to now in the group of over-
weight people who do not actually need insulin anyway.
I do feel, and I think Dr. Crout made the point yesterday, that
we do not have the evidence on that study to extend the therapeutic
nihilism in applying it to all parts of the body.
Would you agree with that?
Dr. FELIG. I would agree.
Mr. GORDON. Dr. Chester, please proceed.
Dr. CI-TEST~R. Two, that education of physicians lags well behind
the knowledge developed through research. Unfortunately, drug com-
pany literature provides the major source of information for many
physicians.
Three, the lack of adequate patient education in their understand-
in~ of diabetes and the hazards of oral, hypoglycemic agents.
Four, the failure adequately to impress the patients with sufficient
understanding of the importance of a calculated isocaloric diet and
their failure to comply in this respect.
Five, the case of using oral medication compared with the in-
jection of insulin.
The UGDP study clearly demonstrated that standard doses of
oral hypoglycemic agents `did not effectively reduce levels of blood
PAGENO="0063"
COMPETITIVE PROBLEMS IN THE DRVG INDUSThY 13~13
sugar over a 5-year period~ These data confirmed j?revious ~stiic~ies
which disclosed that the succesS rate in managing diabetes with tol-
butamide at the end of 5 years was only 13 percent.
Mr. GoBr~oN. Dr. Chester, what precisely does the 13 perceiit rejire-
sent? Does it mean that in only 13 percent of the ca~es w~s the ~ol-
butamide successful in lowering blood sugar at the end of 5 yeai~s?
Dr. CHESTER. At the end of 5 `years. There may be early ~uccess if
one measures levels of blood sugar. This is called primary faili~re.
That is, there is no response `*~thin 1 month. In ~fiect blood su~ar
levels are not at the range that one desires withifi that period. Tl~en
there are secondary failures in patients who initially appear to re-
spond. Again, this is determined by levels of blood sugar. Sub~e-
quently, over each year there are more and more called secondary
failures, so that by the time the 5-year period arrives, at least in
this study, only 13 percent were still responding.
Therefore, we have a drug with limited effectiveness that pro-
gressively loses its effectiveness.
Mr. GQRDON. Well, there is no evidence that lowering blood sugar
prevents tha vascular complications resulting from diabetes. Is that
not correct?
Dr. CHESTER. Yes.
Mr. GoiwoN. Then, what are we actually accomplishing when ~ve
lower blood sugar?
Dr. `CHESTER.' Well, we do accomplish a variety of' things. If the
blood sugar level becorqes exces~~ve, then the amount of urine, ~
et cetera, are passed out into the urine, and the patient not only lobes
tremendous amounts of water and becomes dehydrated, but may suffer
from some of the loss of the electrolytes.
Second, continuation of poorly controlled diabetes, again measured
by levels of blood sugar, may be followed by a variety of very serious
manifestations. One is diabetic ketoacidosis, where presumably, l~e-
cause of lack of insulin and other factors, large amounts of the fat
are broken down, mobilized, converted to a number of substances
known as ketone bodies. As these substances accumulate, the patie4it
may become unconscious and death may follow.
There is a comparable state in which the blood sugar reaches e~-
tremely high levels, perhaps in the range of 1,000 or above, where
`extreme loss of water becomes critical. These patients may becon~e
unconscious and die within a relatively short period of time if nçt
treated adequately.
There is also t~ie question of whether or not keeping bl6od sug~r
at given levels will prevent infection. This is difficult to document.
What we do know, however, is that the individual with diabetes whp
develops infection, unless we treat the diabetes vigorously aud simul.!
taneously treat the infection,' the patient i~ likely to suffer and ma~y
die.
S~, there are reasons to try to reach given le~eis of blood sugar.
The difficulty is that in general no one of us knows what the optimu4'i
level may be. It is extremely difficult tO restore the blood' ~ugár levels
to those that supposedly normal people wOuld carry throughout the
day witho~t the risk of develop~ng extremely low blood sugar or
hypoglycemia, which in turn may damage the brain and cause other
problems.
PAGENO="0064"
13314 C0MP~TITIVE PR0BI~1EMS IN TI DRUG INDUSTRY
Mr. GoiwoN. Would any of the other witnesses care to comment?
Dr.. SIMs. .1 thin1~ that it is fair 1?o sa~ that jn~Uh~ as~ we now' a~d~
minister it to a'p~tient, particularl~r one with a high~de~rBe of iflsulin
resistance, falls fa~ shOrt of re~ro~dt~eifl~ i~h~ elegance of corilroi that
the nQrmal body accomplishes. }So', we have to kee4~ ~ reser~atiou in
our minds as to whether, when arid if the day cOnne~ ~vhen we can
more precisely regulate iusuli~n administration tht~u~h an artificial
paucreas, multiple injections, Or some such, the results with respect
to vascular disea~e might not. be quite differeflt. I am just arguing
against a general attitude and nihilism that might lead to some of
the complications that Dr. Chester mentioned.
Mr. GORDON. Any other comments?
Dr. Chester, please proceed.
Dr. ChESTER. These data confirm previous studies `tha~t `disclose
that the success rate. in mauag~ng diabe1~es with tolbutamides at the
end of 5 years was Only 13 perdent~ Eelap~e or secondary failure is
recorded as 22 percent within 5 years. After 6 to ~T years of thc~rapy
with oral hypoglycemic agents, only 6 to 12 perceilt remain well cOn-
trolled. Yet, despite the ineffectiveness of these hypoglycemic agents
and their demonstrated relationship to increased rñortality from
cardiovascular disease, these drugs are still widely use4 in the treat-
ment of matllrity-onset diabetes. I have noted previotisly the reasons
for the failure on the part of physicians and patiefits to heed, the
warmng of the hazards and ineffectiveness c~f. tliis~group,,of drugs.
There appear to be three apprOaohes tO this ~robleifl. `I~hese include
an immediate restriction of their use through firm warning arid label-
ing via the FDA, a long-term educational process, and the develop-
ment of more rigid' drug testing requirements.
Mr. GORDON. May 1 interrupt at this point?
Is it your opinion that warnings and strong labeling by the FDA
will have an effect on the use of these drugs?
Dr. CHESTER. I would say yes,. provided the labeling is done on
the patient's drug box or bottle. Although I think it is fair to say
that the insert in the drug package should have these warnings, by
and large physicians and certainly patient's do not see them.
Mr. GORDON. hr the light of previous experience, do you not think
that something stronger and more dramatic might be more effective?
Dr. CHESTER.. Yes. I would put a label right on the box or bottle
that the patient has, indicating that this is a hazardous drug.
Mr. GORDON. Maybe a skull and crossbones?
Dr. CHESTER'. Yes. I would put on a skull and crossbones.
Mr. GORDON~ I am kidding, by the way. Are you serious about that?
Dr. `CHESTER. Well, something close to It.
Mr. GORDON. In the light of what you know about these drugs, and
if the drug firms were ~only now seeking approval to market them,
would you approve' them as being safe and effective for the' purposes
claimed?
Dr~ OH~sm~. No. If they' are related to the sulfonyl~~irea group or
to thebiguanides, I can not see that any modificatipns~iu the drug or
prolonging their half-life, or `making them more potent, will resolve
this problem.' ` `
Mr. GOEDON, So, those drugs that are now in the NDA process,
which are now being considered by the Food and Drug Admini~tr~-
PAGENO="0065"
~OMPI~i1ir]~ P~ Si,IN~i~ffl D ~7S~y 4~3815
tion, ~he~g1yb~ide~, ~the
npprov~t~kte~on~s th áa~n th~ mavk~t~owy~ôu tain1y~if1~ttrn~er~
stand it, wou1duiotappxo~ the~othe~ ~nes.
Isthat a~ ~o~e~tassiin~ption?
Dr. ~Oi IRJ~That is tw~eet.
Mr, xo~ If, as you~tate~ tho~o~al hyp~g1~c~inics ~re a~pot~ntiai
hazard~tó health, ~nd if *e are~no1isurethatthey~re eff~etiwe;~what
is these~se~f nsingthern?
D*r~ U sa'ER.~WeU~ there are people~who,~ cfor~ ~reasonsanentiohed,
retuse to ~take~ins~1in, or, a~e teohandicapp~dnto ixs& it. But ~p~tren-
th~tic~J4y,:we~ ha~e `been faced with the hanclieappedcp~tie~for
many years b~fore wech~d the oralagents. There ~re ways to ~d-
minister bn~iiuindn the .handica,pped ~if its ruse is inthcathd.
For example, there acre ~ywinge~ for the &ithlni~tratiom of insulin
that ~~icb~ ~ssed by p~ople w~th extreme'y ~i~itedi~i~ion, ~They ~ome
~under a vasr~ety,of trade names, Essential~y th~y ~eønsist~of am in~ulin
syringe inwhich you canriock the Tplunger~so that~itcannot5go b~rond
a, given.~dos~ ; so that if as j~atient were to ~eceiire~O units~ of :ins~lin,
for example, you could set the plunger at that ievelz E~enwith~lin~ted
visionrthepati~nt~c~n ~titMraw the in~ulin from the vialand can
ject ~it ~nto thessthigh~ith little difticuity.
I think the other thing that we ~imist consi~er~ ~nd `this isrpar4i~u..
larly tmeinrla,~ge i~iics~ such ~s ~ou~s, that ~e frequ~n4Jy do not
have, good., o~nta~t with re~ati~ves who might ber'wi~c1ring `to ~i~arn~this
r~techi~iqme. fot~1y,~~ enois~gh f1~ortsis rn~td~ t~ ~clo this. `The
p~actioing~ physi~iawhas anc adiva~utwge. He~ofben knows the sfa~1s~ly
w~ll~ a~id .1 m~gh~c ~te p~r~thet~ca1~iy, that ~hiie~i~a,pta~ice, ~i1is
wasnos,prQb~rn.rL w~sable tQ ~get~am.i~Iy ~1~mb~rsrto ~ th~w~to
~adm~iister insulin.
`4sgeneies ~no~u communities cam ~ftens pro~id~ ~
~ we~bave as~Visict4~g ~N~isirse tAs~ociatjojr~in Ci~vel~,nd )t~hat~v-ifl
perfuche~ice.
M~. O~RDeN. What pu~zies n~e is ~f these du~igs are harmful' to
~ they ~ to ~p'le who
are ~
Dr.~ Gs~i~. Y~, tb~yare~a~d I woulcics~y't~at~nyt ~ime~~ase
them, there is a calculated risk. The patient and thefa~aily ~sheuldfr
this ~i*.
Mr. (~-OawOw. ~In other woths,~being handie~ipped ~or being blind
doeS not iaaclte you Wimnneto. the msks.
Dr. Uu~ ~n.~t~hat is ~c~rreot.
Mr. GORDON. Doctor, go àh~ad.
Dr. G~I~s~En,R~con ew1atic~is: Quo :Jmmed~ate warnh3g to all
phy$ieiansof~'th~ ~u~ds by, a:huUGthi ~fr~m the~fl~L stfttimg1~he
.foflo~wb~,g: ,~a) ~~rsu~'t~tbiy ek1~ci4a,ted iso~akric~ s~l4~t ~ser~s ~as the
cQrne~o~c ~ re ~
a~le~t~a1 ~l~i~~ry `~a ge~nt,~pr~ ~
tutcdd~L ;(~)~jJ~ quater,~e~ls ~of blood ~g~r~a,uuot jtaj~a(~d
~ith tl~ ~ ~ in es~b~ence of r~yn~pt,oZn~ ~~ss~i~atec1. ~i~h
S ~ ii~sti~ute4; ~d) ~
~mie~nt~re, aip~t t~a1c l~~a~J toc h~~dr~ho.uid (bet ws~d,
~after cthe ,~p~i~tt1~s ~ of ~tl!4s f~ct, ~
5~-~-_7~__5
PAGENO="0066"
13316 COMPETITIVE PROBLEMS IN THE DRUG I~DVSTRY
under the following circumstances: One; If the patient is handi-
capped by serious visual loss or other physically incapacitating dis-
orders; and two, if the patient refuses to use insulin.
The drug companies should be required to include the above facts
on the package inserts of medication, despite the fact that physicians
infrequently read them. Some method of identifying these medica-
tions as hazardous must be developed for patient protection.
Two: Long-term educational effort. Medical school educators, dm1-
clans who care for patients in university and community hospitals
must emphasize the facts known about these drugs to medical stu-
*dents, house officers, and physicians in practice. Efforts should be
made to reach the last mentioned through post-graduate courses and
through the development of self-educational units in an attempt to
providQ more reliable and scientifically based information to counter-
act th~ biased and often inaccurate statements issued by pharmaceu-
tica~ cOinpanies and the throw away pseudomedical periodicals.
~,vital step in the educational process is the need to encourage and
support the young investigator. Greater availability of research and
training grants through the National Institutes of Health or other
Government agencies should be encouraged. For, it is through the
development of such investigators and teachers that the many prob-
lems related to diabetes may be resolved.
Three: Adequate long-term trials before drugs are released for use.
Most dimgs, and this applies to the oral hypoglycemic agents, were
initially tested for their ability to lower levels of blood sugar in
animals. Search for toxicity was made as well. These studies were
short in duration. After short-term trials in man were made by able
investigators and clinicians, the drugs were released. Subsequent
long-term studies of these drugs were retrospective and dealt only
with their ability to alter levels of blood sugar and lipids. The TJGDP
study was the first well-controlled prospective study and was designed
to determine the role of thes~ drugs in the development of vascular
disease. Thus, many years elapsed before medications, which were
commonly used1 were found to be hazardous to health and to possess
very limited effectiveness. Standards for long-term studies must be
developed by the FDA to insure adequate clinical trials of drugs be-
fore their release.
The steps inçlicated above am~e likely to be met with severe outcry
and resistance by pharmaceutical companies and scientists and clini-
cians who do not accept the conclusions of the TJGDP study. Con-
tinued support of the medical socIeties, particularly the American
Diabetes Association, would be essential.
Restriction in the use of theoral hypoglycemic ag~nts would ~ig-
nificantly alter modes of care for the patient with diabetes. To begin,
it would needfully provide ,a great emphasis on the impOrtance of
dietary management. In many instanc~s with adherence to diet, ade-
quate reduction of blood ~uga~ and removal of, symptoms would fol-
low. Physicians or their assistants would have to instruct patIents in
the use of insulin when dIet, ai~ne did n~t suffice.: Thus, mo~re teach-
ing wQuld b~ n~eded, for each patient. Perhaps more teaching re-
lated t~ mecb~nisn-~s inv~lyed in the producti~n'. of the dis~ase, the
need for preventing infections manifestations of hypoglycemia, and
PAGENO="0067"
COMPETITIVE PROBL]~MS IN TW~ DRUG INDUSTRY 13317
other measures would be taught as well. Since the cost of insulin is
considerably less per patient than oral hypQglycemio agents, t1~ere
would be a decrease in total cost.
The issue of the clinical use of research information is exemplified
by the mixed reception of the results and recommendations of the
TJGDP study. Why, one may ask, are there delays in the trans-
mission of research data to it~ clinical applications?
There are several reasons: One: Early research data may be pre-
sented initially to select groups in research societies ~nd published in
journals which are read by only highly trained speeiahsts. In ackli..
tion, most articles are not published for at least 6 months after tl~ey
have been submitted. Two: Further delay occurs because of the need
for clinical testing. Three: When the information is finally released,
there are varying degrees of receptivity and understanding. Here we
deal with a number of variables which include initial training and
continuing education of physicians. Medical educators, both basic
scientists and clinicians, and medical societies must play an important
role in narrowing the gap between delivery of research information
and its clinical application.
Corrective measures in this regard are most likely to be effective if
medical students, fellows, and house officers in training are a4e-
quately prepared to receive audi evaluate research data. This requires
improvement in the teaching of basic science, biostatistics, and clini-
cal pharmacology during medical school and postgraduate training
programs. As a teacher of students and physicians in training dur-
ing their formative years, one is aware of the need to stimulate them
to share in the joy of learning. Such an effect develops and fosters
intellectual curiosity, critical thinking, and the self-discipline re-
quired for continued intellectual development throughout their c~-
reerS.
During their period of formal training, they will recognize the
need to continue their education once they embark upon their careers
as practitioners. Reading current literature, attending medical meet-
ings, utilization of self-educational material, and attending specific
postgraduate courses are effective approaches. Physicians should be
urged, `if practicing in groups, to exchange information and ideas
with peers. Journal clubs and conferences could be developed. As a
former pi~actitioner, I found that becoming a part-time teacher at a
university affiliated hospital was an excellent learning ~xp~rience an~1
a considerable stimulus to encourage my own intellectual develop-
ment. Medical schools should encourage suitably trained physicians to
participate In clinical teaching.
The task of communicating with the well-established pi~actitioner is
more difficult. Those who are well trained in various major specialitie~
generally keep abreast of new developments in their area of interest
and expertis~ through many of the educational methods previously
mentiOned. TJhfortunately, there is another group of physicians who,
because they are either overwotked or inadequately trained, find o~
take little time to i~ead or attend educational meetings, and rely upor~
ill-informed pharmaceutical ~corbpany ~ep~'esentathes and thedical
thro~áwayS for their sources. of information. Malr3r of them. observe
that because of theiv lack of~ scientific backgro.und and ti~e~ tren~e~id~ot~
PAGENO="0068"
13318 co~P~Ti~rIvE PftGB~EM IN ~L~I~2E ~EU~G I~AUST~Y
burst ~f new h~for*mation tha~t they cam~ot under~taud and pro~b
from eu~rea~t n~e~1icai ~i~beratu~re. They are thus oe~ri~y pmq~e~d to
accept new research data which are dinically .ap~1icabie. ~s a re~
sn~t, they s~re~t eqi~ppex~ to be~critie~ü od~ someof the. dai~ms ~y&rug
eompan~ies o~E.,the eff&~theness~ o~;v~rio~lits forms of thera4~y.
It is di~ult ~for m~ to envisien n~a~or correetive asures ~or
this group. Obviously they ehould be uiged to M~tend postgraduate
courses in wtrióh efforts woul~t be m~de to hring them abreast of cur-
rent understanding of disease and thera~py. The Academy of ~euera1
Practice has made ~effo~ts to promote asich courses. Medictd ~chools,
medIcal societies at local and national levels must share in this edu-
~ational process.
Thank you.
Mr. Go~mo~r. Thank you ~rery ¶m~h, Dr. Chester.
There is one more ~caastion I woukl like to ask you, buct I thiiik I
shall save it for later because .1 think That the four of you may wish
to discuss it. The next question is: ~Iavayou read the p~roposed label-
ing and what am your comments on it?
But I shall wait, and maybe we can talk ~bout it as a group.
Dr. Felig, would you please give your statement?
~TA1~FII~E~tT OF P~BI1IP FELIG, ED., Th~)F~SSO~R UD ViCE OKAIB~
IVIAN, DEP T~/IE~T ~O!F mTE~NAL ~V ~CiNE, TALE ~R1V~-
1SXTh' SO~!OOL OP U~CINE
Dr. F~o~ I am p~1ease& to have this opportunity to participate in
these heariicigs on the o'r~al i~iypog1ycemic drugs. Over 5 years have ~ow
elapsed scinee~the ithtis~l presentation of the findings of the University
G-toup Diabetes I~rogram indicating an increased risk of do~h from
cardiovascular disease in patients treated with tolbutamide or phen-
formin. ~Sinee that time, there has beeti considerable debate and con-
troversy in ttlie medical profession as to the validityof these findings
and their ho~p1ications with respect to the treafment of diabetic
patients.
My &iscussion will ~focus on the ~foliowing areas: One, those aspects
of the pharmacdlo~y tand ~linical app~lic~tions of the oral hyp&g'iy-
comic agents in ~hich there is fairly uniform agreement amo~g pro-
ponents as ~w~ll ~as opponei~tts of the'IYODP study; two, the impact
which the fi'xtd4i~gs ~f the IYGDP study have had on medical practice;
and three, the `mechanisms by which the prescribing `hthits of physi-
cians maybe faltered.
Virtually all e~perts in the field of diabetes agree that `the oral
hypoglycemic agents are drugs of convenience. They are convenient
because they may be `taken oraliy as opposed to the injections of
insniin. Moi~e iinprtantly, `they are coneenient because `they do not
requfre the ~elf~discipliue and m~iianoe ~iuhei~ent in `a
reduci~ug dietary regimen. `In contrast `to the effiects of ins~liu in'the
patient ~vith dUabetic coma, the cmi h~pogly~erPic ~gei~s are not
lifesaving kkrflgS~ Fm~therrnore, ~ic convincing e~ridenoe ~ available
which indicates i~hat megniotion ~f blood sugar 1y oral 4gents rataPds
or preveilts ctbe iox~g~~term tlegenera~tive e plicat3m~s of diabetes
~hi1i mal ecttbe~eyes, kidney, or `nerwu~ s~stezn. .: ..
PAGENO="0069"
COE~ETIYE~ PEOBT~EM~T i~ TR~ Wti~E~r iN~USti?RY i3~31~
M~. ~ Yoiis~ty ti~iuV the~re~ i~ no e vi~ci~g evid~ae that ~
tm4Iing~ blood ~ig~r prev~nt~ or~ r~ta~s~ the vaseular co~~lieation~
resulting from diabetes.
b titat e~se,~ehen,i~ there. aaiy sense. i~ u~'ng druga that ai~e harm-
ful t~ acco~npli~~i soni~th~iiig that s~e &on't kno~w is hei~fui~
]Dr~ FtLIth Patie~ts may dei~& some benefit from oral agen~t~ by
virtue of~ their ets on the acute o~i short-term consequences of. the
high blood siig~r. When the blood sugar rises to the point of causing
excessi~ve urination, as~ Dr. hester has indicated, the depletion. of
certain, essential body m.ine~rais~ electrolytes,. is harmfuL Tn that ~ir~
cmasta~nce,. there are patients who, because of their total unwiiihg-
ness to use insulin, could benefit from t~Iae ora' drugs, in the sense that
thely couJ4 have symptomatic reiief~ So, in that short-term or limited
sense, one could. ascribe a~ therapeutic benefit from these. d~gs~
Whether or not that outweighs the consequence or the potential risk
is something that is the essence of ~iinical judgment.
Mr.. GOnioN~ We have asked Dr. ~khn Davidson, who. is director
of the diabetics clinic in Atlanta., Ga., what percentage ~f the peo'~l~
using these drugs should actually be using them. He stated that
maybe 1 percent of the people who are using them should be using
them; or, 99 percent should not be' using them~ We asked Dr. Brad~ey
who is one; of the proponents of the use of this. drug-or at least
so he seems to be-the same question. He said that about 80 perc~nt
of the people who were using them should not be using them. Tl~en
I asked T)r.. Winegrad the same. question on the telephone,. and he
estimated about ~0 percent of the. people... ~ yo.u see,. you have fr~in
80 percent to 99 percent of those who a~e using them that.. should
not be using them.
Would you gen.tlëmen care to. make any estimates,. given your
experience, given, the fact that maybe 11/2 million people are using
these drugs?
Dr. Fi~o. I am firmly convinced that there is overutilization,. and
we think that the' fignre of 80 percent represents a very conserva-
tive estimate of overutilization, We estimate that probably somewhere
in the n.eigbborhood of 9) percent should be treated. with means
other than the oral agents.
Mr. GoRDoN. Would the other.s. agree with you?
Dr.. CHESTER. I WOuld.
Dr. t~tNnn I would think 90 percent or more..
Dr.. SIMS.. I think it is a game of selecting a figure. But it is worth
en~phasizing that,. if there is a. minute percent in' which the dri~g
is indicated and th'e arugs are allowed to., remain on t,l~ .rn~rket fQ.r
that reason', the realities are that it will c~ntiuue. to' be used on many
more patients. I believe it is for that reason that the FDA and. others
have to play an active role in education, as. was emphasized at the'
lmarin.~ yeste.rday. S
Mr. &ORDON. Dr. Felig, please proceed.
Dr. Fi~i~. It is thus clear,. I think, from what the other experts
hç~re have said., as well as from what is generally recognized, that
these drugs arei useful in a very limited number of patients with
aduJt~-onset diabetes namei~y, those with symptoms due .~ `an elë-.
vated' blood sugar in whom dietary measures have failed and in
whom insulin is impractical or refused by the patient. While some
PAGENO="0070"
13320 GOMPETITIV~ PROBL~MS ~ `T~E DRUG I~ThU~TRY
experts would inchde patientS ~i~h' an elev~ted blood sugar who a~'e
asymptomatiC, ~therë is tmiversal `agrement that these drugs are
overprescribed in the United States.
All of the abOve w~s iii faCt well recognized before the UGDP
study i~aJ~ ~eported.. The effect `Of the UGDP has been to add evi-
denie of a telationship between oral agents and increased cardio-
vascular niortality. This relationship has been considered conclusive
by some, persuasive by others, and at the least possible by all, in-
eluding the most severe critics of the UGDP. Given the fact that:
One, these agents are drugs of convenience; two, they are overpre~
scribed ; `three, they may' increase cardiovascular mortality; and four,
tha1~ the practice of medicine is usually governed by the axiom
"Primum non nocere"-"above all, do no harm"-one may question
whether the ~flndings of the UGDP study have resulted in a change
in the clinical treatment of diabetes. Unfortunately, the answer is
very definitely no. The most recently available data reveal that the
total prescriptions for oral hypoglycemic agents increased 5.5 per-
cent between 1972 and 1973. This represents a total of over 19 mil-
lion prescriptions costing over $100 million and involving over 11/2
million patients.
Mr. GORDON. Can you explain why the use of these drugs has in-
creased in the face of the known results of recent studies-human
and animal-that show that these drugs are harmful?
Dr. FELIG. I think it is difficult for me to assign a specific factor
or factors. I think that what we are dealing with has been an over-
ridding tendency to use a convenient method which both the physi-
cian and the patient are likely to be more willing to tolerate or to
follow. In addition, we have the influence of a `very vocal group which
has been so severely critical of the UGDP that the effect has been to
totally mute any of their own concerns regarding the overprescription
of these dri~gs. So, I think what we have is the practicing physician
faced with a choice between different methods, one of which is more
convenient than others; a~id, lie is being bombarded with informa-
tion that could be reassuring regarding his convenient method be-
cause the data suggesting that this may be hazardous is constantly
being attacked.
Mr. GORDON. How about advertising?
Dr. FELiG. I think when we talk about the data being attacked, it
becomes difficult to separate the constant criticism of the UGDP by
those who attack it from a seemingly scientific viewpoint and fail
to point out that it is overprescribed'on the one hand, from those who
are actually advertising the drugs. Given the profusion of literature
to which the physician is subjected, much of which is not really scien-
tific but a pseudoscience, one can' appreciate the quandary of the prac-
ticing physician. He may not have `the opportunity or perhaps does
not avail himself of a more dispassionate form of instruction, so as to
make adequate or appropriate decisions.
Since all agree that these agents are overprescribed and, at the
least, possibly toxic, it is apparent that the experts in the field of
diabetes have failed to appropriately influence the clinical manage-
ment of this disorder. To rectify this situation, I would propose the
following:
PAGENO="0071"
COMPETITWE PROBL~M~ IN. TII~ DRUG INDUSTI~Y i3~21
One: Leading proponents as well as critics of the TJGDP study
should meet f~r the purpos~ of issUing a joint statement in wh~ch
the primaoy o~ diet ana the ~bvious need fo~ ~rest~iction in the ~ise
of oral hypoglycemic ag~its is clearly ~pelled out.
Mr. GoRDoN. What kind of a 1oint statement?
To w~hoin woulcLit b~ directed?
Dr. FELIG. I think we have witnessed, now, for th~ last 5 sears,
since, the report of the UGDP study, that physici~hs W~io had b~en
very critical of this have tende~l to band together ahd Deiease st~te-
ments as ~the Comwittee for the Care of the Diabetic Patie~t,
et cetera. ~This gr9up, I tith~k, would be one wJ~ieh is so clearly iden~
tified as pritical of the TJGDP, that I would hp~è they would be~p~rt
of a joint statement together with other individuals who have been
proponents of the TJGDP, or who have not attacked it, so as to
come to some joint statement regarding the overall situations in
which these agents should be prescribed.
It is interesting that in the criticism of the IJGDP, the severe
critics do not generally raise an argument as to the situations in
which the drug is indicated, but restrict their argument to the ques~
tion of whether there is absolutely incontrovertible data that ,th~se
agents will be harmful. They should in fact be addressing themselves
to the facts before us; namely, that we have a situation in this coun1~ry
in which a potentially toxic drug is being widely overprescribed. If
one assumes to be, or in any way is willing to be called an expert in the
field, he has a responsibility which goes with that designation; namely,
to influence the prescribing habits and over-all practice of medicfrie
by his colleagues. I think this is where the field of diabetes has been
remiss, and in particular those who have been critical of the UGDP.
They have failed, as I think all medicine has failed, to rectify a sit~a
tion which all agree is not optimal from the standpoint of the patient.
Mr. GORDON. When Dr. Bradley testified here, he acknowledged that
these drugs are vastly overused. I do not know whether he used the
word vastly, but I am putting that in.
Nevertheless, it appears that in his attacks on the UGDP study, he is
essentially promoting the use of these drugs.
Is that a correct conclusion from what you have stated?
Dr. FELIG. I would think that any group or statement that tends
to accentuate the criticism of the TJGDP and is not accompanied at the
same time at least by an equally forcible statement indicating that
these drugs are overprescribed, will have the effect of perpetuating
the use of the agents; or, probably more likely, they would promote
their utilization.
So, it becomes very difficult to divorce comments from such critibs
of the TJGDP from an effect which is very similar to that which
would occur with a drug promotional type of statement.
Mr. GORDON. I conclude from what you state-and I ask you ~f
this is. a valid conclusion-that it is really the responsibility of the
critics of the UGDP to insure that there is some rectification in
prescribing habits of physicians today.
Dr. FELIG. I think it is the responsibility of all experts in the field.
That responsibility becomes that much more manifest and incumben~t
PAGENO="0072"
1 çQ1~I~2V~ L~ p~
ind~i~t~ th~t ~thb ~qi1~t4on'.
bbfié. If is a1tha~g~iu&
ha~ b~en ~aM ~
~ ~a~ièht to dO ah~ythin~ I~ th~ skin~
~t ev~F~y ~p~iertt 1 s~&~ii1d~ r~e'~to tak~
ii th~ w~ in ~c~ich I pi~e~h~ th~ a~tOt~ii~tlv~s..
If I s~ ~6u1~I~yóii 1i1k~ th~~ n~è~ thbi~th o~t this n~d~iáin& th~t
has to he rnjected with a needle I think we ea~t all p~edi~t what
pro ort On ~oti~lli r~fh~é a pathxftil iiij~cti~ e$ery
day. So, I L t1~i~thé~ c1 i~ptibm.
of .1 t1~~ p~±thft~iht~ i~ext
of th~ dhi~ Of! th~ T~
st ~ieftt r~i~1~th~at r~i~ents'~
a th~è ~tti~ftt~p'art w~hkh is
~A~ffi~+ f6i~
~ th&ad-~
~tc~iis~ tI~L~
e~ti~l~ frbtWYP~StiI1d!Y.~
~ e~rit ~oMh~) U~PP~t1id~3r~
an~d ratity ~
)Hb*é~er~sd 1ot4~ `~iich'
a ~t~thffiM~ attadki~1~ ~`or
~ñ~I resWt~s~a ~
of tlies~ agthits whi~h chataetetizes `curreilt ~medi-
shOu1d~ be p1ac~e& on dietary marr~gbn1~nt ~rather
~ tl~& ii~t~r d~ioti `~of n~d~a1~ ~t~idM~~ ai~!d~1n ~poSt~
course~ o~i th~ti ~thi~fit"o~ di~thMt~
shO1i1~tl `be th~d~i~k~ on' de~eiOØ~g~ in~!~ro'~ed~
iti~ènt cdti~liaii{èé anç1'~succes~' in' ad~eitht~ `to~
t1~ie' la~bèlin'g of oraF ~
a' w~rning tha~t é~id~i~ has beei~
catdib~va~iil'~t mort~Iity;
restrictM ~tÔ ad~iltLo1~set
hate fa~le~d ari&"ins1ili~ is re~
the propDsed 1ab~ll~i~' t~ti~ the
~honid ii~f~r~i' the
pat~it shotdd p~ir-.
vai~t~ges, if any, that he is a~
dr~
W1~ d~y6i~ ~ ~
PAGENO="0073"
w r~ ~ ~
bo~h ~°r the p~tie~t
si~ st~ement f~Ea~ he
~othe~
~t~Q ~%1~O ~i~itt~ IQte
~ 4Ljc1~4 ~s~e
~
Mr. GoitroN. Wotdd not this also be another way of ~u~tt~pg ~wn
on the use of these drugs ~
Dr. FELT ~ ~ ~ J~ ,ti~i~ that one ~~ou1d run the ri~sk-
if we were to have such a situation of written informed consent
~ts a roq~i~ei~nt ~or ~ p~r~t~ct4~r ~ ~[ ~t~k we 14 i~i a
~ev~v~ ~is1~ ~ i~ ~t~ii~ ~yj~~h tre~t~nt ~i other cIrcumstajices
`tQt~ily ~l~e4 to the orai agents; where, ~or exarnp~Ie a potenM~lly
toxi& antibiotic could ~be a4~inièfered or is consid~rèd ~pp~priate
treatment for a particula~r i~J~o~i. ~ i~g~t ~I~çl ~ ~iti~tion
wjie~e treat~er~t j~ interferred with because of the need to obtain
~formed consent.
~[r. ~O~1~DON. Any other comments?
Dr. L~RNER. I feel the sa~the way. It ~think iih~t it wonid ~be ~e~y
good from ~he point of view of mini~ni!zing ~the use of the drug, `hut
it ~pu'ld be ye~y ~1i~fficii~t from `the point o~ view of the ~ne~à1ity
o1~ ~he~ituati~on~if it were applied a~ro~s the board. how would you
de~j~ which di~gs to app~r this to, an~t whidh ones not?
Mr. GOIWON. *Tell, it~ ëoiild be used in ~ertain drsigs ~which ~re
known to be to~i~. ~Fo~ ~a~ple, ~cb d~p~gs ~ ~ç~ip~pl~ico1,
wh~i~b ~is alsç .ya~tly o~vEsmsed? oç clindi~myc~in, or ~l~corn~zeip, ~w1Mch
~ic y~tly pycli~c~1 Pl~~e f~t th~t ~oi~ a~pply theip to a ~éw drug~or
~ ç1r~g~ 4q~s p~t ~ ~i~e~~iy ~ ~s ~p
to every drug.
~r. ~i'~ç~G. `~;
situatiau
o~pin~ ~9, a mec4nis~ip of
tq ~ ~ ~tr~g, 1Ô~t ~ ~1i~AT~ a
~et 1~fprn~ed c~~se~at
IOVi~e
PAGENO="0074"
13324 cOMPi~nTIvE `PROBI4EMS IN THE DRUG INDTJSThY
Dr. LARNER. This would, I think, present a difficult choice situa-
tion. I mean, what would you do with-all dtugs are potentially
toxic, and what would you do in the case of digitalis where-
Mr. GoRDoN. But the benef~ts may outwe~h the risks. It may be
toxic, but the numerator representing the benefits is such that for the
purposes claimed, digitalis ha~a favorable r~ti'd. Now; for the~ drugs,
I do not know. Clindamycin, as you know, ison the line. These drugs
are vastly overused. Some drugs may not be vastly overused. I do not
know.
Dr. Sims?
Dr. SIMs. `I think it would be more consist~nt with the whole idea
of peer review, which is prominent today, to have a physician simply
justify in the record use of the particular agent under the circuim
stances. I am reminded of an informed consent form that appeared
in Science, years ago, by Greenberg, I think it, was, for a hernia op-
eration. He listed all of the possible, hOrrible things that could
happen, and indicated it would have to be sigued by the patient's
lawyer and another-in-law as well. I believe that if informed con-
sent was required for everything, we would end up in a difficult
situation.
Mr. GoRDoN. Dr. Felig, would you proceed?
Dr. FELIC. There has been much discussion in the: lay press and
medical journals of the need to maintain the physician's freedom of
choice in the treatment of his or her, patients. I believe that our
overriding concern as physicians is to do no harm. As experts in
the field of diabetes, our primary obligation should be to improve
the lot of our patients by influencing current treatment practices
rather than perpetuating a situation which is at the least wasteful
and at worst causing an unnecessary shortening of lifespan in adult-
onset diabetics.
Mr. GORDON. Thank you very much.
Dr. Lamer, would you proceed with your statement?
STATEMENT OP JOSEPH LARNER, M.D., PH. D,, PROFESSOR AND
CHAIRMAN, DEPARTMENT OP PHARMACOLOGY, AND DIRECTOR
OF THE DIABETES AND ENDOCRINOLOGY CENTER, UNIVERSITY
OP VIRGINIA SCHOOL OF MEDICINE
Dr. LARNER. I am responding to five points which Senator Nelson
wrote in his letter of June 19, as follows: Point number one, the
proper labeling of the oral hypogTycemic drugs in the light of the
studies recently conducted with these drugs.
Having reviewed the literature, I have come to the following
conclusion which is quoted from chapter 71, written by myself and
R. C. Haynes, Jr., of a standard textbook in pharmacology, Goodman
and Gilman's textbook, fifth edition, to appear in September 1975.
The sulfonylureas should be used only in subjects with diabetes of the
mattirity-onset type' who cannot be treated with diet alone or who are un-
willing or unable to take insulin if weight reduction and dietary control fail.
The physician must realize that he is using these agents only to control
symptoms associated with hyperglycemia and that dietary control with or with-
out insulin is more effective for this purpose.
PAGENO="0075"
COMPE~ITIV~ PEOBL~MS IN `~IIE DRUG INDUSTR1~ 13325
The major complications and life-threatening 4isord~rs~ associ~ted
with diabetes are heart disease, kidney disease, blindness, and limb
gangrene, There is no evidence that sulfonylureas ameliorate or pre~
vent these d1~orders. While in marry instances in medicine the physi-
cian must prescribe for ivert symptoms, we all prefer to correct the
~und~rlying problem if possible. Unfortunately, this is not presently
possible with diabetes without additional basic and clinical investilga~
tion. There is no evidence that sulfonylureas will assist in the un4er-'
lying problem. Since these agents would appear to relieve primarily
the symptoms of hypoglycemia, one should restrict their use until
~e~s costly and perhaps safer measures have been used-diet with or
without insulin.
For this reason, I feel that there should be stronger labeling of
the oral hypoglycemic drugs in the package insert. With regard to
the nature of the labeling, I feel that the stronger 1972 FDA draft
is preferable to the weaker 1974 draft for the reasons just 4is-
cussed.
Mr. GORDON. Dr. Lamer, this is a question actually for the pa~iel
rather than for you alone, but I would like you to take the lead in
this.
Yesterday the Commissioner of the Food and Drug Administra-
tion acknowledged that phenformin has even a more unfavorable
benefit to risk ratio than the other oral hypoglycemics. In an article
that appeared in Controversy in Internal Medicine Dr. Albert
Winegrad and two others wrote that the biguanides have no role in
the treatment of diabetes mellitus. In addition, the Director of the
Bureau of Drugs, Dr. Crout, yesterday could see no justification ~or
this drug to be on the market. That's phenformin.
How do you feel about this?
Do you find any medical justification for that particular drug to
be on the market?
Dr. LARNER. Well, I would generally agree that probably there is
no-at the present time-medical indication for phenformin that I c~n
think of. That would be my feeling.
I do not see any justification for phenformin being prescribed
at all.
Mr. GORDON. Dr. Chester?
Dr. CHESTER. I would agree. It is not only hazardous; but it is al-
most totally ineffective.
Mr. GORDON. Dr. Sims?
Dr. SIMs. I thinl~ that first of all we ought to speak specifically
about phenformin and not the biguanides as a group.
Mr. GORDON. I am talking about phenformin, which is the only
biguanide on the market now.
Dr. Sm~s. There is a tendency~ to condemn the whole group. Phei~-.
formin does some very interesting things and some of them resemble
the effects of exercise lowering insulin, and whatnot. Further re-
search may develop new drugs of this class which will do what or~e
wants without the side reactions. So, I would not say that it should
be a blanket condemnation of. all of that type of drug. But, on tl~e
other hand, my own feeling about phenformin is that the labeling
should be shortened to four words: Not for internal use.
PAGENO="0076"
13326 ~or1~rIv~ r4~S
Mr. (~rnoN. In other words, take it off the a~aarket.
Dr. Smrs. Big~ht.
Mr. OQED0N. Br. Felig~?
Dr. Fc~. I agree with the other speakers. Th~ usef~iness o~
phenformhi is extremely limit~ed~ if it exists at all. I ticnk th~re ~s
no question t1i~t the risks e~oeed those that we ha~e with any ~other
i~orm oi~ ~ntM~abetic treatment. I do not think it w~d be any loss
~f this were remoi~ed frmn the r1e(.
Mr. ~ Ana yon ~gree with Dr. Grout that the drug ~shouliI
be rexncr~red from the market-~-is that correct~
Dr. ~ I agree with its extremely limited usefulness.
Mr. GORDON. Would the record show that four witnesses agree with
Dr. Grout.
Dr. iiarner~ pien~se proceed with your statement.
Dr. LARNER. Two, the effect of these studies on medical practice.
To my knowiedge these sthdies have had a variety of effeet~s on
medical practice. The total utilization of this group of agents, how~
ever, has not seemed to durnge mi~veh. For example, when the results
~of the atudies were initially unnoanced, some ~physicians ?changed
their patients to other sulfonylurea analogs not realizing that the
fundamental pharmacology should be quite similar to the drugs
~studied. This obviously demonstrates the need for additional post-
graduate tr~ining and education of some of the medical community.
some physicians a'ocepted the results of the study and some ques-
~tioned the design and control nature of the experiment. This con-
troversy has undoubtedly been apparent to this committee. On the
whole, these studies indica* that the use of oral hypoglycemic
agents should be limited to the small percentage of patients with
diabetes for whom other therapies have pr~wen impossible to carry
out.
Three, the availab~iity of scientific evidence, if any, which demon-
~trates the benefits of oral hypoglycemics.
I know of no evidence that directly demonstrates that the oral
hypoglycemics are life-saving or life.proionging: in the therapy of
diabetic patients.
The major therapeutic problem in diabetes is no longer the acute
ketoacidosis which used to be the cause of death before the intro-
duction of insulin. Rather, it is the long term or chronic vascular
complications of the disease. In other words, the major problem
now is the well recognized thickening and other damage to the
blood vessels throughout the body leading to kidney disease, heart
disease, blindness, and gangrene in the limbs. We still do not
know the answer to the following fundamental question, "If the
blood glucose level in the diabetic patient o~nld be controlled
as precisely as that of a nondiabetic through the use of an insulin
delivery system yet to be developed, would there still be vascular
complications ~" In other words, we are dealing with a situation
here, where there is a fluctuation as a result of three meals per day
of the amount of insulin delivered from-in a very regulated man-
tier, and, to date, we have not been able to duplicate this situation
in the diabetic patient, such as it exists in the normal. And the
PAGENO="0077"
C~1~IVE ~ IN THE DRUG INDVSTR)~ 13Z27
question is if ~e ~xuJd duplicate it theoretically, could we prevent
the vascular complications.
rri~ second part to this is; alternatively, is there some factor or
factors involved oth~r than proper insulin delivery which leads to
these harmful ~ffects on the blood ysssels? Basic and clinical invcsti~
g~tors are working on this question with respect to msulin at prescnt.
Mr. Goiwo~ Can you give us some more information about the
kinds of research being conducted in this area
Dr. L~nN~i~ Yea, very simply, ~or example, the question that is
n~w being investigated in diabetic humans is whether to return to
the early methods of therapy, using several injections of rapidiy
acting insulin, for example three injections of the rapidly acting
insulin, coordinated With meals, leads to a better situation with
regard to the prevention of thickening of the blood vessels and the
vascular complications.
In other Words, biopsy studies are being done to investigate
whether an insulin delivery system of three injections or multiple
injections of rapidly acting insulin, rather than a single injection
or several injtectlona, or slower acting insulin are more effective im
prertnting the vascular complications than the therapy with long-
acting insulins.
Now~ this type of questioning is being done in humans and
analogous and even more sophisticated experiments are being done
in animal systems. So, my point is, we don't yet have the answer
to that question with insulin. And, insulin is itself a direct hormone
replacement therapy. And for this reason, if we don't have the
answer with insulin yet, we certainly don't have the answer with
the sulfonylurea drugs.
What we need is accelerated research in this area to answer this
question. Fortunately, we have enough information now to be able
to phrase the question in a sound way as an either/or type of ques-
tion. Either it is the insulin delivery system, or It is not. And we
should be able to get a yes/no answer on this situation. Until we
do, we can't go forth, in terms of other applications, until we under-
stand the theGry.
For this reason, I feel there is no direct evidence that these or~al
a~gcnts tt~re beneficial, that is, in the sense of life saving or life
prescr~v~ng.
Mr. GoRDoN. You seem to emphasize "directly." Is there any in4i-
rect evidence, whate~ver that means
Dr. LARN~R. Neither direct or mdirect. I didn't mean to distinguish
between them.
Numbei~ four, the problems of tran~lating the results of basic re.!
search deve~o~ed b~ medical scientists to the practice of medicine.
This is a very broad question, and we could spend a great deal of
time discussing it~. Briefly, I ana of the opinion that scientists today
a~re more bware than ever before of the importance of a~piying
their fundamental stut~ies to the, practice of medicine.
For `~xan~ple, in my own field, pharmacology, there has been a
st~ong dev~elepment in the area of c~nical .pharmacolqgy which a~k-
dress~s itself t~ this problem: gamely, the application of iundamental
PAGENO="0078"
13328 COMPETITIVE PROBLEMS I~ TEE DRuG INDUSTRY
laboratory findings to the patient in order to understand and treat
the disease process.
For example, the sulfonylureas have been used clinically for about
20 years, yet a great deal of information regarding these compounds
is still lacking. The metabolism of these compounds in patients and
their precise mechanism of action are still very poorly known. These
have been complicated problems and require additional studies in
both animal systems as well as patients.
Scientists are very interested in coordinating such diverse efforts
and studies. I feel that clinical research work in this area should
be further nurtured, but that it must be balanced by a broad base
in fundamental animal research as well.
And other aspects of the subject which you think might be helpful
to the subcommittee.
I feel strongly that the time has come in terms of the oral hypo.
glycemic agents to restudy their efforts in animals and patients. It
is my feeling that since recent animal studies are proving of con-
siderable interest in terms of the actions of these drugs on organs
such as the heart, adrenal glands, and liver, it would be wise to re-
study these compounds in animal systems during the time their
clinical use is reevaluated in order to see whether we can gain an
understanding of the mechanism of the c~rdiovas~ular deaths or
even reproduce them in animals.
Here I note with particular interest two recent pieces of data in
animals: One, the summary statement of the work of Wissler et al.
which states that in rhesus monkey fed an average American diet for
74 weeks containing 20 milligrams per kilogram tolbutamide, there
were present in the coronary arteries two times more frequent and
three times more severe atheromatous changes than in the coronary
arteries of control monkeys; two, the work of Hsu et al. from our
Department of Pharmacology at Virginia which demonstrates that in
heart, adrenal medulla, and other organs, sulfonylureas inhibited cate-
cholamine release from the nerve endings of the antonomic nerves.
Thus the function of the autonomic nervous system, which provides
the involuntary control for many of the organs of the body, is sig-
niflc~ntly influenced by these drugs.
Therefore, I feel that it is time to caution physicians about the
use of these drugs, and to restudy them in the clinical and basic
laboratory much more extensively.
Mr. GORDON. Dr. Lamer, thank you very much.
With respect to the Wissler study in rhesus monkeys, what con-
clusions can be drawn from this for humans?
Dr. LARNER. Well, I think, the obvious warning can be put on
that these may be potentially harmful drugs., that they may affect
selectively, the coronary arteries, that these changes in the artery,
may lead to malfunction and difficulty in the heart.
I think the warning is obvious. I think that more studies need to
be done, both of anatomical nature, and of a functi6nal nature.
These studies reported here were of an anatomical nature, in ~which
the structural changes were pointed out. And I think they must also
be accompanied by studies in which the function of the heart i~ also
studied, so that we will have some more information.
PAGENO="0079"
COMPETITIVE PROBLEMS IN TI~E DRUG INDUSTRY 13329.
But I definitely think that, since these were done in primate~,, a
species which is closer to human species than the rodents and so
forth, they definitely should be taken seriously and consid~red
seriously.
Mr. GORDON. Thank you very much.
Dr. Sims, would you proceed with your statement? 1
STATEMENT OP ETHAN A. H. SIMS, l~D., PROFESSOR OP MEDICINE,
COLLEGE OF MEDICINE, UNIVERSITY OP VERMONT, BURLING-
* TON, VT.2
Dr. SIMs. Mr. Gordon and members of the committee a ritual
of hornblowing seems to be in order at the beginning of these state-
inents, so I will mention that I have had experience with a number
of diabetic patients over a consi4erable period at Yale New HaVen
Hospital and in Vermont, though not as many as has Dr. Chester.
I am a member of the workshop on obesity of the National Dia-
betes Commission and of the advisory and editorial group for the
Fogarty International Center conferences on obesity. The backgroi~nd
of a lot that I have to say is contained in the volume from the cen-
ters based on the last conference, which is to be released this sumther
by the Government Printing Office. I do not claim to be an exp~rt
in anything except in our research work persuading volunteers to
gain weight.
I would like to acknowledge a major contribution to my written
statement of my wife Dorothea, who is a Fellow in Health Care
of the Radcliffe Institute, and is working on diabetes education, and
also of my son Nat, who has been writing a history of the UG~~
as his undergraduate thesis at Harvard. They have both been ddIng
their best to educate me.
I have included a brief summary at the beginning of my *rit~i
statement, but instead of that I will read a restatement of SQR~e of
the points which I believe should be emphasized. To my ki~ów1e~dge
they have not been emphasized at these hearings before.
I would like just to list the main points, which I want to be' sure
to get over. To my knowledge, they have not been emphasized in
these hearings previously.
One: Obesity is now recognized as a factor predisposing to n~n-
insulin dependent diabetes in those who are genetically susceptible.
Untreated obesity represents a long-term risk in relation to cardio-
vascular and also other diseases.
Two: Insulin, in addition to its well-known action of lowerbig
blood sugar, is a hormone which promotes the deposition of fat.
Three: The intense preoccupation with one aspect of the UGDP,
the cardiovascular mortality, and the accompanying sometimes ac~i-
monious debate has blurred Our perc~ption of the fact that at least
50 percent of the maturity onset diabetes in the study were over~
weight and underexerciseci. and that both th~ sulfOnylureas and insu-
Tin work to make them fatter. This is a threat to their well-beir~g,
`See prepared. statensint, page 13676
2 Dr. Sims on sabbatical leave at the EndocrIne Division Tufts-N. B. Medical Center
Boston, Mass. * * *..
PAGENO="0080"
c6i~i~i~v~ ~ ~
an,~cT ft ireA~~ ~ ~v~d1 t~bii~1ié~d ~k f~t~r ~1~àk~ ~a~d ~r~td~tr th~d
otj~ diseases.
Mr. GORDON. t~r. ~iiuis~ &th `S,TthI e~f5l~1ii hd~v iIÔ~th iii~li~ àkid th~
oral hypoglycemics promote obesity? You have just said th~t they
do, is that right?
Dr. SIMS. Eveh in thn~lI ~IñOUiIt~ i~~thih tu~ ~ the i~M~se ~f
f~tty acids from ~the fat deposits in both .experime~tal and spon-
taneous obesity ~y p1 ovidnig ex~~ 1h~lrn, ~ith~ b~ ~n~rtlM or
by ~i~rbig ~u1~ `~Ms ~ii?~'h~ ~ in~t444~ ~sè~ ~w~i~t gain
is enhanced. I will go into this in a few minutes in a bit frio~Ye ~
With alJ due, rpspect to Dr. Max Miller, who is here today in the
ai~c~~ I *~L1~t like to sa~ thkt the 1*e~ttheftt oft~ons s~lectèd for the
Ut~t)'i~, ~*hi~ re~e~entèd the pr~tiling o~5tMxis o~ i~&)~ ~re öift o~f
da~tè iió~v. Th~ d6 i~dt i~ciiide ekcrcisc or ii~teñ~it7e e~lu~c~ttioi~, or
s~v~f~i~ óther iiê~ver in~Tho~ds O~ hth~ethe~tt, s~th~ of *hich can
1~lly ié~r~é 1~lie övèft diabetic ~1tato.
W~ rAu~t cO~c1t1de th~tt neith~r ~l~on~yl ~, si~id~ thé~r iñci1~i~e
t1~ê ~ ~et~i~th!i of thkhlitt,, hth~ in~ii1h~ it~èlf ar~ ii i~at~d f~r the ~ti~e~t-
of the o~rci~~eigfrt ~h~d T emphà~si~e o$~&~~Wcight, th~tiittty-
oi~ét di~~tic.
f~ht: Muèh of the db~Iei~i of ~tathtit~r-ôh~t di iê~be~ Is ~
~iice dt oi~tr A iè~rld~n affli~eht po~t-WOi~d Wa~ ~[i iIfêst~I~~ A1tef~-
ir~g thi~ i±th~r ~r~4~fitII3t revt~rs~ the di~b~ti~ ~tat~, ~ p~'oIohged
use ~f th~ sulfonylureas will not.
It~'ik~: 1~±crcis~ or ifiet~ed liwei ~of physical acti~it~ is ~ ~ans
of eV~iitiOñ ~thd t~e~t~nè1~t which has bé~n s~td1~ ighor~d at these
thã~ ~hei~evth~ diet is m~enMan~d, as i~ the pro-
for thC dr~l a~ciit~, It ~hott1d b~ as diet ~nd
~sed i~Mi~i tMt~. Tt~i~ is bee~i~st~ Offi~ is
with both the e~e~gy out~ht
~hefi ~tith!it.
bfl~f c tO eThIior~te ~ bit.
hi ~ i~ thO ifi~uiii~i
the ~ècOnd
of the 1'/2
usii~ll~ o~tet-
of those in the TJGD~P ~*&è O~rer 2~ pe~cei1t
ii *Cight. Arid ~ think, Dr. Chester, did ~ou
of ~O ~efceiit o~te~rweight in ~tOth d1ei~e1kn4
ther~ i~ ~t i~tiiCë to ~ihO a~Wlon
itt~t1 thO fà~L The th~u1in In. the blOod is
it is ifi1h4w~te hi the f~tb~ Of the rCsisttthce
df fOod. This bdiI h stress oh the
ul~lth~t~l~7 bC fOllo*Od b~ f~i1ti~e ~thd
Eleiit dWbétes.
~ 1ra~rO tO khb~ wheFe h~ is *ith r~pect
the disCFd~r. AM ~e ~l~o hhve to k~dw
~ h~e i~ ~i Wci'glit. WO i~i1Ow h OitF *~Fi~ iii Verththit
with normal volunteers, who have no fámi~Iy history p~ di~b~ Oi~
obe~ity~ ttntl whti h~ve atgtreed to d~hb4~ã~tOI~ ~a~hI ~th~ht, that this
insulin resistance may develop secondarily to the obesity ~td d~t~F-
eating. It is also completely reversible.
PAGENO="0081"
thM~t%~v~ ~ i~ ~ ~ 1~L
~ We â~~i it~t~ôi~ f~ô~i th~ ~ ~ i~w~igM p~1~ts~
ai~ ~dWè~t1, th~ii~ h~s~thti ~k~i~~fiM ~tThtiis ~ Th ~ thiI~ ¼~h~1
to ~ ~ ih~tek~ th~ bbé,~4%~7 ~i~i&~ ~tir~ t~ a14 ~
the insulin resistance of our patient.
The T1(Th~ ~Wd~e~ ~iMv~çI ~eati~ th&b~h~t ~tct~iiy4~ Wh&b
ifi the grtht~ gi~reh thlbut~thiM~è ~idt hi the tW~ gi~tei~ his~t1i~
tLIt ~f the fii~e ~i~Ot~~s th~rC W~ ~ 1o~ &f ~bd~ñ 8 ~ of 1~
1YOd~t ~i~ht Mitiafly, ~ ~ F~suit ~f the ~oth~rat~ dMtM~r r~stieth~Ei
*hi~h they iv~ ~fl gIw~i. The ~i&~bo gthi~p M~d t4~ ~h~bt~1$Yi
group, maintained this loss throughout thiS study, ~ in th1~ th~
j~hMifot~ñih gi~oiip ~t~s ae~afly ~e betthr thah the ~ee~rô~ ~
other h~h~1, ~tients in ~Oth the tblbutkthikI~ ~iid the tM lh~411$n
g~b~nj~, n&~t ô~i1y i~egaihE~4 the *eigM iMitially iost~ but ~IhM i*~iI
a~hó\re their ~*igiiiaI Mseline. So, the~e i~ a ~OnMd ~Me tfiff~enee in
the ~ttdy betwe~ñ the Wei~ht o~ tlthse recei~tiñg hi lih o!i~ ~u~fbrI3~1-
i1~S ~s O~DdSE~d to the piaeebO g~up.
What *te the other opti~ons ávsii~b1e to tis iti tcY7& Ati~I I s~e the~
a~ ~fo11ôWs:
hi the S tCitih~r 1~ he~iMg beTh~'~ thi~ eo~nrti44itee 1a~t f aft,
Thy. J~oMi I)a~ik1eb~i g~akte his ek~e~iei~~Ce ~n thM*iM~ o~l ~ge~th~
f~tô~i ~Oap~ttietits ~t the G~a&y s~pit~d h~ Athtutk. As y~u
~ he ha~ 6rt~d ~rthe~ on hiS expe14~Ce in The M~y ~
issue of the Journal of American MediCSi ASs~~~hiti~ii, ~d I eug~e~t
thta't this ~rtic~~ be part of the record ~f thie heaI4iig. ~y a eom~re-
hOnslve and ri~'Oroi~ks FegiMCi~ Which ifréhided ~5 hourS `of edu ion
~ ~atiefft, he ~ beeti able to achieve substantta~ *eight redttcti~n
hi ?ti~ to ~9O percent oT the ~tiêhts and has eSsefitiaIi~r 4~é~Pe~5d their
OVe~t d~ab~tic sthte. This i~ ~ôMCChifrg q~tè ~h~ei~t f~Oin ~the t6k~
p~c~ibthth Of ~ tiiabetic diet of Which nio&t of tie l~ 4~een ~
He inà1nt~4W~ that ~iM d~abetié patients Who ~rC ~5~~ht ~*~h
t~hè~ ~esent thethS~i~r~5 ean be éôfrttO'l~et1 *ith~ift tt~ Of his~uihi, ~f
they a~f~e ~vefi SuCh ~ regimen, ai~d, renieniber %h~tt thifl `cdtiti4t~s
peFcent of the g!t~oiI~ bi niattit~it~-onset di~e~ee that ~
tk1kiii~ nbok~t.
Yes~ei~d~y sOffithod~ `asked ~S, `well, if inSui~hi ~ Sb cbtit~ahid~
cated in this group, of patients, shouldn't we have a package `wai~h1n~
~oi' iM~sulin as weil, alk~st it'S tiSe hi the e~~t diabetic ~
yoti th~nk %b~tit ~it~ th~±e ~aily `ShoWd be SuCh a *aFfthi~, So, ~t.
~*heI1 Co~miS'SionSr Set~M~4dt ~e%S dO*e `with the b~al ~geht~,
he can rewrite the pk'cka,ge 1i&bei for ltisttli~.
Se'~hd, there k~e MO~ë rlW$±~ue friè~nS of ~Chiès~in~ WM~ht hs~.
Dr~ ~a~1~Soft M~d `orher~ have ~O~tiM~eS hi~i~tMted tfferap~ of
seriously obese d~~b~Vi~ ~vit1!i ,brWf ~1!aSth~ig. T1ker~ ti~e !i~oW iie*
t~i~iqueS ~ ifro'd~f~e4, so-c~l1~ed protein sparhig i~vatio~tt tha't can
eiMfip~lieh *thg'ht iose ~4~tthont a da~nagii~ l'OsS Of h6~y prdt~it~.
riltiese regimens so~netimes very dramatically reverse the WCrt di~-
)~e't1e stkt'~ a'S ~wCi~l. thOy ht~ve been prb~en in e~f~iy pi'ot work to be
a fttl adjnt~et ~kr iMithil w~ight loss. I woiu!ld i~a~i~e that
be ~t~nO frtid~r s14 ~risiOn ~atid We hive ffitI~1I ffi6~e to
a~Milit theM.
th~e i~ ti ~$vhele ~ie* 1~eld Of ha~1~1r~l ~
tt~ kp~lied i~O b~OTh thhig' ~nd ~i~sic~i ~ `~h~h ~a~t bob
p~t4~etit ~it~od~ry b~S~c lif~et~de. r On~h~i~l~ 1~l~e bhtth~ ~
PAGENO="0082"
13332 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
many of the problems ought not to go to the physician working with
the limited resources he may have available or to the pharmaceutical
house. ,A. lot of our current problem is a reflection simply of our
American lifestyle.
Mt. G oic. May I interrupt for just a second? Do I understand
what you are saying is that an additional danger of these oral
hypoglycemic'agents-I mean their very presence has been a danger-~-
because it has taken the attention of the doctor and the patient
away from the essentials of diet and exercise to a much less rigorous~
regimen of just taking pilis?
Dr. SIMS. Thank you for stating it so well, That is precisely what
I mean. The presence of this option over the 20 years that it has
been a~railable h~s been undesirable just for that reason. In the
preinsulin days, when Dr. Allan did not have that particular op-
tion, he did very well with diet in this type of patient. I am pleased
also to see that you mentioned exerciSe. That is about the fourth
time' the word has been mentioned in any of these hearings.
Mr. GORDON. And one other point: Am I also correct in that you
are also saying-I am trying to summarize this in my own words-
that insulin and the oral hypoglycemic agents are really treating
or at least being used to treat symptoms and not the basic problem
which would require a change in lifestyle, which would include
diet and considerable exercise?
Dr. SIMS. Precisely. Consider the problem, say, of a relatively
young housewife who has had a couple of babies and gained a lot
of weight. Unfortunately she has selected the wrong parents, who
are both diabetic, and her grandmother was obese. If she develops
giucosuria, the odds are that the average dietary effort in the busy
physician's office will not correct it. She is already running an ele-
vated blood insulin and has an increased insulin response. If then
we give her a shot of insulin every day, we are instituting a regimen
that will just progressively make her gain more and more. And
ultimately the increased weight is going to interfere with her well-
being and probably will have a greater negative impact on her sur-
vival than might the toxicity of the oral agents itself had she been
given them.
Now, the fourth option is exercise. I emphasize it as a potent
means of treating a patient, although I am well aware that the pa-
tient applying to a large hospital' clinic, elderly or far advanced
in his disease, is not going to join the squash team.
Mr. GORDON. Dr. Sims, I might point out to you that there are
certain hazards in exercise, too. One being the broken' bone that `I
have in my foot. That is the result of playing tennis.
Dr. SIMS. Perhaps, Mr. Gordon, if you should have been exercis-
ing more, maybe your metatarsal bone would have stood up under
the strain.
`To resume, support for the use of exercise is given by some work
by ~a Dr. Bjorntorp in Sweden, who measured the insulin response
in obese, middle-aged men before and after a course of physical
training, even though he urged them not to lose weight. The insulin
response to glucose was markedly reduced. In' other words, exercise
alone did much to decrease the insulin resistance which is a major
problem, in~ the maturity onset diabetic. The r effect, of exercise~ is
PAGENO="0083"
COM~TITIVE PRQBLE~S I~ TEE' DRUG INDTJSTR~ 13~33
something that every insulin-depend~nt diabetic know~ well. A lo1~ of
formerly overweight patients have learned to rely on physical ac~
tivity to maintain their weight loss.
I believe that these options should be emphasized in `the package
labeling for the sulfonylureas. And `on page 9 of my Aill statement
I have written out a suggestion for altered labeling.
Mr. GORDON. That will be included in the record, All your state~
ments will be put into the record.
Dr. SIMS. I would like,, finally, to say that what we have b~en
talking about is changing people's lifestyle, which is a very d~ffi-
cult thing. Some may say it simply cannot be accomplishe& Btit I
suggest that our lifestyles have been reversed once, largely as a
suit of advertising, and they might be reversed again, by education.
And I was very pleased yesterday to see the degree to which ]~r.
Schmidt and the staff of the FDA. are concerned ~vith this aspect~ of
their responsibility.
Thank you very much for the opportunity to emphasize these
points.
Mr. GORDON. Thank you very much, Dr. Sims.
I must say that exercise was one aspect that was really not empl~a~.
sized in our previous hearings. I brought up the subject of exerc~se
when Dr. Schmidt was testifying some time in September, but I do
not think that we spent very much time talking about it.
Dr. SIMS. I remember that you mentioned Dr. Jesse Roth's silg-
gestion that exercise actually did have some long-term effects, a~id
that Dr. Schmidt demolished the idea.
I think that one of the `problems with evaluating exercise as a
modality of treatment, is that it is hard to measure, and also it can~
not just be prescribed like a dose of an oral agent. But I think t1'~at
we ha~re the techniques now to run a prospective study, perhaps an-
other UGDP study, which will include the variables of exercise arid
of vigorous weight reduction like that produced in `the Grady H~s-
pital program. I think that we would see results which would make
the meager benefits achieved for the p~tients in the various groups
of the `[JGDP seem insignificant.
Mr. GORDON. Is this being emphasized at the University of Ver-
mont?'
Dr. SIMS. We say and do a lot about it, yes.
Mr. GORDON. I have just one more question. This is' a question
about the labeling. It is addressed to the four of you. And that' is:
Have you read the proposed labeling and, what are your comments
on'it?
We can start with Dr.' Felig and go from left to right.
Dr. FELIG. I am pleased to see, in terms of labeling, that the FDA.
is apparently making a `stand to change the labeling, and I ~o
believe there has been some change, as regards to, previOus ~ugges-
tions with regard `to the iabe]ing; namely, that this would' apply to
the entire group Of oral `hypoglycemic ag~nts rather than "be re-
stricted to phenformin and tolbutamide, but also the ot1~ier sulfonyl-
urèaic agents.
I am concerned about the question of the `indication in cl~abet~c
`patients without qualifying the fact that it should be re~t~ioted,
mainly of symptomatic diabetic patients. I recognize that there are
PAGENO="0084"
1~4 ~ ?~&~ i~
tho~ thai ~biiM b~ll~v~e t~h~tt ~ ~ Ii~th~it t~f$~sod ~ng~u', kL th~
~ ~Yf ~ ~iki~'ht b~ ~ai~ i~Mi~b Thit ~L ~W~1~d pr~f~r to
see, based on the available evidenoB, ~ ~4ffiifl~t th 1~e ~kthehng that this
be ~4itfki~ly i~or th~ rn$~~tic dieb~ti~ patient, b~caiise it is
o~iy h~ tha~ th~t~isthnc~ that Ic ioall~ hai~e ~d~ncc tif th~ benefits.
There is a lot of idonce of tl~ risk, bet wo ~re conøerne& with nsk~
b~ofit i~tio~ and ~e ~itght to ei~phas~ze the utilization, of these drugs
only in situations where one can provide e1~i&cncc of benefits. Sympto-
ffi~t~c ~ii~f `Wo~ktd be sidi~re~ a benefit, witho~tt any question, ~nd
tM~ i~ .WIYy I i~ot~d f~vo~r ~ iabeiling whioh emphasizes the iitiportalice
of 4~Stri~thd utili~tiotn t~ the sy~ 1~matic diabetic pntient~-~-cleatly,
~c~tl~iet'e ~iiot hi~s 1~ai~cd end wiiore insulin is reftsed by the putient.
~. Go~oow. frcidenthdly, we ~rill send a ~opy of yo~ir comments
to the ~bod ~nd Drug Administi~ation to be included in their record
bef~re they issue the final order.
Dr. Lamer?
* Dir. ~ Well, in general, I am very pleased that the move-
ment to insert the labeling is now going on, and it presumably will
`be consummated, and I, in general, agree with the labeling as it is
~r!t~a up.
I wouM feel ~ little bit more comfortable if perhaps something
spe~ifie toald be ssid in the labeling about warning physicians ad~
ministering these agents to patients who have demonstrated cardiac
preblein~, ~or `etamp1e~ with abnormal eiectrocs~rdiagra~ms ~nd so
forth, I would like to see a little bit more of that type of warning.
Mr. GORDON. Dr. Chester.
Dr. Ciiws~i~. I agree in general, but there are two things that
disturb me, and one, on page ~9, the very first sentence: "The Oom~
rniseioner also ~ontludes that, a patient population exists for which
these `drugs, properly labeled, can be considered as safe and `e~fec~
I ~woUld take issue with, "safe" and would indicate that the e~ec~
tiveness is limited.
And the other thing that bothers me----and I `cannot find in this
document-is how the patient will ever See the label. Will it be on
the bOttle with a skull end crossbones?
Mr. GORDON. Maybe that should be made more explicit in the pro.
posal.
Dr. CH~ss~tR~ I would think so.
Mr. GO~RDtN. We shall send that on to the. FDA.
Dr. Sims?
Dr. SIMS. I have already described some ways in which the label~
ing could be modi~fied to include kiention of `other pre~erable options
for treatment. The question has been raised as tie whether the FDA
has `the right to dictate to the physician how he wiU manage his par-
ticular patient. Another question is whether, if specific priorities and
options are outlined, the physician would then become m~di'co
legally liable to suit if he does not follow them. I believe that tIie~~
fears are a distortion, The FDA, in seCtion ~O5 of the Food and
Drug and Cosmetic Legislation, is given the responsibility to deter-
~biue~ to jhsar~~ rather, the `eftlcacy of a drug. Now, e1~cacy is a rela-
tive thing," and if there are other options which are better, the ~riig
PAGENO="0085"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13$35
no longer can be considered efficacious. So, I thinl~ it is appropr~ate
for the FDA. to list the options. Also, regardless of any warnin~ that
is issued, the physician is in the ultimate position to say, `I am
~tware of your warning; it does not apply to my particular patient,
because of such-and-such condition." He can write that on the record
~and will not be vulnerable for legal suit, if his reasoning is valid.
So that I think that the two fears are not grounded, and I think
that furthermore, we all have the problem of educating patients ~nd
getting them to go along with this. I think we could regard~ an ap-
propriate warning as a useful adjunct in our own education àf the
patient. I think that the package labeling should be written i±~ a
form. and language that the patient is able to understand. We are
~entering an era where patients with chronic diseases are no longer
satisfied to be passive sheep waiting on the word of the doctor.
Rather, they are assuming more the role of a client of the physician
working together with him toward the management of their lifelQng
problem.
Mr. GoRDoN. On behalf of the chairman, Senator Nelson, Senator
Abourezk, and myself, I want to thank you very much for coming
here and for your very informative contribution to our record.
Thank you very much, gentlemen.
The hearing is recessed, subject to the call of the Chair.
[Whereupon, at 11:50 a,m., the hearing was recessed, to reconvene
~subject to the call of the Chair.]
PAGENO="0086"
PAGENO="0087"
APPENDIX
EXHIBITS PROVIDED FOR ~hrE HEARING RECORD BY THE
SUBCOMMITTEE ON MONOPOLY
[From the Journal of the American Medical Association]
REPORT OF THE COMMITTEE FOR THE ASSESSMENT OF BIOMETRIC ASPECTS O~'
CONTROLLED TRXALS OF HYPOGT~YCEMIC AGENTS
Sec.
1. Introduction.
2. Clinical trials In general.
3. UGDP trial,
3.1 Methods.
3.1.1 The selection of patients.
3.1.2 Randomization and allocation of treatments.
3.1.3 Data collection.
3.1.4 Methods of data analysis.
3.2 Findings.
4. Other studies of hypoglycemic agents.
4.1 Keen et al (The Bedford study).
4.2 Paasjkivi.
4.3 Feldman etal
4.4 Tzargournis and Reynerston.
4.5 Summary
5. CritIcisms of the trials of oral hypoglycemic aget~ts.
5.1 Main issue in criticisms of the UGDP trial.
5.2 Selection of patients.
5.8 The TJGDP mortality findings.
5.4 Failure to adapt dosage of drugs to Individual need.
5.5 Discontinuation of tolbutamide and phenformin In tTG-DP study.
6. Data from the UGDP and Bedford trials.
6.1 TJGDP data.
6.1.1 Randomization by ser within clinics.
6.1.2 Cardiovascular failure Fates.
6.1.3 Multiple logistic model.
6.1.4 Analysis with respect to adherence to as~igned treatment.
6.1.4.1 The extent of the problem.
6.1.4.2 Statistical analysis.
6.2 The Bedford trial.
6.2.1 Cardiovascular events.
6.2.2 Mortality data,
6.2.2.1 Failure rates and death rates,
62.2.2 Multiple logistic model.
7. Conclusions.
7.1 Protocol.
7.2 Conduct of the study.
7.3 Methods of analysis.
7.4 Findings.
References.
Appendix A Use of the logistic model.
Appendix B Relative allocation method.
Appendix C Survival modeling method.
Tables.
13837
PAGENO="0088"
13338 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
COMMITTEE FOB THE ASSESSMENT OF BIOMETRIC ASPECTS OF CONTROLLED TRIALS
OF HYPOGLYCEMIC AGENTS
~JOHN P~ GILBERT, PHD RODOLFO SARACCI, MD
Office of Information Technology Sezione di Biostatistica e
Harvard University Epidemiologia Clinica
Cambridge, Mass ~J~oy~tqrio di Fisiologia Clinica
4e1 p~
Pisa, Italy
PAUL MEIER, PHD ~J~VIN ZELEN, PHD
Departp~ient of Statistics Statistical Science Division~
Univ~sit~ ~ ~J~iç~gq State Un1yer~it~ of New Yç~k
Chicago at B~~aIo
CHRISTIAN L. RIJMKE, Mt~ Cox~ ~ ~ BS (Chairman)
Afdeling Mediscbe Statistlek Department of ~lpidemlology and
Vrije Universiteit L~ublic Health
Amsterdam Yale University
~lb
njiyz~ ~ ~E J~TE~ $~c~~r
PETER ARMITAGE, PuD BERTHOLD SCHNEIDEn, DPIIIL
Department of Medical Statistics Department fur Biometric und
and Epiden~jolog3r Medizinisp~ I~9~p~tik
London School of Hygiene and ~ eej~~~
Tropical Medicine Hannover, W~t Ø~i~y
London
TResearch Associate
1THEODORE HOLFORD, P~ ~ ~. ~ MP
Department of Epidemiology and Departm~ ()~ ;~4w~
Public Health Q~
Yale University chicago
New Haven, Conn
1. I~F~?B~DIWT~ON
In a report that was published as a supplement l~p ~ ~v~pber 1970 i~,ie
of Diabete.s~, The University Group ~ ~ ~J$P~?) concluded ~t~t
in patients with adult-onset diabetes, "talbutamide and ~ ~a~y be less e~tc~
tive [in pro1~~~ i~ ~ ~4~t ~ * ~t 441~1~ i~ ~ cardiovascui~tr
mortality is concerned.~ ~, 2~) ~ tJj~ s~b~p jl~re~t~ wi~4i ~c~itamide, 12~7~~
died from cardiovascular causes, as compáre~ ~ ~4% ~ ~r~er in the o~lWr
treatment groups. As a result of t~es~ ~ jJ~e j5~o~ ~4 Drug Adn~i~i-
stration (FDi4) ~ec1 t~1 ~Ø4~ ~ ~e u~4 ~nly in pat1~s
w~ ~ia~l ln~, ~1~%bJ~ ~ 4~ çp~l~ n~4 ~ coi~t~o~W by diet a1~e,
and who, for some good ~ ~ ~ t~t~çl ~ i~L.
The findings of the UGDP and the action of the FD4 1~a4 ~t 4~amatic imRaet.
Tolbutamide had been in gei~p~ ~ ~ ~ie trq4 ~ ~ ~etes since jiI~q~it
1956 and was thought to be a safe and, ~ff~ct~y~ ~ug, ~ T~TGp~ iieport was
t~ully scrutinized by many, and, though 4e~e~4e4 by ~ ~ya~s severely ~d
~xhaus~~ ~m~1~iq~4 ~y ~isr~ 4 ~ç~p ~f ~ cpi~ti~s s~ç~4 legal action ~o
enjoin the FDA from issuing a labelin~ p Icy ~b~t ~vpu)4 d~s~Q1~rage the ~ pf
tolbutamide.
In 1971 the UGDP reported that treatment ~i~h iex~fi~r~1im hydrochlori4e
also resulted in an excess cardiovascular 4p~4i ~r~i~e aji~4 i~4eed, an qx~e~s
overall death rate. (3) These findings have not bee~i ~~4e~y ~~141ssed, and ~$ir
impact on the treatment of diabetes ~
The UGDP study is the largest controli~l i~n~i,ç~,l 4~a~ af ~ hypo4y~c~z$c
agents to date. Other studies of these agents are in ~ ~th prelithmn~ry
results, however, that appear to differ from those of th~ ~ The Natipi~al
Institutes of Health (NIH), which has ~ ~ ~ ~e~l~t the need Qf a
reviewofevidence~ available in all the trials, ~o~i~g1y, çn J~tne 9, 1972, b~ie
director of the NIH at that' time, Robert Q. Marston,~1~~X~, ~rQte as follow~ .~to
the chairman of the group presenting this report:
References at end of article.
PAGENO="0089"
i'~ ~kj~ ~Mr~I~ ~
At my request, on SepteM1~ 14, 1~i', ~ ~óz~xñrs~ Cirnimers, Associate Di-
rector of N~H. fQr. Clinical pare,. i~uv~ted the PresMent q~ t~ ~metj~i~ Soc&~r~
P~~S~5J! ~ ~ t~ á~o1~t~a d~d~ñ1i~t~e to çth~sider t~ie bi~m~trnc a~pe~ts
o1~ ~ ~ri~ti~o~ dr~t1 hi~k!~Mi~ ag~ii~s. T ~ii~ ~ifdrüi~t~ ~J~iht~ t~ co~-
mittee has now bqen 4~pointe~ and that you Jaav4 á~5e~d to ~et a~i~ ~h&irñi1~1L.
f~t~±~i~ of ~EH ~ I~iIiS~ th~C~e~ a~f~ f~t4i1n t~IA~. fact t1kiI~. ~
fdit~ n11ffft~i~ Ath~iç~th~ ~ 4~ déth~IØ ~ ~ ab~i~i~t 1~e~giyeernt&
a~ t~h~ ~ tMrt~&~i~e o~1~ th~ ~t1~ o~ ~ ct~u~~ d~èr ii.~
ni~ffl:1~ d~äl~fi&ai~ ieiyJi~iftgeat~d~ ~Xth ortif 14y~~o~iycethic ~geiith. it is~
n~is~tboüt tWoa1fnt~e~ t1~Lè f~~ft~ ~Ôt~P ~i~iiet~es PrOgrän~ ifrst réporf~ci~
tJ~t ~tfi~ or~t~i d~1i±~1~ ~ I~i1~1~ i~ë~e~s~ tile death rat~e from cai~d~o~
. ~ ~aii~e~ . E~t~tu~ df t1~ê Wk~è djf~i»=áI i~. d~C tli~ d~n~gs in t~e tre~1~
m~h~O~ dib~t~ it i~1D~1$61~tth1t t~h~t~ tl~e ~ ási~êct~ o~ I~1i~ evidence eo~i-
cE~nli1~g thea~geñ4~ ~ todátcef~j1 ~
~Ibe& th~ ~tht ls!üli of tb~ tY~4~i~ ~hk1~ d~ehd~ in ~r~at ii~ea~ure oii~ the. bi~
mt~1~d~ át~o~ th~ ~ ~
1. to make an in-depth assessmetit O~ t~1ie s~it~e I~lf~ o~ t~ tJ(~D~
st~ad~ that lii ~ri~tiTh~ Of th~ ~loñ~tth ~S~ofs of t1~e ~ési~±t,, cOnid~uet,~ anc1~
axt~1~ of. the~ ifl1~;
2~ i~o iMk~i a~ ~thiiiar a th1E#nt~ of otii~ ~oñi~rOI1e~d tI~IaIS ~f or~d ~ypuglyct~m~c
ag~êftt~.
~ ctt~e~ i~ iirg~1 ItO ntittzè all tile sour~~ it ñOed~ ~O ar$~te at a
sa~ctö1~~ a~s~e~; aiid~ to p~pare a r~p~t foi~ p flcát~on.~ flte Cbmmift~e~
sfro~ild~ f~t f±ee Ito obIt~iil eipe~t help lii ~rOi~ai~fn~ t~hi~ i~ë1~Ort aiUt Ito call c~n.
repre~entatives of ~rthi~iit d i~l1tr~ a~ ~O t~aiTh~: A1thQU~h nO i~riOr a~-
pi~ovhl 1~ Ith~ ~ i~ r~4il1r~ed~ *~ sh~all extiect to be ke~tt informed of the
cdil~1li~tOn~ a~ WeSt de~e1to~.
~Phe c~oththit~tee- w~ apj~~'Oliffed b~y the 1~rO~1deil1~ of the 1~iil~t1~ic S~Oelety ai~d~
fl~'Wrë1~t~ it~~fkn~s.
The' ~hU ththiñ~ttth~ fr~It on ~ o~t~io~i~ d~llthig~ thO pe~1dd~ f~Ofn A~1~tst X9~
tc~ Ci&~thtiet' 1~4'. h1 dtIti~nt, tIle ~1tii~dj~ean iri~hlbe~j?It n~t OflC~ ~ ~ gt~U~, and
tlt& t!~ ~ëtr~' dI& ilkèWl~e. lS~ th~t dont'~e ot th~~ ne~tIl~g~ ~l ttt~tlO~fS ~*ere
hdid *iIth J~A~ ~W]~d ate t~It1~itár *IItIi Itb~ clthldal Itti~tl~ of ~E~i' ~l~e~X&
a~,tit~ ~*f~i~ to fhu1thtf~h's of'tithe, tItré' ~OththiIt1teé w~ abl~ to lilIttu~ o1tl~t a
o~! t1~e' tSeO~l~ *1~o ~trEt io~l~ëa1t~l~O' lIt tfti~ iie~i~i~. It ~t1~h~e's th t~rd ~iec1at
tha~±~ fbi' h~li~ gF~eir Ity ~o~Ile~rf ~, d4~; 1~ttI~ ~et!oiI14~ It~ô t!eftf; ~l'~ra~
ln~ieth~, h4t~'; i~. 3. 3~th¼Itt, 1ttA~ ~ti~; ftI~t~ ~ ~ItU. ~ J~O1ln ~.
O~lW~ali, lfP; ~t~tnle~ ~ 1~h~; tht~ 1~o1~e~ ~ltzE~t','fttD. -
The full committee visited the Coordinating~ Qenter of the 1~GD~' at ~alIt~--
m~e' a'th~ à~ snhcOthlthltte'e' ñia~d~- ~ ft~i~th~t~ i~i~lt ltd t1O~' We ~r~ltse~ t~Sdd i~i
rtttklOmtzltlg ttth~ a11~odOthSii of ti'ea~tilitht~ dfrti~tia~d f~. 1~i1rnf, ~ j3~t~, th~
ditt~Ie~tOi' Of the dilnafh~ (~ntOr ilmi iii~ st~tt p4~M~ e~t~hs1'$~d tt~ta-
tidlis thid~ off~in'~tl~ d~thL of ~-tV~f it~~i~it ~arr~ it~niii; ~11~ P~It~~ ~ltfs~- kfra11~'
m~d~ê dl~ta ~t~dI~t~%le t~bth- tl~ ~t1ttf~ tfi~t' i~e dItd~ ñis~ ~Q1i~gtI~ cdiItt.
Sth~t~tthtfttd~ ~*ItsttW wth~e ~ThSd ltb It1~ ce~athrs at i~?o~Itdn at1tft~f1~eihi1aft thnt~
par&i~*tItet 1~n th~ T?(!i1~ tti'IaIt. -
SIlidè~ ~ of tIle r(i1~P Is sItItl'14 iii pi'ogi'e~s~ ft is ndf pO~s$~hI~ to' a~ss~s~
the ~n*~ out&lne of the trial. hr perltletrhlr, mittcft of the o~ta .011 nOflfataIe'veni~
still r'èi11kI11~ ~ñ~\ibhisIl~& `~he~ C n11ittd~ saw' as ith th11M tO~k It~e4 lCstl~thio~
of the excess cardiovascular mortality In the subjects who had.-recei~of ltoThnt~ -
ann~e'. ft rc~1e*C~ in d~teii repor1~s of th'e U~i~ on tli~ ~ nfl~~, 1Mse~-
lifle dalta,, thict ndrthlLtl results (1, 2); the ortgiiial c~nt~ oil *Ili~ these se~
pdrtu wete b~l~IecT ; ailci the cOmthe1i,ta1~~ that' hiS ap~OarC'd' in ltIl~ lite$ittti~e i~
to' tile' en~I of i~rhi'eli M,4~ ~The ni'ottaftty' dkta e~rCrc~d~ a ~ei'i~id 0
8.5 ~`C&ts efld111~ In' ~ctebet~ i~; tIle' ~o111111j~tltC'e hitS se~i1 la1~i o'~ethf]t ~g~in~es~.
bnt si4Ce tbe~ *e~ i~ot fttaalt the dçltiuled dat~I o1~, ~h1cb Itht~ ~k~éist itit~d ~`e±~-
no't ~ft~tIf~d~ hjñ~a~l~~' rd.~OtIt thi. gehfdrlullI' (3~)--w'a~ e effli in
teitt~t ~$1tt I~i ~tieW of'-tilef~t. ithl~t~til~ i~h~ttl IMI~I~ on' thit1~ ~01~jtetlt. ~n'~
still il i~i~Itsile~l, tilO cOñuMtthe' CUd 11ot r~iloth' til~ 1~it~tc d~tt~t oh tild ~et~
of~ thI# tf~atn~ènt: 1~o~ a ~lfnffikr rettstin it 1Th~s ~fren c011~ lli11Itt~i'1 ~~tt~ntioi± to
tile' iI~Ow ñol thtl'é ts in the thllftCt*ulthTh ~thd~4~ -
~he ~Othth*I~ hit's ~ th~ itIl~il'ett tlcIli~c~é avM-thht~ ~f t1It~ einlt o
I~ oil o1~frCr cblitrdflèd trMTh of Orth hfl~o~i~eiMd a)~f1f~ ~ ~`e~rdil'c'e I
mitd~ tO' MI of these, `Iknlf the' thatih eta~s o~ tilts' i~CtiO*t Its `On' tll~ stu'df' b~
the U47~DP and the Bedfe~4 ~tpdy çrganized by Keen. (57 6) -
References n~ Qt ~t~c1~
PAGENO="0090"
13340 cOMPETiTIVE PROBLEMS. ~ THE DRUG INDUSTRY
2. CLXWIC~L TRIALS IN GENERAL
The evidenëe considered by the committee *aS almost wholl~ derived from
randomized clinical trials, and it"is appropriate to be~ib *ith, ,`sothe general
remarks about this type of study;
The effects of drugs, whether bOneflcial or adverse, can be asse~sëd in various
ways. A traditional approach is to present a small number of case reports that
ai~e judged against other dlinieal experieifce. This kind of cOmparison between
a few observations on a new treatment and a larger experience on standard
treatments, may be convincing when the new drug has a clear-cut effect. More
otten, though, a new trOatment produces a small improvemezit as compared
with the standard treatment, or there is a relatively large number of variables
that affect the outcome of therapy. `In such situatiOiis; c~së, stttUles~ l~owever
carefully carried out, do not provide clear evidence of the imp~bvefzient What is
needed is a controlled study using groups of explicitly defined patients who are
comparable in all relevant respects, or whose potential lack of comparability
c~n be allowed for in the analysis of the data.
Serious attempts to conduct large-scale controlled trials can be traced back
to the 19th century or earlieu~. (7). The essential ingredients of present-day
trials, however, are found notably in those planned during and after World War
II, particularly those for the treatment of tuberculosis a~nd cancer, and for
prophylaxis against infectious diseases. (8--12) We may identify for special com-
ment three aspects of a clinical trial to which much thought has been given;
the assignment of treatments to patients, the assessment of the outcome for
each patient, and the analysis and interpretation of the results,
It is very desirable that assignment of treatments to. patients be done by a
random mechanism, the most convenleint form of which is a table of random
numbers. Randomization ensures that groups are unlikely to differ materially
in any prognostic factOr, known or unknOWn. `More specifically, it enables the
investigator to determine the probability that observed differences in outcome
between groups are due to sampling fluctuations rather than to real differences
in treatment effects. Only when this probability is small can we feel confident
that the treatment effects are really different. Without randomization there is
no guarantee that differences in outcome are not due to the investigator's
tendency to assign certain treatments predominantly to patients who have a
poorer than average prognosis-a tendency of which be might be quite unaware.
A further advantage of randomization is that it facilitates the use of methods
for maintaining "blind" assessment, although it does not necessarily' ensure
their success.
If the response to treatment is thought to be influenced by one or more quali-
tative variables-such as sex, clinic, or stage of disease-a stratified system
of allocation may be used to ensure that the treatment groups are balanced
for these variables. Alternatively, simple random allocation may be. relied on
to produce near-equality of the groups for these particular variables, with a
post h'oc adjustment of the treatment comparisons in the subsequent analysis.
Experience has shown that the assessment of the response of a patient to a
specific treatment may sometimes be influenced when either the patient or the
investigator knows which treatment is being given. Even if such influence did
not apply in a particular instance, it might be very difficult to be confident of
this; hence * * *
The analysis of the results of a clinical trial centers on estimating the mag-
nitude of treatment effects and assessing the precision of these .estimates. The
analysis will need to take account of concomitant variables and to adjust for
ai~iy large discrepancies in base line characteristics arising despite the randomi-
zation. Furthermore, there might be interactions between treatments and various
characteristics of patients, i.e.. a tendeilcy for the differenc~s between the
e1~fects of particular treatments to vary with difference categories of patients.
In evaluating the results of trials, one must bear in mind the important
role played by sample size in the ability of a trial to detect a difference of a
given size. In trials of chronic diseases, where special importance lies in the
rate of mortality or in the incidence of particular episodes of morbidity, the
accuracy of the results will Increase both with the number of patients entered
into the treatment groups and also with the length of the follow-up period.
When a trial with. a relatively small number of patients ~r a short follow-up,
References at end of article.
PAGENO="0091"
~OMPETITIV~ ?RO~LEMS IN TH~ l~EUG INDUSTRY 13341
or both; fails to confirm a diff~renceapparently révealéd bya larger trial; ~he
discrepancy may well be e~~lained by the relativel~r large rándom~errOrs Inher-
~nt in a small study.
Any clinical trial imposes an administrative burddn on the iiavestigator in
addition to the effort that he would in any case have to give to the eare of ~iis
patients. Much of the additional burden ii accounted for b~ the need to pro-
duce careful and unambiguous records of all the relevant clinical obsetvations.
The random allocation of treatments does not in itself cause much extra work
during the trial, although a good deal of effort may go Into the pre~arat~on
of a detailed plan for randomization and blind assessment. Many large-sétde
trials can only be mounted as collaborative studies, since any one medical center
Would provide an inadequate number of patients. Multicenter trials give rise to
special problems of coordination, and there must be a clear prOtoéol for the
Study. It is necessary to ensure that principal Investigators from different
centers meet regularly and to establish a co-ordinating center that monit9rs
the standards of the study, that receives the records as they become available,
and that analyzes the data at regular intervals. One of the moSt difficult prob-
lems in multi-institution trials is to have an adequate quality control system
for the data. The processing of data from a clinical trial, particularly a multi-
center trial, is also a substantial task that must be properly handled to ensure
the efficiency, and indeed the success, of the trial..
Randomized clinical trials pose ethical problems. Some of these are commOn
to all medical research involving human subjects, but others are specific to t~ils
particular form of study. Three important questions are (1) Is it ethical to
assign the proposed treatments to patients according to a study design dra*n
up by the investigator, even if the patients have given informed consent? (2)
What are the criteria that should allow an investigator to depart from an
assigned treatment? 0 When and how should a trial be stopped, or its desi~n
be modified, if one of the treatments seems to differ markedly from the othE~rs
in either adverse or beneficial. response?
These and similar questions have received much attention (13, 14) and *e
cannot discuss them fully here. A few points, however, are particularly relevi~nt
to the studies under consideration. The investigator's belief is well-found~d.
Investigators will differ both in their readiness to undertake a randomized
trial and in their reluctance to continue in the face of accumulating data st~g-
gesting that a difference may exist in the response to the treatments, In vegard
to this latter decision, statistical evidence about the possible size of the dift~r-
ence is relevant, as is also a consideration of the sequential nature of the an~tl-
ysis, which may well place exaggerated importance on transitory random fiuctu~
ations. Different people will make different assessments of the evidence that
may be available from other studies, many of them perhaps nonrandomized, a~d
of, the risks and benefits of continuing or stopping the trial. No criteria wfll
satisfy everyone. No matter bow long a trial of this sequential type continues,
some will criticize it for going on too long and others for stppping before su~fl-
ciently conclusive evidence has been obtained. In any attempt to review t$
propriety of particular decision to stop using all or some treatments in a trial,
one must bear in mind the range of decisions that might reasonably and prop-
erly, be reached.
3. THE U~OP TRIAL
3.1 Methods
Patients began to be recruited for the TJGDP trial in 19E1, and curre~it
plans are to continue follow-up through August 1975. The objectives and mCt~..
ods of the trial are described in the published reports (1, 2) and the followir~g
account is confined to a review of a few salient features.
8.1.1 Selection of patients
The method of recruitment of patients varied somewhat among the 12 ce~i-
t~rs involved. Some patients were obtained from diabetes clinics or through
referral by physicians, and others through special screening procedures. Pit-
tients were considered as suitable candidates for th~ trial if diabetes bad Jeen
recognized within the preceding year. From these candidates, all. those wl~o
met one of the following conditions were excluded: (1') those who did not shoW
a pos~tive diagnosis by a standardized glucose tolerance test; (2) patients with
a history of ketoacidosis; (3) those who did not remain free of kOtosis durir~g
References at end of article.
PAGENO="0092"
1~3842 COMP~Y1~ITiV~ ` ?ROBLEMS IN TINE DIjIJG . I~WU~TRY
a one~n~unth pE~riQd ~f treatment b~ diet ~ a1~e oir who, dmiu~g tht~ o~e-inciith~t
period o1~ o~vatio~i,~ were j~$g~d ua~*i~b1e q~r unwU~Ung to ~o1Lo~w the study
proto~oI ; and (4) thos~ whO had any serious conctition th~b1~ in ~ jUdgrnent~
of~ t1i~ ~lin~ie~ p~ys$~oia~n~ implied a~ 14~ e~p~tanay o~ Lees than five years..
A~ a i~su1t,.. th~ t~P si~bje~t& may be though o1~ as a gr~p o1~ patients with~
aduit-oiis~t~ nouketotii~ diOetes~ Thete~ wa~ a ,prepoaderan~e~ o1~ w~rneu, wbo~
made .ui~ 7~% oJ~, the tota1~ t~iabete~ ranged~ in severL1~y trom asymptomatic,..
with nç g.lyeosu.ria, to symptornatc4 with prma~ent glycoauvia. and markec~
byperglycem~t~ A description of the patients ia g4veu by the UGDP (4~. pp~ 777-
7~3) in the form of several tables that present dlstributiou~sof base4ine charac-.
teristicS of those in the study.
Tn the ~ many of the critexia~ for oxeluding subjects were, well defined.
There wnr~ also sjtua.tlons, however, in ~bich the clinic physician had to use
his .iudigment-~-fbr exantp1e~ in screening to obtain patients with a. mi~hnum life~
expectan~y. of ~ve yearn1 [n view of this a~d of thc faet that patients were~
drawn from , various sources, it, would ba expected that clinics might differ
s~stematieally in the charactcr~sties of,the subjects selected.
3:1.2 ~t~ot d3UZ ltt~id~in `~f t4~ni~nt~
Tild T1~J~ ttM~' ~v~i ~tt ngt~c1 at( a baiatn~d ~eslgh, ~tr~tifi~d by bIock~ of
1~O th~ i4 ~ tient~ witlith ~lthies but without other res~r1etious oa.
the pattern of assignment of tthat~h~±1t to thts, Tiilt1aIi~ d~ring 19d1 In
~ ol~ se~et1 ~flttf~s, the fbilrtreamttettts-rarlaiile-dose'itisiilhr (IvAn), standard-
ddne lnsnthi (`Th~Pti), tolBnthinlcle', ~lid p'laeebo-were afiochted randomly to
~Lttit~tLt~ itt lil6cks of 1~-f~Mn' subJects In each of the four n~eatments Ill ran-
dotti otder. t~i i~2~-i~i6~ phettf~rlttin WaS added to the ti~atttiettts at five new
cthttc~ ~is~ Well a~ at Otte of the original ess~eii and, in ordOr to achieve overa1i~
~atit~ itt tIlo tdta~ number of patients assigned' to each treatment, the block
s1~e' Was fL~é'd' at f4, with each i~Tocit containing si~ subjects receivIng ~hen-
to~iilitt, dilti IWO `redei~ittg' each of the four oIlier ti~eanneuit~n
For purposes. of administrative .efficiertcy~, ihdivithml patients receiving toT-
htttattMe `dr' pionel3tj were' ttot assigned inilqiiel~ idCtttiftOd me'dleation, but Were
siippl~ed' as folloWs: For the folbnttmlde assignments, numbers 1' to 24 were
s~l1t at raiidbni IntO tWo gtoil~~ Of T~ dflO grOU~ oi~ numbers heing assigned to'
iith~e'bO and the reniaftider to bottle'S' that Would be' used for toThutamide~ 1~acb
of th~ th'st 24 51Thjt~c't5 receivIng ~lacclbO Or t~tbtitanil'dt~ itt a giuren clinic Was~
aliOtt*d a te~ltrtttte' bottlO Muhber, the se~tience' then' being re~~ettted. Bottles
2~l th~ougb 4S Were used for patientS assigned to toibtitamide' itt the clinics that'
also tl~ed phenfornitti.
As g eotisecjttecee of this orrangP~xient for the distributloli of medication,
sotnetiftics two and Ss most three subjects it~ fi gIven ~Tii1i~ Were. Sttppiied with~
i~e'fitl~5l hoti'ie" niftObCr~. The administrariva ath'snta'ge of Slits, sehe'me is that
e'flclI ~linfe' coiUti be' giten a~ ittltthi' sttpply of 4S w1ique1~ labeled fnedications
a~nd cOttid~ order ttdthtioitai supplies, aS rie'ed arOse, WithOut burdening the'
ccnttal pharmae~ With rè5pe'ri~ihilft~' fOr' moie' than ~kio separately lab~Ied~
ic~tlOriS.
The ,orafl~ gfren nthd~lcttfIoits in the tolbitiamide' Stticly Were' In tablet form.
~1~'II~ ~ntt~tititi~tibff of phenfarrnin itt The ~~eOlid pttrt of the study required' a
change in the method of administration, since phenformin is supplIi~d as' gvaatr1e~*
filled capsules. In this part of the sftd3~ all control medication for new patients.
was given as capsules. Tolbutamide was still supplied as tablets but, unitnown
t~ tl1t~ ~au~t1C1pa~thig ël1nI~~, pl~eeho iii the' f6rsi of t~bletS WS's ttet ~i~ett in the
Iif0r~thhk e'thties. t~eW bottle n*thbth~S ~49 tn 72) We~e' uSe'd for the capstti'e's,.
bitl thW sttnte' the'tbod of resupply was
In executing this plan, liStS Of ord~u~ed ti'Oaftnent~ S igatiterits we..1jre~s11~ed1
in advance for `each clinic by the Coordinating "Cente'r, Randont permutationa
of 16 from' tl~ tables given by Cochran and Cox (15) Were used for the treat-
thettt alloCS'ti0ttS Iii the' ftrtit sh~ ~thiiCn; and' tile ~SILd tthres' (1w)' Were em-
pTh~e't1 for' th clinIcs itt Wili~il ~hettf~WeitS waS admtnisrered. The assignments'
*e're, enteted th a mg book, and spncs was left on each liSt for entry of the
thtnie Slid Identifying nuthbet of the ~ttthint and the datit of `a i~rirnent Pc
f~e'tiit~tc tnfthttton of. tte~tittenC, asSignment re'~ttesis' could be macto by the
clittit tO ste' ~dotdinhti~ig Ce'nt~t att~ f4i~d ~y telephotte', itt which eSSe a himifo~
nuniber of ittdi~rh1ua'ls had authority, to revorti the pame of the patie'nt mt the
References at end of article.
PAGENO="0093"
~OMP~PI~VE ~ ~ ~ I~Rt1G 1~D~t~V~ ~
:approp~i~tte line ~f ~ ~ ~ ~ aiadl r~o~rt ba~k ~ ~n~s&~eeteU tb~r~p3~ . as
sh~wii km the 1~st, that 4s,~1t~r Ls~r ~!r I~v!tR o~r~a ~øt~I~ flk~rnb~. C~a~ft~r~
1et~s wer~ i~angk~d tbs~i~e~t~. ~ the a~M~M~a&~t ~e~i~lis
might ~rne by rn~a41, &~d th~ ~ ~Ye ~i~j~Ottt~d ~ in ~t1~ ~an~n~r~ All 4i~it-
~rnent ~assignm~fflis ware inad~e in ~t1a~ sa~nence mid ~ttt in ifre mt~a~ioti
iist.~
Once treat4nen13sw~e ~ssign~d, the~a~ was thitiiMad by t~e clinic. titbi
th~rapies net being "ttlind," rei~piired `n~ ~nr%her ~en~ld~ration~ Th ~1T~ cas~ o~
orally given medication, however, l~he traatinent was i~ntified ~only by ~ botUe
~nurnber.
~.1.3 Data coZlction
Ollineal wnd laboratery &aba were ebtained ini~i~fly and nt ~nhsequeat i,navt-
erly follow-up visits. Provision wee niade for ~gat4th~g death re~orts from the
clinic. phy~ieigzi wko prepar~fl1 a auiaa~ry 4f the ~na4iqa1 record a~cI ef the ~t-
tendiiig pliy$icinn's description of the ~circtunstances ~nrrouud4~g the death. 4
copy of the death eeztiticate and also a ~eopy of the ant sy rejart, if ~ne
exthted, were funitsbed to a central pa Leonsisting of ~ap internist and. a j*th-
ologiat. This panel ~nade the final decision as to the eanse of death and ~I~y
dad so without knowledge of iibe treatment group to wlfleh the patient bed been
assigned.
3.1.4 Mctkofs of data anni~~sis
As the UUDP rnudy progressed, it, appeared, nne~pectediy, that the ,~a:tL~nts
treated witb~tolbutamide ~bad.an excess i~ortality fro~,ear4ioxaseular canses.
~This called for statistical technigues tO provide ways of eval~ating the~ao~gni-
tutle of treatment. differeneeshi.wortality.
g2 T~st.-The first naethod was to perform a X2 test to eom~are the prier-
tion who died In the placebo-treated,groi~p with that in each of th~Ather.grO~ps.
Life table aeml~jsis.-,In a study such *~s the tJ~ThP trial, sii~bjec~s are en-
rolled over a period of time, so that when the stu4y ends, the ie~~tb %of ~o~1p~-
up has varied from individual toindividnaL Xu o~d~r to Mjnat. f~r these cli~tor-
~enees, the i~v~sUgat~rs used a life table method. This involv~s, e~timatim~g, fi~r
each treatment group. the survival curve, that is, the ~pr~por~iQ~n, pf
surviving a given number of years.
Multiple logistic rmo4o~.-Since tlie~4eath rates were affected by many factors,
the distribution of which differed somewhat from one treatment group to an-
~otber, it wasr desirable ~to lInd a metbsi of a&}usting fer these tti~erenee5 in
base4ine variables. A. multiple Ipgistic model waS apteynd fer this ~purjm~ese.
`The ruse of. such a uwdel can be ~i~ded as an a*n~at to take into noe4
`simultaneously the covaviables that affeet the auteoa~e. The projsability of death
is ~presseU as a fmsction of the base-hue rl?iefliables, and the dMa are ij~ed
to estimate the parameters that appear in this function. The statistical umtL~ed~s
that were applied were those appropriate ,for large pamples.
Monte Carlo monitoring procedure.-~i'o rpro~Me a ttatistiCal b~si~ ~er com-
paring drug~pIactho dl'fferenees In mortalit~ asihe study js~ceeeded~ bounda~ies
~Cor thiS dffferenee `weSe censtruoted b~ siatc,lation. The
to simmth~tbe ` rtd1i't~ dtfhrences that wo*ld ObSe~ivedh~i4)tbe ~n~taigty
rates from the 4~5~t~1961 VS `life table `been in effect for igroups ~vcritli the same
age, race, `~thd sex as `the `indttidutts iii the study gisiups. ~Uppcr and lo~wer
bounds were set to the ~differenee in ileath rates over 1~imc in ~u~h a way i~hsrt
there was a ~rebaI~iIlty of only .05 that a greater diffarence wonld~be obse~red
if the 5~-~14~61 death ratashadprevailsd.
Liloelihrood ~ further method used d,y the ~UG~),P 40 ffionitor
the umber of deaths was the calculation of relative betting odds (flBO~ This
I~ a i~ayes1an stati5t1~àI~pr~ce~tLrerby ~l4dhaat dttempt `was linatte to rin~~~o-
rate a prior belief iii a hypothesis ~b~ut the difference ~botsme~n ~the ekrutu1w~i'vc
mortality, as ealeiflated from the life table, Of the titug-t~roated groap ~ the
corresponding rate for `plaeebo~treated group.
3~2 Fiadin4is , , . . , ` , .
By October 7, 196~, a total of 8~ deaths had occprred in the,Aur4zditiinl 4r~at~
went oupai~ ,f~C these~J~ ~erQ~due ~o ~ 4tó~sc*larennses. ifl~e, uns$~er
cf `ps~iients i~tlt5*y assigned to ~ rnsut~gpoups:au4 jhe i~ardent
c~4ea*bs airesbown he:~topipottion of~Th14e I. .
lU~ferences at end of article. .
PAGENO="0094"
1.3344 COMPETITIVE PROBLEMS IN TIlE DRUG INDUSTRY
A 12 analysis of these data indicated that there were differences, not statis-
tically significant, in the proportion dead from all causes, but the tolbutamide-
treated group bad a significantly greater proportion of deaths from cardio-
vascular causes than did the placebo-treated group. Analysis by the life table
method confirmed these results. Since there were differences in the base-line'
characteristics of the patients in the various treatment groups, however, the
question arose as to whether the differences in cardiovascular mortality rates
might be adequately explained by differences in the incidence of risk factors~
The conclusions from the use of the logistic model (17) indicated that the ex-
pected number of deaths due to cardiovascular causes in the tolbutamide group
if it indeed had the same cardiovascular mortality as the placebo group, was
only 10.7, whereas 26 cardiovascular deaths had been observed. Further con-
firmation of these analyses was obtained from the Monte Carlo monitoring pro-
cedure and the likelihOod calculations. ence, the investigators concluded ~-that
there was an excess of cardiovascular deaths in the tolbutamide group-an cx.-
cess group that could not be explained by differences in the base-line variables.
A closing date of Jan. 6, 1971, was used for the analysis of mortality data
in the group receiving phenformin and those receiving other treatments at th~
phenformin clinics. A total of 47 deaths occurred, of which 37 were due to
cardiovascular causes. The treatment groups in this case are the groups in the'
clinics where phenformin was used. The number of patients and the percent
dead in the various treatment groups (3) are given iii the bottom portion, of'
Table 1.
The investigators concluded that there was an excess of cardiovascular deaths
In the phenformin group, and further analysis showed that the, excess could not
be explained by baseline differences in the groups at risk.
~he, relatively few pubiished finding on nonfatal untoward events in the
`IJGDP trial show only minor `dIfferences among the treatment groups, (3) and
`these data will not be considered further in this report.
In 1969, a decision was made to discontinue treatment with tolbutamide. In
contrast to the controversy that this action of the UGDP investigators provoked,
there has been relatively little discussion of the decision, taken in 1971, to dis-
continue treatment with phenformin, and this latter step is not considered in
detail in the' present report.'
4. OTHER STUDIES or HYPOGLYCEMIC AGENTS
In this section, fotir other controlled studies of hypoglycemic agents will' be
reviewed. Uncontrolled studies will not be discussed because it is extremely
difficult to tell which of the effects observed in such cases are due to the treat-
ment and which are due to the selection of patients and their assignment to
the treatment groups. The studies under discussion are identified by their
authors. `
4.1 Keen et r1 (5, 6) (The Bedford Study)
The ~ubje'cts for this trial were `people in whom the capillary blood glucose
level, measured two hours after a 50~gm~glucose load, was between 120 and .200
.mg/100 nil. Of the 248 persons ~dentifled in this way, 228 were recruited
through a screening program and 20 from a glaucoma study. The subjects are
described as being borderline between norman and diabetic, and presumably had
milder disease, on the average, than those included in the UGDP study. The
latter had an average two-hour blood glucose level of 229 mg/lOU ml, but this
was not necessarily strictly comparable to that obtained In the Bedford study
since the glucose load and conditions of the test were not identical in the two
studies. ` ` , ,
The subjects studied by Dr. Keen and his colleagues included 129 males and
.119 femalea, so that the percentage of females was 48, considerably lower than
the 71% in the .TJUDP study. The average age of the males was 55.4 and of the
females 58.9 years. These ages, were higher by 1.3' and 6.8 years, respectively
than those of the corresponding groups in the UGDP. All subjects entered the
trial, effectively,, on one of two dates~ June 1, 1962 or Jan. 1, 1964, and all were
studied at one center. ` ` ` ,
Half of the subjectS were treated' with~ tQi1n.~tamide, 0.5 gm twiee `daily, and
the othev half with placebo tablets. Ii~'additi0~i, dne half of each of `these'groiI~s
was recommended to limit carbohydrate intake to 120 gm daily and the other
References at end of article. , , "~` " ~` , ~` , ,
PAGENO="0095"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTIfr 13345
half simply to reduce their intake ~f table sugar. The dose of tolbut~m1de W~s
two thirds of that used by the UGDI~. The sfibjeets were allocated to the trea~t-
metit groups by a method of randomization that was based on the use of ~ürn-
bers read from a telephone book. here was no stratification according to risk
factors prior to the randomization.
At the time of entry to the trial, information was obtained on age, sex, wei~bt~
clinical history of arterial disease, blood pressure, and blood glucose level. ~ol-
low-up examinations were conducted at Intervals of every six months except
on three occasions when the interval was `one year. The cardiovascular coxt~po-
nent of the follow-up examination included the administration of the I~ose
questionnaire and the taking of an electrocardiogram.
TwO types of outcome have been considered in the an~yses: (1) death, eit~ier
from cardiivascular causes, as identified on the death certificate, or from all
causes; or (2) cardiovascular events, which, in addition to death from car~Iio-
vascular causes, include cardiac infarction, angina, worsening of the BOG, on-
set of claudicatlon, and stroke. ~he trial was planned as a double-blind study,
A list `of treatment assignments was a~cailabIe to the principal investigator dud
was occ~siinally consulted by him when it was thought that the welfare of the
patient required it. rrhe principal investigator is confident that the decoded in-
formation was promptly forgotten by him and did not influence his assessment
of the patient's outcome.
The findings reported by Keen and Jarrett (6) on cardiovascular events at
the end of the seven years of study are given in Table 2.
The authors noted that in each treatment group, the frequency of ~ardio~as~
cular events was, as expected, higher in the subjects who were thought a priori
to be at higher risk. They found no evidence of a treatment difference in the
h~gh;ri~s1t, group but "In the low risk indlviduals~ the rate oi~ events in the tolbut-
aniide-treated group is about half that in the placebo group, a difference sigr~ifl~
cant at the 2% level." (6) They further conclude, "a significant degree of jiri-
mary protection against cardiovascular events can be conferred by tolbutam~de
in mildly and moderately hyperglycemic people."
Mortality data from the same study are presented in a report by Keen. (5)
At the end of'eight. years from the beginning of the trial,' 25 deaths had been.
observed in the placebo group and 24 in the tolbutamide grofip, 14 of the forther
and 12 of the latter, being due to cardiovascular causes. Both total death rate
and that from cardiovascular causes were at approximately the same level in
the two treatment groups. The total death rate of 19.8% was approximately
double that observed in the UGDP study. One important factor in this differetice
is the relatively high ~rdportion of subjects over 70 yearC ofagO in the BedfOr~
study, as shown in Table 3 Another might be that in the Bedford study thOre
was no selection based on the likelihood of a five-year survival, as was employed
by the UGDP.
The data of Table 3 show the higher mean age of the Bedford subjects as
compared with those in the UGDP. The percent over 70 years of age is as high
as 238 in the former and only 5.9 in the latter.
Table 3 also provides an instance of a differefice in the distribution of ba~~-
`line variables betweOn `the two treatment groups of the Bedford study. Of the
placebo group, 29.8% are over 70 years of age as compared with 17.9% of the
tolbutarnide group. The difference is statistically significant at the 5% level..
In section 6 of this report, an analysis will be given to take such differences in
base-line variables into account.
4.2 Paasikivi (18) `
This Is a study of hypoglycemic treatment in 178 survivors from a first ~
cardial infarction. A further 92 patients who had been treated for an infarction
during the same period were excluded fOr various reasons. The antthypoglycemic
agent was tolbutamide, which was tested against a placebo.'
Even or odd birth date determined whether the patient received placebo or
tolbutamide. The maximal' dose of tolbutamide given ~vas I gm; this was alse
the usual dose since it was given to nil' but 28% of the patients, who mos~Iy
received 0.75 gm/day. The period of follow-up ranged froth 1 `to 5.5 ~ears, the
average being 2.9 years foi~ the tolbni~amide groupand 3.0' years for the plaee~x~
grOup. ` ` ` ` ` `
Sixteen patients of 83 (19%) died in the control group, and t8nf 95' (14'~)
died in the tolbutamide group. All deaths were considered to have-been due te
H ` !`` ~ `~
References at end of article.
PAGENO="0096"
i8~46 QQ~ET~TI~ ~ ~ `1~ p1~3~ J~1~~Y
~r~t~e ~ ~ ~1~? th~~t ~2 ~o~a4~s, ~5 ~ço$~s ~ ~ t~j~J~e-
trE~:tt~U ~ ~o t1a~t..ft~ ~e Iy~t~g~ 4i~i~e~t~ ~ i~
,o~ l:oibth;~~i~4e ~ ~ 4xeap~~r~cj~1o ~e ~a ~Uo~1 ~e~t
~of to1bpt4~uj~e trç~~t ,o~ ~ ~r~v~1 ~ter 4J~ ~31m~ ~ ~j~ç~4e of a
`first invocardiai infarction. Aftei~ five ye~u~, ~ ~~i'e ~ ~
~i~nc~ ii~ ~s ~ ~ T~4s ~S~3dy i~i~er cqi4rp~s nor
~or~trac1i~1~t1ie ~Y~J? ~ ~Jae ut~j~oii ~up~er d~ç~i ~ .~t
ofle cf ~a ~1ty~pn~et4ia~et ,~4 t~ ~p~t~pt~ J~i~ tq ~e~h~ç1 ~
~çpo~ed tø a ~ti~ely e~m~U c~o~e fQr ~ Øioj~ter t~n~ ~]~i ~~l~ei in t~e
TJGDP st~4y.
4.3 Feldman et al. (19)
Ree~areh ~dbj~ts in this study were ~5O ~aiabul~tory pa~jent~ with flewly
~diseo~rered a~yrnptoi~ati'c diabetes, Who were between tl~e .~ges çf X5 ~d
`years, :~ree of other' diseases, ~nd not taking dri~gs kp~Qwn to i~~uence carho-
Jiydra~e niet~Olism adversely. `They were ~ariç1onily as~g~ied t~ tolb~tnmi4e (1
~gm daily), phei~forpiin (100 mg "d~ily), or p1~tç~bo treatinei~t to test w~et1&er
the orally glvun thiigs'were ~ffect1ve'1n pi-evé~t~ng or po~t~~il4g overt 4iai~etes
Inthese s~s~eçts. ~hu study began in Deçern~er 1064, ~ the ~p~b~Juh~d ~tta
eover the `fi~t ~ of ob~er~ation. ~ average a~e ~t ~ was `44~4
years and therei~avebeen only `t~o `deaths. Conse~uent~y, `t~e~ ~ ~e insi~-
cient as yet to throw light o~i ~e re]atiye ~ort~i~y rates ~assoct*te4 with the
4iffei~~nt treatments.
~ (MO)
A~prospee~e~st1~y Of b37 p ientswith~prernature c~roiiary'~a~'tery: disOase
was `begnu in1~65to~e~alua*e~theiuduence ~f~henforrnin `(~50 to 10O~g daily)
on mortality 4rom eardlovaseular disease. In a snbgroup of 104 patiei~ts, ran-
d~inized with ~spe~t to' ~henform1n `treatment, or to diet alone, nine deaths
~ecnrre~ 8g~50~ e~tvobpat4eirts `and~slx deat7hs~ * *4*
to~kave~di~betes were excluded In this study, as were obese
individuals, so that the ~eneralizatton eftheseresnlts to~maturlty-onset diabetics
is~ du t~us.'(Itdst~Is~ hiiipattant~to ia~te that there were only ~b9ut~50papien~ts
in eaoh~ group4n ~th1s4 study. ~be sinalldi4!t~erenees in the si~or1*lity rates for
dttfereflt ` eat1n~1~t~ ~b rgdd~iJ~y ~the UGDP ~heu~there were about ~00 mdi-
vklualslu iea~ih Atrer~tmeat4~graup could ~not be `deteeted *lthLthi~h probabilthy
wben~grou~ dfL~0dduals4areAstn~dldd.
~.5 ~n~irp~y
~ ~v4~v~g ~as ,a ~b~~1e he~e~pe ç~f jibe. cont~$led eU~b~al trials of. orAl
~ thatc the unly' ~eiortality de4a that are. ~xieu-
~sivef~l~i17fpr onripnppo~r 4arethese ~om the UUDP ~stndy:~and those from
the Eedford trial.
Th~e ~ .1~n e~t~~4t the~t~IDP
trial. ~The~ii~~al ~ ~4~t~in, ~44Z) b~çr~ ~) S~~-
er, (~3)~~ud ~ 1~i ç~4iç~i, ~
rials and ç~i entgri~,~j~qut ~. U(~J~? ~4riAi4 ~id~ t~ey tue 4~e~le4 ~p4~4y~ çui
these ~eponts ~or~sQu 1. ,~ ~ave~p $~d ~ie~ ~i1p&i~t~ç~n
of two rejOIn~ders. The fir~t was by Cornfield, (~) ~my~io ~~Ø.~se4 ii~n~alf to
refuting the criticisms of Schor and also some crIticisms that ,had t~ppeared in
the report by Feinstein. The scond rejoinder article by Proi~it ct al (~95) was
spe~i~ea~1ly ~*e~to~ ee~i4er ~ Selt~er~s communica-
tion. `
~a~iy ci the 4ci~ms~t~a~e ~made 4jn ~these ~artic1es would ~apply, ~some-
times even more stroi~gly, 4~o~ be ed4~emmd~ stUdy~but~itiw~s the~ ~ trial~that
~ t~C~ie~d~iCS4 efl~iI~gs1Qb4~4s) la~tor~triaJ rso oonj~ter to
~ kbe~ efnlAess sa~etyf$amni~1e, iu1tis~wp-
pie te~4b~4~th ~c1a~si~nau ~c1u ~i~e ~nethod~ ~at ~e&t0~th~lt, ~1mau4&be
eef~*i~, si14~ii~ed, `
~ p~S~d~ft~mns;
a re~v1ew of both trials by members of the committee will be included in~tmL*~e-
4uent s~t1~ps.; ` ` ` ` ` ` `
References at end of article.
PAGENO="0097"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY~ 13347
5.1 Main issue in criticisms of the UODF trial
The primary issue' of cont~ern In the published criticism of the UGDI~ is
whether or not the evidence pointing to toxicity of the oral agents is v~1id.
Thus, in Seltzer's discussiom of the design. of the trial, (23) all of the nine
points he raised bear on this question to an important degree. In the fol1o~ing
account, most attention is therefore given to the UGDP mortality findings, but
in addition, reference Is made to the selection of patbmts, the dosage ~che4ule
adopted, and the decisiOn to discontinue the use of tolbutamide and phenfor~nin
in the UGDP study.
5.2 selection of patients
The first point raised by Seltzer, and also discussed by others, concerns the
selection of patients. Crjftclsm of the criteria used embraces the recruitment
of subjects known to have concurrent disease (including cardiac disease), the
inclusion of some who did not have diabetes, the exclusion of those judged (on
somewhat vague criteria) to have a life expectancy of less than five years, and
the inevitable arbitrary exclusion of those who proved uncooperative.
The determination of criteria for admission to the study depended on ethical
as well as many practical cotisiderations, and was inevitably, to some extent,
arbitrary. It is almost never practicable, and rarely desirable, to make trE~at-
nient comparisons in a strictly random sample from some defined populatio~l of
subjects. To be useful for clinical purposes, however, the' study patients sbo1~lld
be so well described as to be identifiable by the cllntcian and should also. be
among those for whom the competing tJ~erapies are used or considered.
The choice of specific selection criteria adopted by the UGDI? was a respor~si-
bility that was shared with zaedical experts and is not a topic on which this.
comniittee as a whole claims primary competence. It is important to recognize,.
however, that criticism of the choices made is largely irrelevant, to the prim~ry
issue raised by the critics. ~or example, the concern about possible tolbutamide
toxicity would not really be lessened If it could be shown that the study g~oiip,
contained some nondiabetics. A drug found toxic in such subjects would r~ot
likely be counted safe for persons with well-documented mild .diabetes either.
The criteria for inclusion or exclusion do influence the efficiency ~f the study,
and the extent to which its findings can be generalized, but have littlä beari42g
on the issue of toxicity. We turn to criticisms that are more important in tl~s
regard. ,
5.3 The T]GDP mortality findings
The implication of the UGDP mortallt~ ±esults is that the oral hypoglycemi~s
are responsible for an increase in cardiovascular mortality, but that they ~o
not affect mortality from other causes; Several kinds of criticiSms have been
raised about this interpretation, of which we conSider' the following to be t1~te
niost important.
a. Although the total death rate' was higher in the tolbutamide group than ~n
those receiving placebo, the difference was not significant. Correspondingly, the
death rate from noncardiovaseular causes was higher in the placebo group than
in the tolbutamide group. As O'Sullivan et al (24) have commented, "Interpret~-
tion of a study showing no increased risk of * * *
If there were subtle cues that could lead to somewhat different recording of
signs and symptoms for different groups, it is conceivable that deaths of u4-
certain cause might be more likely to be assigned to a cardiovascular cause
in tl~e tolbutamide and phenformin groups than In the others. It will be appr~-
elated that the review panel nsed in the UGDP study had no independent aCcess
to primary "objective" data, but depended on data already Structured `and tç~
some extent interpreted by the clinic physician. Under these circumstances it is
not too surprising that in only 2 of 89 cases was there a major disagreement
between the panel and the clinic physician. The use of a review panel was an
indispensable choice, especially for monitoring possible differences in procedure
among clinics. However, its independent contribution to the actual assignment
of cause of death should not be, thought of as large. The Uç~P took unusuall~#
strong measures to minimize the possibility of bia~d evaluatlop and took care
to use well-defined end point~. in arrl~ing at ,a diagnosis of cause of death.
Nonetheless, the possibility of this sort of biased recording cannot be ~~ule4
out completely. `
References at end of article.
50-592-75--------7
PAGENO="0098"
13348 COMPETITIVE PhOBLEMS IN ~I'HE DEUG INDUSTRY
Our view of this criticism of the UGDP findings is that it has some weight
(although we do not interpret it as a criticism of the action of the UGDP) and
that the to~dc effect of the oral h~çrpog1ycenilc's cannot be affirmed with the ceF~
tainty that would' be present if total mortality were significantly' different.
b. The excess mortality appears clearly In only a few of the clinics. This
might suggest a peculiarity or defect connected with the study methods em-
ployed there, and this~ would have to be understood before any reasonable inter-
pretation of drug effects could be made, We have considered the question of
whether the differences in results between clinics are such as to cast doubt on
the meaning of the IJGDP findings. We recognize that a clinic in a middle
class suburban area is likely to have patients different in many ways from those
of an Inner city environment, so the fact that clinics diffet is in itself not at all
surprising. It would at least call for an exp]i~nation, nonetheless, if a toxic effect
were clearly dis~ernible in one set of clinics and a contrary effect in others. We
present data in sectIon 6 (Table A.3) that beàrC on this point. Looking at the
failure rates for females and comparing placebo with tôlbutainide groups, we
note that there were seven clinics In' which there was at least one cardiovascular
least in one group or the other. The patients receiving tOlbutaniide bad the
higher rate in six of these. In the `case of males, the tolbutamide rate wa~ the
higher in five of seven inst&nce~. We conclude that the excess mortality is not
in fact confined to a few clinics and that this * * *
As mentioned previously, the study of Paasiklvi gave findings that cannot
be appropriately transferred to the UGDP population ifi view of the differences
in dosage of tolbutamide, duration of study, and population at risk.
The study of Keen and his colleagues, however, `deals with a j~opulation of
borderline diabetics somewhat comparable to the UGDP group except that they
were mostly ascertained by screening. Since the investigation is still under way,
we can consider only the findings currently' available. Keen (5) ,found that the
death rates for all causes and, for cardiovascular causes were essentially the
same in the tolbutamide' and placebo groups, but that the* various pathologic~il
outcomes that he designated collectively as cardiovascular events were signifi-
cantly less common among low risk subjects receiving tolbutamide than among
comparable subjects receiving placebo.
The resources available In the Bedford study did not permit as thorough
an investigation as was possible in the TJGDP. The randomization of patients
was carried out without the detailed attention to documentation that a major
trial demands. There was restricted coverage of background variables, and all
the usual safeguards for the maintenance of "blindness" could not be ensured.
Finally, as the work is unfinished, a definitive analysis has still to be produced.
The provisional data that Dr. Keen has kindly sent us are reviewed in section 6
and do not throw doubt on the UGDP findings, in regard to deaths from cardio-
vascular causes. We have regarded the data on deaths as more relevant for
comparison with the UGDP and also more clearly defined than the data on
cardiovascular events.
d. A fourth criticism that has figured prominently in the literature is that
the randomization did not succeed in allocating to the treatment groups patients
who were comparable with respect to base-line risk factors. Since we have had
access to the original data, we have been able to carry out an anlysis that was
designed to test whether in fact the differences in mortality in the tolbutamide
and placebo groups could be explained by the base-line differences. Our findings.
which are given in section 6, take into account the differences between centers
and the differences in length of treatment, as well as the base-line variables.
They support the view of Cornfield (17) that there is no evidence that the base-
line differences arising from the randomization contributed in any importanj~
way to the finding of adverse effects from tolbutamide.
5.4 FaiZure to adapt dosage of dr~gé to i~uUDkZnal t'teed
Feinstein (21) has noted that the oral drugs "were given in unsatisfactory
dosage to many people who did not need them," and others have made a similar
criticism. It Is true that the use of a fixed dose of drug, which was also the ap-
proach adopted by Feldman et al (19) mId I~een et al,(C) limits the generaliza-
tion that can be made about therapeutic effects, but since the dose of tolbutanilde
is about equal to the aIrerage recommended for therapeutic use, an evaluation
References at end of article.
PAGENO="0099"
COMPETITIVE PROBLEMS IN TI~E DRUG INDUSTRY 13349
of its possible toxic effect is highly relevant. Moreover, the prOblem of whe~ber
a $u~ject with mild diabetes who would not normally take any hy~oglycE~inic
drugs can avoiA some vascular compUcattons. pf the disea~e by doir~gso is one
that the trial was deslgueçj tQ iflum1~ate o~r sQive. We do not already. have the
answer, to this; what is understood well is that certain patients. re~ujre h~po-
gly~emic drugs for current needs. ~t is another question as to ~hetber these
patients, and those, with milder disease, can produce a . prophylactic e1~ect
against vascular abnormalities by taking .hypoglycemjcs In an attempt to m4iu-
tam strict control of thejr disease. ThIs is a matter for research and not for
the simple implementation of current therapeutic practice..
5.5 Disoontinncftion of tolbi~ttamid~ a4id phenformi,n in the TIGDP StlAdy
The action of the UGDP in `discontinuing the use of tolbutamide and p'hen~or-
mlii has been criticized by those who beUeve that the trial of these treatm~its
should have been continued in order to obtain more definitive results. It wo~i1d
have been easier to interpret the findings if there were more data on mortality.
We recognize that the pr~cise point at which suspicion of to~ictty outweighs
the need for scientific Information is uncertain and that the choice might have
been made differently by another equally qualified group of observers. Altbot~gh
we are not in a position to défen4 the timing of the UGDP decision in this
matter, it is clear that ethics would dictate that a decision abOut witbdra~ra'l
had to be made before all important questions concerning the effect of the drug
were resolved. We do not criticize the UGDP investigators for having made the
decision when they did. Nevertheless, the result of that decision is to leave us
with some residual uncertainty about the meaning of the findings, a point that
is well understood by the UGOP investigators themselves.
6. DATA rno~ TITE UGDP AND nEDFORD ThIALs
The directors of the tTGDP and Bedford trials have klndl3t' made available
certain data that we requested from them In order to review evidence concer~-
big the death rate of subjects taking part in contrOlled trials of oral bypogl~-
cemic agents. In the case of the IJGDP, the data of interest extehded to the
time at which the drug was discontinued. Events subsequent to that would cast
light on the effects, If any, of previous use of the drugs~-a question to whb~h
we do not propose to address ourselves. tn the case of the Bedford trial, data
are still being accumulated, and we have examined those available `up to June
1972. These must, of course, be regarded as provisional. In both trials the data
bear on many questions of great interest that we did not consider since they
had limited relevance, if any, to our charge.
A simple method of studying data from a long-term clinical trial is to estimaI~e
failure rates for various population groups. Failure may be taken to be an~~
adverse event; commonly, as in the present context, it is interpreted as deat]~.
The failure rate for a group after a certain length of follow-up is .the rate at
which the survivors are then dying. If the failure rate for a group is constant
throughout follow-up (so-called exponential survival), its value, Y, may be
estimated by Y=lc/t, where k Is the number of deaths in the group and t, tb~
number of persons-periods at risk, each subject contributing a survival perioc~
or, if death has not occurred, a period of observation.
Approximately, log Y may be regarded as normally distributed with a mear~
of mY and a variance of 1/k. S
The failure rate takes into account the length of time for which each subject
has been exposed to risk and can be made specific both for demographic char~
acteristics of the subjects and for risk factors of interest. In the present context
we have chosen a three-month period as an appropriate unit of time in ~alculat-
ing exposure to risk.
Simple and informative as the faIlure rates are in many cases, they become
unwieldy and increasi~giy variable as subjects are cross-classified j~ more and
more ways. We have therefore made use of the logistic model in order to carry
out a more detailed analysis of the UGDP and the Bedford data..
6.1 tTGDP data S S
In thIs section, we consider a problem relating to randomization and we pre-
sent our analyses based on failure rates and on the multiple logistk~ model, We
also report analyses designed to take into account the extent of adherence to
treatment. . S
PAGENO="0100"
13350 COMPETITIVE PROBLEMS IN ~HE ~D:RtjG INDTJSTRY
6.1.1 11ando.mi7.~ationby ser ij.,ithin clinics
In con~parIng i~iortality in the treatment groups' In relation to background
variabie~s, the ~J.~GD1? investigators presented data that showed `that the excess
ot cardiovascular deaths `in the tolbutamide group was. particutarly marked
among the females. The' mortality was 1O.6% in `the tolbutamide group as
against 2.1% In the placebo group; the corresponding `rates for ma1e~ were
17.5% and 11.1% respectively. In `the course of reviewii'ig this finding in the
individual clitiics we discovered a puzzling' anomaly concerning the distribution
of the two sexes to the fQur treatment grofips within clinics~
Table A.1 shows the numbers of patients of each sex allocated to each treat-
ment gronp within each clinic. The proportion of males all cated to placebo was
surprisingly high in Boston and in, Seattle. The discrep~'~ncy in Seattle alone
wOuld ~~resent an unusual event in random allocation X2~=1i.31 on `1 df; P
-.001) "and the results taken as a whole are also anomalous (X~=33.33 on 12
df; P-.001).
These unexpected findings do not in themselves explain the cardiovascular
mortality diffd~ences. In an analysis discussed later in this section, adjustments
are made for sex and clinic as welt as other covariables, and there is no sub-
stantial change in the apparent effect of toIbutamide treatment on cardiovascular
mortality. A more important point is whether thOse findings provide evidence
of a breakdown of the randomization procedure-a contingency that might have
grave implicatiolls for the credibility of the whole study.
The randomisation procedure used by the UGOP has already been described
briefly in sectIon 3.1.2. In an. attempt to find an explanation for the peculiar
allocation by sex within clinics, the committee reviewed the randomization in
detail. We were given access to the log books in which th~ `Coordinating Center
maintained records of the allocation of each patient to a treatment group and
were impressed by the quality of the documentation that the investigators
provided. We were not able to find an assignable cause for the surprising allo-
cation of the sexes to treatments but have no reason to think that the stud~r
has been compromised by a b]eakdown ih the randomization of patients to the
treatment groups. Because of the imbalance of sexes in the * * * however, allow-
ance for tbi~ has been made in our analysis. In institutions such as Seattle, in
which no cardiovascular deaths occurred in either the placebo or tolbutamide
groups, there would, of course, be no effect due to the imbalance. In general,
however, aU analyses of the data should be adjusted simultaneously for sex
and clinic.
6.1.2 Cardiovascular failure rates
Cardiovascular, failure rates in the TJGDP study are presented in T~hles A.2
through A.4. The rate for the tolbutamide group is 5.471,000,' quarter-yearS
(Table A.2, top) and this is significantly higher than the rate for the placebo
group. In the next two parts of Table A.2 the rates for the' treatment groups
are presented separately for the two sexes, and the differential bCtween the
placebo and tolbutamide groups is substantial and significant for females (4.4
vs 0.8) hut less marked and nonsignificant for males (7.5 vs 5.0). The number
of subjects at risk is smaller for the males than fOr the females, and the chance
of detecting treatment differences is therefore greater for the latter group.
The results are consistent with the view that the tolbutamide rate is higher
for both sexes, but if the males were considered in isolation, the evidence in their
case would not be strong. ~urther, it is the older women who in ,patticular
show substai%t~ally different rates ~or the two treatment groups. Athong women
ov~er 53 years of `age receiving . tolbutamide, the rate is 8.6~% and for those re-
ceiviiig placOJo, 1.4%' For younger women the' corresponding `rates are 0.6%
and 0.5%.
In Table, A.3 the failure rates are presented by sex ~thd treatment group
at each clinic. these are the data that have already been referred to in section
5.3 `to make the, point that excess mortality in the tolbutanhide group was `not
conflende to a feW `clinics. In the case of the females it was observed at the
clinics in Boston, Minneapolis, Williamson, Cincinnati, Cleveland, and Birming-
ham. Of the remaining six dines, four bad leSs than 200 quarter-years of pa-
tent eyposnre and showed. no deaths from' ,ca~d1ovascUlar causes in either treat-
ment gron~ ~ñ the part of th~ table showing data, for both. sexe~ combined
ft .i~ sqe~a ~that in seven clini~ the failure rate in the tplbutamide group. was
huigher `and in t*o It was lowei~ than `in the placebo group and that in three
there was no information.
PAGENO="0101"
.OOMPE~ITIVE~ ~RO~L1~MS IN ~ ~H1~ PRIJG INDV~T~1T l3aM
0.1.3 1kf~Ztipie Zbg~$tic moZeZ ` .
We have ü~ed the sair~e model as was employed by the UGDP Investigators,
in which th~ probability of death Was expressed as ~t function of the treatn~ent'
and of the. base-line variable$. We have, however, added additional variab1~s, 1
to take account of the time between enrollment of the subject and completion
of the study, and 11 to account for the inflaence of the clifltes. The' lengthy
list of variables that was assembled in this way is shown in Table A.5.A b~ief
account of the method of analysis based on the multiple logistic model is given
in appendix A. (26). The analysis leads to the findings reported in Table A,6.1.
As shown in the upper portion of Table A.6.1, the potential length of follow-
up, that is, the length of time from admission to the study to the time of a~ial-
ysis, proved, as one would expect, to b~ a highly significant predictor ol! cardio-
vascular death, the valt~e' of X2 on 1 df being 23.56. This variable was not in-
cluded in the analyses done by UGt~P. Many of the demographic i~arlables and
risk factors studied by the UGDP were also highly significant predictors of par-
diovaseular death. After adjusting for the UGDP variables, treatments, ~nd
potential length' of follow-up in the analysis, no additional significant improve-
ment was made by adding the clinic effects; ho~ver, the UGDP base-line
variables as a group still remained highly significant. It is worthy of note that
the clinic effects remained significant after adjustment `had been' made for
demographic variables and treatment. It was the `additional a~jUstiflent by
means of the variable length of follow-up' that reduced the clinic effects t~p a
nonsignificant level. Although the finding of clinic differences would not be ~ur-
prising, since they might be due' to differences in the patient' populations or
clinical practice, this indicates that most of the dif~erei~ees are ~expTained by
adjuattig for the different length of follow-up~
The most Important point in this analysis is whether or not the adjustments
for cOvariables could be responsible for the treatment differences observed. Qur
analysis indicates that the treatment effects ba~re been changed very little by
this adjustment. Tolbutamide treatment vs an `average' of other treatme~its,
~dju~ted for a subset of the demographic variables and time of potentie" foll&w-
up, showed a X2 of 12.14 on 1 df. In comparing the tolbutamide treatment with
the other treatments, it is apparent that this contrast accounts for almost `the
entire treatment Offect, and thus there is no significant `difference bet~ceen `the
insulin treatments and placebo.
When the data are presented separately for males and' female In the next
two parts of Table A.6.1, the comparison' of tolbutamide with the remaining'
treatments results in a 12 of 2.56 on I df ifi the case of males and 9.06 itt 1 df in
the ease of females. The effect of tolbutamide may further b~ c~mpared' with
the placebo treat'thOnt alotte in such a way `that the ,ixmnlin grcmps u'lso supply
information on the demographic variables.' Under these circwm~tanees, the 12
that tests for `thO adjusted effect of the tolbutamide treatment te `O~5 for fliales
and 11.70 for females. These result'S indicate that the effect of'tolbiif,amide treat-
ment is signifieayit in fethales but not in. mflles. `
Table A.6.3 shows that the coefficient `foi~ the tolbutainide 1~reafment' eft~et
is 2.1158 in females and 0.35~8 in males, and that the standard errors are
0.7094 and 0.4983; respectively. This implies, as noted before, `that the effect is
significant' only in females. The effeCt in females, however, is `hot significantly
different from that in males.
`The analysis by meafis of the multiple logistic modOl `confirms the princijal
finding' from the simpler study of failure' rates, namely, that the cardiovascular
death' rate `Was higher in certain patienth receiving tOlbutamide thafi in~ `th~se
receiving placebo. This result was definite in the case of femai~s; it "they wOll
be true also of males, but the" evidence iii that group is not st~tistically algntfi
cant. The multiple logistic analysis indicates that the difference `in death `rates'
remains after adjustment has been made for the effect of vario'i~ base-li~ie
variables and cardiovascular risk factors. `
6.1.4 AnaZysfs'with respect to adherence to essiyned frcatrnent ,` ,`
The UGDP' pi~otocol sp~ëified fl±ad doses for the pacebi, `tôlbutafni'de and `in-
sulin standard treatthents. AltOrntions' were permltted'only if `the patiebt Oeoi~ld
itot be safely niainl~aifled On the assigned"medicatiOn ~chedule,"~ ~1odification of
the dosage on th~'basis of elevated lbood glucose 1evel~ alone' was not permitt~d.
Adjustments of `the `dose `for the `patients taking variable-dOse ittsulin,~bowev~r,
fleferences at end of article.
PAGENO="0102"
J.3352 COMPEITITIVE PI~OflLEMS IN THE DRUG INDUSTET
~eould be made on the basis of the observed blood glucose ievel~ from the short~
times `of the glucose tolerance test. Since some patients did not adhere corn-
:pletely to the assigned treatment, they may have gone for periods of time
without any medication or with a modified. dose, or they may have switched to
`another therapy.
It is clear that the interpretation of the UGDP data could be Influenced by
such variation in the assigned treatment. The UGDP analysis and the analysis
discussed In the preceding part of this report are based on the assigned treat-
ments. In this section * * *
Caution must be taken, however, in the interpretation of these results. It Is
quite possible that adherence is related to bnportnt base-line or other unknown
variables for some treatment groups and not for others. If such were the case~
subgroups having a. particular pattern of adherence rnigbt not yield fair compari-
sons of treatment. The analyses presented In this section are designed to ac-
count for. the known base-line influences. owever, without the use of random-
ization to form treatment groups, there is no assurance that an unknown prog~
nostic variable is present that affects adherence patterns selectively for differ-
ent treatment groups and thus invalidatOs the treatment comparisons.
6.14.1 The ea~tent of the problem
Table 4 summarizes the number of patients who continued taking their
assigned treatment for the entire follow-up period, and the number who, for at
least one quarter, changed to other treatments or none. rlhus, for the 205 pa-
tients initially assigned to the placebo group, 76 (37%) continued receivi~g
placebo for the entire period of follow-up, and the remainder, had at least one
quarter of nonassigned treatment as follows: 1 (0.5%) received tolbutamide;
7 (3%), insulin at a variable dose; 92 (45%), no treatment; 4 (2%), tolbut-
amide and no treatment; and 24 (12%), insulin and no treatment. (One pa-
tient did not fit any of these categories.) An int.eresting point is that 168
(82%) of the patients initially assigned placebo were receiving either the pla-
cebo or no medication for the entire study. Since the initial treatment . groups
were assigned to their treatment by chance, these patients could be regarded
as representative of the UGDP patient population. Tbu~, over the average fol-
low-np time of 6.15 years, a very large proportion of the patients could be
maintained without medication.
Another consideration in evaluating the extent of the problem of adherence
is the proportion of follow~up time Individuals continued receivii~g their initial
therapy exactly as prescribed in the protocol, Table 5 classifies the patients by
the proportion of their total follow-up time spent receiving treatment initially
assigned. In `order to compare the extent of adherence of receiving standard-
dose insulin with the adherence of other patients treated with, insulin, a dose
modification of variable-dose insulin after the initial titratiop dose was regard-
ed as a "modification." Note that 26% (218/828) of the entire population were
100% adherers for the total follow-up period and some 23% of the patients
were receiving the initial treatment less than 50% Qf the total follow-up time.
Table 6 summarizes the total follow-up time (patient-years) with treatment
exactly as assigned. with the assigned treatment at a mo4ifled dose, and with
other treatments. Note that for each of the treatmetn groups, 14% to 16%
of the follow-up time was spent receiving no medication at alL Fprther, the
proportion `of foliow~~up time that patients spent receiving the fixed `dose of tel-
butainide was 58% and receiving the fixed dose of insulIn, 55%, It is interesting
that for 25% of the follow~.up period, the tolbutamide patients were taking
a dose other than that specified by the protocol; similarly, for 30% of the
follow-up period, the standard-dose insulin group was taking an altered dose of
insulin.
6.1.4.2 ~tatistieat analysia
The `statistical analysis of the TJGDP data in relation to adherence to treat-
ment is divided into two portions. The first pari uses a lionstandard method that
was developed for the ptoblem at hand and will be called the rek~tive allocation
method. It takes lute cOnsideration (1) time spent receiving no medlcati~n, (2)
time spent receiving ~nodlfled doses of the initially assigned therapy, and (3)
time receiving other than the initialb~ assigned' medication. The second. method
of analyss is called the Rnrljival `ntode7~in~j method and is based on techniques
recently developed by Cox (27) for modeling survival data when base-line
References at end of article.
PAGENO="0103"
COMPEPITIVE PROBLEMS IN THE DRUG INDUSTRY ~133~3
(concomitant) variables that affect the outcome ar~ present. Both ana~yses
led to the conclusion that women receiving tolbutamide ilaire higher total inor-
tality and higher cardiovascular mortallty than women receiving placebo. This
holds especially for the older women (over ~3 years of age).
Relative allocation method-The basic idea of this method is explained in
appendix B. The problem is to allocate the number of patients, and the de~tbs,
to the treatments when individuals have not been receiving the initially assign-
ed treatment for the full follow-up time. The method of relative alloc4tion
a ssigns numbers of both patients and deaths to the treatments in such a way
that they are proportional to the length of time the patients have been taking
each treatment. Suppose a subject has been in the study for ten years, ba~f of
them with the initially assigned treatment and half with no treatment. This
subject would contribute half an observation to each of these categories. If the
subject had died, half a des th would be allocated to each of the two categories.
The sum of these allocations defines an effective sample size, n', and an effe~tive
number of deaths, d'. We then define as follows:
death rate =0' =d'/n'
If time of follow-up is allocated in a similar way to the various treatm~nts,
and T is the total follow-up time for a subgroup, then
failure rate Y' = d'/T
Appendix Table A7d presents data on total deaths (d'), cardiovascUlar
deaths (d") and effective sample size (n') by initially assigned treatment ~nd
by treatment received; and in Table £7.2, the calculation of 0' and of ~` is
illustrated.
Table A.7.2 summarizes the death rates and failure rates corresponding to
assigned treatment without modification, treatment modified by change of dose,
and no treatment. The cardiovascular mortality associated with tolbutau~ide
is highest among the four assigned treatments, regardlcss of whether the tr~at-
meuts are pursued with modification ~r without. A comparison of the ca$io-
vascular mortality in the tolbutamide vs the placebo groups results in ~tatist~cal
significance at P~~015 (no treatment modification), P = .00 (doses chang~d),
and P=~.50 (no drug) ; using the Fisher test for combining tests of significatice,
one finds that the overall result is significant at the P=.007 level.
Since there appeared to be a randomization anomaly with regard to the allo-
cation of treatments with respect to sex, it is of interest to examine the eff~cts
of dose modification for each sex. Table A.7.3 summarizes the cardiovasci~lar
mortality by sex and dose modification. It is clear that the largest difference in
cardiovascular mortality between the placebo and the tolbutamide groups occurs
in the female group not having any dose modification (P=.004). A simi~le,
overall statistical analysis can be carried out on the mortality rates given In
Table £7.3 by ranking then for each of the six dose-modification groups (ro~Vs)
and assigning to treatments within a group the ranks 1 through 4. Since those
receiving tolbutamlde have the higb~st mortality in five groups and the next
highest in one group, this group has a rank sum of 5(4)±3=23; the rank sums
for the other treatments are as follows: placebo, l~; standarçl-dose insulin, ~2;
and variable-dose insulin, 13. The probability of obtaining a rank sum eqpal to
or higher than ~3 if there were no difference between the treatments is P~.O07.
(This probability is the ratio ~f the number pf ways of obtai~ipg a rank sum
equal to or greater than 2~ to the total number of posslbhhitie~, ie, 28(6(6) )/
(24) (6)). If one were to make a tWo~tailed statistical test, the P value woUld
be multiplied by 2; je, P~.O14.
Another way to analyze the effect of adherence is to partition the data 4c-
cording to both (a) dose modification and (b) whether the patients adhered to
the initially assigned drug for the complete follow-up period. Table A.7.4 sum-
marizes the cardiovascular death rates according to these twO variables. The
highest death rate is found In tb~ tolbutamide group who were 100% adherers,
and had no dose modification. The comparison of placebo vs tolbutamide for this
subgroup is significant at the P=.O03 level. The comparison of placebo vs t~l-
butamide in the ease of the other three subgroups is not significant. The cardio-
vascular death rate for the three tobbutamMe s~bgr~ups ~ho either did not
adhere completely to the medication or bad a dose modification is 7.6/9$.9=,08.
A comparison of this proportion with that for the subgroup that adhered CoUl-
pletely and had no dose modification (21) gave signjficance at the P=.04~2
References at end of article.
PAGENO="0104"
13354 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
level. This is in line with the view that if tolbutamide do~s indeed Increase the
risk of Cardiovascular death, taking less of it should lower the death rate.
It thus appears that a significant number of the cardiovascular deaths In the
tolbutamide group are associated with patients who took the drug for every
quarter of the follow-up period without any dose alteration. An attempt has
been made to examine this further by subdividing the groups involved. Table
A.7.5 presents data separately for males and females. Note that the relative allo-
cation of cardiovascular deaths to the female placebo group totals 2.5, (The
total number of cardiovascular deaths among females was 3, of which the rela-
tive allocation method assigned 1.2 to no drug treatment and 0.4 to insulin.)
A comparison of placebo vs tolbutamide in the group of females who had 100%
adherence and no modification of dose results in significance at the P-.01
level. None of the other comparisons of placebo vs tolbutamide are significant
at P-.05. The data cannot be meaningfully partitioned if all four treatment
groups are kept separate. One way to make a finer subdivision of the covariables
is to compare tolbutamide with a composite of the other three treatments. Table
A.7.6 summarizes the cardiovascular deat~i rates with respect to both sex and
age at entry, using the cut-off age of 53 years (53 years represents the median
age at entry). From this table, it is clear that there are too few cardiovascular
deaths in the younger women to justify any comparison of tolbutamide with the
control treatments. However, this comparison in the older women who adhered
100% and did not modify the dose is significant at the P=.03 level. There are
too few patients in the male subgroups to be able to detect real differences
by making simple comparisons, although in five of the six instances the talbut-
amide group has the higher rate.
SurvivaZ modeling method-The analysis in the preceding section was car-
ried out by simple partitions and comparisons of the treatment groups. To take
account of the effect of institutions, demographic variables, and base-line van-
a1~les, however, more sophisticated statistical methods are required. In this
Section the UGDP data are analysed by means of a statistical technique recently
developed by Cox (21) and modified. (28) The method relates failure rates to
both the treatments under study and concomitant variables (institutions, base-
line variables, and demographic variables).
The method, as used in this analysis, took into account the proportion of
time each patient was receiving the assigned medication, the time of treatment
with other protocol medications, and the time during which no medication
was used. (See appendix C for details.) A preliminary analysis was carried
out for both total and cardiovascular deaths by means of the following con-
comitant variables to determine which were important: 4 treatments, 1 variable
representing no treatment, the 14 UGDP base-line variables, sum of the initial
glucose tolerance tests, 4 variables aSsociated with interaction between treat-
merits (including no treatment), and 12 institutions. The results of the analysis
shoived that a mbdel could be used that included, in addition to the treatments,
7 base-line variables (sex, race, age, digitalis history, electrocardiographic ab-
normality, presence of irrterial calcification, sum of the glucose tolerance tests),
arid the :F2 Institutions. A final analysis was then done with the usc~ of a model
incorporating these variables. It was done Independently for males and females
to allow for dIfferential treatment responses, and was carried out separately
for total deaths and cardiorascular deaths.
The final results of the analysIs can be expressed as a ratio ut adjusted
failure rates of different treatments. The term adjusted refers to the failure
rates after allowing for the effects of the concomitant variables. The data can
be summarized by presenting the natural logarithm of the ratio of failure rates
and its associated standard deviation. Table A.7,7 exhibitS these quantities,
comparing each treatment ~vith the placebo. Also given is a compjirison of
placebo vs no medication.
The conclusion is that for women lii the. tolbutamide group, as compared
with placebo, there is an e~eessive mortality, both cardiovascular and total
deaths. The difference is more dramatic (P~.008 for the cardiovascular deaths,
although it is also significant for total deaths among women (P~.04). Thus.
this analysis supports the conclusions reached in the previous section that
tolbutanuide, as used in this study; produces an excess mortality of cardiovascu-
lar causes in women, when compared against placebo. The data do not support
the same conclusions for men, but one possible reason is that the smaller number
of patients in the male group results in lack of sensitivity to detect differences
of moderate magnitude.
References at end of article. . .
PAGENO="0105"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13355~
6.2 The Bedford trial
The data supplied information on both cardiovascular events and mortality.
6.2.1 Oardiovascular events
In the analysis of the data from th~ Bedford study the authors devote ma~ór
attention to "cardiovascular eventa." These have been described by Keen ~nd
Jarrett (6) as "a mixed bag" and Indeed do raise problems of classification~ in
that the events are not mutually e~i1~t~Ive and are ascertained partly by q~es-
tionnaire, partly by electrocardiographic evidence, and partly from mortality
data. They include reported episodes of angina and intermittent claudication,
and this could cause ambiguity, since reports of pain may be greatly infiuen~ed
by variations In the mood of the subject and in the style of inquiry. In view
of the lack of a formal procedure to ensure blind evaluation, the results of si~ich
analyses do not lead to firm conclusions.
6.2.2 Mortality data
These will be examined by two methods: the estimation of failure rates a~id
death rates, and the use of the multiple logistic model.
6.2.2.1 Failure rates and death rates
These two rates differ only in their denominator, which is person-periods in
the first case and persons in the second. The numerator in each instance is t~ie
number of deaths. The two rates are highly correlated for this set of data, since
there was little variation in the length of exposure to risk.
In Table A.8.1 the influence of. binary background variables on the death r~te
is shown. The rate is increased by hypertension, hyperglycemia, and arter~al
disease. It is higher for females than fQr males and higher for those over 45
years of age than for younger patients.
In Table A.8.2, placebo and tolbutamide treatments are compared, taking into
account the background variables one or two at a time. In no case is the~e
adequate evidence of a difference between the two treatments.
In Table A.8.3, the death rates (by treatment group and sex) are given fpr
more finely divided age groups. The reason for this is that as shown in Table 3,
the proportion of older subjects is higher in the placebo groups, and age is~
therefore a potentially confounding variable in the comparison of treatmeht
effects. Owing to the relativ.~1y small numbers of subjects in the individual age
groups, the rates are somewhat irregular. The effect of age is marked, but the~e
is no evidence of a difference between the treatment groups.
6.2.2.2 Multiple logistic model
In this trial, all the patients were entered Into the study at essentially two
different points in time and not over a period of time, as in the UGDP. rllhis
feature enables some simplification of the analysis of the data. To adjust for
the differences in the length of follow-up~ and possibly for other differences as
well, between those who entered at the two different times, an indicator variable
was included in the logistic analysis to distinguish the two groups of subjects.
The variables included in the logistic analysis of these data are given in Table
A.9.1.
The results of the analyss Of the fledford study deaths due to cardiovascular
causes and deaths due to all causes are shown in Table A.9.2. The variances
introduced in this analysis were significant in predicting death either from all
causes (X2'=81.02 on 5 df) or from cardiovascular causes (X2 = 55,16 on 5 df).
Adjusting for these variables, however, did not change the basic conclusio~i
reached from the unadjusted analysis that the death rates did not differ sig-
nificantly according to treatment.
7. CONCLUsIONS
In this section we summarize our overall findings of the TJGDP study with
respect to the protocol, ~he conduct of the study, the methods of analysis, an~I
the findings.
7.1 Protocol
Question. Was the target population lor this study an appropriate one?
Answer. Critics have pointed out that certain patients were required to ac-
cept treatment that would not normally conform to clinical practice, and their
References at end of article.
PAGENO="0106"
13356 COMPETITIVE PROBLEMS I]~ THE DRUG INDUSTRY
argue, therefore, that the target population was unsuitable. Such a claim, how-
ever, overlooks the important but ili-understo~d prophylactic aspect of the
trial, in which certain treatments were given to patients who, initially at least,
could safely go without drugs, in order to test whether the common vascular
complications of diabetes could be prevented. The issue was the testing, of Cer-
thin possibly preventive treatments rather than the implementation of certain.
standard therapeutic regimens.
Question. Was the decision to in~lude phenformin in the study justified?
Answer. In the e\rent it proved to be, since valuable information was obtained
about the limitations of that drng. Its use, however, greatly complicated an
already difficult study. It is clear tl~at one of the problems of a long-term clinical
trial iS that potentially interesting therapies may develop while the trial Is In
progress, and the natural desire to include them may divert resources.
The omission of a history of smoking was a blunder.
7.2 Uond'uet of the study
This was necessarily a lengthy and complex trial, and a substantia1~ pioneer-~
lug effort was needed to mount it succe~fully. We have raised a question of
whether the randomization was properly carried out. The only evidence that it
might not bi~ve been is the data on the allocation of treatments according te
the sex of the patient. Against this possibility are two * * *
7.3 Methods of anaZysis
The UGDP investigators sought to examine their data from a number of
different points of view, and in so doing tbe~ made use of some relatively un-
familiar and exploratory statistical techniques. In some cases these methods
would not necessarily have been chosen by Other grottps of statisticians faced
with the same situation, but since the results of all the analyses tended to point
in the same direction, there would be little advantage in discussing at length the
weight to be attached to the different analyses.
The likelihood calculations seem to us to ttdd very little to the other analyses.
The results are rather difficult to grasp and require rather arbitrary weighting
to be given to the likelihood of different hypotheses. The method does not take
concomitant variables into account.
The Monte Carlo monitoring procedure was a major attempt to overcome
the selective effect of a secjuential analysis of the mortality data. The investiga-
tors were concerned lest they had paid undue attention to contrasts between
treatments at a particular moment when extreme fiu~tuatlons might have oc-
curred. Their method was ingenious, and although minor points of criticism may
be raised, we do not think that these materially affect the issue. (Some of these
points might be (1) the use of national mortality data, with death rates higher
than those in the study population; (2) the use of an `~average1' survival curve
for all patients in the sImulation; (3) the adding of life table death rates at
different ages to obtain the death rates during intervals; and (4) the arbitrari-
ness of the linear boundaries. 1~or an alternative approach to the sequential
analysis of survival data, using internal comparisons only, see Breslow and
Haug. (29)) The detailed outcome of such a monitoring procedure is of no
great importance. The decision to stop the use of tolbutamide must have depend-
ed on considerations of various sorts, among which the monitoring procedure
provided a contribution-no more than that.
The UGDP did not try to determine whether interactions were present in their
data. This criticism was raised by Feinstein and is valid.
7.4 Findings
Although we have concerned ourselves almost entirely with Issues related
to the possible toxicity of tolbutamide, we wish to point out that one of the
valuable aspects of the completed TJGDP trial will be the provision of data on
the long-term treatment of adult-onset diabetes with insulin, It is already clear
that the benefits from this treatment are not dramatic, and the only worthwhile
information about them will have to come from the relatively precise methOds
of a controlled clinical trial. In this sphere, the TJGDP trial has no competitor.
Indeed, we would generalize from this and point out the * * *
On the question of cardiovascular mortality due to tolbutamide and phen-
formin, we consider that the UGDP trial has raised suspicions that cannot be
dismissed on the basis of other evidence presently available.
We find most of the criticisms levelled against the UGDP findings on this
point unpersuasive. The possibility that deaths may have been allocated to car-
References at end of article.
PAGENO="0107"
`cOMP~EnTIV~ PRO1~LRMS IN TEE DRUG IND~lSTRY 13~57
~liovaseular causea preferentially in the groups receiving `oral therapy e~dsts,
and, in view of the "nonsignificance" of differences in total mortality, some
reservation about the conclusion that the oral bypergiycenaics are toxic *flust
remain. Nonetheless, we consider the evidence of harntftlness moder~Lte1y
strong. The risk Is clearly seen in the group of older women `a~ shown in ~able
A.2. Whether it acects all subgroups'. ~f patients cannot ber decided on the basis
of the available data. owing to the small number of deaths involved in these
subgroups.
There remains the question of generalization of these findings. A's has been
frequently pointed out, the conditions of drug use in this study were, to some
extent, abnormal. Tolbutamide dosage is varied in practice, aiid the pa1~ient
unable to maintain adequate `control with lolbutamide could be shifted to in-
sulin. A good deal rests, then, on the matter of whether tolbutamide is act~ally
toxic~ If this should~ be admitted, it `is hard to see. how it could be regai~ded
as a reasonable therapy, even when given in variable rather than axed dosage.
If, however, this finding is rejected, there remains the question of whether
tolbutamide, although ineffective in this fixed-dose regimen, might be an effective
therapy as ordinarily used. The UGDP gives up direct answer to this question,
but the dose of tolbutamide ordinarily employed varies only moderately.'
There is also the question of the extent to which the UGDP subjects rea~on-
ably represent the population of maturity-onset diabetics who. are candidates
for oral therapy. Little of the . commentary available to us raises questions on
this point, and we assume that the UGDP population is representativeS o~ a
large fraction of the maturity-onset, non-Insulin-dependent, diabetic populat~on.
In conclusion, we consider that in the light of the UGDP findings, it remains
with the proponents of the oral , hyperglycemJ~cs to conduct scientifically ade-
quate studies to justify the continued use ~f such agents.
REFERENOnS
(1) A study of `the effect's of hypoglycemic agents' on i~ascular compllcation~ in
patients with adult-onset diabetes: I. resign, methods and baseline results, Uni-
versity Group Dia'betes Program. Diabete$ 19(suppl 2) :747-783, 1970.
(2) A stu'dy of the effects of hypoglycenile `agents on vascular complications in
patients with adult-onset diabetes: II. Mortality Results, Universlty Group
Diabetes Program, Diabetes 19(suppl 2) :7~7-830 ;l970~
(3) Effects of hypoglycemic agents `on vascular complicationa in patients with
adult-onset diabetes: IV. A preliminary r~~ni~rt on nhenformln results, Univers~ty
Group Diabetes Program. JAMA 217:7.77-784, 1971.
(Jr) A ;~tudy of the effeets of hypoglycemic agents on v'ascular complications In
patients with adult-onset diabetes III. Clhiical implications if UGDP results,
University Group Diabetes Program. JAMA 218:1400-1410, 1971.
(5) Keen H: Factors influencing the progress of atherosclerosis in the diabet~c.
Aeta Diabetol Lat 8(suppl 1) :444r-4E6. 1971.
(6) Keen H, Jarrett RJ: The effect of carbohydrate tolerance on plasma lipids
and atherosclerosis in man, in Jones' 113' (ed) : Atherosclerosis. Berlin, Spr~ngér-
Verlag, 1970, p 435.
(7) Bull JP The historical development of chinièal therapeutic tri~'l's. J Chron
Dis 10:218-248, 1959.
(8) Burdette WJ, Geh'an BA: The Planning and Anal~ijsis of Clinical Studi~s
Springfield, Ill, Charles C Thomas, Ppblisher, 1970.
(9) Hill AB: Statistical ilfethods in Clinical and Preventive Mediç~ine. Londo
Oxford University Press, 1962.
(10) Lasagna L: The controlled clinical trial: Theory' and practice. J Citron
Dis 1 :S58-367, 1955.
(11) Witts U: Medical Surveys and Clinical Trials, ed 2. London, Oa~fo~d
University Press, 1964.
(12) Zubi~od C'G: Multiclinic `trials in cancer chemotherapy. Can Med Assoc ~T
~7 :101-103, 1967.
(13) Beecher HE: Research and the Individual. Bost'o~, Little, Brown & CQ~
1970.
(14) The Institutional Guide to DREW policy on protection of human subject~~.
National Institutes of Health, US Dept of Health, Education, and Welfare.
PAGENO="0108"
13358 00 ErnTIv~ PROBLEMS IN TI~E DRUG INDUSTRY
(15) Cochran WG, Cox G: Ecoperimental Designs, ed 2. New York, John~ Wiley
& Sons, 1957.
(16) A Million Random I~iyLts With 100,000 Normal Deviates, Rand Corpora-
lion. Glencoe, Ill, The Free Press, 1955.
(17) Cornfield J: The University Group Diabetes Program: A further statis-
tical analysis of the mortality findings. ,TAMA 217:1676-1687, 1971.
(18) Paasikivi J: Long-term tolbutamide `treatment after myocardial infarc-
tion. Acta Med Scand, suppl 507, 1970, pp 1-82.
(19) Feldman 11, Crawford D, Ela'shoff R, et al: Progress report on the prophy-
lactic use of oral hypoglycemic drugs in asymptomatie diabetes: Neurovascular
studies. AdvMetab Disord 2(suppl 2) :557-567, 1973.
(20) T'aagournis M, Re3~nertson R: Mortality from coronary heart disease dur-
ing phenformin therapy. Ann Intern Med 76:587-592, 1972.
(21) Feinstein AR: Olinical biøstatistics: VIII. An analytic appraisal of the
University Group Diabetes Program (UGDP) study. Olin Pharmacol 12 :167-191,
1971.
(22) Schor S: The University Group l)iabetes Program: A statistician looks at
the mortality results. JAMA 217:1671-1675, 1971.
(23) Seltzer 118: A summary of criticisms of the findings and conclusions of the
University Group Diabetes Program (UGDP). Diabetes 21:076-979, 1972.
(24) O'Sullivan JB, Mahan CM, D'Agostino RB: Critique of the UGDP mor-
tality analyses. New York, Academic Press, to be published.
(25) Prou't TE, I~natterud GL, Meinert CL, et `al: The TJDGP controversy:
Clinical trials versus clinical Impressions. Diabetes 21 :1035-1040, 1972.
(26) Cox DR: Analysis of Binary Data. London, Methuen & Co, Ltd, 1970.
(27) Cox DR: Regression models and life tables (with discussion). J B Statis-
tical Soc Br 34:187-220, 1972.
(28) Kal'b~eisch JD, Prentice RL: Marginal likelihoods based on C~nx's re-
gression and life model. Biometrika 60:267-277, 1973.
(29) Breslow N, Hang C: Sequential comparison of exponential survival
curves. J Am Statistical Assoc 339 :691-697, 1972.
APPENDIX A
Use of the logistic model ~,.
The logistic model has bten recognized as being very useful for studies in which
there are only two outc~ries,. for example, death or survival (-6) In this use of
the modelit is assumed that the probability of death, P, depends on m independ-
ent variables, Xi, X2, * * ~`, Xp, according to the relation
~ 1
where
Ab0+ b1X1+ b2X2... +bmXm
On each subject in the UGDP trial, the data available were the m independent
variables and an outcome variable that was given the value 0 or 1 according to
whether the patient survived or died The multiple regression equation was
fitted to relate the probability of death to the independent variables A maximum
likelihood procedure was used to find estimates of the regression coefficients
b0,b1,. . . bp
Groups of people such as those receiving a treatment or those from a particular
clinic were incorporated into the model by the inclusion of an indicator variable
that, for a given individual, took the value 1 if the indivudual was in that group,
and 0 otherwise In order to avoid redundancy, there must be one fewer variable
for clinics than there were clinics, and so for other sets of categories To allow
for the varying lengths of follow-up, potential length of follow~up (ie, the length
of time between entry into the study and the end of the study) was entered as a
covariable in the regression.
As a test of the various covariables in the logit regression, the likelihood ratio
x2 was computed. The likelihood ratio x2 can be computed for a set of parameters,
1~, by comparison with a set of parameters j3'~' to which, under the null hypothesis,
PAGENO="0109"
COMPETITIVE PROBLEMS IN TIlE DRUG INDUSTRY 133E~
p constraints have been applied If the maximum likelihood e~tin1atos of th~ set
of parameters are j3 and 13* respectively, then -2 in [L(13*)/L(13)1 is asym~toti-
cally distributed as x2 on p degrees of freedom, where L(.) denoted the likeli~iood
function When investigating the regression coefficients them~e1ves, one can use
the fact that they asymptotically have a' multivariate normal distribution with
a variance-covariance [1(13)1_i where
`(13) mXm
This en~I~les one to, obtain estimates of the `standard deviation of the estin~ate~
of the regression par~~eters~
Table £5 gives a lint of the yariables tl~at were considered in the analy~s of
the TJGDP data and Table AA~1 summarizes the fmndings,on them. `The es~ates
of the regression coefficients when ~d1 of these, variables have been includec~ are.
given in Table £6.2. As often happens when one does multiple regression Mth'
many parameters, there are redundaucie~, ~o that a relatively small subset o~ the
variables gives nearly as good a prediction a's the' entire set. In looking fo~' an
appropriate subset of variables, one still Includes the variables that are of g~eat-
est interest, in this case, the treatment effects. Sex was also included becau~e of
an interest in the treatment effects for dach sex. It has already been noted that
after adjusting for treatment variables,, demographic variables, and t1m~ of
follow-up, the clinic differences were not significant, and so the variables for~
clinics were dropped. Other demographic variables ~were added to the regression
until the maximum of the likelihood did nut differ significantly from the maxi-
mum of the likelihQoe when all the' ~ariablès were 1nclud&~. `The Ord~r in w~iich
the var!âbies tvere entered' into the regression depended `on `th~ absO1ut~ value Of'
T (see Table A.6.2), the large values being entered first. The subset `of v~rla~blev
thus identified (age, sex, systolic biood,pressure, electrocardiographic abnor~nal-
ity, cholesterol level, and arterial calcification) is indicated in Table AS and
these were used in the further analysls~ A regression analysis with tbIs'subs~t ~f,
variables other than sex was also done separately for each sex. ~1'he results o~ the
analysis using the subset of variables are sumtharized in Table 9.6.1, as we~I as
in Table A.6.3,. ~his analysis indicated that the treatment effects may be different
from the two sexes' The harmful dfeet of tothutamide treatmeilt ~`5 ihost appafent
for women, although the effect for men, is not significantly different from tlia~io~
women. It is not clear whether thC'~eskT1th fth~ tolbutamide apply only to women
or whether the effect on won~en is were obvious because of the larger num1~ers.
involved. , ` `
Table A.6.4 g1s~es the number of deaths observed in each' treatment gr~mp,
broken down by age and se~ `along with th~e,number e~pected OR: ~ ~J! tl~
model using the subset of yariable~ just mentioned. It appears that the m?d'el
does reasonably well in' predict'ihg tbe~niimber of death~',iil each group.
The variables used in the analysis of the Bedford study data are showr~ in
Table A.9.1. The a~aly~is was done for all cáüses"and `cardio~ascular' cause~ of
death;' Ond~the rO~ulf a are ~umlnarized in Table £9.2. The regression eoeffici4nts
ol)fained when all the variables are included are shown in' Table £9.3.
APPENIMX B
Relative allocation method , `
In this section we outline th~ rudimnEthts of"the `re'ative allocation (RA) method
of analysis. Define
~S~"= Total fol1ç~w-up time f~r the at~~ incUvi dua~ (a= 1,2,..., u); `.~
1S~= Total follpw up time for the ç~th indIvidual ç~t~ the ~ lrea~ment (m=
0,1,2,3,4); ` ` ` ` ` " "
wj~=Sja/&~B~Oldtivetime On i~h ~tteatment~fo .~iiy'qa~;
öa f1 if ath individual is dead (or if cardiovascular death) `
~ 0 otherwise. `` ` ` `a",
We shall denote the treatments placebo, tolbutamide, standard-dose insthin,
and variable-dose insulin by the sub~cr~pts i~ 1,2,3,4, and i=0 will refer to no
treatment. Therefore, the 0th patient ih the study supplies the vector ~f in~o~m-~
tion (2a,Wia,öa)j~ 0,1,2,3,4.
"1)he~relative allocation `number of deaths for the i~ trOntmen.t is deflne4 ~y
d,' ~
PAGENO="0110"
i3360 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Similarly the effective number of observations for the ~t1~ treatment is
The proportion of deaths for the jtI~ treatment is theh estimated by'
The ~} can be defined to take on the value unity depending `on i~hethër one is
calculating mortality for total deaths or cardiovascular deaths.
The estimate of the proportion of deaths associated with a treatment is usually
ddne for a subset of patients according to whether the patients belong to the sub-
set or not. If' C defines such a class, then the 1~A deaths and sample sizes are
d~"(C)=~ ~
flj'(C)=~Wja `
d~'(C) /n~' (C)~
If there isno difference between the treatments for patients b~iOhgh~ to a
particular class, the `probability of dying while hi the study does not depend on
the treatment, ,ie,
P{~a(C)1}O(C), .
*here aEC. Consequently (conditional on the' ~w~} ~being fixed), we have
and
*~sar tO~'(C) -o~'(C) ~`2 {A~~~(C)'± A~1.(C) -2A~,(C) }.,
where `
O ,~,2~e(C)(,j_O(C))
=~ wjawja/[nj'(C)nj"(C)'l.
a~C
Approximate tests of ~ignificance can be made by taking
``[O~' (C) -oj' (C) ]Nvar'~O~' (C)'~-O~' (C)1 `
tie have a standard normal distribution. *. . . . " ;" `~
In an~analegous ~ay the failuro rate For `the ~ treatment and p~tie,ntsbe~onging
to class Cis defixicdby . .
x~'(C) =~o~'(C)/t~(C),
where
t~(C)'E ~,/n~'(C)
a C
is the .~ssociated average follow-ii~p time. Thus~ conditional on' ~he average
~olIow-u~ tiinn, the variance of a differen~e between two failure rateC is
var tX~'(C) ~-X~'(C') =e~?'{B~~(C) ±B11(C) -~-2B~1(C) },
~where . 0 0
B~1(C) = A~1(C)/[t~(C)t1(C)1.
41~PuNDIx c *` 0 0
M~1üz~ ikW~tIvod `.
The thodei used Thr the survival .n~oduling method expresses the lo,garitba~ of
~the failure rate for the a patient as
log ~~=log x(t)+~1 Xsaa'~l, 2, .. . fl,
PAGENO="0111"
COMPETITIVE PROBLEMS J~ TI~ DR~JG INDUSTRY 1,3$61
where X (t) and {~3~} are unknown parameters to be estimated and {x,als== l,~,...,
p and p covariables associated with the ath patient. The x-variables for the four
protocol treatments were taken to be equal to the proportion of time the a~~'
individual was on the particular proto~o1 treatment. That is, if the first fo~ir ~3
coefficients refer to the treatments in t1~e stt~ndard order, then X~=WSa for
s=l,2,3,4. If a patient receiving standnrd-~do~e insulin had an altered dose of
insulin, this was regarded as contributing jnformation to the variable-dose insulin
group.
The estimates of ~ (s~ l,2,a,4) corre~pon4 to the logarithm of the ratip of
the failure rate of the 5th treatment to that of the period for which no medication
is taken. The differences ~ estimate. the logarithm of the ratio of the 5t~i
treatment to placebo. These ratios are "adjusted" ratios that have been adjusted
for base-line and demographic variables as well as institutions. The mo d~l is
based on the work of Oox, (p7) utilizing a modification suggested by Kalbf1e~sch
and Prentice. (p8)
REFERENCER
TABLE 1.-UGDP STUDY: TREATMENT GROUPS AND CAUSES OF DEATH
Percent dead
Number All Cardiova~cuIar
Treatment group of subjects causes c~uses
~
Clinics using tolbutamide:
Placebo 205 10.2 4. 9
Tolbutamide 204 14. 7 12. 7
9. 5 6. 2
Standard dose insulin 210
Variable dose insulin 204 8. 8 5. 9
Clinics using phenformin:
Placebo 64 9. 4 3. 1
Phenformin 204 15. 2 12.7
8.8
Standard dose insulin 68
4. 6
Variable dose insulin 65 6.
TABLE 2.-BEDFORD STUDY (6): CARDIOVASCULAR EVENTS AND TOLBUTAMIDE
Placebo . Tolbutamide -
With event With event
N~mberof~-- . -- Numberof - -1--~
subjects Number Percent Subjects Number Pertent
~
Both risk groups:
Both sexes 123 . 46 37. 4 125 34 7.2
62 19 0.6
Male 69 ~5 36.2
Female 54 21 38.9 63 15 3.8
High risk group:'
Both sexes 34 39 55.9 41 21 51.2
Male 14 10 71.4 14 11 78.6
Female 20 9 45. 0 . 27 10 ~7. 0
Low risk group:
Both sexes 89 27 30.3 84 13 15.5
Male 55 15 27.3 48 8 16.7
Female ~4 12 35.3 36 5 13.9
~
1 The high risk group consisted of those who, on coming iritothe trial, had cIearolink~al evidence of cardiovascular dis~ase
or clinically signiticant hypertension.
TABLE 3.-BEDFORD (5) AND UODP $TUDI~S: AGE DISTRiBUTIONS
Bedfor
d study
UGDP
study
Placebo
Tolbutamide
---
Placebo
-
Tolbutamide
-~-, -~-
Age Nun~ber Percent
Number Percent
Number Percent
Number
Percent
20 to 29 7 5,6 6 4.9 7 3.4 3 1.5
30 to 39 7 5.6 12 9.8 25 12.2 23 10. 3
40 to 49 22 17.6 22 17.9 51 24. 9 45 2Z. 1
50 to 59 25 20. Q 34 27.6 60 29.3 74 36.3
60 to 69 27 21.6 27 22.0 46 22.4 51 25.0
70 to 79 30 ~4, 0 20 16. 3 16 7. 8 8 3. 9
80 and over I ~.6 2 1.6 0 0 0 0
All ages 125 100.0 123 100.0 205 100.0 204 100.0
Mean age 58.6 55.4 52.2 53.0
PAGENO="0112"
8
-3
TABLE 4.-NUMBER OF PATIENTS BY ASSIGNED TREATMENT AND BY TREATMENT SUBSEQUENTLY USED FOR AT LEAST ONE QUARTER
0
- Standard dose Variable dose Tolbutamids and Variable dose insulin
- Placebo Tolbutamide insulin insulin None none and none
Number -Percent Number Percent Number -- Percent ~Number Percent Number Percent Number - Percent Ptumber Percent
Subsequent treatment -
Assigned treatment
Placebo
Tolbutamjde
Standarddoselnsulin -
VariaLs1~doseinsulin
~
Total
;
-
- 76 37
0
0
0
-
1 0.5
106 52.0
-~1 0.5
0
0
0
- 50 24.
0
7
8
51
82
3
4
24
40
92
75
104
119
.
45
37
50
58
-
4- 2 20 -. 12
G...~ 15 7
4 .2
- 3
1204
204
210
204
!~
~
.3
~
~
~
~
76 9
-
108 13
50 6
-
148
18
390
47
-
11 3 - 35 5
-
-
822
--- -
I One patient omitted. Did not fit
any of these categories.
-
.
-
-
0
C12
PAGENO="0113"
-4
Go
TABLE 5.-NUMBER OF PATIENTS BY INITIAL TREATMENT I AND PROPORTION OF FOLLOWUP TIME Ot~ THAT TREATMENT
0
L~J
Proportion of total fol
lowup time on initial treatment
.
-
~.
~
0
001 to
0.25
026 to 0.50
0.51 to 075 0.76 to
O~99
1.
11
Initial Treatment
Number
Percent
Number
Percent-
Number Percent
Number Percent -- Number
Percent
Number
Percent
Total
02
Placebo
Tolbutanilde
Standard dose insulin
3
3
6
1.5
1. 5
3.0
19
18
21
9
9
10
19 9
17 8
23 11
21 10 79 39
15 7 74 36 -
41 20 &~ - 33
64
77
50
31
38
24
205
204
210
.~+
Z
Variable dose insulin2
3
1.5
23
11
35 17
47 23 69
34
27
13
204
Total
15
1.8
81
10
94 11
124 15 ~291
35
218
: 26
823
1 Without dose modification.
2 Figures reflect proportion of time on initial-titration i
nsulin dose.
-
-
-
-
Q
z
02
PAGENO="0114"
13364 COt~L'I~IVE PROBI~M~ I~ T~ DRTJ~ INDUS~RT
TABLE 6.-PATIENT FOLLOWLJP TIME (PATIENT-YEARS) BY ASSIGNED TREATMENT AND TREATMENT RECEIVED
Assigned treatment
Standard dose Variable dose
Placebo Tolbutamide insulin insulin Totals
Per- Per- Per- Per- Per-
Treatment Received Number cent Number cent Number cent Number cent Number cent
As assigned 693.8 65.0 712.6 58 709.3 55.0 5 304.0 24.0 2, 419.7 48
Other treatment: 2
Placebo 319.3 25.0 319.3 6
Tolbutamide 3.3 ..3 3044 25 4.6 .4 3.3 .3 315.3 6
Variable dose insulim - - 43.3 3.0 32.9 3 389.0 3Q.O 3 789.8 62.0 1, 255. 1 25
None 198.6 16.0 180.9 15 198.0 15.0 177.4 14.0 754.8 15
Total 1, 258. 3 1, 230. 5 ~, 300. 9 1, 274. 5 5,064.2
I This figure represents total patient years on initially selected dose of insu'in.
This includes assigned treatment but with modification of dose.
* This figure represents total patient years on with insulin at a dose different from the initial dose.
PAGENO="0115"
COMP~TI9rIVE PROBLEMS IN T~D~ DRUG I~DTJ~TRY 1~336~
)-
~ 0)+~ ><
~ ~
~ II
2
0)
00
0~
2 ~ 00Q~ 00
2
~
~ 1:~0~! ~ ~
C,)
~
2 t)~0r~ooo,
p
C,)
~ =)00_~~ C~
~ I)) C')~I*~ `~)c~J
LU
2
G) ~ - L)~C0C~J P~
U.)
2 ~co
~
U.) CC)
2 )~0) 00~~C~
><
>-
00
2 Q~C~JI)CC)o0
.~!
00
C)
H'''''
CO
a)
00 ,,,, ```
>
PAGENO="0116"
13366 COMPI~TITIVE PROBLEMS IN THE DRUG INDTJBTRY
TABLE A.2.-UGDP DATA: CARDIOVASCULAR FAILURE RATES FOR EACH TREATMENT GROUP
Treatment group
Number of N
patients
Total time
umber of at risk
taUures quarter-years
Fai~ure rate
(in thousandt)
Both ~exe~
Placebo
Tolbutamide
Standard dose insulin
Variable dose insulin
Males:
205
204
210
204
10
26
13
12
5, 033. 6
4, 922. 2
5, 203. 9
5, 098. 1
2. 0
5. 3
2. 5
2. 4
Placebo
Tolbutamide
Standard doseinsulin
Variable dose insulin
Women:
63
63
57
46
7
11
5
2
~
1, 389. 2
1, 491. 5
1,340.0
1,137.5
5. 2
7. 4
3.7
1.8
Placebo
Tolbutamide
standard insulin dose
Variable dose insulin
142
141
153
158
3
15
8
10
3,644. 4
3, 430. 7
3, 863. 8
3, 960. 6
. 8
4. 4
2. 1
2. 5
TABLE A.4.-UGDP DATA: CARDIOVAS~ULAR FAILURE RATES FOR EACH TREATMENT GROUP BY AGE,
AND SEX
Treatment Group
Number of
patients
N
umber of
tai~ures
Total time at
risk, quarter-
years
Failure rate
(in thousands)
Males:
«=53~rold;
Placebo
Tolbufamide
Standard dose insulin
28
26
28
1
5
0
664. 6
656. 6
722.5
1. 5
7. 6
0
Variable dose insulin 2~
«=S3yrold:
Placebo 35
Tolbutamide *31
Standard dose insulin 29
Variable dose insulin 25
~53yrold:
Placebo 85
Tolbutamide 71
Standard dose insulin 7~
Variable dose insulin 82
«=S3yrold:
Placebo - 58
T~lbutamide 70
Standard dose insulin ~ ~74.
Variable dose insulin_ .~_ 76
.
~
~
,
~
~
1
~
6
6
5
1
1
1
1
1
2
14
7
9
521. 7
~
724. 5
834. 9
611.6
615. 8
~
2, 187. 8
1,793,8
2,048.6
2, 118.9
~
1, 456. 6
1, 636.9
1, 815. 2
1, 841.7
1. 9
8. 3
7. 2
8.1
1, 6
. 5
.6
.5
. 5
1.4
8.6
3. 9
4. 9
PAGENO="0117"
COMPETITIVE ~ROBI~EMS `IN TUE DRUG INDUSTE~ 133~7
TABLE AS-VARIABLES USED IN THE LOGISTIC ANALYSIS OF THE UGOP DATA
Time: Length of thee from admission to study to time of analysis
Treatments:'(Coded 1 if the patient was assigned to the treatment and 0 otherwIse).
Tolbutamide~
Insulin (standard-dose).
Insulin (variable-dose).
Demographic variables and risk factors: ("demographic variableg").
Age.2
Sex (1=male, 2=female).
Race (1 white, 2=nonwhite).
Relative body weight.
Systolic blood pressure.2
Diastolic blood pressure.
History of use of digitalis (1=yes, 2=no.).
History of angina pectoris (1=yes, 2= no).
Significant electrocardiographic abnormality 21 (1=yes, 2=no).
Serum cholesteroLa
X-ray evidgnce of arterial calcification 24 (0=no, 2=yes).
Fasting value from baseline glucose tolerance test.
Serum çreatinine value, mg 100 ml.
Visual acuityforboth eyes(0= >20/200 for both eyes; 1 = «=20/200foreither eye).
Clinics:' (Coded 1 if the patient was in the clinic and 0 otherWise).
Boston.
Minneapolis.
New York.
Williamson.
Cincinnati.
Cleveland.
Baltimore.
Birmingham.
Chicago.
St. Louis.
San Juan.
1 There is 1 fewer treatment variable than there are treatments, and 1 fewer clinic variable than clinics. This avoids
redundancy; in effect, the treatments are compared to the missing treatment (placebo) and the clinics to the mis~ing
clinic (Seattle).
2 These variables constitute the subset referred to in tables A. 6.1 and 6.3. Sex was not included as a variable ifl the
analyses done separately for each sex.
a Major or minor Q-waves, S-I depression, T-wave inversion, complete heart block, teft bUndle-branch bIdcK,~or yen-
tricular tachycardia.
4 Evidence of arterial calcification noted in both Of 2 independent readings of the same set of soft tissue X-rays of the
right lower limb.
PAGENO="0118"
13868 cO~fl'ETITIVE P~ftOBLEMS I~ TH~ flRUG ~NDUSTEY
TABLE A.6.-ANALYSIS OF CARDI~VASCULAR DEATHS IN1 IH6 IJGDP TRIALS USiNG THE LOGISTIC MODEL
Likelihood
df ratioXa p
1
a
3
14
3
11
3
3
11
11
14
1.
2
6
19
3
1
2
1
2
706.03 .001
23.56 .00].
10.46 .02
10.68 .02
95.87 .001
11.55 .01
12.30 0].
27.68 .01
10.91 .02
11.98 .02
14.33
30.53 .01
82.56 .001
11.66 .001
0,32 -
83.05 .001
27.19
12.31 .01
8.49 .01
0.17
12.14
3.82
159. 53 .001
35.45 .001
5.07
0.35
Both sexes:
Constant
Time
Treatments
Treatments adjusted for time
Demographic variables and risk factors
Treatments adjusted for demographic variables
Treatments adjusted for demographic variables and time
Clinics adjusted for treatments and time
Treatments adjusted for clinics and time
Treatments adjusted br demographic variables, clinics, and tin~e
Clinics adjusted for demographic variables, time, and treatments
Clinics adjusted for demographic variahies and treatments
Demographic variables adjusted for clinics, time, and treatments
Tolbutamide treatments adjusted for demographic variables, clinics, and time
Other treatments adjusted for demographit variables, clinic, time, and tolbutamide
treatment
Subset of demographic variables adjusted for treatment and time 1
Other demographic variables and clinics
Treatments adjusted for subset of demographic variables and time
Tolbutamide treatment adjusted for subset of demographic variables, time, and insulin
treatments (tolbutamide vs. placebo)
Insulin treatments adjusted for subset of demographic variables, time, and tolbutamide
treatment (insulin vs. placebo)
Tolbutamide treatment adjusted for subset of demographic variables and time (tolbuta-
micle vs. other treatments)
Insulin treatments adjusted for subset of demographic variables and time (insulin vs.
other treatments)
Men:
Constant_.
Demogr4phid variabtes and time
Treatments adjusted for variables and time
Tolbutan,ide treatmtnt adjusted for insulin treatments, demographic variables, and
time (tolbutamide vs. placebo~.
Insulin treatments adjusted for demographic varibles time and tolbutamide treatment
(ihsullh Vs. placebo)
Tolbutamide treatment adjusted for demographic variables and time (tolbutamide vs.
othdr treatments)
Insulin treatments adjusted for demographic variables and time (insulin vs. other
treatments)
Women:
Constant
Demographic variables and time
Treatments adjusted for variables and time
Tolbutemide treatment adjusted for insulin treatments, demographic variables, and
time (tolbutamide vs. placebo)
Insulin treatments adjusted for demographic variables, time, and tolubuamide treat-
ment (insulin vs. placebo)
Tolbutamide treatment adjusted for demographic variables and time (tolbutamide vs.
other treatments)
Insulin treatments adjusted for variables and time (insulin vs. other treatments)
6
3
2
2
6
3
2
2
2.52
2.56
4.72
551.79 .001
67.66
12.23 -
11.70 .01
252
9.06 .01
0.53
1See table A.5.
PAGENO="0119"
-4
m
-4
-4
Cr) -
-4~
-
Ci~ C)
C) (~
C~ -4
~ C,)
Cr)-4
C ~
ci
-4
-4
0
-4
0
C)
~
~
~ ~a ~-°~
~. :~ ::~:F~-~:
C C) N)
OO(~)
~N)
PAGENO="0120"
~9MPETITIVE PROBLEMS IN THE DBUG INDTJSTRY
~ C' CJ c'J co~ c c~oo Lt)(~
C'C~ C%~~ ~C~4C~ ~C~C)U)C) ~C%4~C~J
~ C~) -- -
~ -~°` ~
-~ co u~ ~, ~ ~, ~
~ r-~ c~c~Ø ~c~3
~ ~ (~, ~ ~ ~ ~,,
- c~c~o~ ~
~ ~2°'~
0)
~
C')
>-
C,,
~o c~ ~ ~ CC ~ ~ ~(C)
~ coc ~`o~ c~c~ c~
~4~'4 C?)C~CC~~~ id~©~
C)
C,,
C
0
~ 00 ~ r-~ ~
0
~ - - - ~ - -
r
E
0,
C)
PAGENO="0121"
COMPETITIVE PROBLEMS IN TEE DRUG INDUSTRY 13371
C) (CC) (0
-~C) C)C)~ QC)C)C) C)C)CC) C)C)~C)
(Cc~4C)cC r-C0)'- 0)~-~C) 0J~.(Cc~) 0)C)0)
C)C~4C)C) C)C)~'-4 C)C)C)C) C)~C)C) QC)C)
C~J (C) ~(C C)'-~ C) C) (0 C) C) (01~ (C C) (0(000 ~)Ø) C) C)
C)
C)(CC)C) C) C)C)
~ C'4 (C 00 c~ C) 0000)(C
C) C)C) C) - C) C) C) C) C) -C) C)C) C) C)
C~JN~ ~C)C'1C~ C%J(C(~4 O0~0(0 00(000(0 0C~J04C)J
(C C) LO) C) ~ F~ C) C)
C0~-~CO(0 C)~C)C) ~C)C'~ C)C)C)~C) C)C)C)~C) C)C)L~)C)
C) ~ C~ C'4 o0~o0O 04 (C) 040- (0 C)~ CO
L000 C) (000(0 C) (0(0 (C) (0(0 C)
~
_C)__ C)C)_C) C)C)_C) C)C)_C)
~ 04 (004 `-04 04 C) 04 04~ 0)00
0404(4 ~-~-)
PAGENO="0122"
13372 COMPETITIVE PROBI43~MS IN TUR DRUG INDUSTRY
TABLE A.6.2.-UGDP DATA: ESTIMATES OF THE COEFFICIENTS, THEIR STANDARD DEVIATIONS AND TI FOR THE
LOGISTIC REGRESSIONS USING ALL VARIABLES
Variable
Coefficient
Standard
deviation
T
Age
Sex
00692
-.3956
0.0213
.3919
3.25
-1.01
Race
Relative weight
Blood pressure Systolic
Diastolic
Digitalis use by history
Angina by history
Abnormal ECG
Cholesterol
Arterial calcification
Glucose tolerance test
-.6432
-1. 1122
0143
-.0125
-.6748
-.6519
1. 3907
. 0041
27~0
0011
.4560
.7768
. 0076
.0154
.5117
.5171
.5693
. 0026
. 1877
.0031
-1.41
-1. 43
1. 89
-.81
-1.32
-1.26
2.44
1. 58
1. 47
.36
Serum creatinine -
Visual acuity -
Boston
. 6990
-. 1592
1.3005
. 8727
. 6095
1. 2477
. 80
-. 26
1.04
Minneapolis
New York
1.5444
.8680
11989
1,3323
1.29
.65
Williamson
Cincinnati -
Cleveland
1.0567
2, 3536
.8507
1.2473
1. 3089
1. 4507
.85
1. 80
.59
Baltimore
Birmingham
Chicago - -
St. Louis
San Juan
Time
Tolbutamide
insulin:
-.1780
.6587
1. 1578
4500
.3470
.4362
1. 2412
1. 6263
1.3571
1.3484
1. 5386
1.5304
1. 1805
. 4470
-.11
.49
.86
.29
.23
2.42
2. 78
Standard dose
.2424
.4939
.49
Variable dose
-. 0005
. 5257
-. 001
Constant
-8. 8522
2. 9440
-3, 01
1 Estimate divided by standard deviation.
TABLE A.6.4.-UGDP DATA: NUMBER OF CARDIOVASCULAR DEATHS OBSERVED AND PREDICTED FROM THE
LOGISTIC MODEL
Treatment
Placebo
Tolbutamide
Standard dose
insulin
Variable dose
insulin
IViales:
«=53:
Observed
Expected
>53:
Observed
Expected
Total:
Observed
Expected
Females:
«=53:
Observe&
Expected
>53:
Observed
Expected
Total:
Observed
Expected
1
0.521
6
3.640
5
2.600
6
9.310
0
0.952
5
4.398
1
0.447
1
3.130
7
4.162
1
0.898
2
4. 940
11
11.910
1
3.260
14
10. 829
5
5.350
1
1.107
7
6. 543
2
3.577
1
1.057
9
7. 365
3
5. 838
15
14. 088
8
7. 649
10
8.422
PAGENO="0123"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13373
TABLE A.7.1.--UGDP DATA: RELATIVE ALlOCATION OF TOTAL AND CARDIOVASCULAR DEATHS AND EFFEC1hVE
SAMPLE SIZE
Assigned treatment
Standard Variable
Treatment received
Placebo Tolbutamide dose insulin dose insulin
Totals
Treatment as assigned: 1
d'
d"
Dose modification or other treatment:
12.9 19.9 6.9 26.9
6.4 17.9 3.9 5.6
115.6 119.5 65,7 53,9
46.6
33.8
354.7
Placebo:
ci'
2.8
2.8
d"
n'
1.0
48.6 -
1.0
48.6
Tolbutamidd:
.
d'
0 7.7 0 0
7.7
d"
n'
0 5,7 0 0
.5 51.4 .8 .6
5.7
53.3
Variable dose insulin:
d'
None:
1.2 .4 12.5 38.5
1.2 .4 8.6 54.6
6. 9 10. 7 111. 1 3 121. 9
~2.6
14.8
250. 6
d'
4,1 2.0 .6 2.6
9.3
d"
n'
Totals:
1.4 2.0 .5 1.8
33.4 22.4 32.4 27.6
5.7
115.8
d'
ci"
~1.0 30.0 20.0 18,0
10. 0 26. 0 `13. 0 12. 0
89,0
61.0
205,0 204.0 210.0 204.0
823.0
I d' indicates total deaths; d", deaths from cardiovascular causes; and n', effectiva sample size.
2 These figures are associated with the initially selected doqeof idsulin.
3 These figures are associated with the subsequently selected dose of insulin.
TABLE A.7.2-UGDP DATA: DEATH R
ATES A
ND FAILURE
RATES BY ASSIGNED TREATMENT AN
D DOSE GIV~N 1
Dose given `
Assigned Treatment
Placebo
Standard dose
Tolbutamide insulin
Variable dose
insulin 2
CàrdiQvascUlar deaths:
As assigned:
0'
Dose modified:
0'
No drug:
0'
Total deaths:
As assigned:
0'
Dose modified:
0'
No drug:
0'
`
0.06
.9
.02
.3
.04
.7
.1~1
1.9
.06
.9
.12
2,1
`
0.15 0.06
2.5 1.0
.11 .08
1.9 1.2
.09 .02
1.1 .3
.17 .11
2.8 1.8
.15 .11
2.5 1.8
.09 ` .02
Li .25
0.10
1.8
~04
.6
.07
1.0
.13
~.3
.07
~.1
.09
~.5
I Death rate (8')is effective number of deaths/effective sample size; failure rate (A') is deaths per hpndred patie~it-
years of follow-up.
2 Combining both modification groups results in 0'=10.2/175.9=0. 0., X'=0.9 for cardiovascular deaths; 0'=15.4/
175.9=0.09, X'=1.4 for total deaths.
PAGENO="0124"
13374 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE A.7.3.-UGDP DATA: CARDIOVASCULAR DtATH RATES BY ASSIGNED TREATMENT SEX, AND DOSE
GIVEN 1
Assigned treatment
Standard dose Variable dose
Dose given Placebo Tolbutamide insulin Insulin
Females:
As assigned 1. 4/79. 8=0.02 10. 8/780=0. 14 3. 9/52. 8=0.07 4.6142. 4=0. 11
Dose modified 0/33. 6=0 3/39.7=0. 08 3. 7/76. 4=0. 05 3,6/94. 3=0. 04
No drug 1. 2/23. 6=0. 05 1.2/19.6=0.06 0.4/23.5=0.02 1. 8/20. 9=~Q. 09
Males:
As assigned 5/35. 9=0. 14 7. 0/41.5=0. 17 1/14. 0=0.07 1/11. 5=0. 09
Dose modified 1/15. 0=0.07 2. 7/11. 7=0. 23 3. 9/33. 7=0. 12 1/27. 6=0. 4
No drug .2/9.8=0.02 0. 8/8.6=0.09 0. 1/9.0=0.01 0/6.7=0
I Tables give ratio of cardiovascular-deaths (relative allocation) to number of patients (relative allocation).
2 Combining both dose modification groupsfor IVAR results in the following: females 0 =8.2/136.7 =06; males
0 =2/39.1=.05.
TABLE A.7.4.-UGDP DATA: CARDIOVASCULAR DEATH RATES BY ADHERENCE AND DOSE MODIElC~TION
Treatment
Dose Standard Variable
Modification Placebo Tolbutamide dp~e insulin dose insulin Totals
.
100 percent ad-
herence :1
No 4/64=0.06 16/77=0.21 4/51=0.08 4/27=0.15 28/219=0.13
Yes 1/12= . 08 4/29= . 14 3/50= .06 2/55= .04 10/146=0. 07
<100 percent
adherence:
.06
No 2.4/51.6= .05 1.9/42.5= .04 4.6/60.1= .08 1.6/26.9= .06 10.5/181.1=
Yes 1.8/36.6= .05 1.7/22.4= .08 0.9/15.7= .06 2.6/66.9= .04 7/141.6=05
Totals 912/164.2= .06 23.6/170.9= . 14. 12. 5/176.8= .07 10.2/175.8= .06 55. 5/687.7= .08
1 100% adherence is defined as taking the initially assigned drug for every quarter of follow-up.
TABLE A.7.5.-UGDP DATA: CARDIOVASCULAR DEATH RATES BY ASSIGNED TREATMENT, SEX AND COMPLIANCE
-
Compliance1
.
Treatment
Placebo
Tolbutamide
Standard dose
insulin
Variable dose
insulin
Females: -
100 percent adherence-no dose
modification 0/41=0 9/51=0.18 1/28=0.04 3/21=0.14
<100 percent adherence or dose
modification 1.3/72.3= . 02 4. 8/66. 6= . 07 6. 6/101. 2= - 07 5. 2/115. 8= . 04
No drug ).. 2/23.6= .05 1.2/19.6= .06 0.4/22. 5= .02 1.8/20.9= .09
Males:
100 percent adherence-no dose
modification 4/23= . 17 7/26= .27 2/22= .09 1/6= . 17
<100 percent adherence or dose
modification 2/27. 8= .07 2.7/27. 2= . 10 2. 9/25.7 = . it 1/32. 9= . 03
No drug 0.2/9.8= .02 0.8/8.6= .09 0.1/9= .01 0/6.7= 0
ICompliance Is based on both adherence to the assigned treatment group and use of the prescribed dose.
PAGENO="0125"
COMPETITIVE PEO~I4EMS IN `ri~ ~uo ~irs~n~ 13375
TABLE A.7.6.~-UGDP DATA: CARDIOVASCULAR MORTALITY RATES BY SEX, COMPLIANCE, AND
MEDIAN AGE AT ENTRY
Compliance Control' Tolbutaàdde
F
Females:
«=53 yr old:
100 percent-no [3M 2
<100 percent or DM
No drug
>53yrold:
100 percent-no DM
<100 percent or DM
Nocirug
Males:
0/45=0
2.6/155.9=. 02
O.4/39=.0t
~
4/45=. 09
1O.5/133.4=.08
3/28=11
0/22=0
1/33. 41.03
0/11.79 0
9/29=~. 31
3. 8/33.2=r. 11
1.2/6.9=~.17
«=53 yr old:
100 percent-no DM
<100 percent or DM
No drug
>53 yr old:
100 percent-no DM
<100 percent or DM
No drug
1/23=. 04
1/35.6=. 03
0/16.7= 0
6/28=.21
4.9/51=. 10
0.3/8.8=.03
3/10=~. 30
1.8/10.5=. 17
02/4.99.04
4/16=4.25
0.9/16. 7=~. 05
0.6/3.7=s.16
`
Control is made of of placebo and standard- and variable-dose insulin treatment groups.
2 DM indicates dose modification.
TABLE A.77.-UGDP DATA: TREATMENT' VS PLACEBO ADJUSTED LOGARITHMS OF RATIOS OF FAILURE RATES
BY SEX FOR CARDIOVASCULAR AND TOTAL DEATHS
ToIb/Plbo ISTD/Plbo IVAR/PIbo None/~'lbo
,
Cardiovascular deaths:
Female:
Log ratio 2. 64 1.79 1.75 3. 69
SD2 99 i.o5 1.03 3.21
P3 .008 .09 .09 .16
Male:
-.29 -4.48
Log ratio - . 094 -.
SD .59 .69 .86 2.08
.73 .03
P .87
Total deaths:
Female:
Log ratio 1. 12 - 23 . 64 -h-. 51
SD .55 .63 .56 .78
.26 .52
p .04 .72
Male:
Log ratio -.30 .037 -1.49 -Los
st .5 0 .86 1.18
.08 .16
p .56 .94
*____*__ -,_-,--_ ,---,-,__ ~ ~_~_ ~,___ ,,___ ~ *____*_ ~,,--,_:.--.- .~_ ~______~___~ ~ *.___~. *~___ ~
1 ToIb indicates tolbutamide; Pibo, placebo; ISTD, standard-dOse insulin; and IVAR. variable-dose insulin.
2 Refers to estimate of standard deviation of log of ratio of failure rates.
Refers to test of significance (two-tail) using normal theory.
TABLE A8.1.-BEDFORD DATA: INFLUENCE OF BACKGROUND VARIABLES ON DEATH RATES
~
Factor
Hypertension:
No 27/193=0 14 1- 54
Yes 15/53=0. 28 3. 41
Blood glucose level:
«=139 mg/100 ml 17/124=0. 14 L 55
>139 mg/lb ml 25/123=0. 20 3.28
Arterial disease:
14 1. 54
No
Yes 12/30=0. 40 5. 02
Sex:
Male 20/1290.15 376
Female 22/119=0 18 2. 08
`Age:
«=45 yr 0/54=0
>45 yr 42/194=0. 22 2. 52
,
j is the number of deaths/total.
j)i is the number of deaths/totOl time followed up (6-mo periods).
PAGENO="0126"
13376 cO~?ETITIV~. PRO~LEM~: IN THE DRUG INDUSTRY
TABLE A.82.-BEDFORD DATA: DEATH RATES FOR PLACEBO AND TOLBUTAMIDE GROUPS, BY BACKGROUND
VARIABLES
Variable
Placebo
Tothutamide
p* -
Hypertension:
No --
Yes.
Blood glucose level:
<139 mg/100 ml
>139 mg/tOO ml
Sex:
24/125=0.19
17/102= .17
7/23= . 30
10/64= 16
14/61= 23
~
18/123=0.14
~
10/93= 11
8/30= 27
7/60= 12
11/63= . 17
Males (>45 yr old)....
Females (>45 yr old)
Arterial disease:
11/52= . 21
13/46= 28
9/46= . 20
9/50= . 18
Yes:
Females
Males - - ~
No:
2/7= 29
6/8= .
1/6= .17
3/9= .33
Females
Males -
Hypertension:
Yes:
11/50= .2
5/60= . 08
8/56= . 14
6/52= . 12
~
Females
Males -- -..-=
No:
5/15= .33
2/8= .25
4/23= . 17
4/7= 57
Females
Males
Blood glucose level:
<139 mg/tOO ml:
Females
Males
>139 mg/100 ml:
Females
Males
8/42= .19
9/60= . 15
3/28= .11
7/36= .19
10/29= . 34
4/32= . 12
5/39= .13
5/54= . 09
4/32= . 12
3/28= .11
5/30= . 17
6/33= . 18
is the number of deaths/total.
TABLEA.8.3.-BEDFORD DATA: NUMBER AT RISK, NUMBER OF DEATHS, AND DEATH RATE BY AGE, SEX,
AND TREATMENT
Age, years
Males
Females
Placebo
Tolbutamide
Placebo
Tolbutamide
20 to 29
30 to 39
40 to 49
50 to 59
60 to 69
70 to 79
80 and over
All ages
0/4=0
0/6=0
0/10=0
1/19= .05
6/14= .43
6/12= .50
3/3=L0O
0/4=0
0/4=0
1/14= . 07
3/18= . 17
3/12= .25
5/8= .62
1/1=1.00
0/3=0
0/1=0
1/12= .08
3/6= .50
1/13= .08
11/18= .61
3/4= ,75
0/2=0
0/8=0
0/8=0
3/16= . 19
5/15= .33
7/12= .58
0/1=-.0O
16/68= .24
13/61= .21
19/57= .33
15/6~= .24
PAGENO="0127"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY i33~7
TABLE A.9.1.-VARIABLES USED IN THE LOGISTIC ANALYSIS OF THE BEDFORD STUDY DATA
Group: A variable indicating whether the individual entered the study in 1962 or 1964.
Tolbutamide: A variable indicating that the patient was receiving tolbutamide.
Diet: A variable indicating that the patient was receiving dietary advice
Tolbutamide-diet interaction: A variable indicating that the patient was receiving both dietary advice and tolbutaniide.
Variables:
Sex.
Age.
Arterial disease history(O=no, 1=yes)
Significant hypertension (0= no, 1 =yes)
Blood glucose.
TABLE A.9.2.-ANALYSIS OF CAUSES OF DEATH IN THE BEDFORD TRIAL USING THE LOGISTIC MODEL
Source df Likelihood ratiokt
Deaths from all causes:
Constant - 1 62. 68 (p-<.OOl)
Group 1 0.00
Tolbutamide adjusted for group and diet 1 0. 90
Diet adjusted for group and tolbutamide 1 0. 02
Tolbutamide-diet interaction adjusted for group, diet and tolbutamide 1 0. 57
Variables adjusted for group, diet, tolbutamide, and tolbutamide-diet interaction 5 81. 02 (p<00~)
Tolbutamide adjusted for group, diet, and variables 1 0. 02
Diet adjusted for group, tolbutamide, and variables 1 0. 19
Tolbutamide-diet interaction adjusted for group, tolbutatnide, diet, and va~iabIes 1 0. 01
Deaths from cardiovascular causes:
Constant 1 118. 17 (p<.O0~)
Group 1 0.14
Tolbutamide adjusted for group and diet 1 1. 02
Diet adjusted for group and tolbutamide 1 0. 73
Tolbutamide-diet interaction adjusted for group, diet, and tolbutamide 1 0. 66
Variables adjusted fo~ group, diet, tolbutamide, and tolbutamide-diet interaction. 5 55. 16 (p<.tI0~)
Tolbutamide adjusted for group, diet, ahd variables 1 0. 06 I
Diet adjusted for group, tolbutamide, and variables 1 0. 22
Tolbutamide-diet interaction adjusted for group, tolbutamide, diet~ and variables 1 0. 03
TABLE A.9.3-I3EDFORD DATA: ESTIMATES OF THE COEFFICIENTS, THEIR STANDARD DEVIATION, AND Ti FOR T~E
LOGISTIC REGRESSION ON CARDIOVASCULAR AND ALL DEATHS
Cardiovascular deaths
Standard
All deaths
-
Standard
Coefficient deviation
T
Coefficient
deviation
T
Sex -
-0. 19~0 0. 4231
-0. 46
-0. 0856
0. 3776
-0. 23
Age
Arterial disease history
Significant hypertension
Blood glucose
Group
Diet
Tolbutamide._
Constant
.1069 .0207
1. 1197 . 4908
.7126 .4386
-.0009 .0091
.8525 .7124
.1842 .3922
-.0967 .4039
-8.4326 1.8741
5.16
2.28
1.62
-.10
1.20
.47
-.24
-4.50
.1259
. 7148
,4399
-.0109
.7279
-.1557
.0448
-7.4192
.0194
4798
.4119
.0084
.6668
.3541
,3634
1.6450
6.4
1. 4
.1.07
-1.3
1.0
-.44
.1
-4.5
1 T is the estimate divided by the standard deviation
PAGENO="0128"
THE EFFECTS OF LONG TERM THERAPY WITH ORAL HYPOGLYCEMIC AGENTS ON THE
ORAL GLucosE TOLERANCE TEsT DYNAMICS IN CHEMICAL DIABETICS
(M.H. Tan,* M.D., CA. Graham, M.D., R.F. Bradley, M.D., R.E. Gleason, Ph.D.,
and J.S. Soeldner, M.D.)
From the Joslin Clinic and the E. P. Joslin Research Laboratory, in the De-
partment of Medicine, Peter Bent Brigham Hospital and Harvard Medical
School, and the New England Deaconess Hospital, Boston, Massachusetts
Supported by U.S.P.H.S. grants AM-09748, AM-04146, AM-05077, the Joslin
Diabetes Foundation, Inc., Boston, Mass., the Upjohn Company, Kalamazoo,
Michigan, the Eli Lilly Company, Indianapolis, Indiana, and Pfizer, Inc., New
York City, New York
Presented at the 33rd Annual meetng of the American Diabetes Association
on June 23, 1973 in Chicago, Illinois
A13STRACT
The effect of fixed doses of oral hypoglycemic agents and placebo upon the
blood glucose, serum insulin, triglyceride and cholesterol responses during oral
glucose tolerance tests done annually for up to 4 years follow-up was studied,
in a double blind manner, in 5 groups of mild chemical diabetics. The drugs
used were chlorpropamide (100 mg O.D.), tolbutamide (500 mg b.i.d.), phen-
formin (50 mg C.D.), acetohexamide (250 mg O.D.) and placebo. Each subject
wag given an individualized diet aimed at attaining and maintaining ideal
weight, Comparison by Clii square analysis between the placebo group and each
of the drug groups showed: (a) no significant differences with regards to the
number of subjects with normal glucose tolerance test in each of the tests and
(b) no change in the insulin secretion dynamics. Comparison between the
initial test and each of the subsequent tests within each group showed: (a) a
greater number of subjects had normal glucose toelance in test 2 in the placebo
group, test 2 and 3 in the tolbutainide group and tests 2-4 in the chiorpropa-
mide `group; (b) no change in the insulin secretion dynamics except in the
chlorpropamide group where there was an increase insulin/glucose ratio in
test 2 and (c) no change in the fasting serum triglyceride and cholesterol
levels.
INTRODUCTION
A goal in the detection of the early stages of diabetes mellitus is the hope that
prompt therapeutic intervention may retard the clincal manifestations of the
later stages of the disease. Both remisson of the disease (1,2) and improve-
ment of the carbohydrate tolerance (3-16) following the use of oral bypoglyce-
agents have been reported. Some groups reported the improvement of the car-
bohydrate tolerance to be associated with incrased insulin secretion (4~-7)
whilst others indicated otherwise (8-16). A lowering of fasting serum lipid
levels in diabetics treated with oral hypoglycemic agents has also been reported
(5, 10, 17-20). Most of the above studies were performed after relatively short
term (weeks to months) therapy with the oral hypoglycemic agents, The pr~s-
ent study reports on the effects of long term (up to four years) therapy with
fixed doses of oral hypoglycemic agents and diet upon (a) the glucose toler-
ance, (b) the insulin secretion dynamics, and (c) the lipid dynamics during
oral glucose tolerance tests of chemical diabetics.
METHODS AND SUBJECTS
Over a six year period, 365 mild chemical diabetics participated, with in-
formed consent, in a double blind prospective study to evaluate the influence
*Fellow, Medical Research Council of Canada.
13378
PAGENO="0129"
COMPETITIVE PRO~LEl\~S IN THE DRTJG IN1~USTR~ 13379
of fixed doses of oral hypoglycemic agents and proper diet on the nati~ra1
history of the disease. Each of them was asymptomatic, had normal fas1~ing
blood glucose levels, but had two abnormal oral glucose tolerance tests prior
to entering the study. They were randomly assigned to four groups, each
group taking a different drug. In each group one out of every four subjects
was placed on a placebo (fig 1). The drugs used were: Chiorpropamide 100 mg
daily, Tolbutamide 500 mg twice daily, Plienformin 50 mg daly and Acetohexa-
mide 250 mg daily, Drug adherence was assessed by history during the follow
up visits every three months. Each subject was given an individualized diet
aimed at attaining and maintaining ideal weight as defined by the Metropolitan
Life Insurance Company-1959.
In this prelminary report, oniy male subjects who had at least two tests ~ind
complete data (namely, glucose, insulin, cholesterole and triglyceride values)
in the initial (test 1) and subsequent tests (test 2-5) were included. As shown
in table. 1, there were 37 in the placebo group, 18 in the chiorpropamide, 28 in
the tolbutamide group, 23 in the phenformin group and 14 in the acetohe~ca-
mide group. As the follow up years increased, the number of subjects in each
group decreased.
Each subject had an oral glucose tolerance test (100 ing dextrose) at the
beginning of the study (test 1) and then annually during the follow up ye~trs.
Each subject followed his usual diet which contained 100-200 grams carbo-
hydrate and took no medication on the day of the test. Each test was done
after an overnight fast and begun between 0800 and 0900 hours. Blood samples
were obtained by venipuncture prior to and at 30, 60, 120 and 180 minutes after
ingesting the glucose. Blood glucose was measured by Hoffman's method as
modified for the AutoAnalyzer (21). Serum insulin was assayed by the double
atitbody method of Soeldner and Slone (22). Serum cholesterole and triglycer-
ides were measured by the Technicon AutoAnalyzer method (23).
The criteria for an abnormal oral glucose tolerance test are those used in ~he
Joslin Clinic. A test was judged abnormal if any one blood glucose value at a
given time interval exceeded that listed below: Fasting = 100 mg/dl, 30 mm. =
160 mg/dl, 60 mm = 160 mg/dl, 120 mm = 120 mg/dl and 180 mm-hO mg/dL
The criteria for normal fasting serum cholesterol and triglycerides are thOse
of Goldstein et al (24).
Statistical analyses were done by Chi Square, paired and unpaired Students'
t-tests as indicated.
COMPARISON OF THE GROUPS AT THE BEGINNING or THE STUDY
The placebo group did not differ significantly from each of the treated groups
in mean age and percent ideal weight at the beginning of the study and in
subsequent years (table 2). The percentage of obese subjects n each group was
also comparable. A comparison of the percent ideal weight of the subjects in
each between the initial test and each of the subsequent tests was made by
paired `t' analysis. No significant difference was noted except in the phenf9r-
mm group. In this group there was a significant decrease in percent ideal
weight in test 2 (100.2±2.9 p<0.01) and test 5 (98.2±3.7 p-
-J
C')
LU
>
I-
-I
LU
LU
~1
C',
.~ ~
~ a).-
~ ~- >
LU
C,,
COMPETITIVE
E
PROBLEMS
LU
C,,
~
~ ~*~V)>
LU
C')
a)
*~ ~
~
LU
C',
.E ~
-S -a;.-=>
~- .~
LU
C,,
~
z
IN THE DRUG
00
~
r.~c~O~,-a OOIg,u,C')
r~oo O~ r-~a)o,CO
d' ~c~c
do ~
~~C',C~4 u~OO0~4
~ ~C%~
0
0 -
0
~
0
PAGENO="0156"
13406 COMPETITIVE
PROBLEMS IN THE DRUG INDUSTRY
LU e.JOO C~) O~©
C,, ~ c*.J ©-
c `~
~ ~U)
~ g~jr-.c~
LU ~ oooO~c~
C/) Ø©~©
~ U)'~ U~oO
E ~ ~ ~
L&.
I ~ ~
~
I
~
-J
~ E ~ ~
> C ~ U):~U)._ .
~ U) ~
a) .C
C'~C. ~
LU 00 U, 00
> (0 OC)) - oD C) C~) -
-J
LU
~ ~ ~
z
PAGENO="0157"
COMPETITIVE PROBLEMS IN THE~ DRUG INDUSTRY 13407
E
I'
LU
C,,
~ Co.;
0~~ 00
0000 C~J
00 -
0~ (0 ~
00 CC)
CC) 0000
00000000
00
00
(0(0 00~-~
CC) 00 CC)CCJ
00 0)0) ~
0000000
0
r~ o~ r~
(0(00000
0
00
0
~ CC) CC)
CC) 00
00
0)
~0) CC) CC)
0) ~ 00
~~~~(DCC)
1~- (0 0
o 0~ -
C~J ~
F~ 0)l~ 00
~00) ~
00
~(ooor~
00 CC~ C'J (0
~ 00CC. 0)
~Ø0) CC)0)
I-.
LU
00
U-
0
C,)
Cl)
C~)
-J
C,,
0
C-)
-J
C,
00
0
0
-J
00
00
01:
>~
C,)
>-
00
-J
>
>
00
(`-I
LU
>
01:
-J
LU
00
LU
-J
00
I-
.0
00
LU
C,)
00 CC:
0(0 ~5 0
PAGENO="0158"
PROBLEMS IN THE
LU QOO~-~
`-°
~ ~ ~
~ c~''> ~
C~J
`0,- c))00~r.~
= ~ ~ I~ N. 0000
~
c'J
0) -
~
~
.~ ~ ~
~ ~
~ 0)>
c,J
>
>
E
0
U.
0
E
13408 COMPETITIVE
Cl)
=
I-
LU
C.)
C'.,
>
0
0
C.)
C,,
0
C.)
-J
CD
0
0
0
-J
00
0
LU
Cl)
>.
00
-J
>
>
C')
LU
>
-J
LU
LU
-J
00
DRUG INDUSTRY
00
*00
0
C..)
~
0) ~
PAGENO="0159"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13409
LU 00 C~JC~)
C,, ©O
~
a)
~ .~ ~ 8~
E ~ ~
~
~ ~
0
- z
~
~
0
Q >~
~
~ ~
I
~
0
_J~ a)
~
~ .&~th*~ ~
I- ~ 0
r~. ~ r% ~
C~J~C~) `~J~OC~-~
PAGENO="0160"
13410 COMPETITIVE PROBLEMS
LU
C,,
IN THE DRUG INDUSTRY
(00)0) (00)0))
0)0)_ 00-0)
0)0)
0)
z
LU
C,,
00~~';
LU
C,,
LU
C,,
~
00
LU
C,,
~
~
z
E
Li.
4)
5
00000)0
0 -
(0(0000)
0)00-
0) 00 r~
0)
000000
00000)) 0)J
00
0)) 0000 C))
(0(0~ 00
(0(i) ~ 00
~,j ~J C') C~)
0)00
H
0)
00***
0)
0)
0)) C') 0)
- C') 0)
CO
=
I-
LU
0)
00
-J
C.)
CO
>
0
0)
00
C.) o
CO 0)
C.) 4)
-J
CO
0) 0
00)
0 =
CO
00<
LU
CO
>.~
00
_J
0)
>~;
00 0.
C')
LU
>
I-
-J
LU
00
LU
-J
00
C') 00(00))
0)0)0)-
(000(0(0
000)0)00
(00)
00 00 0)
0000 C') 0)
H
LU
C,)
~- `O.~
0000
~()0)0)0) >(0)0)0)0)
00
PAGENO="0161"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13411
00 00 ~ ~`.
o~ N) 00 N. 00 ~ CC) - CC)COCC CL)
*~ C)) 0000 C) 00 C) 0)000)00 0)0 N. -
E ~ ~.
.2 ~
00 00 00
2
00 c~ o
C.,
~ E ~ ~.. a.'.
>.
2 00 00 00
0)
00
- C)) U) N) CC) CC ~ CC) CC) ~j ~ 000NJ NJ
*~ 00 C) 000)00 r~ 0)000)0 0)0N.
00 ~ o~
C,, I
~C 2
-J ~ E d ~**~
~
= U) C)COC 2 00~OO00 0L~.
>. . 00 00 00
co
~
I. 2 00 ~
00 -
LU
-J
00
CC CCCC) U) U)
~ 00 0)000000 N. 0000 CC CC
8 E o~*~ ~ 00~~*
22
0 00 C') ~ CC) CC N. (ON. ON) (CC) NJ 00
0 NJ C) 000000 C) COON. 00 N. 00CC -
- 0 00 00
.CC: ~:~: ~:~: ~: ~:~:
co LJJLU *-
`.~ 0CC)u5(I) ~0 ~ OCC)U)C#)
2
PAGENO="0162"
13412 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
C~) ~ (0 (000 u~
00 C~ 00000 e.i 000 ~ 0 r~ r~c~J
0
2
~ E o°~ o~
~
0 0
0)
-
00 ~ ~*** a...
C)
I ~
0
00
~; ::::,: ::i:,:; :::::
2 0 0
~ ~
00 ~
-J
00
0 0
~ ~
0 0 0
HE~~ ~ ~
0 00
.~* o,0 0:0: ~ 00
LULU ~ LULLJ ~UJ~'LU
z 0 CI) ~ CI) - Z 0 ~ c~ Z 0 c'~
-i 2
PAGENO="0163"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13413
[From The New York Times, luly 8, 1976]
A.M.A. AIDE LET UPJOHN USE LETTER To SElL DRUG
(By David Burnham)
WASHINGTON, July 7-The chief executive officer of the American Medical As-
sociation wrote a letter early this year to state and county officials of the ~tsso-
elation minimizing questions that had been raised in the association's own maga-
zine about the safety of a widely used diabetes drug.
The official, according to confidential A.M.A. documents, then permitted the
1,100 salesmen of the largest manufacturer of the controversial drug to use his
letter in their sales talks despite a warning that such use violated A.M.A. pQllcy
and might prove embarrassing.
Details about the distribution of the letter, written by Dr. James H. Samn~ons,
executive vice president of the A.M.A., became known as the Food and Drug Ad-
ministration published today a proposed rule requiring a new label for the drug
stating that it may lead to death from heart disease.
The drugs are often called oral hypoglycemic drugs. They are taken by mouth
by an estimated 1'/2 million adult diabetics to lower blood sugar level. The pills
are believed to represent a $100-million market for the pharmaceutical industry.
Dr. Sammons sent his letter concerning oral hypoglycemics to the executives of
state, county and medical speciality societies on Jan. 28.
The letter said that the Feb. 10 issue of The Journal of the American Medical
Association would contain both an article and an editorial raising questions about
the drugs. I
Dr. Sammons said that the article, by a committee of the Biometric Sodiety,
supported an earlier critical study of oral hypoglycemics that had been challenged
by some other scientists. He reported that the editorial, written by Dr. Thomas
Chalmers, formerly with the National Institutes of Health, alleged that the drug
might be associated with as many as 10,000 to 15,000 unnecessary deaths a year
in the United States alone.
"A considerable body of expert scientific opinion contradicts these published
findings," Dr. Sammons wrote. "Diabetic patients should not be influenced by press
reports, and should continue on whatever diabetic management program their own
physician had prescribed."
Shortly after Dr. Sammons wrote his letter, the Upjohn Company requested
permission to reprint it for use by Upjohn salesmen. The company manufactures
two oral hypoglycemic drug products under the brand names Orinase and Toll-
nase. The Upjohn product Orinase has been for years the most widely used o~ the
oral hypoglycemic drugs.
An A.M.A. staff lawyer said in a memorandum dated March 18 that the "policy
of the A.M.A. is that the association's name may not be used for trade purpo~es."
"Permission to reprint A.M.A. materials has not been granted if there is any
indication that the name of the association or its materials will be used in any
manner that might directly or indirectly be construed as an endorsement b~ the
A.M.A. of a particular product or manufacturer," the memo said.
The staff lawyer, Betty Jane Anderson, said that if this policy was waived,
"Dr. Sammons should be cautioned that the use made of the letter by Upjohn
salesmen may cause embarrassment to him personally or to the A.M.A.
COMMENT BY COMPANY
The lawyer added that "Upjohn's purpose could be better accomplished by
having an article presenting the other side of the controversy published in The
Journal of the American Medical Association." A notation at the bottom o~ the
memorandum indicated that a copy of it had been sent to Dr. Sammons.
A spokesman for Upjohn, reached at the company's headquarters in Michigan,
said that the A.M.A. had granted the request for use of the letter, and that copies
of Dr. Sammon's letter had been given to each of the company's 1,100 salesMen.
The spokesman was unable to say how frequently the letter had been used.
The warning that the Food and Drug Administration proposed requiring on the
PAGENO="0164"
13414 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
label of the drugs would be set in boldface type and would say: "Oral hypogly-
cemic drugs may be associated with increased cardiovascular mortality as corn-
pared to treatment with diet alone or diet plus insulin."
Cardiovascular mortality means death from disorders of the heart and cir-
culatory system.
ADVANTAGE AND RISK
The warning would indicate that the drugs should be used only for adult pa-
tients not totally dependent on insulin, whose blood sugar cannot be controlled
by diet alone and who can not or will not take Insulin.
The warning would also say that the doctor should inform the patient of the
advantages and potential risks of the drugs, and that the patient should partici-
pate in the decision whether to use them.
The F.D.A.'s warning would apply to a number of oral drug products in addi-
tion to the two manufactured by Upjohn. These include: Diabinese, Pfizer, Inc.;
Dynelor, Eli Lilly & Co.; two forms of a drug called DBI, Geigy Pharmaceuticals;
two forms of the drug Meltrol, USV Pharmaceutical Laboratories, Inc., and the
drug Tolbutamide from Premo Pharmaceutics Laboratorities, Inc.
[Excerpt from The Pharmocological Basis of Therapeutics, Pifth Edition, Goodman and
Oilman]
ORAL HYPOGLYCEMIC AGENTS
History. An important event in the history of the treatment of diabetes mel-
litus was the introduction of orally effective hypoglycemic agents. Janbon and
coworkers (1942), in the course of clinical studies on the treatment of typhoid
fever, discovered that a sulfonamide (p-amino-benzene-sulfonamido-isopropyl-
thiadiazole) induced hypoglycemia. Janbon's colleague Loubatiêres (1957),
made the fundamental discovery that the compound exerted no hypoglycemic
effect in the completely pancreatec,tomized animal and suggested that the action
was the result of stimulation of the pancreas to secrete insulin. There was no
practical application of these findings until Franke and Fuchs capitalized on
the discovery that the antibacterial agent carbutanvkie lowered the blood sugar
In patients treated for infectious diseases. These workers demonstrated the
apparent usefulness of carbutamide in the treatment of diabetes mellitus. Soon
thereafter, the compound tolbutanUde was introduced. This substance is not
antibacterial, is less toxic than carbutamide, and soon became popular for the
management of certain diabetic patients. Tolbutamide is a member of the class
of oral hypoglycemic agents designated as sulfonylureas.
Another group of compounds, the biguanides, was developed independently
of the sulfonylureas. Historically, the development began with the discovery
In 1918 by Watanabe that guanidine is hypoglycemic in rats. Guanidine and its
substituted derivatives were found to be too toxic to be therapeutically useful.
Diguanides, two guanidine molecules joined by a chain of methylene groups,
were more effective and less toxic than the substituted guanidines. SYNTHALIN A,
a potent diguanide, was given clinical trial in diabetes, but it also was found
to be too toxic for therapeutic use. Finally, phenformin (Ungar et al., 1957), a
member of the biguanide series (derived from two molecules of guanidine with
elimination of ammonia), was found to have an apparently acceptable toxicity,
and this compound has since had widespread use.
STILFONYLUREAS
Chemistry. A number of sulfonylurea compounds exert hypoglycemic activity.
The commercially available preparations are tolbutarnule, acetohewamuie,
toZa,~iamide, and civiorpropamide, which have the following structural formulas:
PAGENO="0165"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13415
,CH2~-CH2-CH2
H3C_~))-SO2~NH-C-NH~N\
CH2-CH2----CH2
Tokizamide
CI_ SO NH-~-C--NH -(CH2)2CH3
Chiorpropornide
All of the effective compounds are arylsulfonylureas with substitutions on the
benzene and the urea groups. In the case of tolbutamide, the aryl group is tolyl
and the urea substitution is butyl. Tolbutamide differs from the antibacterial
compound carbutamide in having methyl instead of amino on the benzene ring.
This substitution accounts for the loss of antibacterial properties and for the
reduction of toxicity.
Mechanism of Action. The sulfonylureas stimulate the islet tissue to secrete
insulin. The evidence, coming as it does from a variety of experimental and
clinical studies, unequivocally supports such a conclusion. Administration of
sulfonylureas increases the concentration of insulin in the pancreatic vein in
cross-circulation experiments. Recipient animals, diabetic or nondiabetic, ex-
hibit hypoglycemia in response to. the infusion of pancreatic vein blood from
donor animals treated with sulfonylureas but not to the infusion of mesenteric
or femoral vein blood from the same animals. Sulfonylureas cause degranulatlon
of the p cells, a phenomenon associated with increased rate of secretion of ii~su-
lin. Clinical studies demonstrate that the sulfonylureas are ineffective in com-
pletely pancreatectomized patients and in juvenile-onset diabetic subjects. On
the other hand, they are effective in maturity-onset diabetic patients in whom
the pancreas retains the capacity to secrete insulin.
Tolbutamide
Acetohexamide
PAGENO="0166"
13416 CO~ETITIVE PROBLEMS IN THE DRUG INDUSTRY
Although the molecular mechanism of action of these agents is not understood,
several pertinent observations have been made, Hellman and associates (1971)
concluded that labeled tolbutamide is restricted in its action to the extracellular
space and does not need to enter the p cell. The invoked release of insulin is
immediate and is intimately related to the action of glucose; the drug may
sensitize the cell to the normal secretagogue (Widstrorn and Cerasi, 1973).
Sulfonylureas do not increase the secretion of glucagon.
Extrapancreatic effects of the sulfonylureas have been noted in various organs,
and certain of these may potentiate the effects of insulin. A reduction in the
hepatic uptake of endogenous insulin has been described (Marshall et of., 1970).
Tolbutamide enhances the antilipolytic action of insulin in adipose tissue. This
appears to be related to an altered effectiveness of cyclic AMP rather than to any
change in metabolism of the cyclic nucleotide (Brown et of., 1972; Fain et al.,
1972), and an inhibitory effect of the drug on cyclic AMP-dependent protein
kinase has been observed (Wray and Harris, 1973). Other reports indicate a
variety of influences on cyclic AMP metabolism in different tissues (Brooker and
Fichman, 1971; Kuo et al., 1972; Lasseter et of., 1972); their significance is diffi-
cult to assess.
Duration of action, fate, and excretion. The sulfonylureas are absorbed from
the gastrointestinal tract and hence are effective when given by mouth. The most
important difference among the sulfonylureas, for clinical purposes, is in their
duration of action; in increasing order they are tolbutamide, acetohexamide,
tolazamide, and chiorpropamide.
Tolbutamide can be detected in the blood within 30 minutes after oral adminis-
tration; peak concentrations are reached within 3 to 5 hours. The drug is bound to
plasma proteins. Tolbutamide is oxidized in the body to butyl-p-carboxyphenyl-
sulfonylurea, which is a major excretory product. The half-life of tolbutamide is
about 5 hours. Two or occasionally three doses are required daily.
Acetohea~amide is rapidly absorbed, and maximal hypoglycemic activity is
observed about 3 hours after ingestion. The total duration of action is 12 to 24
hours. Much of the activity is ascribable to a metabolite, hydrocyhecamide, which
has a plasma half-life of about 6 hours; the parent compound, acetohexamide,
has a plasma half-life of 11/3 hours. In persons with normal renal and hepatic
function, more than 80% is excreted, largely as metabolites, in 24 hours. Two
doses are usually required daily.
Tofaza~mide is slowly absorbed; the onset of hypoglycemic action occurs at
4 to 6 hours and persists at a significant level up to 15 hours after a single dose.
Tolazamide is metabolized to a number of hypoglycemic substances that are
largely excreted by the kidney. For most patients controlled by tolazamide, a
single daily dose is sufficient; a few patients require administration of the drug
twice daily.
Chlorpropanvtde is also rapidly absorbed from the gastrointestinal tract and
is bound to plasma proteins. In contrast to tolbutamide, chiorpropamide is not
metabolically altered to any significant degree and is excreted very slowly in
unchanged form. The half-life of a single dose is about 36 hours, or seven times
as long as that of tolbutamide. With daily doses of 250 to 500 mg, blood concen-
trations may not be expected to reach a plateau before 3 or more days. Chlorprop-
amide is administered in a single daily dose.
Toxicity. O'Donovan (1959) analyzed the incidence of side effects to tolbut-
amide in 9168 cases. The total incidence of side effects was 3.2%; the drug was
withdrawn in 1.5% of the patients. The reactions have been classified as hemato-
logical (0.24%), cutaneous (1.1%), and gastrointestinal (1.4%). Of the 22
subjects exhibiting hematological abnormalities, 19 had a transient leukopenla;
in 9 instances, the leukocyte count returned to normal despite continuation of
the drug. Paresthesia, tinnitus, and headache may also occur.
The total incidence of untoward reactions is about 6% for ch~orpropamide
(hematological, 0.6; cutaneous, 3; gastrointestinal, 2; and jaundice, 0.4%). The
jaundice is of the cholestatic type and is usually transient. Hyponatremia has
been reported in a small number of patients treated with tolbutamide and
chlorpropamide.
Experience with acetoheo~amide and toks~zamide suggests that the frequency
and the kinds of toxic reactions are similar to those encountered with tolbut-
amide and chiorpropamide. Hematological (leukopenia, agranulocytosis, throm-
bocytopenia, pancytopenia, and hemolytic anemia), cutaneous (rashes, photo-
sensitivity), gastrointestinal (nausea, vomiting, rarely hemorrhage), and hepatic
PAGENO="0167"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13417
(increased serum alkaline phosphatase, cholestatic jaundice) reactions have
been reported.
Hypoglycemic reactions, including coma, may occur (Seltzer, 1972a). While
they are usually not severe, several fatalities have been reported. Hypoglycemic
episodes may last for several days so that prolonged or repeated glucose ad-
ministration is required. Reactions have occurred after one dose, after several
days of treatment, or after months of drug administration. Most reactions are
observed in patients over 50 years of age, and they are more likely to occur in
patients with impaired hepatic or renal function. Overdosage or inadequat~ or
irregular food intake may initiate hypoglycemia. Drugs that may increase the
risk of hypoglycemia from sulfonylureas include other hypoglycemic agents, sul-
fonamides, propranolol, salicylates, phenylbutazone, probenecide, dicum~rol,
chioramphenicol, monoamine oxidase inhibitors, and alcohol.
Sulfonylureas should not be used in a patient with hepatic or renal in-
sufficiency because of the important role of the liver in their metabolism an4 of
the kidney in the excretion of the drugs and their metabolites. Intolerance to
alcohol reminiscent of the disulfiram reaction has occurred occasionally in pa-
tients taking sulfonylureas.
These `agents are also not recommended for use in pregnancy, but only s$rse
data have been reported on this point. Teratogenesis in animals has been ob-
served to follow the administration `of large doses.
A cooperative clinical trial in 12 university-based clinics (University Group
Diabetes Program; UGDP) was established in 1961 to determine if the cor~trol
of blood glucose concentration helps to prevent or delay vascular disease in non-
insulin-requiring diabetic patients. About 200 subjects in each of five thbra-
peutic regimens were treated with diet and either placebo, a standard dose of
tolbutamide, a standard dose `of insulin, a variable dose of insulin, or a standard
dose of phenformin.
During a period of over 8 years of observation, there were 120 deaths, in-
cluding 87 from cardiovascular causes; while 10 to 12 cardiovascular deaths occur-
red in each of the placebo or insulin groups, 26 such deaths (a significantly bi~ber
number) were recorded among the patients in each group taking oral hypoglyce-
mic agents. The overall mortality rate was correspondingly higher in these two
groups of diabetic patients. The conclusions of thjs study were that the comb~na-
tion of diet and either tolbutamide or phenformin was no more effective in pro-
longing life than diet alone; furthermore, it was felt that diet and either tol-
butamide or phenformin may. be less effective than diet aYone or diet together
with insulin in preventing cardiovascular mortality. (See University Group Dia-
betes Program, 1970; Knatterud et al, 1971.)
Since the UGDP report, a flood of comments and reports have appeared, J~oth
critical (see Seltzer, 1972b) and supportive of this massive study. However, no
warning has yet been included in the package inserts for these drugs, and there
have now appeared a "second generation" of even more potent sulfonylureas (gli-
midine and glibeiu,lamide), which are in clinical use in Europe and elsewi~ere.
Additional studies have continued to indicate an increased incidence of seri-
ous difficulties in patients taking oral hypoglycemic drugs. More episodes of
ventricular tachycardia and ventricular fibrillation were noted in such diabetic
subjects, usually during the early stages of acute myocardial infarction, al-
though there was no difference in the number of deaths (Clayman, 1974; Soler
et a!., 1974). Patients taking oral hypoglycemic agents in England have I~een
reported to have twice the incidence of myocardial infarctions observed in ~ub-
jects being treated with diet alone (Boyle et al. 1972; Hadden et a!., 1972).
Furthermore, at the instigation of the Director of the National Institutes of
Health, the Biometric Society appointed a committee to review the UGDP re-
port. The committee concluded that the shortcomings of the study do not in-
validate the observations and conclusions, the most pertinent of which are
described above, and that other studies do not contradict that of the UGDP.
(See Chalmers, 1975; Report of the Committee, 1975.)
Preparations and Dosage. Tolbutamide, U.S.P. (0RINA5E), is marketed in the
form of 500-mg tablets. The sodium salt (1 g) is also available for administra-
tion intravenously for diagnostic use. Aoetohe~rarnide, U.S.P. (DYMELOR), is atail-
able in 250- and 500-mg tablets. Tolazarnide, U.S.P. (T0LINAsE), is supplied in
100-, 250-, and 500-mg tablets. Uhlorpropamide, U.S.P. (DIABINE5E), is marketed
as 100- and 250-mg tablets.
PAGENO="0168"
13418 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The usual daily dose of tolbutamide is 1000 mg, while 2000 mg is the maxi-
mally effective total dose; corresponding dosages are 500 and 1500 mg for
acetohexamide. Tolazamide and chiorpropamide are usually administered in a
daily dosage of 250 ing, while 750 to 1000 mg is maximal.
Therapeutic Uses. The sulfonylureas should be used only in subjects with
diabetes of the maturity-onset type who cannot be treated with diet alone or
who are unwilling or unable to take insulin if weight reduction and dietary
control fail. The physician must realize that he is using these agents only to
control symptoms associated with hypoglycemia, and that dietary control with
or without insulin is more effective for this purpose. There is no evidence that
the oral hypoglycemic agents prevent cardiovascular complications from diabetes,
and the best data available suggest that the incidence of such complications Is
increased in patients taking these drugs. This is obviously too high a price for
the convenience of an oral agent, unless all other measures have been exhausted.
In general, the likelihood of adequate control with an oral hypoglycemic agent
is inversely proportional to the dose of insulin required to maintain the patient.
When the insulin requirement is in excess of 40 units per day, the chances of
success are relatively low. The sulfonylureas are of no value in the juvenile-
onset type of diabetes, in which the pancreas has lost all or nearly all of its capac-
ity to secrete insulin. However, whatever the age of onset, in unstable, ketoacidotic
diabetes, sulfonylureas will not provide adequate control. such patients require
insulin, and attempts to control them with oral therapy are dangerous and
doomed to failure. Deaths from acidosis and dehydration have occurred in
patients with unstable ketotic diabetes in whom regulation was attempted with
sulfonylureas.
There is no fixed dosage of sulfonylurea to be used in diabetes mellitus. Treat-
ment is guided by the individual patient's response, which must be frequently
monitored with chemical determinations, because the requirements change from
time to time.
The mildly diabetic patient, whose insulin requirement is fewer than 20 units
daily, can be started on the usual dose of the agent chosen, and at the same time
all insulin is discontinued. The dose is then adjusted up or down, depending on
the patient's response. In the instance of chiorpropamide, about 3 days is required
to attain steady-state concentrations in blood. Consequently, upward adjustments
of dose should be made at 3-day intervals. Patients of advanced age should begin
with about half the usual daily dose, for some are very responsive to sulfonylureas
and may develop severe hypoglycemia after usual doses. During the period of
initiating treatment, all patients should test their urine four times daily and
communicate the results to the physician daily.
The patient who requires more than 20 and fewer than 40 units of insulin
daily should be started on the usual dose of the chosen agent and his insulin
dosage should simultaneously be reduced by 50%. Phereafter, guided by the
patient's response, insulin dosage is progressively reduced and eventually dis-
continued. Sulfonylurea dosage may need adjustment.
The patient requiring more than 40 units of insulin daily should be given the
usual dose of the agent chosen and his insulin dosage should be reduced by 25%.
Insulin is then cautiously withdrawn and eventually discontinued, and sulfonyl-
urea is adjusted according to the observed response. It is to be emphasized that
the chance of success is relatively poor. In the patient who requires more than
40 units of insulin daily, it may be desirable to carry out the attempted transfer
to the sulfonylurea therapy in the hospital to provide assurance against develop-
ment of dehydration and acidosis.
Stimulation of the pancreas of the maturity-onset diabetic can often maintain
these subjects under ordinary circumstances. However, when insulin require-
ments are increased, as fever, surgical interventions, or trauma, the sulfonylureas
are inadequate and the patient must be given Insulin to carry him through such
critical situations.
Weight reduction is of the greatest importance in the treatment of diabetes.
A vigorous effort must be made by the patient and the physician to reduce the
patient's weight as an integral part of diabetic treatment, irrespective of the
drug chosen.
Patients whose diabetes is not controlled by sulfonylureas from the initiation
of treatment are said to experience "primary failure." Patients whose diabetes
is regulated for a month or more after beginning sulfonylurea treatment, fol-
PAGENO="0169"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13419
lowing which inability to maintain control developes, are said to experience "sec-
ondary failure." The incidence of this type of failure may be very high, regard-
less of the agent chosen.
In patients with pancreatic islet-cell tumors, the blood glucose concentration
drops rapidly after intravenous injection of tolbutamide and remains low for
about 3 hours. A similar effect is not observed in other hypoglycemic states, and
tolbutamide administration can thus be used as a diagnostic test. Serum in~-
munoreactive insulin determinations should also be performed. Care is necessary,
since fatal hypoglycemia has occurred.
In addition, reports have appeared of the successful treatment of reactive
hypoglycemias due to a variety of causes with sulfonylureas (Anderson and
Herman, 1971).
BIGUANIDE5: PHENFOEMIN
Chemistry and Preparations. The only commercially available preparation in
the biguanide series of hypoglycemic agents is phenformin. Its structural formula
is as follows:
~
Phenformin
Phenformin Hydrochloride, U.S.P. (DBI, MELTR0L), is marketed as 25-mg tablet$
and as a 50- and 100-mg time-disintegration capsules.
Mechanism of Action. The biguanides differ significantly from the sulfony-
lureas in the mechanism of their hypoglycemic effect. Thus, phenformin does
not act by stimulating secretion of insulin by the pancreas, hypoglycemia is not
readily induced in normal human subjects, the concentration of insulin in the
plasma is not increased, and the morphology of the fi cell is uninfluenced.
Basically, three actions have been described. In vitro, phenformin, in relativelV
large doses, increases glucose utilization by enhancing anaerobic glycolysis (see
Williams and Porte, 1974). This is thought to occur as a result of, or coincident
with, an inhibition of cellular respiration. As a result, adenosine triphosphate
(ATP) concentrations fall and those of lactate increase. A second action of the
drug is to decrease gluconeogenesis (see Gordon and de Hartog, 1973; Haecke~,
1973). The third and most recently recognized is inhibition of intestinal absorp-
tion of glucose ard probably certain other substances as well; for example,
decreased absorption of vitamin B1, has been observed (Berger et al., 1972).
Phenformin does not act in the normal subject (at least as readily as it does
in the diabetic), presumably because the increase in peripheral glucose utiliza-
tion Is compensated for by an increase in hepatic glucose output.
Phenformin has been used experimentally to correct the hypoglycemia that
may follow abnormally rapid intestinal absorption of glucose (Permutt et al,
1973).
Absorption and Duration of Action. Phenformin is adequately absorbed frota
the gastrointestinal tract. The drug has a short half-life (3 hours) and a corr~-
spondingly brief duration of action. The hypoglycemic effect may be prolonged to
between 6 and 14 hours with the use of timed-disintegration capsules.
Toxicity. Phenformin may cause a metallic taste, nausea, anorexia, vomiting,
diarrhea, or cramps in some patients, particularly if the dose is greater than 200
mg per day. Reduction of the dose or withdrawal of the drug results in prompt
disappearance of the untoward reactions. Weight loss and weakness may som~-
times occur.
The cause of ketonuria during phenformin therapy has been the subject Of
debate. It is most common in patients with unstable juvenile-onset diabetes
treated with a combination of insulin and phenformin. While it may at times re-
flect an insufficient insulin dosage, at other times it is associated wfth normal
plasma glucose concentrations. Therefore, in patients taking both insulin ai~d
phenformin in whom ketosis develops, plasma glucose concentration should l~e
measured before the insulin dosage is increased, to avoid hypoglycemic reactions.
56-592 0 - 75 - pt. 28 - 12
PAGENO="0170"
13420 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The recommended treatment for ketosis with normal plasma glucose concentra-
tions is a reduction of phenformin dosage or an increase of dietary carbohydrate
intake. Increased concentration of lactic acid in the blood without ketosis has
been reported to occur in patients with severe renal or cardiovascular impairment
under phenformin treatment. However, the drug may not contribute to the
lactacidemia, since such severely ill diabetic patients may exhibit lactacidemia
even when treated with insulin. Results obtained with phenformin in the UGDP
study are discussed above.
Diabetic subjects with severe hepatic or renal insufficiency or congestive heart
failure are not suitable candidates for oral hypoglycemic therapy. Almost no
data are available concerning the effects of phenformin in pregnancy, and its
administration during pregnancy is currently not recommended.
Therapeutic Uses. Phenformin is used in the treatment of maturity-onset dia-
betes according to the principles presented above for the sulfonylureas.
The patient is started on two tablets, 25 mg each, one before breakfast and the
other before supper. The dose is increased until control of the diabetic state is
attained or until digestive disturbances limit further increase in dosage. The total
daily dose is usually somewhere between 100 and 150 mg. However, doses as high
as 400 mg per day are tolerated by some patients. A single 50-mg, timed-disintegra-
tion capsule may be substituted as the equivalent of two 25-mg tablets in divided
doses.
It is claimed that about 70% of maturity-onset diabetic patients who are im-
perfectly controlled by either a sulfonylurea or phenformin alone respond favor-
ably to the concurrent use of these agents (Beaser, 1960; Unger et al., 1960). The
fact that the sulfonylureas and the biguaniles act by different mechanisms to
reduce hyperglycemia lends support to this contention (see Breidahl et al., 1972).
However, since the indications for the use of either phenformin or a sulfonylurea
are now severely constricted, such combination therapy should represent the
choice of the physician who has exhausted every other alternatitive.
PAGENO="0171"
COMPETITfl~E PROBLEMS IN THE DRUG INDUSTRY 13421
SE WESTERN RESERVE UNIVERSITY `CLEVELAND, OHIO 44106
August 21,1975
Hearing Clerk
Food and Drug Administration
Room 4-65
5600 Fishers Lane
Rockville, Maryland 20852
Dear Sir:
I regard it as a privilege to have this opportunity to discuss
the proposed FDA labeling on the use of oral hypoglycemic agents.
It is indeed appropriate that such direct action has finally
being taken if the scientific basis for medical practice is to
have any real significance. The heated discussions that followed
the publication of the findings of the longest and largest
prospective controlled clinical trial in the history of therapeuti~s
serve only to point up the critical issues involved in the practice
of medicine today. These issues involve the crucial question as
to the basis for the judgement of the physician for choosing the
optimum treatment for the patients entrusted to his care. Is the
decision to be based on ttclinical impression", anecdotal stories
and wishful opinions, or will it be based on substantial evidence
from adequate, well-controlled clinical investigations? If modern
medicine is to have a firm foundation in basic and clinical science
the wisdom of the Drug Amendments of 1962 resulting from the
thalidomide disaster becomes crystal clear.
The scientific design, merit or validity of the UGDP study has
been amply confirmed by the most intensive and extensive reviews
in the history of medicine. When the evidence of excessive cardio-
vascular mortality first surfaced after several years of the trial~
outside consultants were brought in to review the data independent~y.
After the report was made public at the A.D.A. meeting in St. Loui~
in June of 1970, separate peer review committees appointed by the
American Diabetes Association, the A.M.A. Committee on Drugs and
the Medical Letter accepted the basic conclusions of the study.
Because of the alleged contradictory findings from other studies,
none of which approached in any way the magnitude or relevance of
the UGDP study an elite committee of the internationally based
Biometrics Society over a period of two years reviewed all these
controlled trials and in their report published on February 10, 1975
the J.A.M.A. caine up with essentially the same conclusions and
recommendations. This report like the UGDP reports will stand as
nonumental and classical papers in the history of clinical medicine.
3partrnent (if ~i4edicine
kesith' Hospital
PAGENO="0172"
13422 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
August 21, 1975 Page 2.
Whatever other fragmentary data that has appeared since 1970
have only served to support the general conclusions. It is
interesting to me that the total efforts of the Committee on Ca~'e
of the Diabetic has been directed to the criticism and denigration
of the UGDP study and that no sound scientific evidence has been
brought forth since the introduction of the oral hypoglycemic
agents in 1955 to show any long-tern benefit whatsoever. It is
a sad commentary that despite the expenditure of some 10-15
billion dollars by the diabetic public throughout the world over
these twenty years, the drug companies and their adherents have
failed to come up with any studies that adequately prove that
any reduction in morbidity and mortality has resulted from the
long-term use of their drugs. If the logical concept of scientific
proven benefit over risk as defined by the laws established by
Congress in 1962 is to have any meaning for the twenty million to
30 million diabetic patients 111 the world the strongest possible
warnings clearly stated should be implemented as soon as possible.
The UGDP studies were published in 1970 and yet by August 1975 no
clear labeling warning has as yet been issued by the government
agency responsible by law to protect the American public.
I would appreciate having a transcipt of the hearings of August
20th, 1975 when they are available.
Very truly yours
Max Miller, M. D.
Professor
PAGENO="0173"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY 13423
PRESEN1~ATION OF
SIDNEY H. WOLFE, M.D. AND ANITA JOHNSON,
PUBLIC CITIZEN' S HEALTH RESEARCH GROU~',
to the
FDA HEARINGS ON PROPOSED LABELLING FOR DIABETES DRUGS
August 20, 1975
It is now more than 5 years since the findings of the University
Group Diabetes Program (UGDP) were presented at the American Diabetes
Association annual meeting. Despite the findings of increased
cardiovascular mortality in patients taking oral diabetes drugs and
their lack of efficacy, their use has continued. Over the last 5
years, however, several of the largest diabetes treatment centers
have swung strongly away from using these agents.
The question is, why are 1 1/2 million American adult onset
diabetics being given drugs costing more than 100 million dollars
per year, killing an estimated 10-15,000 patients per year and
not having any demonstrable benefits in treating diabetes?
There are three reasons which appear to explain this irrational
state of affairs:
1. The pride of physicians
2. The inadequacy of most diet programs for diabetics
3. The profits of the drug industry
33 Pride
As candidly stated this year by Dr. Frank Davidoff -- Professor
of Medicine a~ diabetes specialist for the University of Connecticut
Medical School -- and probably unspoken but thought by thousands
of other physicians, pride plays a major role in the continued
prescribing of these drugs by doctors in the face of evidence to
the contrary. (Testimony - Sept. 19, 1974, Senate S~nall Business
Subcommittee).
"It is one thing to challenge the safety of a drug . . . . But
to be told that the drugs we had been giving to diabetics for 12
years were unnecessary, beside the point, in a word ineffective,
was, as I see it, a more serious blow to our professional pride.'
We submit that the pride of doctors is standin9 in the way of
giving the best treatment to their patients and that this is
irresponsible medicine if not malpractice.
~.L ~
According to Dr. Ethan Sims, diabetes specialist and Professor
of Medicine at the University of Vermont Medical School,
"If we grant that there are 1 1/2 million patients reportedly
taking oral agents and that 50% are grossly overweight, we have
750,000 patients with diabetes and obesity who are probably also
less physically active than they should be. If we assume that
90% of them are not exposed to any vigorous and comprehensive
regimen such as that at the Grady Hospital, 675,000 are left with
their obesity essentially untreated and 4 Out O~ 5 are taking an
agent which increases their obesity. The taking of oral medication
lulls both physician and patient into believing that something
worthwhile is being accomplished, while the options which could
PAGENO="0174"
13424 COMPETITIVE PROBLEMS IN THE DRtTG INDUSTRY
-~-
make a fundamental difference in a patients life and survival are
being neglected." (ibid, Senate Hearings)
Dr. John Davidson, Director of the Grady Hospital (Atlanta)
Diabetes Unit which has been very successful in getting diabetics
off of the oral drugs, has said,
"Why do so many physicians have little success in treating
it? (the problem of obese diabetics]. * . . In my experience, at
least 80% of diet therapy failures are due to physician failure,
not patient failure." (ibid, Senate Hearings)
3) Profits
In a letter to the FDA, requesting to speak at this hearing.
B .R. Allen of Upjohn commented that:
"The proposed class labelling of oral diabetes drugs.
is, in our view, inappropriate, uninformative, and hence mis-
leading.
Translated into English, Upjohn -- which has about 40% of
the $100 million American market for oral diabetes drugs -- doesn't
want its leading money-makers -- tolbutamide (Orinase) and tolazamide
(Tolinase) -- to suffer in sales andprof its just because scientific
studies show the drugs are ineffective and extremely dangerous.
That the extraordinary profits of Upjohn, Pfizer, Ciba-Geigy
and Lilly -- who have cornered this several hundred million dollars
per year worldwide market for these drugs -- has had a major effect
on the irresponsible delay in ending the massive misuse of these
drugs is not arguable.
In addition to sponsoring hundreds of "educational" symposia
around the country -- with academic facades -- intended to assure
doctors that these drugs are 0.R., the drug industry has kept the
AMA alive with infusions of advertising revenue and political
contributions and, in return, the AMA has written reassuring
letters (Dr. Sammons) to physicians about these drugs.
To demonstrate how much of a curtailment of profits would
occur if the abuse of these drugs were ended, cxnsider the experience
at Cleveland Metropolitan General Hospital. At the peak of use
of the oral diabetes drugs (1970) at that hospital, expenditures
per year were as follows:
Tolbutamide Phenformin
Q~naseJ~ _________
$32,376 $7,857
After use was restricted, the 1975 figures
were:
$6,966 $138.00
a savings of $38,000 or 78% of the $48,527
these drugs.
In the face of pride, profits, and inadequate dietary manage-
ment, aided and abetted by court delays, the FDA has somehow found
it possible to delay for almost 5 years finalized labelling changes
for these drugs.
Anita Johnson will discuss our specific criticisms of the
present version of proposed labelling (the third in 3 years) but
I would make the following addition.
Chiorpropamide
jpiabinase~
$8, 294
(projected by the hospital)
$3,371
1970 expenditures for
PAGENO="0175"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13425
It is not enough to have labelling -- albeit strengthened --
on a product which the patients doctor has already decided to
prescribe. Pat lenta oust be given information in the doctors
off ice which will allow them to participate in the serious decision
to use these drugs. Specifically, future use of oral diabetes
agents should be preceded by a written informed consent form
including, but not limited to the following information:
INFORMED CONSEWF FOR USE OF ORAL DIABETES DRUGS
1. 1 have participated in a program of dietary control
and physical exercise including at least 25 hours
of instruction.
2. This program did not succeed iD weight reduction or
control of blood sugar and Dr .___ __ told me
that insulin was the preferred drug if one had to
be used,
3. I refuse (or am physically unable) to ta)~ insulin.
4. I am aware of the increased risk of cardiovascular
death from taking oral diabetes drugs and of the
animal study showing that one of them (Tolbutanide)
causes a significant increase in coronary artery
disease and that therapeutic efficacy has not been proven.
In light of the above I agree to take
(oral diabetes drug).
Date Patient's Signature
Recommendations for the Label
1. The antidlabetic drugs should he indicated only for patients
with symptoms from high blood sugar, whose symptoms cannot be
controlled by diet or insulin. They should be used by patients
who cannot be controlled by diet ~nl~' if such patients also cannot
inject insulin.
The label as proposed by FDA approves use in a broader group
of patients those whose "symptoms cannot be controlled by diet
alone and in whom insulin cannot be used because of patient on-
willingness, erratic adherence to the injection regimen, poor
vision, physical or mental handicap, insulin allergy, employment
requirements or other factors." The FDA label essentially condones
use in symptomatic patients when insulin is merely ~cgf~_njfflt to
use. Mere inconvenience is not a legitimate reason, in our view,
to sustain the known risks of th~se drugs. The FDA label also
indicates use in patients without cymptoms.
2. The label as proposed by FDA grants an indication to patients
with high blood sugar who do not have symptoms The Food, Drug
and Cosmetic Act requires that all indication~ be supported by
"substantial evidence," evidence from "adequate and well -controlled
investigations, including ci inical investigations ` Obviously,
since these patients l~ye~ no short-term symptoms of diabetefr~, such
PAGENO="0176"
13426 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
-4-
as dizziness, polyuria, etc. the drugs are not effective in
treating short-term symptoms. Some diabetes doctors believe that
they prevent long term symptoms of diabetes such as vascular
deterioration. However, there is no scientific substantiation of
this belief.
3. The label should state that there is substantial evidence
that these drugs are ~ effective for the prevention of the long-
term symptoms of diabetes. FDA has proposed new regulations for
all prescription drug labels which contains the following re-
quirement:
If there is a cormeon belief that the drug may be
effective for a certain use or if there is a coimson
use of the drug for that condition, but the pre-
ponderance of evidence related to such use indicates
that the drug is ineffective, the package insert shall
state that there is a lack of evidence that the drug
is effective for that use, a. 1.112 (3)(d),
40 ~. ~ 67, 15392ff.
The UGDP study demonstrated that these drugs are ineffective
in preventing the long-term effects of diabetes.
4. The Warning section should state that these drugs ~
associated with increased cardiovascular mortality, without
referring to the TJGDP study. The proposed warning states that
the drugs !fl~ be associated with increased cardiovascular
mortality and that "This warning is based on the study conducted
by the U.G.D.P." together with details of that study. If the
UGDP reference is retained, the other studies which confirm the
tTOD? findings must also be cited. Otherwise, the warning will
give the false impression that the UGDP findings are isolated
and unique, which they are not.
Since the UGDP study, laboratory studies have pinpointed the
mode of adverse action on the heart. This mode of action has been
confirmed in humans by catheterization studies. Another study has
described a significant rise in ventricular fibrillation in
patients on these drugs. Two epidemiological studies have shown
an abruptly increased mortality among diabetics since these drugs
were introduced. Three retrospective clinical trials confirm the
UGDP results, as do two cohort studies, the Kanarek study, based
on patients at the Joslin Clinic, and the Palumbo study based on
patients at the Mayo Clinic.
5. The Warning section should not include a statement that there
is controversy as to the need for the warning. The FDA proposed
warning says: "Despite controversy regarding the interpretation
of these results, the findings of the UGD? study provide adequate
scientific basis for this warning, thereby including a statement
of controversy. The 1972 lawsuit, Bradley v. Weinberger, raised
this issue. Bradley, who had the habit of prescribing these drugs,
sued to prevent FDA from putting a warning on the label, or, if
that failed, to get a statement in the warning that there was a
difference of opinion among experts concerning the need for a warning.
The case was never decided on the merits. However, FDA contested
Bradley stating that warnings ahould not contain disclaimers which
would encumber and dilute the warning. Now FDA is reversing its
position and including a "controversy" statement. We agree with
FDA's earlier view that this statement is unwarranted and from
a health point of view, counterproductive.
Without these changes in labelling, the addition of informed
consent and immediate finaliz~,~ of these improved regulations,
the FDA, to the delight of the drug companies, will be condemning
American diabetics to a continuation of the needless death end
waste of precious h~al~h do,Liars,
PAGENO="0177"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1342~7
Oral Hypoglycemic Agents
* Are Worthwhile
ROBERT F. BRADLEY
Joslin Clinic, Boston, Massachusetts
The controversy evoked by the University Croup Diabetes Program (UCI)l'
results reported in December, 1970,1 and a few months later,2 has quite prnptrly
rekindled the interest of clinicians in the need for intensive dietary thvra~)y *`~
the adult maturity-onset diabetic and has provided another example of po.ssth!i
insidious effects of foreign compounds administered to humans. It has al..n
exhumed the more basic controversy, namely, that relating to the benefits, if any.
which can be gained from the rigid metabolic control of diabetes niellitns. a~
reviewed previously in the first edition of this text.3
The UCDP study by all odds had the best designed and the most lauclalk
objectives of any yet undertaken. Unfortunately as they "cyeballcd" the dali
week by week and month l)y month and saw first a cluster of deaths in patiiut~
treated with variable doses of insulin and then a somewhat larger cluster in thme
treated with tolbutainicic, the biostatisticians held sway. Clinicians, shaken by
their lack of expertise in biostatisties and forgetting that the study was nit
intended to evaluate mortality results,' bo~vecl graciously to the intonationS Of
those who extrapolated the data to the maturity-onset diabetic populatn)n at
large. The tragedy of this issue rests both in the possibility that the implicatu)ie~
of the UCI)P study are entirely correct, arki the equal pOSsil)ility (bitt they ii~
completely invalid, i.e., that the observed cardiovascular events resulted front a
ICj)OS itomy Of individi tals treated ~vith tolbutait tide or phenforinin Win) Were
greater cardiovascular risks at baseline.'
Regam(lk'ss of the many arguments that have l)een presented pro niitl Coil.
one must keep in mind the preliminary nature of the results, namely, that th~
total cardiovascular deaths Oceinritog in the UCI)P stu(ly represented only `~
pCJ~ cent of those diabetic pal it'nts Comprising the entire study pOptllittU)fl.
The sulfonylureas (carbutaniidc', tolbntaniide, chloipropanndc, acettth°'x.t
midle, tohoziunkle, glybenzeyclaznide) and biguanido's (phenforinin, inotfriin!l.
and bti fominin ) lower 1)100(1 glucose levels 1))' (Ii 1rmiitg flI('ClOtniSiflS. Soidi 80
effect of these oral hypoglycemic agents used singly or in combination itas boon
404
PAGENO="0178"
13428 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
J1OJWRT F. IU1ADIEY 405
well documented in appropriately sekctcd hyperglyccmic individuals and
accounts for their ~vidcspread usage in the United States and other countries
during the past 17 years. In many of the more responsive patic'nts with maturity-
~insct diabetes, normal 1)100(1 glucose levels are more readily attainable than with
(1U~t or insulin.
Despite numerous reports during the early years of their clinical use that
these compounds would effectively lower 1)100(1 glucose leVelS in 50 to 75 per cent
of matUrity-(~nset diabetics who ~~`crç not insulin dependent or kctoacidosis-prone
and whose diabetes began at age 40 or older, physicians familiar ~vitli insidious
long-term problems of the diabetic have from the beginning been concerned.
that patients SO treated might be less well controlled and more prone to premature
(le~'e1opmcnt of these complications than individuals treated with insulin. Experi-
ence has provided ample evidence that for one reason or another oral hypo-
glycemic agents lose their effectiveness at varying but relatively short intervals
of time after initiation of treatment. The rate of such "secondary failure" depends
upon many factors, including patient selection and dietary cooperation, thera-
peutic objectives, and the manner in which the oral hypoglycemic agents are
used. A lucid presentation of "primary" vs. "secondary failure" ~nd the effect their
definition has upon long-term "failure" has recently been published.8 The element
of convenience for middle-aged and elderly people is obvious, but always has had
to be balanced agains.t the increased cost for those who took more than minimal
doses and the possibility that physicians and patients alike would rely too heavily
upon their effectiveness, so that diet would be either ignored or less carefully*
lulloweci..
If benefit is to be anticipated from lowering blood glucose levels as ~vell as
reversing lipid, protein, and other metabolic abnormalities associated with insulin
(l('fkit, what should be the 1)100(1 glucose levels attained? From the early days of
their use, many sets of criteria have been utilized by those involved in the study'
of (lial)etie patients. In general, these have fallen into two categories: (1) those
who consider the oral hypoglycemic agents effective despite blood glucose levels
in excess of normal, provided the symptoms of diabetes have been relieved and
remain so, and (2) others, such as Marble and his associates, including this
auth ior, whose objective has been normoglyeemia and aglycosuria, in accordance
with the criteria originally published by Camerini-i)avalos et al in 1957S (Table
1). In defense of the former is the fact that in many maturity-onset diabetic
))at icuts whose 1)100(1 glucose levels remain elevated despite dietary adherence,
the addition of oral hypoglycemic agents lowers 1)100(1 glucose levels to a degree
comparable to that readily obtained with insulin and relieves syniptoiiis, so that
little would be gained by. insisting upon a more rigid standard of metabolic
control. Recognizing that evidence for time lx'nefits of tight metabolic control
renm:nns controversial, the adoption of such standards would seem to be reason-
able. On the other hand, if it is true that protection from long-term complications
is attainable only tlnough the more rigid control of 1)100(1 glucose levels, tIme
latter criteria slmoulci be applie(l, ~ind if tIme standards are not aLt mined, more
rmlenth&'ss application of diet and insulin if necessary is required. At l)meseiit (lie
fimmulanmemital eommtrovers~' continues to be that related to the possible benefits of
such control Considerable new t'vidcnce is availal)le today', unfortunately no
PAGENO="0179"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13429
OIIAI. I IYI'OCI.YCEM IC ACE~TS AIU WOItTI ~ !"~`~
Table 1. Josli:a Clinic Standard for Bk~d Cb:co.~' Control in Diabetic Patients
`i:rc:tcd wit/a Oral Ilypoglyecm ic Coin pounds: Standards of Control
0:1(1 Dcgrcc of (:o,zirolf
ft
ELATION TO
FOOl)
GOOD
FAIlS
Tilood
Urine
l3loocl
Urine
.
Sugars
~`nig./iOO an?.)
Sugar
(per cent)
Sugarf
(mg./lOO niL)
Sugar
(per cent)
Fasting
110
Trace
130
0.1
All
1 hr. p.c.
2 hr. p.c.
150
130
0.3
0.1
*
180
150
0.5
0.3
oIhe~
ca~cs
3 hr. pc.
110
Trace
130
0.1 -
°For purpose of classification as to degree of control 70 per cent or more of vahws mant
conform with standards listed in the table.
~ These standard values are the highest acceptable.
lChtcosc as determined by the Soniogyi-Nclson procedure.
mqre COnClUSIVe than the data of 20 years ago. Meanwhile the practicing phvsi.
clan is busily and assuredly attempting to lower blood cholesterol levels, l,ut i~
not certain how assiduously to work toward lowering blood glucose levels, and
if so, how much.
A major concern has been the possil)ility that these oral agents might pro.
duco or allow the earlier development of the following: (1) islet cell failure with
consequent decompensation of endogenous insulin function and greater activity
of the diabetes; (2) infections; (3) neuropathy; (4) cataracts; (5) onset or annie
rapid progression of niicroangiopatlay; and (6) greater morbidity and/or nmr~
tality from accelerated macroangiopathy, particularly that involving the coronar>~,
cerebral, and peripheral vaseulature. Because no good evidence had ever be~a
presented to suggest that oral hypoglycemic agents were "antidiabetic,' which
would mean that, they were inherently capable o delaying or pres'c~tit~g tluse
more serious problems, the only reasonable benefit to be expected would _lw
consequent to a net increase in the effectiveness of endogenous insulin, as inclI
cated by lowering of the 1)100(1 glucose level over and above that obtainable svit~s
diet. Thus, if siga~ifieant lowering of 1)100(1 glucose levels could not be attained
and then maintained, there would be no basis for using any of the currently
available oral hypoglycemic agents. The extent to wllicll oral hypoglycemic agents
have succeeded or failed with regard to these potential problems will be briefly
reviewed.
ISLET Ceu. FUNCTION
That snlfonylureas lower blood glucose levels primarily by increasing insuli~1
secretion from the paIlcreils la:is beeis dcfin~tely provcd.G I Iowcver, it )I~IS IUI1
been contended by some investigators that extrapancreatie tffects of st~lf0i Yl
reas, particularly those related to hepatic glucose release, contribute to (lac dEecl~
upon glncosc.G Recently considerable data have conrirnscd extrispancreatic 8cftC)lI~.
which are demonstrable in the absence of insulin at cellular sites and with cOl~
PAGENO="0180"
13430 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
JIOBEIIT P. IThADLEY. 407
centrationof drug cornpatil)Ic with the clinical setting.° In addition, a portioii of
the cifects of sulfonylurcas may relate to inhibition of cyclic AMP pliospiso-
dicsterase, with a consequent net increase in cyclic AMP.'°
One must also keep in mind the significant, and sometimes serious, hypo-
glycemia induced by sulfonylurea drugs. None of those currently available is an
exception. Some of these enhanced hypoglycemic effects of the sulfonylureas
have been related to the coincident use of other drugs such as salicylates, mona-
mine oxidasc inhibitors, phenylbutazonc, sulfonamides, sulfisoxazole, sulfapi (`na-
zole, coumarin anticoagulants, and phenyramidol.
Occasional patients on sulfonylurca drugs are suspected of having a rapid
loss of endogenous insulin function because it appeared necessary after a short
period of treatment to give insulin to control the diabetes. Such observations
Ol)VIOUSIy have suggested that the sulfonylurca might have accelerated the deple-
tion of pancreatic insulin. However, studies in animals chronically treated with
sulfonylureas have not supported ti us concept. Rather, there has been consistent
histologic evidence of an increase in the number of beta cell mitoscs, hypertrophy
of the islets, and an increase in mass of islet tissue.6' 9
No data have been presented to suggest that biguanidcs "wear out" the
insulin mechanism. The means by which l)iguanidles lower blood glucose levels
in (liabetics, but not in normal humans, remains unccrtain~ By whatever means
they net, these substances lower blood glucose levels in clial)C~iC il)diVidUalS
having some available endogenous or exogenous insulin, albeit more gradually
than is noted in responsive diabetic individuals following sulfonylurcas. Available
endogenous or administered insulin simply appears to be niore effective when
phmeuiformin or other biguanides are administered. Tn the absence of any demon-
strable effect of these compounds upon the pancreatic islet cc'll, it is not surpris-
ing there is no evidence thus far that (lial)etes is worsened ITiCtal)oli('ally by their
iulnuinistratiOn.
A number of studies have suggested that sulfonvlureas or biguanides may
ameliorate "chemical" or "latent" dialx'tcs6 (as dc'fined by the American ))iabctes
A~sociat ion''). A ill tough inconclusive, observationS of no adverse (`fk'cts have
now accumulated for a sufficient number of years to allow one to assume that at
h'ast no worsening of the diabetes is likely to be Produced.
lxrEc-rJoN
At one time the incrc~scd susceptibility of the diabetic to invasive local and
systemic infection accounted for an important portion of the niorhidity and mom'-
talitv among (lialmetics, \\`jthi ifflplovc(l control of diabetes following the avail-
ability of insulin and the proper tms~ of ~nitilaotic treatment, infections ifl the
(hal mt't ie 110w 1)OSC 01 mcli less of a prol koi
Recent studies have helped to clarify the iSstl(' as to whether the diabetic is
indeed muon' suseeptibk' to infection Defects in host dc'ft'nsc' (`all be r('hiit('d to
11 ic d `grt'es of hivpt'rglvc'inia and/or ketoacidosis. `" `~ Ui ~`Oi it r( ihled dial n'tes (if
short (lflratiOil may not 1w associated with, a great likelihood of infection, bitt
when present over a period of weeks and mouths tin' patient becomes more sums-
(`(`1ltibi~'. `I'lnms fur no studies have (`heady (lefimicd the critical degree or duration
PAGENO="0181"
COMPETITIVE PROBLEMS IN THE DRuG INDuSTRY 13431
408 OHAL IYPOCI.Yc;EM IC A(;Et~Ts Mm ~ORTI Iwl nu.:
of uncontrolled diabetes, but inctabohc COntrol IS well accepted as a (und.uuental
Palt of the preventive prograiii in avoiding fungal infection and acLive (uherctt-
losis, as well as l)acterlal infection. Perhaps the most common and easily
clen)onsLral)k example clinically is the persistence of Candida ii ifections 1)r0(1OC-
ing vulvovaginitis in the female 011(1 balonitis in the male, responding poorly to
specific therapy such as nystatin or gentian violet hut (Iramnatically improving
with the cessat )fl of or marked improvement in excessive glycosuria anti hyp~r-'
glycemia. Such improvement is readily shown to occur in the diabetic patient
responsive to oral hypoglycemic therapy, with rapidity of improvement com-
parable to that obtained with insulin. Similar responses can be obtained in certain
p~ttiettts with nearly normal endogenous insulin reserve upon application of diet
and, of course, with the adnunistration of insulin.
To date there has been no indication in patients responsive to oral hypo-
giycernic agents, when pioperly combined with reasonable dietary adhcren~e,
that flC\V infection or aggra~'ation of existing infection has occurred as a result of
the use of oral rather than insulin therapy.
NEUROPATUY
No controlled studies have proved that metabolic control prevents the devël-
opment of ncuropathy. Flowever, neuropathy, particularly the more severe types
such as amyotrophy, anesthetic feet, Charcot joints, and so forth, classically
develops in the adult who seems to have had a short duration of diabetes, l)Ut
who was unknowingly hyperglycemic for a period of time, or in the ~atient with
known diabetes of longer duration in whom therapy was inappropriate or inade-
quate. Although diabetic patients may at times have an exacerbation of sympto~ns
due to neuropathy following treatment of any kind, and although on oCcasion
hypoglycemia may itself induce neuropatimy, the usual clinical observation is that
following improved metabolic control by whatever means, many manifestations
of neuropathy improve sooner or later. -
Recent biocIieniieal data demonstrating the presence of the poiyoi patli~vity
in nerve suggest a mechanism by which increased ambient glucose concentratiOns
in the Schwann cell activate the formation of sorbitoi,'4 so that nerve function
may be compromised. To date the critical circulating blood glucose levels for
activation of this pathway have not been clearly demonstrated, hut its possible
major mctal)olic role in the production of neuropathy provides further evidence
for tile desirability of keeping blood glucose levels as close to norma! as is
readily obtainable.
CATARAcTS
At least two morphologic types of cataract OCCUL in diabetes. These are: (1)
the snowflake, flocculent, or mnetai)nlic cataract, occurring mainly in juveniles
with grossly uncontrolled (liIll)(tCs; and (2) the senile cataract due to sclerosis
of tile lens nucleus, indistinguishable from that ~eeu in the nonchabctic and the
COUullOflest type Ol)Served in adult (hal)etie 1)atidllts. The inure rapid maturation
of senile cataract in the diabetic than in the nondiabetie has recently been ic-
PAGENO="0182"
13432 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
flOJ3EIIT F. DflADLEY 409
emphasized by studies showing grossly poorer dial)etic control in patients having
cataract extraction than among tho average patients attending a diOl)etie clinie.~'
These obseivatioiis, and demonstrations of the appropriate enzymc~ in the lens
of man for activation of the polyol pathway' by existing byp(1glyCemiiL,1~' `~ lend
considerable support to the long-held Opinion that poorly controlled diabetes is
a factor in the rate of maturation in ~emlc cataract as well as in the developt~icnt
of metabolic or snowflake cataract.35' ~
Although of lesser importance, the ~vcll known relation of refractive changes
in the eye of the diabetic to changing blood glucose levels would be still another
basis for adequate blood glucose control, at least in terms of the quality of daily
living.
M1CI1OAr~GIOPATKY
The microangiopathy of diabetes involves small blood v~ssels, particularly
capillaries supplying many tissues. The possible value of careful metal)Ol ic control
in protecting the inclivi~lual from clinically important retinal and/or renal vascular
disease has been the subject of major controversy `fo~;25 years. The UCDP study
was directed in major part towitrd seeking an flnswcr to this question. Thus far,
neither the prospective UCDP study nor otbe~ studies, all wholly or in part
retrospective in nature, have proved conclusively that significant benefit is to be
gained from any tighter control of the metabolic components of dial)ctes than is
necessary' to avoid the symptoms of diabetes, ketoaciclosis, and so forth. Detailed
reviews of this subject have either supported no relationship between careful
metabolic control and the pre~'e~tion of inicroangiopathy,'9 or indicated that such.
control improves the chances of preventing severer grades of clinical micro-
angiopathy, such as retinitis prohifcrans and/or nephropathy with renal failure.20
Despite the. lack of unanimity concerning this controversy, recent biochemi-
cal data tend to shift the weight of evidence in the direction of favoring tight
lnetal)olic control. In particular, the observations of Spiro regarding the role of
hyperglycemia as a stimulus to the biosynthesis of basement membrane material
of the renal glomerulus,~' and observations of l)aselncnt membrane thickening
shto~ving an apparent correlation with tlur~itiOii of insulin deficit,2 appear to
bolster tie practice of ti ose 1)1 ~vsicians who strive for normoglycci nia.
Meanwhile, although not specifically related to eflects upon microangio-
patliy, sufficient data have accumulated to support the ,`ole of striving for
normoglycemia as assiduously as possible in assuring survival of the fetus of
the diabetic mother.2"
MACl~oANC1orATIlv
Although neuropatl~y. increased susc'ept il3ihity to infect ion, and inicroangio-
path alt' the most s1)(Ciflc manifestations of diabct(s ,,~(`lhit~is at;~l ame of pam'-
ticular concern bc'eiumse of their ~ulvc'rse efi&'cts up()im mmnnv )`01113g('1 i)at i('flts, t he
overall greatest problem in terms of I1lOll'idit\' and ,mimmt'tahitv is that iclated to
involvement of inc'dii,mu ammd larger vessels, espreiallr' the cOronary, der(i)ral all(l
lower m'xtrt'nmity arteries. TIme prtvaknee of suelt vascular lesions is high in the
PAGENO="0183"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
1433
410 ORAL ILY1'OCL\CLMIC ACENTS AI1E \VOIlTIt\VII1LE
Table 2. Causes of Dcatli in 912 Diabetics
(Per Cciii of 701(1! Dcal1is,~
1966_1969*
Va~cu1ar disease, total 74.3
Cardiac ~4.6
Renal 8.0
Cercbrovascular 10.0
Cangrenc, "circulatory" 1.7
Cancer 12.8
Infectious, Non-TOC
All others . 0.0
°Expericnce of the Joslin Clinic.
gcne~al popuhrtion, but cardiovascular disease as a cause of death is nearly
doubled in the diabetic. Table 2 shows that cardiovascular causes account for ~tt
least two thirds of the mortality in the diabetic population as a whole. With such
a high frequency and with so -many factors other than diabetes playing a potential
role, any attempts to a.~sess the possible benefits of diabetes treatment a~c
extremely difficult to evaluate. An additional problem in trying to judge t~ie
effects ~of therapy directed toward improved diabetes control is that in the adult
maturity-Onset dial)etic the duration of hyperglycemia is extremely difficult to
ascertain, except in those individuals who have been subjected to blood glucose
measurements from early in life either because of a family history or as part of
ioutine examinations.
Many observations relate vascular disease of medium-sized arteries to tl~e
presence of hyperglycemia.2' Perhaps the earliest, and certainly, one of the most
striking, has bccii derived from the huge autopsy series of Bell at the University
of Minnesota27 (Table 3), . which points out the extraordinary prevalence of
peripheral vascular disease and gangrene in the hyperglycemic individual. Such
associations appear to ju~tify efforts of physicians within, reason to provide
"metabolic control" of diabetes.
In the clinical use of oral hypoglycemic agents, the assumption has been
that lowering blood glucose would reverse the metabolic abnormalities related
to insulin deficits, such as those in protein and lipid metabolism, much as such
defects are reversible with comparable degrees of blood glucose lowering follow-
ing insulin. As has been summarized else~vhere,28 various types of circulating li1~id
abnormalities are reversible with sulfonyluicas in those patients who have suffi-
cient endogenous insulin, such that blood glucose levels fall to normal following
the administration of one of these oral agents. On the other hand, a number of
Table 3. Results of /tulopsies Following Ailicrosclcrotic Gangrene°
NONDl\1~ETiC (50,733) m~nEric (2130) n.vrio or
(PER CENT) (eRa CENT) ritia~uExcr
----- ~
,1 F M F l)iabetic/No,,dialwtk
Age
20-40
0 0
.3.4 0
All Ages > 40
40-CO
0.1 0.03
14.7 . 14.0
M 53:1
00-80
0.45 0.4(3
24.3 24.0
.
F 71:1
°lroin Bell, E. T.: Amer. J. Cla~. Path, 28:27, 1957.
PAGENO="0184"
13434 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
flO8Ei~F F. BRADLEY * 411
reports have demonstrated persisting abnormalities in cholesterol, free fatty acids,
.ind triglycerides in individuals who were reccMn~ a sulforiylurc'a, lmt on close
inspection of the (btta, "control" was dcterinrncd only by measurements of fasting
blocnl glucose levels, and even these were persistently elevated. However,
,sbnorsnahtms in circulating lipids associated with imperfect management of
~li..btvs using oral hypoglycemic agents have been shown to be reversible by the
a'l(l)tiO') of sufficient iiisulm to improve blrmi,d ~1ucose lcvels.2~ This Ol)5('rVatiOfl
has l)eCn one of the factoms which have supportei.l the use of more rigid 1)100(1
glucose criteria OS an Ol)jCCtiVc in treatment with oral hypoglycemic agents, be
tIny sulfonyhircas or biguanicles. Thus far there is no clear_cut: evidence that
),loocl lipid abnormalities clue to relative insulin deficit arc more or less likely to
Ime preventable or reversible at comparable blood glucose levels w'hcther insulin
or oral hypoglye~mic agents are uscJ.
A real problem has been the tendency for physicians to use oral hypogly~
c'cmnic agents not in an ideal manner, but rather f~r the sake of convenience, with
too little emphasis upon diet, adequate choice or dosage of the agent used, or
p~~p~'r selection of the patient. In such situations. obviously, the performance of
oral hypoglycemic agents should be less effective than that of insulin, assuming
ilmat control of blood glucose and lipid abn~irmalities arc' indeed important in
slowing clown progression of macroangiopathy.
`l'be above observations arc critical in evaluating studies such as the Univer-
~ity Group Diabetes Program (UCDP'~. which recorded more cardiovascular
th'atlms in patients receiving tolbutamicit' or phenformin than in those treated
with diet and placebo, diet and standard dose of insulin, or diet and a variable
dose of insulin. The res~ilts ma~' be interpreted as follows: (1) If it is true that
tilhuitamide, as a rosult of an inotropie effect up~m the myocardium2~ or via some
other .flI('Ch:lnisifl. and an unrelated compound such as plienformnin, tluough some
ummith'ntifk'd mechanism, actually contribute to ~`ardio~'aseular death, the serious-
mn'~s of this particular end point would w~'li~h ~) heavily that the use of these
i~iai hypoglycemic agents should be sumuni:uily discontinued. (2) On the other
imumuil, if these oral agents were secmninglv less dFective because of their improper
umw. time question is \Vhether the results wouhi be iniproveci by correct usag(' and
what time criteria should be for such ~ ,,3-~ Time third pOsSil)ility is that
imm.ulvertent signilicant differences in basdine cardiovascular risk factors
mieomnmted for the less favorable CardiO\a5e~l3m' °~°~Y ~ ~n
treated with tollMmtamidc or phenforimmi ai~I that the sttmcly doc's not prove or
uikpmuve lack of effectiveness for tolhut;itiitdt' imp to the tutU' it ~~`as discontinued
(P1)1mm tin' st nd (October, 1969) or for pluemmternmimm ( dliSCOittimltm('d January, 1971 ).
I In' hitter is more than a mere possibility. for time imiterpretat omis of UC1)P results
I y the investigators, the Anmc'rican Diabet ~`s .~ss,idat ion," the Con ncil on Drugs
`1 liii' Aim u'mica n Mccl ic'al Assoc'iatiomi_~ and the U.S. Food and Drum g Ad ni nist ma-
tiummi'- ~ based upon statistical grounds tlm.ut do not take into acdOmlimt tin' clinical
I'.utkguommn(1 (if kl)(>wledg('. cOmicermming coromi.mry ln'am't disease in the (lid (`te. Time
$ll,ul)\ fI,usvs iii the UCI)P study mmk~' ammv t'~tuapul:ut ion of time results to time
(Im,mi)ttic popumlutiom at large o'xtrm'miit'lv ima.'.irdous. amid a ntmumll)er of Ol)jeC'ti(mmmS
r('uuma,mi Ippareimt to the clinician:
I. lit idact'lm trcat('(l p;mticnts not ii single mimyocordial immfarctiomi ~~`as ru'corckcl
ammmommmg time c'amolio~'ascomlim' (baths.
PAGENO="0185"
COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY 13435
412 OT1AT. Il\L~OCE.YCL~%ttC Ac~:NrS A1IL~ W01~TIIWJU1.O
2. A characteristic feature of cardiovaseuhtr mortality in the diabetic is the
female predisposition to dying, o'qualiiig or exceeding that Ol)served among
Inales.2h UC flP findings of fewer than one half as many cardiovascular deaths in
females as c upared to mniiks in plitceb() treated diabetics indicated a hick of full
eXPI,CSSiOO of tlIC efrCcts of oIi~ibetcs upon cardiovascular mortality in this group;
i.e., the diabetes was milder and/or of shorter duration, or the numberS were
too smiill.
3. Cardiovascular risk factoms at baseline were present ~~ti1i greater frequc'm~cy
in patients on toihutanude than in those treated with placebo or with small fixed
doses of insulin ("insulin standard"). Although univ one factor, blood cholesterol
levels equal to or greator than 300 nig./ 100 ml., reached statistical significance in
its greater frequency in toll)utanhidc treated patients, out of a total of 10 baseline
risk factors, nine occurred more often in the tolbutamide treated than in the
placebo group. On comparison with "insulin standard" treated patients, seven
cardiovascular risk factors were present at baseline with greater frequency in
tolbutarmde treated patients as compared to three affecting mnoic patients in the
"insulin standard' group.
4. The prevalence of coronary heart disease as evidenced by `~signifkant
ECC abnormality" at baseline was extremely low in all treatment groups (Table
4's. especially in view of time higher frequencies of digitalis usage, qf angina
pectoris, and of ECG abnormalities in diabetics of comparable a~c reported in
the litcrature.~ ~ When. the original baseline findings were reported by the
UCDP investigators in 1967, different criteria for ECC abnormality were used,
so that on the basis of the electrocardiogram, a distinctly greater prevalence of
coronary heart disease was reported in time tolbutanmide as compared to placebo,
insulin standard, or insulin variable groups of patients.34
5. The duration of diabetes among patients entering the UGDP study were
indeterminate. I-Iowcver, elevated fasting blood glucose and greater increments
of post glucose l)lOOd levels were found in more tolbutanmide treated patiemmts at
baseline than in any other treatment group. Therefore, more tolbutan mide treated
patients had severer and/or longer durations of dial)ctcs meihitus.
6. For an end P01rmt~ (cardiovascular mortality) having an extremely high
prevalence in the diabetic population in which a number of risk factors, both
Table 4. UCDP Stud y~W6P vs. 1970f
Selected Baseline Cardirnxiscular Risk Factors
INSUlIN' INSULIN
rLAcFtmo T0LIWTAMn)md STAND. VAISIATITE TOTAL
No. Per Ccut No. J't'r Gnu No. Per Cent No. Per Ccnt No. Per Cent
Significant ECC'.
abnormality 1967
30
(15.2)
48
(210)
40
(19.3)
39 (19.3)
157
(~94)
Significant EGG
abnormality 1970
llntory of digitali' inc
histOry of anasna pectoris
-~
6 ( 3.0)
9 ( 4.5)
10 ( 5.0)
8
15
14
~
( 4.0)
( 7.6)
( 7.0)
~
11
12
.16
( 5.3)
( 5.8)
( 7.7)
8 ( 4.0)
10 ( 5.0)
7 ( 3.5)
~
33
46
47
( 4.1)
( 5.7)
( 5.8)
°1tnfcrence 34.
I Reference 1.
56-592 0 - 75 - pt. 28 - 13
PAGENO="0186"
13436 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
flOBEflT F. BflADLF.Y 4j3
known and unknown, were present, the number of individuals reaching that end
point was too small to permit a definitive conclusion.
Sutnnia;y
The benefits of oral liypo~lycemic agents arc limited to those diabetic
patients who arc responsive, in that symptoms of diabetes are al)scnt and blood
glucose levels are significantly and consistently lowered (20 per cent or more)
below pretreatmc'nt values. Under these conditions, benefits may be summarized
as definite or qualified.
DEFINiTE
1. Convenience.
2. In those with diminished vision, arthritis, or other problems, who find
injection of insulin difficult or impossible.
3. For individuals whose diabetes is not controllable by diet a~id whose
employment and/or economic status might be jeopardized by the taking of
insulin.
4. In certain patients with allergy to insulin, in whom desensitization is
dirncult or cannot readily be maintained.
5. In truly responsive diabetics in whom iiormoglycemia is more readily
attained than with insulin.
QUALW1E1)
If lowering of lipid and other metabolic abnormalities related to iflSUlifl
deficit are important in protecting the diabetic from earlier progression of vrtseu-
Jar lesions and neuropathy, as well as from infection, the use of oral hy~)oglyceInie
agents is of benefit in patients who attain "sjgnifk;uitjv" lower blood glucose
values than arc attainal)le 1)V USC of diet alone. Mv criteria for such 1)100(1 glucose
values are that on two out of three occasions the blood glucose, ~v1iencver drawn,
is normal.
Data from the UGDP study have thus far COntTil)Ut(d nothing to the contro-
\Tc'I5)' rcgar(ling the Cth'(hvclleSS of 1)100(1 glucose (OI)tlOl afl(1 are ~uf1\cic'nt1y in
(loul)t as to t1i~.' apparent lesser benefits of tolbutamide audi ph(nformm OS COOL-
pared [0 dicE alone or diet and insulin, so that the results c~ituiot at l)1eSeIlt 1)C
(Xtrap(~latc(1 to the diabetic population at large. They (10 not ~varr~mt discontmua-
(ion of the appropriate routine clinical use of oral hypoglycemic agents.
References
L Ti,' UnivcrsLv (rtu,~ i')alit,s Pograni: A stmlv of (1( (.1f('Ct~ Of i,Vj)O~t\(('1)ic aernts 011
~`asut,iar (impi i:t iu~ ~, pt ii it ~vitI, adult ~ot (1 iahtcs. ii. ~ it lOy results:
i)ialn.is 19( Sppl. 2) :7S~. 1970.
PAGENO="0187"
COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY 13437
414 OflAt~ 1 IYPC C1 .YC1~~t IC AC1~ rs Ant: \VORTI IWI 11L1~
2. Knattcrnd, C. L, Meinert, C. L, Klimt, C. IL, 0.borne. 11. K., and Martin, 1). 1k: Fifeets
of hypoglycemic agezits on vascular complkati'ms i~ patitnt' with adiilt-oi*set (Ha-
bitt's. 1\'. A L)r('ltlIIlry re))Ort no pht'ntonoio ti'~'iIt~. j.A.M.:~. 217:777, 1971.
3. Marble, A. Contr~1 of dial ides lt',.,~eiis or postponis vaseisl:ir compiieatinhls. In l'lflog~r,
F. J., lteioian. A. S., and I"iiiland, M. (eds. ) Coulrorerxij in I,itrr'uzI .~t('dkihC.
Philadelphia, W. B. Saiiiiders Company, 1966, p. 191..
4. Feiiitein, A. 11..: An an:tlytie appraiiai of the tJnivc'rsity Croiip ))iabetes Program (UCI)P)
Study. Cliii. Phannaeul. `Iiivr. 12:1 67, 1971.
5. keen, IL, and Janet t, It. j.: F1ItCIS of oral hy~)'.iycemic agents on cardiovascular (IISeZ1SC.
In Fajans, S. S., and Snssmsu, K. E. (ecls. ) : I)iabcft'x ,\It'!Ii!us: l)iagno~i.~' and Treat-
711(111. Vol. 111. Nev,' York, American I)iabttcs ,\ssociatiou, Inc., 1971., p. 167.
6. Krali, L. P.: The oral lmypogiyecmic agents. In Marble, A., ~VIiiIe, 1'.. Itradley, II. F., and
Kraii, L. 1'. (eds.): Jo.'din's Diabetes Mdllitus. 11th ed. Philadelphia, Lea & Febiger,
3971, p. 302.
7. Baloditisos, M. C., Cleason, 11. E., Bradley, R. F., and Marble, A.: Long-term tolbutamnide
therapy in diabetes. A co~~tioli~'d study of tin' frcjumemicy of cardiovascular disease
mmml other findings. In lInt t~r(~cid, \V. j. 11.. amid Van W'esteriog. \V. (eds.) : ToIbi~ta-
wide . . . after `len Years. l'roc,'ediogs of the Brook Lodge Symposium, Augusta,
Michigan, March 0-7, 1907. Amsterdam, Excerpts Medica Foundation. 1967, p. 270.
8. Camcriui-Davaios, 11., Itoot, H. F., and Marl)1e, A.: Clinical cxpenenccs with carbutarm~ide
(BZ-55). 1)iabetes 6:74, 1957.
9. Lebovitz, 11. E., and Feldman, J. M.: Oral hypoglycemic agents: Mechanisms of action.
In Fajans, S. S., and Sossman, K. E., (eds.): Diabetes MelIitmi.r: Diagnosis and Treat-
ment. \`oi. III. New Yoik, American Diabetes Association, Inc., 1971, p. i'll.
10. Roth, J., Prout, `F. E., Coldfinc, I. I)., Wolfe, S. M., \Iuenzer, J., Crauer, L. E., and Mar-
cus, M. L.: Sulfonyiurcas: Efiects in vivo and iii vitro. (NuT Conferences.) Ann.
Intern. Med. 75:607, 1971.
11. Fajans, S. S.: Classification and natural history of genetic diabetes mellitus. in Fajans,
S. S., and Smmssunan. K. E. (eds.): Diabetes ,\Iellitu.c: Diagnosis and Treatment. Vol.
III. New Ynuk, American Diabetes Association, lime., 1971, p. 89.
12. Bagdade, J. I).: Infections. In Fajans, S. S.. and Sussmnan, K. E. (eds.): 1)iabetcs Mcli jt us:
Diagnosis and Treatment. Vol. III. New York, American Diabetes Association, Inc.,
1971, p. 211.
13. Younger, 1)., and Iladley, W. 13.: Infection and dtal)etes. In Marble, A., White, P., Brad-
ley, It. F. and .Krall, L. P. (eds.): Jo.slin'.s Diabetes ,`mlell4tus. 11th cd. Fhiladelpbia,
Lea & Fcbigcr, 1971, P. 621.
14. Prockop, L. I).: Diabetic neuropathy. In Fajans, S. S.. and Snssman, K. E. (eds.): Dia-
betes ~`m!elIitizs: i)iagnovis and Treatment. Vol. 111. New York', American Diabetes
Association, Inc., 1971, p. 347. -.
15. Clinical aspects of cataract in diabetes. in Caird, F. I., Pine, A., and Barnscll, `1'. C. (eds.):
1)iabctcs an(i time Eije. Oxford, lllacksvell Scientific Publications. 1969, p. 127.
16. \Vinegrad. A. I., Cknients, It. S., and Morrison, A. I).: Jim Fajaus, S. S., and Sussnian,
K. E. (cds.): Diabetes Mellitims: i)iagimosis and `I'rcatinent. \`ol. LIE, New Yomk,
American l)inbctes Assoeiatiøn, lime., 1971, ~. 209.
17. The nn'tabolism of the l(.nS ill rm'latoii to (`us opacities in diabetes. In Caird, F. I., ~irie,
A., and ltamn.ccll, `F. C. (cds.): Diabetes and tIme Eye. Oxford, Biackwell Scientific
Publications, 1969, P. 140.
18. Bradlcs', U. F'., nod Itanmos, E.: `tin' eves and diabetes. In Marble, A., White, P., Bradley,
U, F., and K'rail, I,. P. ( eds. ) : Jo.sIin's Diabetes Mellilims. I ith d. Philadelphia, Lea
& Felmiger, 1971, p. 478.
19. Knoxs'les, II. C., Jr., Guest, C. NI., 1.amnpe, 1. Km.'ssler, Nt., and Skiilniaim. 1'. C.: The course
of juvenile diabel is treated wit ii m:mllmlt'aSml r('(l d mt. 1)iabeti's 11:239, 1905.
20. Ntarl)le, A.: Augiopmthy iii diabetes: Aim uu~nlved prnlmleimi. ( `[`lie lkm:ting Ntcm(mrial Lee-
tore.) Diabetes 10:825. 1967.
21. Spiro, 11. C,: Pnthoph:y~iolocy. Ii. lIin~i mimmical 1)asis of tin. diahet k' mmmkroaimginpathy. In
Fajamms, S. S., and Smmssmaim, I. F. ( isis. ) : J)iolwtr.s ,~hllUmm.s: I)iattmmo.sis (mud `treat-
wemit. Vol. Ill, New York, Amneric:mn I)iai,'tes :\ssm)ci:mti(m:m, Inc., 1971, p. 275.
22. Spiro, It. C., aim:1 Spiro, NI. 1. : Eli mit of (li:mlutes oii liii: hiim.yiitlmesis of time renal ghmmnn:rnlar
basi'mn,'imt mnimmmlmr:mimm. I)inbmti'. 20:6 II. 1971.
23. Kilo, C., Voglm'r, N. J. mmmd \Vilhi:mmimsuii, J. U.: Bcmu'iimmuit umeiuliramu' tl,iekm'i:iimg in diabetes.
Jim J"ajans, S. S.. amid Su'.soma:m. K. F. ( eds. ) : Diabetes ,IImlIitm:v: I)iagmmnsi~ arid
l'tC(Itl?ietlt. Vol. Ill. New \`mmrk, Amnimican 1)imhctcs Assochmtimmn, lime., 1971, p. 289.
PAGENO="0188"
13438 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
ROBERT F. RUAI)I.EY 415
21. ~,Vhife, P.: Pregnancy anti diabetes. In Marble, A., White, P., Bra(lhy, 11. F., and 1(rail,
L. P. (edit.): Jo'Jin's J)kthctcit A~JdIa1u.s. 11th ((1. 1'Iilaclclphia, J~ ~ Febiger, 1971,
p. 581.
2.3. J'ctlersen, J.: Management of the pregnant diabetic. In Fajans, S. S., and Snssman, K. E.
(edit.): Diabcf~s 4~1cI1iti,s: Diognoitiv and Treatunut. \`ol. 111, Ncw York, American
Diabetes Msoeiatiun, Inc., 1971, p.235.
20. Bra(iley, 11. F.: Cardiova~cu1ar disease. In Marble, A., White, P., Bradley, U. F., and KralI,
I.,. P. (edit.): )u.slin's 1)iabek.s ?~!ellitus. 11th ccl. Philadelphia, Lea & Febiger, 1971,
p. 417.
27. Bell, E. T: Atherosclerotic gangrene of the lower extremities in (iiabctic and flOn-(iiai)etiC
persons~ Amer. J. Clin. Path. 28:27, 1957.
* 28. Bradley1 11. F.: Cardiovascular disease. In Marble, A., \Vhitc, P., Bradley, 11. F., and Krail,
L. i': (edit.): Joslin's Diabetes .\fdllhtu.s. 1 itiL ccl. Philadelphia, Lea & Febiger, 1971,
p. 429-430.
29. Lev~y, C. S., Palmer, 11. C.. Lasseter, K. C., and McCarthy, 3.: Effect of toibutamide on
adenyl cyclase in rabbit and human heart and contractility of isolated rabbit atria.
3. Clin. Enclocr'. 33:371, 1971.
30. flieketts, 11. T.: Editorial Statement. October 7, 1970. Diabetes 19(Supt4. 2):iii-v, 1970.
31. A.M.A. Council on 1)rugs: Statement regarding the University (;roup Diabetes Program
(UGDP) Study. Xovember 2, 1970. Diabetes 19( Snppl. 2) :vi-"ii, 1970.
32. Food and Drug Adminisliation Current Drug Inforntat ion Bulletin, October 30, 1970.
33. Bryfogle, 3. W., and Bradley, 11. F. The vascular complications of diabetes mc~Jlitus. A
clinical study. Diabetes ~:150, 1057.
34. }~limt, C. R., Meinert, C. I.., Miller, M., arid Knowles, H. C.: University Croup Diabetes
Program (UCDP): A study of the relationships of therapy, to vascular and other
complications of diabetes. in Butterfield, \V. 3. II., and Van Westerliig, %V. (edit.):
Tolbutamidc . . . After Ten )`cars. Amsterdam, Excerpta Medica Foundation, 1967,
p. 261.
PAGENO="0189"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 134~39
[From the Washington Post, July 9, 1975]
AMA OFFICIAL LET COMPANY USE NAME
(By Stuart Auerbach)
A top official of the American Medical Association allowed drug compa~iy
salesmen to use a letter from him as part of their effort to persuade doctOrs
to continue prescribing a controversial drug that has been blamed for as
many as 15,000 unnecessary deaths a year.
Dr. James H. Sammons, AMA executive vice president, gave his permission to
Upjohn Co. salesmen in March despite an opinion from an AMA lawyer tl~at
it is against the association's policy to use its name for business purposes.
The letter from Sammons downplayed questions about the safety of widely
prescribed anti-diabetic drugs that had been raised in the AMA's magazi~ie,
the Journal of the American Medical Association.
His letter and the opinion of AMA lawyer Betty Jane Anderson were made
available to newspapers by a man who was identified himself as a former A1~tA
employee dismissed in a staff cutback. The documents were verified by AMA
officials.
In a Jan. 28 letter, Sammons warned about 400 medical society executiVes
around the country that the Feb. 10 issue of JAMA would contain a confiri~ia-
tion by the Biometric Society of a 10-year study that showed that some anti-
diabetic drugs, known as hypoglycemics, cause more people to die of heart dis-
ease than they save from dying of diabetes. The society is an impartial group
of statisticians dealing with medical issues.
The original study, done by the University Group Diabetes Program (UGDI?),
was hotly contested by many diabetes specialists when it was released four
years ago.
But a large number agreed with the UGDP findings, and the U.S. Food and
Drug Administration last week ordered that advertising for the drugs n~ust
contain the warning that they may cause death from heart disease.
The drugs are used by an estimated 1.5 million Americans to lower tbeir
blood sugar, and represent an estimated $100 million-a-year business for the
pharmaceutical industry. Upjohn sells two leading brands of the drugs-
Orinase, for years the most widely prescribed oral hypoglycemic pill, and Toli-
nase.
Upjohn said it wanted to use Sammons' letter "as a result of the confusion
from the Feb. 10 JAMA story." It would be used by Upjohn salesmen in dis-
cussions with doctors "should the subject arise."
Sammons' letter reported that an editorial in JAMA wo"l'1 ~y th~ `-`~r~ `~-r~
probably associated with "10,000 to 15,000 unnecessary deaths" a year in the
United States.
The editorial was written by Dr. Thomas C. Chalmers, dean and president of
the Mt. Sinai Medical College in New York and former director of the Natiönal
Institutes of Health's Clinic Center in Bethesda.
Nevertheless, Sammons wrote:
"A considerable body of expert scientific opinion contradicts these publiehed
findings. Diabetic patients should not be influenced by press reports, and shOuld
continue on whatever diabetic management program their own physician has
prescribed."
When Upjohn asked to distribute the letter to its salesmen, AMA attorney
Anderson wrote Sammons, "The policy of the AMA is that the association's
name may not be used for trade purposes.
"Permission to reprint AMA materials has not been granted if there is any
indication that the name of the association or its materials will be used in any
manner that might directly be construed as an endorsement by the AMA of a
particular product or manufacturer."
Later in the niemo, she warned Sammons that the use of the letter of Upjohn
salesmen "may cause embarrassment to him personally or to the AMA."
"Upjohn's purpose could be better accomplished by having an article present-
ing the other side of the controversy published in JAMA," she said.
On March 17, JAMA published a letter by Dr. M. Hubbard Jr., Upjohn's
president opposing the AMA editorial and the Biometrics study, and on May
26 it published a series of letters and articles from other doctors who believe
in the drugs.
PAGENO="0190"
13440 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
THE UNIVERSITY OF CHICAGO
DEPARTMENT OF STATISTICS
1118 EAST 58TH STREET
CHICAGO * ILLINOIS 60637
17 February 1975
Senator Gaylord Nelson
Chairman, Monopoly Subcommittee
Senate Small Business Committee
424 Russell Senate Office Bldg.
Washington, D.C. 20510
Dear Senator Nelson:
I have revts~ed the transcript of my testimony before your Com-
mittee on 31 January and I think it may be useful to restate the
position I was supporting in our exchange.
First, it was entirely proper for the UGDP investigators to
withdraw Tolbutamide from the study as soon as they had substantial
doubts about its safety. However, as a consequence, they could not
reach a firm conclusion about it. Their own conclusion was stated
as follows:
"... the findings of this study indicate that the combi-
nation of diet and tolbutamide therapy is no more effective
than diet alone in prolonging life. Moreover, the findings
suggest that tolbutamide and diet may be less effective
than diet alone or than diet and insulin at least insofar
as cardiovascular mortality is concerned."
I think that statement should be taken at face value, and not as a
polite substitute for a more forceful condemnation of Tolbutamide.
Second, in the light of all available evidence, a decision must
be made about the appropriate legal or regulatory steps to be taken
about the distribution of Tolbutamide. The suggestion made by
Dr. Ricketts seems to me appropriate. Enough doubt about its effective-
ness in ordinary use and concern about toxicity exists to warrant im-
posing some additional reporting burden on the physician who prescribes
it.
Finally, of far more importance than the decisions to be made
about Tolbutamide is the continuing lack of procedures for accumulating
reliable information about the effects of drugs for which the balance
of risk and benefit is uncertain. The answer cannot be entirely a
matter of adding more and more requirements before a drug is released
for marketing. I am not convinced that present regulations have sub-
stantially impeded the adoption of valuable drugs, but pushed too far,
PAGENO="0191"
COMPETITIVE PROBLEMS IN THE DRIJG INDIJSTRY 13441
they certainly could. I should prefer to see drugs made available Øn
the basis of reasonable evidence of safety and efficacy, and to require
that appropriate studies should be carried out until the matter of i~isk
vs. benefit is settled beyond reasonable doubt.
Sincerely yours,
Paul Meier
P.S. I enclose a recent curriculum vitae, in accordance with your
request.
PM/tk
PAGENO="0192"
13442 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
February 1975
PAUL MEIER
Personal: Born 24 July 1924, New York City.
Education:
B.S. Oberlin College Physics, Mathematics 1941-45
Brown University Intersession Program in Applied
Mathematics 1945
M.A. Princeton University Mathematics 1945-47
(Mathematical Logic)
Study and Research in Statistics 1947-48
Ph.D. Princeton University Mathematics 1951
(Statistics)
Professional Career:
1948-49 Assistant Professor of Mathematics, Lehigh University.
1949-51 Research Secretary, Philadelphia Tuberculosis and Health Associa-
tion (part-time).
1951-52 Research Associate, Analytical Research Group, Forrestal Research
Center, Princeton University.
1952-57 Department of Biostatistics, School of Hygiene and Public Health,
The Johns Hopkins University
(1952-53, Research Associate; 1953-55, Assistant Professor;
1955-57, Associate Professor).
1957- Department of Statistics and Division of Biological Sciences,
University of Chicago
(1957-62, Associate Professor; 1962- Professor of Statistics;
1960-66, Chairman, Department of Statistics; 1962-69, Director,
Biological Sciences Computation Facilities; 1968-74, Professor
of Theoretical Biology; 1970-71, Acting Chairman, Department of
Statistics; 1973-74, Chairman, Department of Statistics; 1975-
Ralph and Mary Otis Isham Professor of Statistics and of the
Pharmacological and Physiological Sciences).
1966-67 National Institutes of Health Special Fellow at the University of
London (School of Hygiene and Tropical Medicine, and Imperial
College), on leave from the University of Chicago.
PAGENO="0193"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13443
Professional Memberships:
American Statistical Association, Fellow; Board of
Directors; Vice-President, 1965-67; Chairman, Committee
on Computers in Statistics, 1967; Chairman, Section on
Training, 1974.
Biometric Society, Executive Committee, ENAR, President,
ENAR, 1967.
Institute of Mathematical Statistics, Chairman of Editorial
Board, IMS-University of Chicago Press Monographs in
Statistics, 1961-64 and 1968-
Royal Statistical Society, Fellow.
American Association for the Advancement of Science, Fellow.
American Mathematical Society.
Mathematical Association of America.
Association for Symbolic Logic.
Society for Industrial and Applied Mathematics.
Association for Computing Machinery.
American Public Health Association, Fellow.
American Thoracic Society, Fellow.
American Heart Association, Fellow, Council on Epidemiology.
Consultant Activities and Special Appointments:
1955 Consultant on Statistical Problems in the Pharma-
ceutical Industry.
1959 Consultant on Sampling and other Statistical
Problems in Transportation.
1959-62 Member, Committee on Lung Cancer, American Cancer
Society.
1960-66 Consultant, The RAND Corporation.
1965-70 Member, Special Study Section, Biomathematics and
Statistics, National Institute of General Medica'.
Sciences (NIH)
PAGENO="0194"
13444 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
1967-71 Member, Therapeutic Evaluation Committee, National
Heart Institute (NIH).
1968 Member, Diet-Heart Feasibility Study Review
Committee, National Heart Institute (NIH).
1970- Member, National Academy of Sciences Committee on
Biological Effects of Atmospheric Pollution.
1971- Member, Advisory Board, Environmental Health
Resource Center.
1972- Member, Task Force on Health Considerations of a
National Energy Policy, American Public Health
Association.
1973- Member, Advisory Board of Veterans Administration
Cooperative Study of the Pathogenic Effects of the
Sickle Cell Trait.
1973- Member, ASA Advisory Committee to Statistical
Policy Division - Office of Management and Budget,
Executive Office of the President.
1973- Member, Committee for the Assessment of Biometric
Aspects of Controlled Trials of Hypoglycemic
Agents.
1974- Member, Panel on Airborne Particles and Panel on SOS,
Assembly of Life Sciences, National Research Council.
1974-75 Sigma Xi Lecturer.
1974- Member, Advisory Council for the Department of
Statistics at Princeton University.
1974- Member, Computer and Biomathematical Sciences Study
Section, National Institutes of Health.
Honorary Societies:
Phi Beta Kappa
Sigma Xi
Publications:
(1) "Timing of the distribution of events between observa-
tions. A contribution to the theory of follow-up studies"
(with T. E. Harris and J. W. Tukey), Human Biology, 22
(1950), 249-270.
PAGENO="0195"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13445
(2) "Tuberculosis among diabetics" (with others), Am. Rev.
of Tuberculosis, 65 (1952), 1-50.
(3) "Effects of simultaneous skin tests on size of tubercu~in
reactions" (with others), Am. Rev, of Tuberculosis, 65
(1952), 201-205.
(4) "Variance of a weighted mean," Biometrics, 9 (1953),
59-73.
(5) "On the theory of the indicator-dilution method for
measurement of blood flow and volume" (with Kenneth L.
Zierler), J.of Applied Physiology, 6 (1954), 731-744.
(6) "Analysis of sinple lattice designs with unequal sets
of replications," J. Am. Stat. Assoc., 49 (1954), 786-
813.
(7) "Note on estimation in a Markov process with constant
transition rates," Human Biology, 27 (1955), 121-124.
(8) "Vitamin B12 Serum concentration in 528 apparently healthy
human subjects of ages 12-94" (with others), 3. of Geron-
tolo~g~y, 12 (1957), 32-38.
(9) "Analysis of a bubble method for estimation of and
in whole blood" (with R. H. Shepard), 3. Appl. Phy~s.,
11 (1957), 250-259.
(10) "Safety testing of poliomyelitis vaccine," Science, 125
(1957), 106 7-1071.
(11) "Absorption of vitamin B12 enhanced by D-sorbitol" (with
others), Am. 3. Clinical Nutrition, 6 (1958), 30-33.
(12) "Reconsideration of methodology in studies of pain relief"
(with others), Biometrics, 14 (1958), 330-342.
(13) "Nonparametric estimation from incomplete observations"
(with E. L. Kaplan), .J.Am. Stat. Assoc., 53 (1958),
457-481.
(14) "Variability of critical flicker fusion thresholds in
brain-injured children" (with others), A.M.A. Archiy~
of Neurology and Psychiatry, 80 (1958) , 6 82-688.
(15) Appendix A of Secret Detention by the Chicago Police
(with William H. Kruskal), Clencoe: The Free Press,
(1959) , 35-31.
(16) "Further consideration of methodo1o~y in studies of pain
relief" (with S. N. Free), BiometrIcs, 17 (1961), 576~-583.
PAGENO="0196"
13446 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
(17) "Fluorospectrophotometric analysis on cervical epithelial
cells' (with George L. Wied, Anita M. Messina and Richard
R. J3lough), ActaCytologica, 8 (1964), 61-67.
(18) "An electronic data processing program for cytologic
screening projects for uterine carcinoma" (with George
L. Wied and Linda M. Clark), Acts Cytologica, 8 (1964),
385-397.
(19) "Fluorospectrophotometric analyses of endometrial and
endocervical epithelial cells" (with George L. Wied,
Anita M. Messina, Jose I, Manglano and Richard R. Blough),
Acta Cytologica, 8 (1964), 408-415.
(20) "DNA-assessments on fuelgen stained endometrial cells and
comparison with fluorometric values" (with George L. Wied,
Anita N, Messina and Ethel Rosenthal), J.Lab. Investiga
tion, 14 (1965), 1494-1499.
(21) "Statistical evaluation of the effect of hormonal contra-
ceptives on the cytological smear pattern" (with George L.
Wied, M. Edward Davis, Richard Frank, Peter B. Segal, and
Ethel Rosenthal), Obstet. and Gyne., 27 (1966), 327-334.
(22) "Analysis of morbidity and mortality of children irradiated
in fetal life" (with M. L. Griem and Glen D. Dobben),
Radiology, 88 (1967), 347-349.
(23) "Analysis of morbidity and mortality of children irradiated
in fetal life" (with M. L. Griem, Glen D. Dobben, and
D. J. Mewissen), Radiation Biology of the Fetal and
Juvenile Mammal, AEC Symposium Series, 17 (1969), 651-660.
(24) "Mass field trials of the diet-heart question" (with
members of the Diet-Heart Review Panel of the National
Heart Institute), American Heart Association Monograph,
28 (1969), 1-51.
(25) Statement of evaluation of risks in use of oral contra-
ceptives, before the Subcommittee on Monopoly (Nelson
Committee), Senate Select Committee on Small Business,
24 February 1970.
(26) "Hazards in oral contraception: The case for further
tests," Midway, (1970), 87-95.
(27) "The biggest public health experiment ever: The 1954
field trial of the Salk poliomyelit4s vaccine," in
Statistics: A Guide to the Unknown, Frederick Mosteller,
et al, eds., Holden-Day, Inc., (1972), 2-13.
(28) "Air Pollution and Pulmonary Cancer" (with Bertram Carnow),
Arch. Environ. Health, 27 (1973), 207-218.
PAGENO="0197"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13447
(29) "Effects of low-dose prenatal irradiation in humans:
Analysis of Chicago Lying-In data and comparison with
other studies" (with B. E. Oppenheim and M. L. Griem),
Radiation Research, 57 (1974), 508-544.
In Press and In Preparation
(30) "The effects of diagnostic X-ray exposure on the human
fetus: An examination of the evidence" (with B. E.
Oppenheim and M. L. Griem). Accepted for publication
in Radiology.
(31) "Statistics and medical experimentation," Presidential
Invited Address, Meeting of the Biometric Society (ENAR~,
March 1974, Tallahassee, Florida. To appear in BiometrLcs.
(32) ~`Estimation of a distribution funttion from incomplete
observatiotis." To appear in Studies in Probability and
Statistics: The M. S. Bartlett Festsehrift, 1975.
(33) "Ascorbic acid and the common cold: An evaluation of its
efficacy and toxicity" (with H. H. H. Dykes). Accepted
for publication in the Journal of the American Medical
Association.
(34) "Man as th~ experimental animal," Sigma Xi Lecture, (1974).
To be submitted to the American Scientist.
PAGENO="0198"
13448 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
USV PHARMACEUTICAL CORPORATION
1 SCARSDALE ROAD TU( KAHOL, NI W YORK 0707
HERBERT H. Mc DADE, JR.
PRESIDENT
CHIEF OPLRA1INC OFFICER
March 11, 1975
The Honorable Senator Gaylord Nelson
Chairman - Monopoly Subcommittee
on the Senate Small Business Committee
The United States Senate
Washington, D. C. 20510
Dear Senator Nelson:
As the discoverer and developer of phenformin, USV
Pharmaceutical Corporation has been following the con-
troversy regarding the UGDP study; and, at your
invitation, on October 21, 1974 we submitted to your
Subcommittee the Company's general policy concerning
the appropriate use of hypoglycemic agents, various
criticisms of the UGDP study and comments concerning the
reports of other groups which have studied these agents.
Our comments, as expressed in that letter, remain the
same; however, in reading the transcript of the testimony
before your Subcommittee on January 31, 1975, we feel
compelled to amplify our previous letter.
The testimony of The Chairman of The Committee for the
Assessment of Biometric Aspects of Controlled Trials of
Hypoglycemic Agents reflected the following statement:
"The findings on phenformin, if one can judge
from the absence of criticism, appear to have
been accepted by medical scientists, even if
they have not so far been translated effectively
into medical practice. Yet these findings also
PAGENO="0199"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13449
were made by the UGDP using the methods that
have come. under heavy criticism when applied
to tolbutamide."
A similar implication of absence of criticism of the
UGDP with respect to phenformin appears at the end of
section 3.1 (Findings) of the Biometric Committee's
Report published in the February 1975 issue of the
JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION.
We believe it is necessary to dispute any implication
that the criticisms leveled against the UGDP study are
concerned with the drug, tolbutamide, alone. To the
contrary, the substantive criticisms made, although
perhaps naming tolbutamide specifically, apply equally
with respect to phenformin inasmuch as the structuring
of the two trials under the UGDP was nearly identical.
Further, it should be pointed out that because of the
nature of phenfonnin, the rather late introduction of its
trials, and a variety of other material factors, the
phenformin portion of the UGDP necessarily becomes the
subject of additional criticisms.
There are several reasons why criticism of the UGDP study
most generally is publicized in terms of tolbutamide.
Tolbutamide was the first of the two drugs to be studied
by the UGDP in fact, tolbutamide preceded phenformin in
the study by 18 months. Likewise, the first data
generated by the studies were those on tolbutamide.
Second, and just as importantly, the final Report of the
UGDP on phenformin had yet not been published as of the
date of the Biometric Committee's review. Since it is
not generally recognized as scientifically proper to
comment upon, or criticize, work presented only in pre~
liminary form, it is natural that there had been less
formal criticism on the phenformin aspect of the UGDP
PAGENO="0200"
13450 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
study. Under these circumstances, it was to be expected
that criticism of the UGDP study was more likely to be
framed in terms of tolbutamide; however, this should not,
in our opinion, be interpreted to mean that as to phem-
formin, the results of the UGDP study "have been accepted
by medical scientists."
The Biometric Committee obviously considered important
the fact that the final UGDP Report on phenformin had not
yet been published and, in fact, the Committee did not
consider the basic data on the effects of phenformin
treatment. This is clearly reflected in the following
quotation from Section 1 (Introduction) of the Biometric
Committee's Report:
"The preliminary report on phenformin was
considered in September of 1973, but in view
of the fact that the final report on that subject
was still unpublished, the Committee did not
request the basic data on the effects of this
treatment."
In addition, the Biometric Committee recognized in its
Report that the introduction of phenformin to the UGDP
."greatly complicated an already difficult study."
[See Section 7. 1 (Conclusion, Protocol)].
Based upon these and other factors, it is our conclusion
that the Report of the Biometric Committee does not
resolve the controversy surrounding the UGDP study. It
does not resolve the general criticisms of the UGDP
study set forth in our letter of October 21, 1974, nor
does it resolve the specific issues relative to
phenformin.
We greatly appreciate this opportunity to present these
PAGENO="0201"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13451
views to you, and respectfully request that they be
made a part of the record in this matter.
Very truly yours,
USV PHARNAçEUTICAL CORPORATION
~ \
Herbert H. NcDade, Jr.
President
Chief Operating Officer
HHNcD,Jr. /mem
56-592 0 - 75 - pt. 28 - 14
PAGENO="0202"
13452 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
FD~HASKELL,COL~ ~Cn~te~ ~~a1e2 ~~ena~e
WALTER F. MONDALE. MINN. SELECT COMMITTEE ON SMALL BUSINESS
DIRECTOR (CREATED PURSUANT TOE. RES. 58,8ISTcONORE8S)
WASHINGTON, D.C. 20510
April 1. 1975
The Bonorable Alexander M. Schmidt
Commissioner
Food and Drug Administration
Washington, D. C.
Dear Mr. Commissioner,
During your testimony befor. our Monopoly
Subcommittee on January 29, 1975 Mr. Butt stated that
the Department of Justice declined to file at least
seven criminal cases that the FDA forwarded to them
for prosecution on false advertising issues.
In this connection, I should be extremely
grateful if you would supply the Subcommittee with
the names of the firms, dates, products involved,
FDA's letters of transmittal and the response of the
Department of .7ustice to FDA in each of these cases
as well as for other cases which Mr. Butt may not
have been aware of
Kindest personal regards.
Sincerely,
Gaylord Nelson
Chairman
Monopoly Subcommittee
PAGENO="0203"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13453
DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
PUBLIC HEALTH SERVICE
FOOD AND DRUG ADMINISTRATION
ROCKVILLE, MARYLAND 20852
APR 101975
Honorable ~aylord Nelson
Chairman, Monopoly Subcommittee
Select Committee on Small Business
United States Senate
Washington, D.C. 20510
Dear Senator Nelson:
Thank you for your April 1 letter to Commissioner Schmidt requesting
information concerning certain criminal cases involving false
advertising charges forwarded to the Department of Justice by the
Food and Drug Administration which that Department has decided not
to file.
We have begun a review of our files to assemble the information you
requested and will supply you with an answer in the near future.
If we can be of any assistance in any other way, please let us
know.
Sincerely yours,
>i4~-V( (~4~L//,~.
Robert C. Wetherell, Jr., DirectoV
Office of Legislative Services
PAGENO="0204"
Senator Gaylord Nelson
Senate Office Building
Washington, D.C.
Dear Senator Nelson:
I am enclosing for your information a letter
which I sent to Dootor Schmidt shortly after my testi-
cony before you. It is my hope that in accordance with
this letter some reasonable conclusion to this dispute will
be forthcoming.
I understand that hearings will be held again
before your Subcommittee at which the FDA will be present.
I am requesting at this time an opportunity to testify at
that session so that the proper balance can once again be
provided to the Subcommittee deliberations.
Thank you for your continued attention to this
matter.
ry 1 yours,
N il Ch~\
NLC CF
Enc.
13454 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
CHAYET AND SONNENREICH, P. C.
ATTORNEYS AT LAW
6 FAYETTE STREET
BOSTON, MASSACHUSETTS 02116
February 1W, 1975 wAsHINOToN,D. C. 20037'
PAGENO="0205"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13455
CHAYET AND SONNENREICH, P. C.
ATIORNEYS AT LAW
6 FAYETTE STREET
BOSTON, MASSACHUSETTS 02116
MICHAEL ~N~~NRE~CH (ei~) 357-0202 WATERGATE R00,$UITE 720
~ February 11, 1975 WASHINGTON, 0. C.20037
Dr. Alexander Schmidt, Commissioner
Food and Drug Administration
Department of Health, Education and Welfare
5600 Fishers Lane
Rockville, Maryland 20852
Dear Dr. Schmidt:
It is my understanding that additional hearings are to be
held later this month on the subject of oral hypoglycemic drugs before
the Subcommittee on Monopoly, chaired by Senator Gaylord Nelson. It is
my further understanding that the Food and Drug Administration will be
asked to provide testimony at this hearing. I would suggest a meeting
between my client and FDA officials to discuss in detail proposed label-
ing changes by FDA and to determine whether there now exists labeling
acceptable to both sides. It is my belief that more can be accomplished
at an informal meeting than in testimony before congressional committees
and protracted legal proceedings.
Your testimony on September 20, lg7t~, indicated that there
exists substantial controversy with regard to oral hypoglycemic drugs.
You stated: " I have personally talked with many diabetologists around
the country, and I am sure you know of the degree of controversy that
yet remains about the UGDP Study. If that has not been brought up before
the Committee, I think it really should be because many `experts' do
publicly attack the UGDP study."
I would also like to note the following portion of Mr. Mutt's
testimony that same day: "We must then stand upon the scientific basis
for our decision.. . .That is why in part we are waiting for the Biometric
Society report as well as amending our regulation Section 1.3, to which the
Commissioner has already averted, to settle the legal issue that was also
involved in the case."
The hope was expressed by both you and Mr. Mutt that the
Biometric Society study would settle this matter. Although we have had
only a brief period of time to review this study, I am authorized by my
clients to inform you that, despite Dr. Chalmers' editorial entitled
"Settling the UGDP Controversy," the controversy is in fact not settled.
It is doubtful whether any single review of the UGDP study could settle
the fundamenta]~ questions which have resulted in the controversy concerning
PAGENO="0206"
13456 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
this study. What is needed is a new prospective study or studies to
resolve the issues raised regarding oral hypoglycemics, coupled with
immediate, balanced labeling reflecting the controversy.
The Committee on the Care of the Diabetic intends to pursue,
through the National Institutes of Health, access to the raw data of the
UGDP Study and further to continu~ its opposition to the proposed amend-
ment of Regulation 1,3, as expressed in comments which we filed relative
to this regulation.
I would like to call your' attention to the fact that one
of the most fruitful aspects of this matter which has continued these
many years was the meeting held with FDA officials in October 1973. At
that meeting I felt that real progress was made in achieving labeling
acceptable to all parties engaged in this controversy. I am enclosing a
draft of this labeling prepared by the Committee on the Care of the
Diabetic and request that all parties meet once again at the FDA to
further discuss this matter,
We offer our full cooperation in continuing the dialogue
between the FDA and the Committee on the Care of the Diabetic which
will result in new fairly balanced labeling which reflects the current
scientific status which we feel, and have felt, is urgently needed.
Very truly yours,
Neil L. Chayet
NLC:GF
PAGENO="0207"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRT 13457
U.S. SENATE,
SELECT COMMITTEE ON SM&u~ BuSINESS,
Washington, D.C., May 27, 1975.
NEIL L. CHAYET, Esq.,
Ohayet and Sonnenreich, P.C.
Boston, Mass.
DEAR MR. CHAYET: This is in response to your letter of February 14, 1975
accompanied by your correspondence with Commissioner Schmidt of the Food
and Drug Administration, as well as your prior phone request to me.
It is true that our Monopoly Subcommittee plans to hold another bearing on
the oral hypoglycemic drugs, at which time the FDA will testify on labeling
changes. Although your request to appear once again before the Subcommittee
Is appreciated, this will not be necessary since your position is already part oi~
the hearing transcript. However, should you desire to place additional materia~
Into the record following the final testimony of the FDA, you may do so.
In the meantime, I should be extremely grateful if you would send us ~
complete list of members (and their addresses) of the Committee on the Care
of the Diabetic, the organization you represent.
Your contittued interest in the work of our Subcommittee is appreciated.
Sincerely,
BENJAMIN GORDON,
Staff Economist.
CHAYET AND SONNENREICH, P. C.,
ATTORNEYS AT LAW,
Boston, Mass., July 10, 1975.
MR. BENJAMIN GORDON,
Staff Economist,
U.S. Senate,
Select Committee on Small Business,
Washington, D.C.
DEAR MR. GORDON: I am in receipt of your recent letter referring the reque~t
of the Committee on the Care of the Diabetic (COD) to testify before the
Monopoly Subcommittee in the hearings held this week relative to labeling of
oral hypoglycemic drugs.
As you know, COD has been involved in the labeling controversy for several
years, appearing before administrative, legislative, and judicial bodies. Froi~i
the list of witnesses who appeared before the subcommittee, it would appear
that the testimony offered this week represented only one side of the controversy.
Particularly in view of very recent developments in the controversy, i.e. tl~e
FDA's proposed relabeling of the drugs (Federal Register, Vol. 40, No. 18Q,
pp 28582-28595), COD regrets not having been permitted to testify, as its testi-
mony could have availed the Subcommittee of a more balanced view of tl~e
issues.
I am enclosing, per your request, a list of COD members, Kindly include this
letter on the record of the proceedings.
Very truly yours,
NEIL L. CHAYET.
Enclosure.
Frank N. Allan, Chairman Emeritus, Medical Department, Lahey Clinic, 44
Barnstable Road, West Newton, Massachusetts 02165.
Seymour Alterman, M.D., 1688 Meridan Avenue, Miami Beach, Florida 38139.
Shepard G. Aronson, M.D., 150 East 56th Street, New York, N. Y. 10022.
James B. Ashmore, M.D., Professor of Pharmacology, Indiana University School
of Medicine, Indianapolis, Indiana.
Donald M. Barnard, M.D., Fargo Clinic, Fargo, N.D.
Donald Barnett, M.D., Joslin Clinic, 15 Joslln Road, Boston, Massachusetts
02215.
Samuel B. Beaser, M.D., 31 Bay State Road, Boston, Massachusetts.
Lewis H. Biben, M.D., 618 MedIcal Science Building, 916 NIneteenth StreØt,
N.W., Washington, D. 0. 20006.
Keith Borden, M.D., Chief, Diabetes Products, The Upjohn Company, Kalamazoo,
Michigan 49001.
Angela Bowen, M.D., 1015 West 4th Street, Olympia, Washington 98501.
Robert F. Bradley, M.D., Medical Director, Joslin Clinic, 15 Joslin Road, Bosto~,
Massachusetts 02215.
PAGENO="0208"
13458 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
George Brown, M.D., 2021 Grand Concourse, Bronx, New York 10453.
R. A. Camerini-Davalos, MD., Associate Professor in Medicine, New York Medi-
cal College, Flower and Fifth Avenue Hospitals, 1249 Fifth Avenue, New
York, New York 10029.
John J. Canary, M.D., Director, Division of Endocrinology and Metabolic
Disease, Georgetown University Hospital, 3800 Reservoir Road, N.W.,
Washington, D. C. 20007.
William Castelli, M.D., Department of Preventive Medicine, Harvard Medical
School, Longwood Avenue, Boston, Massachusetts 02115.
David R. Challoner, M.D., Associate Professor, Assistant Chairman, Department
of Medicine, Indiana University School of Medicine, Indianapolis, Indiana.
Mr. Neil Chayet, 461 East 57th Street, New York, N.Y.
John W. Chriss, M.D., Medical Center, 2436 Morgan Street, Corpus Christi,
Texas.
A. Richard Christlieb, M.D., Joslin Clinic, 15 Joslin Road, Boston, Massachu-
setts 02215.
Burton D. Cohen, M.D., Chief, Section of Metabolism, The Bronx Hospital, 1276
Fulton Avenue, Bronx, New York 10456.
Daniel T. Coughlin, Director Government Affairs, The Upjohn Company, Suite
904, 1101 Seventeenth Street, N.W., Washington, D.C., 20036.
DeWitt E. DeLawter, M.D., 500 Prospect Place, Chevy Chase, Maryland.
Harold L. Dobson, M.D., Herman Hospital, Texas Medical Center, 1203 Ross
Sterling Avenue, Houston, Texas 77025.
Henry Dolger, M.D., 11 East 86th Street, New York, New York 10028.
Arthur H. Dube, M.D., Associate Professor of Clinical Medicine, Upstate Medical
Center, Syracuse, New York.
Sigmund Falk, M.D., Chief, Diabetes Clinic, 15 West 11th Street, New York,
New York 10014.
Alvan R. Feinstein, M.D., Professor of Medicine & Epidemiology, Yale Univer-
sity Medical School, New Haven, Connecticut.
Leonard Felder, M.D., 26 Fifth Avenue, New York, New York 10011.
Robert Feldman, M.D., Director of Metabolic Research, Kaiser Permanente
Medical Foundation, 280 West McArthur Boulevard, Oakland, California
94611.
Philip W. Felts, M.D., Assistant Professor of Medicine, Vanderbilt University
School of Medicine, Nashville, Tennessee.
B. Dan Ferguson, M.D., Joslin Clinic, 15 Joslin Road, Boston, Massachusetts
02215.
Peter H. Forsham, M.D., Chief of Endocrinology, Professor, Department of
Medicine, University of California Medical Center, San Francisco, Cali-
fornia 94122.
Daniel Foster, M.D., Associate Professor of Internal Medicine, University of
Texas, Southwestern Medical School, 5323 Harry Hines Blvd., Dallas,
Texas 75235.
Adolph Friedman, M.D~. 1712 Eye Street, N.W., Washington, D. C. 20006.
Gerald J. Friedman, M.D., 850 Park Avenue, New York, New York.
Richard L. Fulton, M.D., 1211 Dublin Road, Columbus, Ohio.
Edward J. Gallagher, M.D., 10090 Main Street, Fairfax, Virginia.
Edwin W. Gates, M.D., 625 Sixth Street, Niagara Falls, New York 14301.
H. Howard Goldstein, M.D., Joslin Clinic, 15 Joslin Road, Boston, Massachusetts
02215.
George Goodkin, M.D., Senior Associate Medical Director, The Equitable Life
Assurance Society of the United States, 1285 Avenue of the Americas, New
York, New York 10019.
Irving Graef, M.D., 791 Park Avenue, New York, New York.
Charles A. Graham, M.D., Joslin Clinic, 15 Joslin Road, Boston, Massachusetts
02215.
Robert C. Green, Jr., M.D., 230 W. Boscawen Street, Winchester, Virginia 22601.
Barnett Greenhouse, M.D., 129 Whitney Ave., New Haven, Conn.
Richard D. Grimaldi, M.D., Director of Diabetes, Bellevue Maternity Hospital,
P. 0. Box 1030, Stop 11 Troy Road, Schenectady, New York 12301.
William B. Hadley, M.D., Joslin Clinic, 15 Joslin Road, Boston, Massachusetts
02215.
Robert R. Hare, M.D., 2250 N. W. Flanders, Portland, Oregon 97207.
Edgar A. Haunz, M.D., Professor and Chairman, Department of Medicine, Uni-
versity of North Dakota, Grand Forks, North Dakota 58201.
PAGENO="0209"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13459
Richard D. Hohi, M.D., Associate Physician, Division of Metabolic Diseases,
Henry Ford Hospital, 2799 Grand Boulevard W., Detroit, Michigan 48202.
H. Morris Horn, M.D. FACP, Texas Diabetis Assoc., 3434 Swiss Ave., Suite 305,
Dallas, Texas 75204.
Jesse D. Ibarra, Jr., M.D., Scott & White Clinic, Temple, Texas 76501.
W. P. U. Jackson, M.D., Dept. of Medicine, University of Medicine, University of
Cape Town Observatory, Cape Towns SOUTH AFRICA.
Wyman E. Jacobson, M.D., Section on Endocrinology and Diabetes, St. Lou$s
Park Medical Center, 4959 Excelsior Boulevard, Minneapolis, Minnesota
55416.
George M. Jones, M.D., Past President, Texas Diabetes Assoc., Clinical Profes-
sor of Internal Medicine, Southwestern Medical College of the University
of Texas, Dallas, Texas.
William R. Jordan, M.D., 1631 Monument Avenue, Richmond, Virginia 23220.
Allen P. Joslin, M.D., Joslin Clinic, 15 Joslin Road, Boston, Massachusetts 02215.
Miles Kahan, M.D., 2405 Avenue P, Brooklyn, New York 11299.
Dorothy Kahkonen, M.D., Associate Physician, Division of Metabolic Diseases,
Henry Ford Hospital, 2799 Grand Boulevard W., Detroit, Michigan 48202.
W. B. Kannel, M.D., Framingham Heart Study, 123 Lincoln Street, Framingbai~i,
Massachusetts 01701.
Norman M. Kaplan, M.D., Associate Professor of Internal Medicine, University
of Texas Southwestern Medical School, 5323 Harry Hines Blvd., Dallas,
Texas 73235.
George A. Kaufman, M.D., 888 Grand Concourse, Bronx, New York.
Mr. Terry G. Kelley, Public Relations Associate, The Upjohn Company, Kala-
mazoo, Michigan 49001.
Mavis P. Kelsey, M.D., Kelsey-SeybQld Clinic, 6624 Fannin, Houston, Texas
77025.
M. H. Kolodny, M.D., 1020 Park Avenue, New York, New York.
George P. Kozak, M.D., Joslin Clinic, 15 Joslin Road, Boston, Massachusetts
02215.
Leo P. Krall, M.D., Director of Education, Joslin Diabetes Foundation, 15
Joslin Road, Boston, Massachusetts 02215.
Bernard T. Kravitz, M.D., Montefiore Hospital Medical Group, 3444 Kossuth
Avenue, Bronx, New York 10467.
Arthur Krosnick, M.D., Coordinator Diabetes, Endocrine and Metabolic Disease
Program, Division of Chronic Illness Control, Department of Health, P.O.
Box 1540, John Fitch Plaza, Trenton, New Jersey 08625.
William Kurstin, M.D., 618 Medical Science Building, 916 Nineteenth Street,
N.W., Washington, D. C. 20006.
Thomas H. Lanibert, M.D., Scripps Clinic and Research FoundatIon, 476 PrOs-
pect Street, La Jolla, California 92037.
Richard L. Landau, M.D., Professor of Medicine, Head of Endocrinology S~c-
tion, University of Chicago, Chicago, Illinois 60637.
Ira J. Laufer, M.D., 45 Gramercy Park North, New York, New York 10010.
Louis Lasagna, M.D., Department of Medicine, Strong Memorial Hospital, Crit-
tenden Boulevard, Rochester, New York.
Ann M. Lawrence, M.D., Associate Professor of Medicine, Section on Endoc-
rinology, University of Chicago School of Medicine, Chicago, Illinois 60637.
Sydney S. Lazarus, M.D., Chief of Pathology, Isaac Albert Research Institute
of the Kingsbrook Jewish Medical Center, Rutland Road and East 49th
Street, Brooklyn, New York 11203.
Paul Leifer, M.D., 1860 Grand Concourse, Bronx, New York 10457.
Rachmiel Levine, M.D., Chief, Endocrinoligy Service, City of Hope Medical
Center, Duarte, California 91010.
Leon M. Levitt, M.D., 95 Buckingham Road, Brooklyn, New York 11226.
Samuel D. Loube, M.D., Washington Internal Medicine Group, 2400 H Street,
N.W., Suite 7, Washington, D.C. 20037.
Glen W. McDonald, M.D., 426 Merkle Drive, Norman, Oklahoma.
Leonard L. Madison, M.D., Professor of Internal Medicine, University of TeZas
Southwestern Medical School, 5323 Harry Hines Boulevard, Dallas, Texas.
Alexander Marble, M.D., President, Joslin Diabetes Foundation, 15 Joslin Road,
Boston, Massachusetts 02215.
Leona Miller, M.D., Associate Professor of Medicine, University of Southern
California School of Medicine, Los Angeles, California.
Merton M. Minter, M.D., Minter Clinic, Nix Professiinal Building, San Antonio,
Texas 78205.
PAGENO="0210"
13460 COMPETITIVE PROBLEMS IN THE I~RUG INDUSTRY
James M. Moses, M.D., Fairlington Professional Building, 1707 Osage Street.
Alexandria, Virginia 22302.
Henry J. Oppenheimer, M.D., 141 North Meramec, Clayton, Missouri 63105.
Mr. Harold Rlvkin, 510 Plymouth Avenue, Minneapolis, Minnesota.
Morris H. Rosenberg, M.D., Suite 702, University Medical BuildIng, 2141 K
Street, N.W., Washington, D.C. 20037.
A. H. Rubenstein, M.D., Associate Professor, Department of Medicine, Section
of Endocrinology, University of Chicago, Chicago, Illinois 60637.
John W. Runyan, Jr., M.D., 951 Court Avenue, Memphis, Tennessee 38103.
Elliot L. Sagall, M.D., 454 Brookline Avenue, Boston, Massachusetts 02215.
Aaron G. Saidman, M.D., 5530 Wisconsin Ave., Chevy Chase, Md. 20015.
Dominick J. Savino, M.D., 27 Prospect Park West, Brooklyn, New York 11215.
George F. Schmitt, M.D., 30 Southeast 8th Street, Miami, Florida 33131.
A. Schonfeld, M.D., 50 Central Park South, New York, New York.
Stanley Schor, Ph.D., Temple University Health Services Center, Broad and
Ontario, Philadelphia, Pennsylvania 19140.
0. Peter Schumacher, M.D., Cleveland Clinic, 2020 East 93rd Street, Cleveland,
Ohio.
Seymour Schutzer, M.D., 67 Cedar Drive, Great Neck, New York 11021.
Zdenko Skrabalo, M.D., University of Zagreb Diabetes Center, Zagreb, Yugo-
slavia.
J. Stuart Soeldner, M.D., Assistant Professor of Medicine, Harvard University
School of MedicIne, 170 Pilgrim Street, Boston, Massachusetts 02115.
Maxwell Spring, M.D., Clinical Assistant Professor of Medicine, New York
Medical College, 5th Avenue at 106th Street, New York, New York.
John W. Stephens, M.D., 2250 N.W. Flanders, Portland, Oregon 97207.
Daniel B. Stone, M.D., 530 Doctors Bldg., Omaha, Nebraska 68131.
Samuel J. N. Sugar, M.D., 4637 Eastern Avenue, Washington, D. C. 20018.
Karl E. Sussman, M.D., Associate Professor of Medicine, Acting Head, Division
of Endocrinology, University of Colorado Medical School, 4200 East Ninth
Avenue, Denver, Colorado 80220.
Lawrence J. Thomas, M.D., 11801 Rockville Pike, Rockville, Maryland 20852.
James Tullis, M.D., Chief of Medicine, New England Deaconess Hospital,
Boston, Massachusetts.
John B. O'Sullivan, M.D., Chief, Diabetes and Arthritis, Field Research Unit,
U. S. Public Health Service, 408 Atlantic Avenue, Boston, Massachusetts
02118.
Kar' R. Paley, M.D., Chief, Metabolic Clinic, Lenox Hill Hospital, 77th Street
& Park Avenue, New York, New York.
Arthur M. Parker, M.D., 380 East 18th Street, Brooklyn, New York 11226.
Jean 0. Partamian, M.D., Associate Physician, Division of Metabolic Diseases,
Henry Ford Hospital, 2799 Grand Boulvard W., Detroit, Michigan 48202.
John W. Partridge, M.D., Editor, Diabetes Bulletin, 2222 N.W. Lovejoy Street,
Portland, Oregon.
Marjorie Peebles-Meyers, M.D., 3790 Woodward Avenue, Detroit, Michigan 48201.
L. Lewis Pennock, M.D., Forbes-Oakland Medical Building, 3347 Forbes Avenue,
Pittsburgh, Pennsylvania 15218.
W. E. Redfern, M.D., Associate Physician, Division of Metabolic Diseases, Henry
Ford Hospital, 2799 Grand Boulevard W., Detroit, Michigan 48202.
Robert L. Reeves, M.D., F.A.C.P., 1015 West 4th Street, Olympia, Washington
98501.
Harold Rifkin, M.D., Montefiore Hospital and Medical Center, 1111 East 210th
Street, Bronx, New York 10467.
Howard S. Schwartz, M.D., 3201 Grand Concourse, Bronx, New York 10468.
Laurence F. Segar, M.D., 8700 Dean Dr. #1908, Ventura, Calif. 98003.
Donald W. Seldin, M.D., Professor and Chairman Department of Medicine,
University of Texas, Southwestern Medical School, 5323 Harry Hines
Boulevard, Dallas, Texas 75235.
Hoibrooke S. Seltzer, M.D.. Chief of Endocrinology, Veterans Administration
Hospital, 4500 South Lancaster, Dallas, Texas 75216.
Thomas Sharkey, M.D., 60 Wyoming Street, Dayton, Ohio 45409.
L. Benjamin Sheppard, M.D., 301 Medical Arts Building, Richmond, Virginia.
Charles Shuman. M.D., Temple University Health Services Center, Broad and
Ontario, Philadelphia, Pennsylvania 19140.
Abraham A. Silver, M.D., 6210 Park Heights Avenue, Baltimore, Maryland.
PAGENO="0211"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13461
Marvin D. Siperstein, M.D., Professor of Internal Medicine, University of Texas,
Southwestern Medical School, 5323 Harry Hines Boulevard, Dallas, Texas
75235.
Janet D. Sherman, M.D., 18181 W. 12 Mile Rd., Lathrup Village, Michigan.
Roger H. Unger, M.D., Veterans Administration Hospital, 4500 S. Lancaster
Road, Dallas, Texas 75216.
Alexander G. Vongries, M.D., Asst. Director, Clinical Development, Ciba-Gel~y
Corp., Summit, NJ. 07901.
Robert C. Warner, M.D., Asst. Prof. of Medicine, University of N. Dakota,
Fargo, N. Dakota.
Harry F. Wechsler, M.D., 737 Park Avenue, New York, New York.
Irving Weckell, M.D., 7 Pont Street, Great Neck, New York.
Charles Weller, M.D., 17 North Chatsworth Avenue, Larchmont, New York
10538.
George Welsh, M.D., III, Director of Continuing Education for Health Sciences,
University of Vermont Medical School, Burlington, Vermont.
Priscilla White, M.D., Joslin Clinic, 15 Joslin Road, Boston, Massachusetts 022~5.
Fred W. Whitehouse, M.D., Chief, Division of Metabolic Diseases, Henry Ford
Hospital, 2799 Grand Boulevard W., Dejroit, Michigan 48202.
David S. Wilcox, M.D., President, Connecticut Diabetes Association, 85 Jeff~r*
son Street, Hartford, Connecticut 06109.
T. Franklin Williams, Monroe County Hospital, East Henrietta Road, Rochester,
New York.
Charles W. Wilson, M.D., Harbor and Olean Blvd., Port Charlotte, Florida 339~0.
Jean D. Wilson, M.D., Professor of Internal Medicine, University of Texas,
Southwestern Medical School, 5323 Harry Hines Boulevard, Dallas, Texas
75235.
Albert I. Winegrad, M.D., Director of Cox Institute, Hospital of the Universlt~V
of Pennsylvania, Philadelphia, Pennsylvania 19104.
Donna Younger, M.D., Joslin ClInic, 15 Joslin Road, Boston, Massachusetts 02215.
Leonard Zimmerman, Group Service Agency, 342 Madison Avenue, New Yo~k,
New York 10017.
Akira Ilorinchi, M.D., 3-33-15 Minimimagome, Otliku, Tokyo, JAPAN.
Bernard Leibel, M.D., 200 St. Clare Street W., Toronto, Ontario, CANADA.
John A. Moorhouse, M.D., Director, Endocrine and Metabolic Laboratory,
Winnipeg General Hospital, Winnipeg, CANADA.
PAGENO="0212"
13462 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
FOOD AND DRUG ADMINISTRATION
[21 CFR PART 310]
[DOCKET NO. 75N-OO62].
ORAL HYPOGLYCEMIC DRUGS
NOTICE OF PUBLIC HEARING AND PROPOSED LABELING
The Commissioner of Food and Drugs is proposing labeling for all
oral hypoglycemic drugs and announcing a legislative-type public hearing
on the issues involved. Labeling for this class of drugs has been the
subject of extended public controversy and legal challenge for several
years. The Commissioner believes that it is now essential to resolve
the outstanding issues in this matter and that it is in the interest
of the public health to consider the views of all parties in achieving
such resolution. Accordingly, this notice proposes class labeling for
oral hypoglycemic drugs that, on the basis of all information available
to the Food and Drug Administration, the Commissioner believes is con-
sistent with the requirements of the Federal Food, Drug, and Cosmetic
Act and reflects current scientific knowledge on the safety and effect-
iveness of these drugs.
The Commissioner invites all interested persons to submit written
comments on the proposed labeling. In addition, the Commissioner's
designee, the Director of the Bureau of Drugs, will conduct an oral
75-415
PAGENO="0213"
COMPETITIVE PROBLEMS IN TH~ DRTJG INDUSTRY 1346~
public hearing to afford interested persons a further opportunity
for the presentation of data, information, and views. In the Com-
missioner's judgment, the subject matter of this notice is of sufficient
importance to justify the use of this additional procedure, as provided
in Part 2, Subpart E, of the regulations governing the administrative
practice and procedures of the Food and Drug Administration, published
in the FEDERAL REGISTER of Nay 27, 1975 (40 FR 23025).
Interested persons may submit comments on the labeling proposed
in this notice by (insert date 60 daSrs after date of pub1icationi~
the FEDERAL REGISTER). In addition, any interested person may submit
data, information, or views in writing any time within 15 days after
the conclusion of the public hearing. It is the intention of the
Food and Drug Administration to conduct the public hearing prior to
the expiration of the time for submitting comments, and the Commissioner
therefore encourages interested persons to submit their comments
as soon as possible, to allow review prior to the hearing.
After consideration of all written and oral comments and all
data, information, and views presented at the public hearing, the
Commissioner will promulgate in the FEDERAL REGISTER a final regula-
tion prescribing labeling for oral hypoglycemic drugs, applicable
PAGENO="0214"
13464 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
to all drug products in this class It is anticipated that
the final labeling will conform with the guidelines for labeling
of prescription drugs proposed by the Commissioner on April 7 1975
(40 FR 15392)
I. GENERAL BACKGROUND
The following new drug applications have been approved for oral
hypoglycenic drugs:
1 NDA 10 670 Orinase tablets containing tolbutamide The
Upjohn Co., 7000 Portage Rd., Kalamazoo, MI 49001.
2. NDA 15,500, Tolinase tablets containing tolazamide; The
Upjohn Co
3 NDA 11 641 Diabinese containing chlorpropamide Pfizer Inc
235 E 42d St New York NY 10017
4 NDA 13 378 Dymelor containing acetohexamide Eli Lilly & Co
Indianapolis IN 46206
5 NDA 11 624 DEl tablets containing phenformin hydrochloride
Geigy Pharmaceuticals Ardsley NY 10502
6 NDA 12 752 DBI-TD capsules containing phenformin hydrochloride
Geigy Pharmaceuticals
7 NDA 17 126 Meltrol-50-100 capsules containing phenformin
hydrochloride USV Pharmaceutical Corp 1 Scarsdale Rd Tuckahoe NY 10707
8 NDA 17 127 Meltrol-25 tablets containing phenformin hydrochloride
* USV Pharmaceutical Corp. *
9 NDA 12 678 Tolbutamide tablets conraining tolbutamide
Premo Pharmaceuticals Laboratories Inc 111 Leuning St South
Hackensack NJ 07606
PAGENO="0215"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 134~5
The class of oral hypoglycemic drugs can be grouped into two
categories on the basis of chemical structure: the sulfonylurea
category (represented by acetohexamide, chiorpropamide, tolazamide,
and tolbutamide) and the biguanide category (represented by phenfortnin
hydrochloride). The mode of action and adverse effects are different
for these two categories of oral hypoglycemic drugs. Accordingly,
separate labeling is proposed for each category of drug.
Under section 505 of the Federal Food, Drug, and Cosmetic Act,
the Commissioner is responsible for. assuring that all new drugs have
been shown to be safe and effective for their intended uses and
that their labeling is not false or misleading. Exercise of this
responsibility often requires reexamination of the safety, effective-
ness, or labeling of drugs previously approved. The statutory scheme
contemplates that new information may require the Commissioner to
prescribe changes in the labeling of a drug, to reveal newly discovere4
limitations on use or warn of previously unanticipated hazards. And
if labeling can no longer be written to assure that the benefits of
use of a drug outweigh the risks of possible harm, the Commissioner
is empowered, and obligated, to withdraw marketing approval.
The Commissioner believes that information about potential
risks of oral hypoglycemic drugs obtained subsequent to their
initial approval for marketing requires revision of their labeling.
PAGENO="0216"
13466 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Specifically, he believes the study of modes of treatment for adult-
onset diabetes conducted by the University Group Diabetes Program
requiree the addition of a warning about possible cardiovascular compli-
cations associated with the use of such drugs. Because of the importance
of this matter and the concerns it has generated among physicians and
their patients, the Commissioner has concluded that it is appropriate to
invite exploration of the issues in a public forum before reaching a
final determination on the wording of the labeling, including the
warning.
The scientific and legal issues relating to the labeling of oral
hypoglycemic drugs have been the subject of protracted public debate.
To resolve the many complex questions that have been raised, it is
essential that the important issues be identified and that public
comment be directed to these issues. The following discussion is
presented to summarize the history of the oral hypoglycemic labeling
controversy, to identify the issues that have arisen during the con-
troversy, and to explain the position of the Food and Drug Administration
on these issues.
II. ORIGIN OF THE LABELING CONTROVERSY
Although insulin and the oral hypoglycemic drugs are both effective
in lowering the blood glucose level in patients with maturity-onset
diabetes, it is not clear that this reduction of blood glucose has
a beneficial effect on the long term vascular complications of diabetes.
~ ~ ~ .__. .*___ .__.___~,__~
In an attempt to answer this question, the National Institute of Arthritis,
PAGENO="0217"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13467
Metabolism, and Digestive Diseases of the National Institutes of Health
sponsored a long term, prospective clinical trial. The study, begun
in 1961, was conducted by the University Group Diabetes Program (UGDP)
in 12 university medical centers. Patients selected for the study were
maturity-onset diabetics who had been diagnosed no more than 1 year
prior to entry into the study and did not require insulin to remain
symptom-free.
-
All patients were given an appropriate diabetic diet and were
randomly assigned to one of four different treatment groups: (1) Fixed
dose of tolbutamide (1.5 grams/day), (2) Fixed dose of insulin (10 to 16
units based on body surface area), (3) Variable dose of insulin adjust~d
to control the blood glucose, or (4) Placebo. Eighteen months after the
study began, a fifth group was added in which the treatment was a
fixed dose of phenformin hydrochloride (100 milligrams/day). Patient
recruitment was completed in 1966 with a total of 1,027 patients in the
entire study and approximately 200 patients per treatment group.
By 1969 the unexpected finding of a significantly higher mortality
due to cardiovascular causes was present in the tolbutamide group
(12.7 percent or 26 out of 204) compared to the placebo group (4.9 percent
or 10 out of 205), the fixed-dose insulin group (6.2 percent or 13. out
of 210), and the variable insulin group (5.9 percent or 12 out of 204).
After evaluating the available data, the investigators decided to discon-
tinue use of tolbutamide in the study because they concluded that no
~-.-,
benefit had been shown for these patients and there was evidence that
.--,- _. ..__.. ,.___.. .__ ..___ .._.__ _.*___~~__*.* ~.- *~. -, _._ ~..~_*__*__._*__*, ___._.__ ,- ~
the long term use of this drug was associated with a serious side effect.
-. -... ..-.-... -. .- ..........-,- .--..~--. .......,.... ,..-,...--. .-- .. .,. ..-. .. .. ,,-.-.. ,-.-,. ~______________..____
56-592 0 - 75 - pt. 28 - 15
PAGENO="0218"
13468 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
A report on the findings of the UGDP was submitted to the Food
and Drug Administration in March 1970. The i~eport concluded that
"the findings of this study provide no evidence that the combination
_~**__.* **.__*. ~.___*__._.. -- __.** _. .*__ --- -- *__*.___ -- ._~_ _.__ *_*___*__.__ _*___*._* ._. ~ ~
of diet and tolbutamide therapy as described and used for mild non-
insulin dependent diabetics is more effective than diet alone. Moreover,
the findings suggest that tolbutamide and diet nay be less effective,
at least insofar as cardiovascular mortality is concerned, than diet
:- .,~. -*--. -.
alone or than diet plus insulin." The Food and Drug Administration
reviewed the report and convened an ad hoc meeting of experts on May 21,
1970, to evaluate the findings. The report was scheduled for pre-
sentation at the annual meeting of the American Diabetes Association on
June 14, 1970. The program and abstracts for the meeting of the
American Diabetes Association were disseminated in May, however, and the
general findings of the IJGDP study became widely publicized in the
press. In view of this publicity, FDA released a statement to the press
on May 22, 1970, indicating that the agency agreed with the UCDP's
stated conclusions and would require labeling changes for the oral
hypoglycemic drugs to reflect results of the study.
PAGENO="0219"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13469
In October 1970, FDA distributed a Current Drug Information Bulletin
to physicians and other health professionals confirming its agreenent with
the stated conclusions of the UGDP study. The agency recommended that
use of sulfonylurea agents be limited to those patients with symptomatic
onset, nonketotic diabetes who cannot be ad ately controlled by
diet or weight loss alone and in whom the addition of insulin is imprac-
tical or unacceptable.
I
The first report of the UGDP study was published in November 1970 as
a supplement to Diab~j~, the journal ~of the American Diabetes Association
(ref. 1). An accompanying editorial statement representing the view o~
the American Diabetes Association (ref. 2) made the following therapeutic
recommendations:
The clearest indication for oral agents is
diabetes of mild or moderate severity in a patient
who proves to be poorly controlled with diet and who
is unable or unwilling to take insulin. In adult-
onset diabetes with hyperglycemia and glycosuria,
symptomatic or not, and in the absence of ketosis,
a trial with an appropriate diet should come first.
If this does not establish satisfactory control,
insulin iè to be preferred to other therapeutic
~
controlling hyperglycemia and the UGDP study -
indicates that it may be safer.
~ _.~~__* ~`_,---- _~_.._ ._ ~ ~ ._____* ~_..*
PAGENO="0220"
13470 COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY
A statement published at the same time by The American Medical Association
Council on Drugs (ref. 3) included the following recommendations:
Although some flaws exist in the UGDP study, it
clearly demonstrates that every effort should be
made by the physician to control the symptomatic,
maturity-onset diabetic with diet alone. Should
this fail, treatment with insulin or oral hypoglycemic
agents should be undertaken. If oral hypoglycemic
agents are selected for therapy the results of the
UGDP study should be kept in mind. Therefore, the
consideration of treatment with oral hypoglycemic
agents should be secondary to the use of insulin.
In May 1971 the use of phenformin in the UGDP study also was discon-
tinued because there was a significantly higher cardiovascular mortality
in the phenformin group (12.7 percent or 26 out of 204) compared to the
other treatment groups. The preliminary results with phenformin were
published in August 1971 (ref. 4). An additional report by the UGDP
published in November 1971 discussed the clinical implications of the
UGDP study (ref. 5).
In June 1971 the Food and Drug Administration issued a Drug Bulletim
outlining changes in the labeling for all sulfonylurea drugs. The Drug
Bulletin stated that diet and reduction of excess weight are the
foundation of therapy of diabetes mellitus, and that when the disease is
adequately controlled by these measures, no other therapy is indicated.
PAGENO="0221"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13471
The Bullef in also stated that the sulfonylurea agents are indicated in the
treatment of adult-onset, nonketotic diabetes mellitus which cannot be
adequately controlled by diet and reduction of excess weight alone and
when, in the judgment of the physician, insulin treatment is not feasi1~le.
From the time the results of the UGDP study were first reported, the
study was subjected to intense criticism by both clinicians and statisti-
cians (ref. 6 through 12). The basic scientific criticisms of the
study were as follows:
1. Patient selection was inappropriate in that many patients
had such mild diabetes that neither oral drugs nor insulin was indicated.
2. Total mortality in the tolbutamide group was not significantly
different from that in the placebo group.
3. Excess cardiovascular mortality occurred in only a few clinics.
4. Randomization was not successful; therefore, the tolbutanide
group was not comparable to the other groups at the outset of the
study with respect to baseline cardiovascular risk factors.
5. With the exception of the variable insulin group, patients were
maintained on a fixed drug dosage, contrary to the principles of good
medical practice.
6. The use of tolbutamide and phenformin in the study was termir~ated~
prematurely, i.e., before definitive results were obtained.
7. The results of the study are contradicted by the studies of
Keen (ref. 13 through 15) and of Paasikivi (ref. 16).
These criticisms were in turn analyzed by representatives of the
IJGDP (ref. 17) and by a statistician who had served as a consultant tp
the IJGDP (ref. 18) and were rejected as a basis for invalidating the
conclusions of the UGDP study. By this time, however, a widespread
PAGENO="0222"
13472 COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY
belief had developed among many physicians that the UGDP study was
somehow flawed in terms of its design and execution and therefore could
not serve as a proper basis for a warning to the medical profession
Uncertainty about the scientific quality of the UGDP study has been
a prominent feature of all critical commentary since 1970 and has clearly
inhibited acceptance by the medical profession of the study's most
troubling finding namely that the administration of either tolbutamide
or phenfornin to patients with maturity-onset diabetes was associated
with an increase in cardiovascular moxtality Undoubtedly one reason
many practicing physicians were surprised ~
tlie findings of the UGDP study is that the reported increase in cardio-
- ~ .._*_._~_ ~ ~_~.___~_____ .*. _~~____.._ ~~-~-- *_.~_~.*_. ___. ~ ___*_*_*._~__.. -
vascular-mortality--though statistically significant--is not of the
magnitude which can be readily detected.by the individual physician in
the course of practice
The Commissioner recognizes that a large number of physicians still
do not accept the position of the Food and Drug Administration as
expressed in the FDA Drug Bulletin, or the position of the American
Diabetes Association and the American Medical Association Council on
* Drugs as expressed *in the -references cited. An outcome of this dis-
agreement was a prolonged legal confrontation that precluded the
inclusion of warnings in the labeling for oral hypoglycemic drugs
similar to those appearing in the Drug Bulletin
III LEGAL CHALLENGE TO THE LABELING OF ORAL HYPOGLYCEMIC DRUGS
In November 1970 a group of physicians known as the Committee *
on the Care of the Diabetic was formed to oppose the proposed warning
PAGENO="0223"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 134~3
labeling for oral hypoglycemic drugs. The group included some of the
country's leading diabetologists.
In October 1971 the Committee on the Care of the Diabetic petitioned
the Commissioner to rescind his position that labeling for oral hypogly~-
cemic drugs must contain a warning of associated cardiovascular hazards.
The committee maintained that the UGDP study cons~ituted an improper
basis for the agency's decision, because it had been criticized on
scientific, clinical, statistical, aç~d other grounds. The Committee on
the Care of Diabetes cited "conç sing data," particularly the studies
of Keen et al. (ref. 13 through 15) and Paasikivi (ref. 16), which, it
contended, demonstrated the safety of oral hypoglycemic therapy. The
committee also insisted that labe~4~g for .these drugs must reflect a
"fairbalanc&'ofscientific~n~n~d cite the alleged deficiencies
of the UGDP study and the controversial nature of its conclusions as
well as the data in controversy.
After thorough evaluation of all the materials submitted to the
agency,' the Commissioner formally replied to counsel for the Committee
on the Care of the Diabetic on June 5, 1972. The Commissioner's letter
responded to each of the criticisms raised by the committee concerning
the UGDP study and the agency's position. The Commissioner reaffirmed
the position of the Food and Drug Administration that an undiluted and
unencumbered warning in the labeling of the oral hypoglycemic drugs
regarding cardiovascular hazards was fully warranted by the available
~
evidence.
PAGENO="0224"
13474 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The agency's position on labeling for these drugs was again stated
in an FDA Drug Bulletin issued in May 1972. Based on the results for
phenformin reported by the UGDP in 1971, the following labeling changes
were to apply to the biguanide drugs as well as the sulfonylurea drugs:
Because of the apparent increased cardiovascular
hazard associated with oral hypoglycemic agents, they
are indicated in adult-onset, nonketotic diabetes
mellitus only when the condition cannot be adequately
controlled by diet and red~iction of excess weight
alone, and when, in the judgment of the physician,
insulin cannot be employed because of patient unwilling-
ness, poor adherence to injection regimen, physical
disabilities such as poor vision and unsteady hands,
insulin allergy, employment requirements, and other
similar factors.
On July 13, 1972, counsel for the Committee on the Care of the
Diabetic requested a formal evidentiary hearing before the agency.
The Commissioner advised the petitioners that they were not entitled
to a hearing since their submission did not meet the statutory
standard of "substantial evidence" and stated that the Commissioner's
letters constituted final agency action.
Soon thereafter suit was filed in the United States District
Court for the District of Massachusetts by a group of 178 physicians,
many of them members of the Committee on the Care of the Diabetic, asking
that the Food and Drug Administration be enjoined from requiring manu-
facturers to include a warning of associated cardiovascular hazards
PAGENO="0225"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13475
in their labeling for oral hypoglycemic drugs (Bradley v. Richardson,
Civil No. 72-2517 M (D. Mass. 1972)). A temporary restraining order
was entered by the court on the sane day. A hearing on the notion
for a preliminary injunction was held before Judge Campbell on
August 17, 1972, and,, on August 30, 1972, he denied an injunction.
Judge Campbell concluded that the plaintiffs had not demonstrated
a reasonable probability of prevailing on the merits since the admini-
strative action of the Food and Drug Administration, requiring an
unencumbered warning, was a reasonable exercise of its statutory duty
and the potential harm to users of the drugs was greater than any harm
to the manufacturers or prescribers. Judge Campbell further observed
that the Food and Drug Administration labeling requirements would not
preclude physicians from exercising their best clinical judgment.
The plaintiffs filed another motion for a temporary restraining
order and preliminary injuction on October 17, 1972, specifically reqtlest-
ing that the agency be enjoined unless the drug warning was redrafted
to incorporate their views concerning the interpretation of the TJGDP
study. The plaintiffs argued that, without such a discussion, the
labeling required by the Food and Drug Administration was misleading
because it failed to reveal the existence of divergent opinion among
experts, contrary to the agency's own regulation, § 1.3 (21 CFR 1.3).
On November 3, 1972, the District Court issued a temporary restraining
order, which became a preliminary injunction on November 7, 1972,
restraining the agency from implementing the labeling.
PAGENO="0226"
13476 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
On July 31, 1973, the United States Court of Appeals for the
First Circuit vacated the District Court's injunction and remanded
the case to the Food and Drug Administration for its further deter-
mination. In its opinion the Court ruled that the plaintiffs
failed to exhaust their administrative remedies regarding the
issues presented. The Court expressed its awareness of negotiations
between the parties to arrive at a mutually acceptable solution
even during litigation, and also expressed its belief that the
remand could well produce the most informed and responsible solution
possible (483 F.2d 410 (1st. Cir. 1973)).
In its opinion the Court of Appeals also noted apparent incon-
sistency between the agency's regulation on the disclosure of
differences of medical opinion in § 1.3 and the substantial evidence
requirements added to the Federal Food, Drug, and Cosmetic Act by
the Drug Amendments of 1962. The Court directed the Commissioner
to consider § 1.3 as it relates not only to the substantial evidence
standard but also to the misbranding requirements of the act. The
agency is revising § 1.3, by order published elsewhere in this issue
of the FEDERAL REGISTER, to bring the regulation into conformity with
these related provisions of the law. As revised § 1.3 does not permit
a statement of differences of opinion in required warnings in the
-:-.- --
labeling of drugs.
PAGENO="0227"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13477
It should be noted that no manufacturer of oral hypoglycemic
drugs has initiated proceedings challenging the Commissioner's
authority to require changes in the labeling of its products, or
~
that the agency proposed to require.
After the Court of Appeals vacated the preliminary injunction
in July 1973, the Food and Drug Administration undertook additional
discussions concerning the labeling of the oral hypoglycemic agents
with interested individuals and gr~ups. In October 1973, the
Director, Bureau of Drugs, and other members of the Food and Drug
*Administration met with representatives of the Committee on the
Care of the Diabetic, the American Medical Association, the Americat~
Diabetes Association, the National Institutes of Health, and manu-
facturers of hypoglycemic drugs to discuss procedures that would
facilitate the. issuance of appropriate labeling. Based upon the
discussion and input from the agency's staff, proposed labeling
revisions were circulated for comments in February 1974 to those
who attended the meeting and to other interested persons. Addressees
were also invited to meet with agency officials, if desired, to
discuss the labeling. Four such meetings were held between March
21 and April 24, 1974. The minutes of these meetings have been
placed on public display in the office of the Hearing Clerk.
PAGENO="0228"
13478 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The responses to the proposed labeling, including comnents received
at these meetings, revealed continuing major differences of opinion over
the scientific validity of the UGDP study and over the asserted need for
"fair balance" and the acknowledgment of "controversy" in the proposed
warning. In addition, the Food and Drug Administration was advised that
a major outside Teview, described below, of the UGDP study by a committee
of the Biometrics Society was near completion.
The agency therefore decided to postpone implementation of the
warning until this review was published. Since the UGDP study was the
basis for the proposed warning, the Commissioner believed that this
independent review of the statistical validity of the study should be
available to all interested persons before taking definitive action.
The review by the committee of the Biometrics Society required extensive
reanalysis of the data in the UGDP study and was mat published until
February 10, 1975 (ref. 19). A more detailed report of the UGDP on
phenformin was also published recently (ref. 20).
The Commissioner believes that sufficient time has passed to~h~y~
permitted all interested persons to study these reports. Since no
maj or new inf ormation in regard to the UGDP study a an
Commissioner believes it is now essential to effect all labeling changes
that are appropriate and necessary on the basis of the UGDP study.
On June 11, 1975, and June 18, 1975, representatives of the Food
and Drug Administration met with representatives of the Committee on
the Care of the Diabetic to discuss late drafts of the Indications and
PAGENO="0229"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13479
Warnings sections of the labeling proposed in this notice. The
representatives of the Committee on the Care of the Diabetic included
one of the plaintiffs in Bradlei v. Weinberger and the plaintiffs'
attorney. The purpose of these meetings was to engage in good faith
negotiation in an attempt to resolve outstanding issues in conformity
with the intent of the Court of Appeals. Memoranda of these meetings
and of subsequent phone calls and drafts of labeling discussed at the
_______*._* _*~.__.*~_ ._ _. *.* --- _._. ._ _._ ._ ._ -- ---- ._ ----
meetings are on file in the office of the Hearing Clerk.
IV. REVIEW OF BIOSTATISTICAL ISSUES BY
THE COMMITTEE OF THE BIOMETRICS SOCIETY
The UGDP study was subjected to intense adverse criticism
(ref. 6 through 12) largely on the basis of its design and the
statistical analysis of the results. For this reason, the National
Institute of Arthritis, Metabolism, and Digestive Diseases, which
financed the UGDP study, sought an independent review of the study.
In 1972 a contract was awarded to the Biometrics Society, an inter-
national organization of biostatisticians, to make an in-depth assess-
ment of the scientific quality of the UGDP study, particularly the
biometric aspects of the design, conduct, and analysis of the trial,
and a similar assessment of other controlled trials involving oral
hypoglycemic agents. A committee of six members was selected to undertake
this task. The committee visited the UGDP coordinating center and two
of the clinical centers to study methods used in the trial, reviewed
PAGENO="0230"
13480 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
published criticisms of the UGDP study in detail, interviewed both
critics and supporters of the study, and made new analyses from the
original data
On the basis of this in-depth review the Biometrics Societ)
committee commented as follows on the major criticisms of the IJGDP
study
1. The criticism that patient selection was inappropriate was
considered to be `largely irrelevant to the primary issue raised by
the critics viz the validity of the evidence pointing to excess
mortality in the tolbutamide~ and phenformin-treated groups The
committee argued that even if it could be shown that the study group
contained some non_diabetics* * * (a] drug found toxic in such subjects
would not likely be counted safe for persons with we1p~ume~t~e&m~.1d.
diabetes either."
2. Wih respect to the criticism that total mortality in the
tolbutamide group was not significantly different from that in the
placebo group the committee concluded that this criticism `has some
weight (although we do not interpret it as a criticism of the action
of the UGDP) and that the toxic effect of the oral hypoglycemics
cannot be affirmed with the certainty that would be present if total
mortality were significantly different
3 In response to the criticism that excess mortality occurred in
only a few clinics the committee presented calculations of the data to
take account of the number of patients treated in each clinic and the
duration of their treatment and concluded that the excess mortality
PAGENO="0231"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13481
is not in fact confined to a few clinics and, that this particular
criticisn should not be taken to detract from the interpretation of
the UGDP findings."
4. The contention that randomization was not successful was
studied in detail by the committee which identified "a puzzling
anomaly concerning the distribution of the two sexes to the four
treatment groups within clinics." The committee reviewed the
randomization procedure in detail and examined the log books contain-
ing records of the allocation of each patient. The committee's
report reads: "We were not able to find an assignable cause for the
surprising allocation of the sexes to treatments but have no reason
to think that the study has been compromised by a breakdown in the
randomization of patients to the treatment groups. Because of the
imbalance of sexes in the treatment groups in some clinics, however,
allowance for this has been made in our analysis." The committee went
on to analyze the data by several different statistical approaches,
including those used originally by the UGDP investigation. The com-
mittee concluded: "Our findings* * * take into account the differences
between centers and the differences in length of treatment, as well
as the baseline variables. They support the view of Cornfield [ref. 18]
that there is no evidence that the baseline differences arising from ~he
randomization contributed in any important way to the finding of adverse
effect from tolbutamide."
PAGENO="0232"
13482 COMPETITIVE PROBLEMS IN THE' DRUG INDUSTRY
5. The criticism that the oral hypoglycemic drugs were given in
fixed dosage was rejected by the committee, with respect to conclusions
regarding toxicity, as follows: "It is true that the use of a fixed
dose of drug, which was also the approach adopted by Feldnan et al.
[ref. 211 and Keen and Jarrett [ref. 14], limits the generalization
about therapeutic effects, but since the dose of tolbutamide is
about equal to the average recommended for therapeutic use, an evalu-
ation of its possible toxic effect is~highly relevant."
6. Concerning the criticism that the use of tolbutamide and
phenformin was terminated prematurely, the committee acknowledged that
"It would have been easier to interpret the findings if there were
more data on mortality." The committee also recognized, however, the
ethical issues raised by continuing these drugs in the study and
concluded: "We do not criticize the UGDP investigators for having
made the decision when they did. Nevertheless, the result of that
decision is to leave us with some residual uncertainty about the
meaning of the findings, a point that is well understood by the UGDP
investigators themselves."
7. , In considering the criticism that the results of the UGDP
study are contradicted by the studies of Keen (ref. 13 through 15) and
of Paasikivi (ref. 16), the committee analyzed these studies in detail.
With respect to the data of Keen and his colleagues, they concluded that,
in their ongoing prospective study, neither cardiovascular mortality nor
total mortality in the tolbutamide group is significantly different from
PAGENO="0233"
COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY 1348~3
that in the placebo group. Because of imperfections in the randomization
process and in the maintenance of blinding, and because of the pre-
liminary nature of the data obtained to date, the committee concluded
that "the provisional data that Dr. Keen has kindly sent us~' * * do not
throw doubt on the UCDP findings in regard to deaths from cardiovascular
causes." In regard to the Paasikivi study, which appeared to show a
beneficial effect of tolbutamide on mortality in the first year in
patients who survived a first nyocardial infarction, the committee
concluded: "This study' neither confirms nor contradicts the IJGDP
findings, as the population under consideration was not one of maturity-
onset diabetics, and the patients taking tolbutamide had been exposed
to a relatively small dose for a shorter time than that applied in the
UGDP study." The studies of Feldman et al. (ref. 21) and of Tzagournis
and Reynertson (ref. 22) were also briefly reviewed by the committee.
Their conclusion was that in neither study has a sufficient number of
deaths yet occurred to permit meaningful interpretation of results.
In addition to evaluating these criticisms of the tIGDP study, the
Biometrics Society committee conducted extensive new analyses of the
UGDP data, taking into account the effect of various baseline variables
and cardiovascular risk factors. These analyses confirmed that cardio
vascular mortality was increased in the tolbutamide group. ~
was ~
agp of 53, but not in males. An important finding was that the highest
death rate occurred in the group of patients who adhered most closely to
the tolbutamide regimen and did not have their dose modified. Also when
56-592 0 - 75 - pt. 28 - 16
PAGENO="0234"
13484 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
the analysis was conducted according to an approach called the survival
modeling method, which takes into account the proportion of time each
patient received the assigned medication, women in the tolbutamide group
had a statistièally significant increase in both cardiovascular and
total mortality. This does not mean that the study necessarily showed
the drug to carry less risk in males. On this point the committee
concluded: "The data do Cot support the same conclusions for men, but
one possible reason is that the smalLer number of patients in the male
group results in lack of sensitivity to detect differences of moderate
magnitude."
In the final section of its report, the Biometrics Society committee
summarized its conclusions:
Although we have concerned ourselves
almost entirely with issues related to the
possible toxicity of tolbutamide, we wish to
point out that one of the valuable aspects of
the completed TJGDP trial will be the provision
of data on the long term treatment of adult-
`onset diabetes with insulin. It is already
* clear that the benefits from this treatment
are not dramatic, and the only worthwhile
information about them will have to come from
the relatively precise methods of a controlled
clinical trial'. In this sphere, the UGt~P
trial has no competitor* * *
PAGENO="0235"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1348~
On the question of cardiovascular nortality
_.___ ~ ___..__ .~.__ - -~- ~ __~__.. _.__ ~ _*~---.;-_.-__-.-__--_
due to tolbutamide and phenformin we consider
that the UGDP trial has raised suspicions that
cannot be dismissed on the basis of other
evidence presently available.
.-.- .~.., ..-- ~ -.-..- .~.-...-.-. ..- .. -~. .....
We find most of the criticism levelled
against the UGDP findings on this point unper-
suasive The possibility that deaths may have
been allocated to cardiovascular causes prefer-
entially in the groups receiving oral therapy
exists, and, in view of the `nonsignificance' of
differences in total mortality, some reservations
about the conclusion that the oral hyperglycemics
[sic} are toxic must remain Nonetheless we
consider the evidence of harmfulness moderately
strong The risk is clearly seen in the group
of older women* * * Whether it affects all
subgroups of patients cannot be decided on -
the basis of the available data owing to the
small number of deaths involved in these
subgroups* * *
PAGENO="0236"
13486 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
In conclusion, we consider that in the
light of the UGDP findings, it remains with the
proponents of the oral hyperglycemics [sic} to
conduct scientifically adequate studies to justify
the continued use of such agents.
V. RECENT ADDITIONAL INFORMATION ON SAFETY OF ORAL HYPOGLYCEMIC
DRUGS
The more detailed report. on the results of the phenformin study
was published recently by the UGDP (ref. 20). In addition to the higher
mortality from all causes and from cardiovascular causes observed in the
phenformin-treated group compared to the other treatment groups,_evidence
was presented that phenformin therapy resulted in increased blood pressure
levels and heart rate, thus suggesting possible mechanisms by which this
drug might influence cardiovascular mortality.
Recently, additional reports relating to the safety of oral hypo-
glycemic drugs have appeared:
1. At hearings before the Subcommittee on Monopoly of the Select
Committee on Small Business, U.S. Senate, on January 31, 1975, Dr. P. J.
Palumbo reported that a retrospective study of diabetic patients treated
at the Mayo Clinic suggests that survival was lower in those patients
treated with oral hypoglycemic agents, compared to those patients treated
with insulin. The full study has not yet been published.
2. A retrospective study of diabetic patients treated at the
Joslin Clinic, reported in a doctoral thesis (ref. 23), can be interpreted
as providing results that are consistent with those of the UGDP. This
study has not yet appeared in the medical literature.
PAGENO="0237"
COMPETITIVE PROBLEMS IN THE DRuG INDTJSTRY 13487
3. A positive inotropic effect, i.e., increased force of muscular
contraction, of sulfonylurea agents on the heart muscle has been demon-
strated (ref. 24 and 25). The increased oxygen requirement resulting
from such an effect could have a deleterious effect in patients with
coronary artery disease. Limited animal studies also suggest that the
sulfonylurea agents may affect the excitability of heart muscle (ref.
25), which could predispose the heart to develop abnormal rhythms,
particularly in the presence of a decreased oxygen supply.
4. Results from a study on the, chronic effects of tolbutamide in
the rhesus monkey by R. W. Wissler et al. (FDA contract 72-114) indicate
there is an increased frequency and severity of atherosclerotic lesions
in the coronary arteries of the tolbutanide-fed monkeys compared to the
control monkeys (ref. 26). The final report of this study is under
review.
While neither of the two epidemiological studies is a prospective
clinical trial such as the UGDP study, the preliminary reports indicate
that further information casting doubt on the safety of the oral hypogl~-
cemic drugs may be forthcoming. And, although the animal findings
cannot be considered necessarily relevant to the issue of excess cardio-
vascular mortality in diabetic patients, they indicate that sulfonylureas
may have potentially adverse effects on the cardiovascular system of
certain animals which can be detected by appropriate pharmacological and
~~icaltest~
In addition to these reports, two critiques of the Biometrics
Society committee report have recently been published (ref. 27 and 28).
PAGENO="0238"
13488 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
VI. DISCUSSION OF PROPOSED LABELING FOR ORAL HYPOGLYCEMIC DRUGS
The judgment of the Commissioner that changes must be made in
the labeling of the oral hypoglycemic drugs to reflect the findings
of the UGDP study is well known from previously published statements.
The Commissioner is therefore proposing labeling in this notice for
public comment and scheduling a public hearing to receive additional
data information and views After consideration of all materials
submitted, the COmmissioner will publish final labeling for oral
hypoglycemic drugs in the FEDERAL REGISTER.
The warning proposed in this labeling for oral hypoglycemic drugs
is based primarily on a thorough review and evaluation of the UGDP
study In proposing the overall labeling the Commissioner has also
carefully considered
1. Published reviews, criticisms, and rejoinders to criticisms of the
UGDP study.
2. Other scientific and clinical investigations of the oral
hypoglycemic agents.. . .
3 The advice of experts
4 Comments submitted to the agency by interested persons
The Commissioner reaffirms his conclusion that the UGDP study is
an adequate and well-controlled clinical trial which is the most
extensive and detailed examination of long term administration of
hypoglycemic agents yet undertaken Although the study has shortcomings
which might be expected in any clinical trial of this complexity
PAGENO="0239"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 134~9
the shortcomings do not invalidate the central finding that there
appears to be an increased risk of cardiovascular mortality
associated with the administration of tolbutamide and of phen-
forniin to maturity-onset diabetic patients, compared to treatment
with diet alone or diet plus insulin. This conclusion has in the
past been reached independently by the IJGDP investigators, the
FDA, and the Biometrics Society committee, and is again affirmed by
the Commissioner. Other clinical trials of these oral hypoglycemic
drugs are not comparable to the UGDP study and provide insufficient
evidence to negate the findings of the UGDP study.
Accordingly, although comments concerning the validity of the
UGDP study and its conclusion will be accepted, comments on this
issue that contribute no new information and only reiterate
published criticisms, which have already been extensively reviewed
by the Food and Drug Administration, are not considered useful at
this time.
-The Commissioner proposes that a boxed warning concerning the
possible increased risk of cardiovascular mortality be included in
the. labeling for these drugs. This warning is based on the findings
of the UGDP study. The Commissioner emphasizes that the requirement
for such a warning does not depend upon an absolute certainty that
the findings of the UGDP study are correct. Prudence dictates that
PAGENO="0240"
13490 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
a warning be issued whenever there is sufficient evidence from
controlled or uncontrolled studies to believe that a drug may be
hazardous or carry a risk and that such warning is necessary for
safe and effective use of the drug by physicians and paUents. The
Federal Food, Drm~g, and Cosmetic Act provides no standard for the
amount or character of scientific evidence required for the issuance
of a warning. The decision to require a warning is a matter of
judgment ~ must be made in light ~of both the availables e
evidence and the opinion of experts who interpret that evidence. The
Commissioner believes that the UGDP study is a validly conducted trial
and accepts the opinion of the Biometric Society committee and other
experts that the increased cardiovascular mortality found in this
trial* to be associated with these drugs cannot reasonably be attributed
to scientific shortcomings in the study. Under those circumstances,
a clear warning is necessary even though a residual uncertainty over
thd correctness of the study may be present. Warnings may properly be
required on the basis of evidence that falls short of conclusive proof.
In conformity with Food and Drug Administration policy that
warnings must be presented in unambiguous, terms without disclaimers
or qualifications that would undermine or destroy their usefulness,
there is no mention in the proposed warning of other studies
involving the oral hypoglycemic drugs. The mention of studies~
in which increased cardiovascular mortality was not found would
serve only to encumber the warning and would therefore not be con-
`sistent with revised ~ 1.3. Comments concerning the principle of
PAGENO="0241"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13491
an unencumbered warning, which have been received and considered in
conjunction with the proposed revision of ~ 1.3 published in the
FEDERAL REGISTER of September 16, 1974 (39 FR 33229), are addressed
in the final regulation published elsewhere in this issue of the
FEDERAL REGISTER.
The proposed warning does, however, contain a statement acknowledg-
ing the controversy that exists over the interpretation of the TJGDP
study and states that, in spite of this, the IJGDP findings provide
adequate scientific basis for a warning. The purpose of this statement
is to emphasize clearly the basis for the warning. Comments on specific
wording in the proposed warning are invited by this notice. The Commis-
sioner advises, however, that he does not intend to reopen consideration
of the principle of an unencumbered warning which is embodied in the
final regulation relating to § 1.3.
The Commissioner concludes that, from the standpoint of patient
safety, it is prudent to apply the possible increased risk of cardio-
vascular mortality for tolbutamide and phenformin to other sulfonylurea
and biguanide drugs in view of the similarities in chemical structure
and mode of action for members within each of these two categories.
This position was endorsed by the Endocrinology and Metabolism Advisory
Committee of the FDA at its meeting on June 28, 1971, but additional
comment at thià time would also be appropriate.
PAGENO="0242"
13492 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The Commissioner also concludes that a patient population exists
for which these drugs, properly labeled, can be consid~4 sfe and
effective. Marketing therefore may continue. The Commissioner is
proposing, however, that this patient population be limited to patients
with maturity-onset diabetes whose symptoms or blood glucose level
cannot be controlled by diet alone and who cannot take insulin for one
or more of the reasons identified in the labeling. This restriction in
labeling has been opposed in the past on the ground that it interfered
with the practice of medicine. The Commissioner recognizes that drug
labeling impacts on the practice of medicine. For this reason the Food
and Drug Administration has an obligation to ensure that drug labeling
is as correct and accurate as possible and meets the statutory standard
of describing the conditions of use under which the drug may be considered
safe and effective. Those limitations on use that properly derive from
a known hazard or potential risk must, in the interest of safety, be
included in drug labeling. This principle is stated in the proposed
regulations on prescription drug labeling (published in the FEDERAL
REGISTER of April 7, 1975 (40 FR 15392)), time for comment on which has
been extended to August 6, 1975, by notice published in the FEDERAL
REGISTER of June 11, 1975 (40 FR 24909).
The Commissioner proposes the appended labeling for oral hypoglycemic
agents of the sulfonylurea and biguanide categories as labeling providing
the essential information .for the safe and effective use of these drugs.
Comments addressed to any portion of the labeling will be considered.
PAGENO="0243"
COMPETITIVE PROBLEMS IN THE DRuG INDTJSTRY 13493
REFERENCES
Copies of all references cited below are on public display in the
office of the Hearing Clerk, Food and Drug Administration, Room 4-65,
5600 Fishers Lane, Rockville, MD 20852:
1. The University Group Diabetes Program, "A Study of the Effects
of Hypoglycemic Agents on Vascular Complications in Patients with Adult-p
Onset Diabetes. I. Design, Methods and Baseline Results. II. Mortailty
Results," Diabetes, 19 (supp. 2):747-830, 1970.
2. Ricketts, H. T., "Editorial Statement," Diabetes, 19 (supp. 2):
iii-v, 1970.
3. AMA Council on Drugs, "Statement Regarding the University Grou~
Diabetes Program (UGDP) Study," Diabetes, 19 (supp. 2):vi-vii, 1970.
4. The University Group Diabetes Program, "Effects of Hypoglycemic
Agents on Vascular Complications in Patients with Adult-Onset Diabetes.
IV. A Preliminary Report on Phenformin Results," Journal of the American
Medical Association, 217:777-784, l~7l.
5. The University Group Diabetes Program, "Effects of Hypoglycemic
Agents on Vascular Complications in Patients with Adult-Onset Diabetes.
III. Clinical Implications of UGDP Results," Journal of the American
Medical Association, 218:1400-1410, 1971.
6. Reeves, R. L'., "Erroneous Interpretation of Data" (corresponde~ice),
Northwest Medicine, 69:470, 1970.
PAGENO="0244"
13494 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
7. Hare, R. L., "Therapeutics--by Headline and Edict," Northwest
Medicine, 70:118-119, 1971.
8. ,Feinstein, A. R., "Clinical Biostatistics. VIII. An Analytic
Appraisal of the University Group Diabetes Program (UGDP) Study,"
Clinical Pharmacology and Therapeutics, 12:167-191, 1971.
9. Schor, S., "The University Group Diabetes Program. A Statistician
Looks at the Mortality Results," Journal of the American Medical Association,
217:1671-1675, 1971.
10. Seltzer, H. D., "Tolbutamide Mortality in the Six Phenformin
Clin~ics," (letter to the editor), Journal of_the American Medical
Association, 218:594, 1971.
11. Bramdman, 0., "University Group Diabetes Program and the Practicing
Physician," Journal of the Medical Society of New Jersey~, 68:909-911, 1971.
12. Seltzer, H. S., "A Summary of Criticisms of the Findings and
Conclusions of the tiniversity Group Diabetes Program (UGDP)," Diabetes,
21:976-979, 1972.
13. Keen, H., R. 3. Jarrett, C. Chlouverakis, and D. R. Boyns, "The
Effect of Treatment of Moderate Hyperglycemia on the Incidence of Arterial
Disease," Postgraduate Medical Journal, 44:960-965, 1968.
14. Keen, H., and R. 3. Jarrett, "The Effect of Carbohydrate
Tolerance on Plasma Lipids and Atherosclerosis In Man," in Atherosclerosis,
R. 3. Jones, editor,. Berlin, Springer-Verlag, 1970, pp. 435-444.
15. Keen, H., "Factors Influencing the Progress of Atherosclerosis
in the Diabetic," Acts Diabetolo~ca Latina, 8 (supp. l):444-456, 1971.
PAGENO="0245"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13495
16. Paasikivi, J., "Long-Term Tolbutamide Treatment after Myocardial
Infarction. A Clinical Biochemical Study of 178 Patients Without Overt
Diabetes," Acta Medica Scandinavica, Supplernent, 507:3-82, 1970.
17. Prout, T. E., G. L. Knatterud, C. L. Meinert and C. R. Klint,
"The UGDP Controversy. Clinical Trials Versus Clinical Impressions,"
Diabetes, 21:1035-1040, 1972.
18. Cornfield, J., "The University Group Diabetes Program. A
Further Statistical Analysis of the MortalityFindings," Journal of the
American Medical Association, 217:1676-1687, 1971.
19. "Report of the Committee for the Assessment of Biometric
Aspects of Controlled Trials of Hypoglycemic Agents," Journal of the
American Medical Association, 231:583-608, 1975.
20. The University Group Diabetes Program, "A Study of the Effects
of Hypoglycemic Agents on Vascular Complications in Patients with Adult-
Onset Diabetes. V. Evaluation of Phenformin Therapy," Diabetes, 24
(supp. 1) :65-184, 1975.
21. Feldman, R., D. Crawford, R. Elashoff, and A. Glass, "Progress
Report on the Prophylactic Use of Oral Hypoglycemic Drugs in Asymptomatic
Diabetes: Neurovascular Studies," Advances in Metabolic Disorders,
2 (supp. 2):557-567, 1973.
22. Tzagournis, N., and R. Reynertson, "Mortality from Coronary
Heart Disease During Phenformin Therapy," Annals of Internal Medicine,
76:587-592, 1972.
23. Kanarek, P. H., "Assessing Survival in a Diabetic Population,'~
thesis, Harvard School of Public Health, Boston, January 1973.
PAGENO="0246"
13496 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
24 Lasseter K C C S Levey R F Palmer and J S McCarthy
`The Effect of Sulfonylurea Drugs on Rabbit Myocardial Contractibility,
Canine Purkinje Fiber Automaticity and Adenyl Cyclose Activity from
Rabbit and Human Hearts The Journal of Clinical Investigation 51 2429-
2434, 1972
25 Levey G S K C Lasseter and R F Palmer Sulfonylureas
and the Heart Annual Review of Medicine 25 69-74 1974
26 Wissler R W J Borensztajn G S Godfrey G Seymour
A Rubenstein C H Ts'ao A Study of the Chronic Effects of
Tolbutamide in the Rhesus Monkey FDA contract 72-114 February 15
1975
27. Bradley, R. F., H. Dolger, P. H. Forsham, and H. Seltzer,
Settling the IJGDP Controversy?' Journal of the American Medical
Association 232 813-817 1975
28 O'Sullivan J B and R B D'Agostino Decisive Factors
in the Tolbutamide Controversy Journal of the American Medical
Association, 232:825-829, 1975.
PAGENO="0247"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13497
VII. NOTICE OF PUBLIC HEARING
The Commissioner concludes that, to permit maximum public
participation in the development of labeling requii~enents for oral
hypoglycemic drug products, a public hearing shall be held to pro-
vide an opportunity for interested persons to present data, information,
and views on the proposed labeling. This public hearing is ordered
pursuant to § 2.400(a) (21 CFR 2.400(a)) and shall be conducted in
accordance with the procedures established in Subpart E of Part 2
of the regulations. The Commissioner has designated J. Richard
Grout, M.D., Director, Bureau of Drugs, to be the presiding officer
at such hearing, to be held August 20, 1975, beginning at 9 a.m. in
Conference Rm. E, Parklawn Bldg., 5600 Fishers Lane, Rockville, MD
20852.
Interested persons who wish to make an oral presentation at
the hearing shall file a written notice of appearance with the
Hearing Clerk, Food and Drug Administration, Bin. 4-65, 5600 Fishers
Lane, Rockville, MD 20852 by close of business August 6, 1975. The
PAGENO="0248"
13498 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
notice of appearance shall state the approximate amount of time
requested by the person for presentation. It shall also give the
telephone number of the person to be contacted regarding the schedule
for presentation. Individuals and organizations with common interests
are strongly urged to consolidate or coordinate their presentations
because of the limitations of time.
By August 11, 1975, the Food and Drug Administration will communicate
by telephone with each person who requested an opportunity to be heard,
regarding the time his or her oral presentation is scheduled to begin
aiid the amount of time allocated for his or her presentation. The
Food and Drug Administration may require joint presentations by persons
sharing common views. The Food and Drug Administration will prepare a
hearing schedule, listing the participants and the time allotted to each,
which shall be filed with the Hearing Clerk and a copy mailed to each
participant.
The hearing will be transcribed. Any interested person may, consis-
tent with the orderly conduct of the meeting, also record or otherwise
make his or her own transcript of the meeting. Each participant may
use the allotted time however he or she desires, consistent with
decorum and order, and may present written data, information or views
for inclusion in the record of the hearing. Any person who desires to
submit an advance written statement may do so in qu~ntuplicate to the
Hearing Clerk. All written comments and statements submitted before
PAGENO="0249"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13499
August 15, 1975, will be reviewed by the presiding officer prior to
the hearing, so that full repetition at the hearing will be unnecessary.
A participant nay be accompanied by any number of additional persons.
If a participant is not present when his or her presentation is
scheduled to begin, the participants following will be taken in ordet.
An attempt will be made to hear any scheduled participant who misses
his assigned time at the conclusion of the hearing. Other interested
persons attending the hearing who did not request an opportunity to
speak will be given an opportunity to make oral presentations at the
conclusion of the hearing to the extent that time permits.
The presiding officer, as well as any other Food and Drug Admin~
istration employee serving with him as a panel, may question any
participant during or at the conclusion of his presentation. No other
persons attending the hearing may question a participant. The
presiding officer may allot additional time to any participant if he
concludes that it is in the public interest, but may not reduce the
time allotted to anyone.
The record of the hearing will remain open until September 5, l~i75,
for the submission of any additional written statements or comments
regarding oral presentations made at the hearing.
No written submission, or any portion thereof, made in response to
this notice shall be received or held in confidence. The administrative
record of this rule making proceeding shall consist of all relevant
FEDERAL REGISTER notices and the documents to which they refer, all
56-592 0 - 75 - pt. 28 - 17
PAGENO="0250"
13500 cO~n'ETITIvE PROBLEMS IN ThIE DRUG INDUSTRY
written submissions made in response to this notice, and the transcript
of the oral hearing made by the Food and Drug Administration. The
administrative record of the proceeding shall be made available for
public examination.
VIII. PROPOSED REGULATION FOR THE LABELING
* OF ORAL HYPOGLYCEMIC DRUGS
Therefore, pursuant to provisions of the Federal Food, Drug, and
Cosmetic Act (secs. 502, 505, 701(a), 52 Stat. 1050-1053, as amended,
1055 (21 U.S.C. 352, 355, 371(a))) anc~. under authority delegated to him
(21 CFR 2.120), the Commissioner proposes that Part 310 of Subchapter D
of Title 21 of the Code of Federal Regulations be amended by adding a
new § 310.510 as follows:
§ 310.510 Labeling for oral hypo~lycemic drugs.
(a) An adequate and well-controlled clinical trial (the University
Group Diabetes Program study) has indicated that there appears to
be an increased risk of cardiovascular mortality associated with
the administration of tolbutamide and of phenformin (oral hypoglycemic
drugs of the sulfonylurea and biguanide categories, respectively)
to maturity-onset diabetic patients as compared to treatment with
diet alone or diet plus insulin. The Commissioner concludes that
in view of the great similarities in chemical structure an~1 mode of
action for drugs within each of these two categories, it is prudent
from a safety standpoint to consider that the possible increased
PAGENO="0251"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 135O1
risk of cardiovascular mortality for tolbutamide and phenformin also
applies to other sulfonylurea and biguanide drugs. Therefore, the
labeling for oral hypoglycemic drugs shall describe properly the
conditions for their use and include a warning concerning the possible
increased risk of cardiovascular mortality associated with such use,
as set forth in paragraphs (b) and Cc) of this section.
(b) Labeling for oral hypoglycemic drugs of the sulfonylurea
category shall be as follows:
DESCRIPTION
(Trade name, established name) is an oral
blood-glucose-lowering drug of the sulfonylurea
category. It is a white, crystalline compound,
formulated as a tablet for oral administration.
(Manufacturer to add structural formula and other
appropriate information.)
ACTIONS
Administration of (drug) appears to lower the
blood glucose initially by stimulating the release
of insulin from the pancreas; the effect is
thus dependent on functioning beta cells in
the pancreatic islets. The mechanism by which
(drug) lowers blood glucose during long term
administration has not been clearly established.
Many patients who at first demonstrate an adequate
glucose-lowering effect with a sulfonylurea
PAGENO="0252"
13502 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
agent subsequently prove to be no longer satis-
factorily responsive, i.e., secondary failure
may occur.
(Manufacturer to supply information about:
1. Absorption.
2. Metabolism and excretion.
3. Plasma half-life and the effect of
hepatic or renal Impairment on blood levels,
metabolism, and excretion.
4. Peak and duration of glucose-lowering
effect, indicating the duration of effect rela-
tive to the class of sulfonylurea agents, e.g.,
shortest acting, longest acting, etc.
5. Mechanism of drug interaction with
agents that Impair or potentiate drug effect.)
PAGENO="0253"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13503
INDICATIONS
(Drug) is indicated to control symptoms ue
to hyperglycemia in patients with maturity-onset
nonketotic diabetes mellitus whose symptoms
cannot be controlled by diet alone and in
whom insulin cannot be used because of patient
unwillingness, erratic adherence to the injection
regimen, poor vision, physical or mental handicap,
insulin allergy, employment requirements, or
other similar factors.
(Drug) may also be used to lower blood glucose
in asymptomatic patients whose blood glucose
elevation cannot be controlled by diet alone and
in whom insulin cannot be used for any of the
above reasons. In considering the use of
(drug) in asymptomatic patients, it should
be recognized that whether or not controllin~
the blood glucose is effective in preventing the
long term cardiovascular or neural complications
of diabetes is an unanswered scientific question.
The use of (drug) may be associated with an
increased risk of cardiovascular mortality as
compated to diet alone or diet plus insulin; see
WARNINGS. For this reason, it should be used only
PAGENO="0254"
13504 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
when the advantages in the individual patient
justify the potential risk. The patient should
be inforned of the advantages and potential
risks of (drug) and of alternative nodes of
therapy and should participate in the decision
ou
The foundation of therapy in the obese
maturity-onset diabetic is caloric restric~tion
and weight loss. Proper zlietary management
alone is often effective in controlling the
blood glucose and eliminating symptoms of
polydipsia and polyuria. Use of (drug) must be
considered by both the physician and patient as
a treatment in addition to diet and not as a
substitute for diet or as a convenient mechanism
f or avoiding dietary rastraint .
Many patients who are initially responsive
to oral hypoglycemic drugs become unresponsive
or poorly responsive over a period of time,
usually 1 to 5 years. (Drug) should be given
only to patients demonstrated to be responsive
to it; see DOSAGE AND ADMINISTRATION for dis-
PAGENO="0255"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 135O5
administration of (drug) may be sufficient
during periods of transient loss of control.
Concomitant Therapykith a Biguanide:
(Drug) may be used in conjunction with phen-
formin to control symptoms due to hyperglycemia in
patients with maturity-onset nonketotic dia-
betes mellitus whose symptoms cannot be controlled
by diet and maximum recommended Boses of either
drug alone and in whom insulin cannot be used
for any of the reasons cited above.
In considering the use of concomitant
therapy, it should be noted that both a
sulfonylurea drug (tolbutamide) and a biguanide
drug (phenformin) have been reported to be
associated with increased cardiovascular mortality;
see WARNINGS. In addition, phenformin can
PAGENO="0256"
13506 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
produce lethal lactic acidosis in some patients.
Thus the use of (drug) in association with
phenformin carries a greater ~
of (drug) alone.
If a judgment is made that (drug) and phenformin
are to be used together in a particular patient,
it should be established that the patient is
responsive to both drugs. This may be accomplished
either by a trial of each d~ug separately or
by adding the second drug and then tapering the
dosage of the first, observing for diminished
control of blood glucose. Once the need for both
drugs is established, the desired control of
blood sugar may be obtained by adjusting the
dose of either drug. The possibility of hypoglycemia
should be anticipated and appropriate precautions
taken. See package insert for phenformin hydrochloride
f or CONTRAINDICATIONS, WARNINGS, PRECAUTIONS,
ADVERSE REACTIONS, and DOSAGE AND ADMINISTRATION.
C~TRAINDICATIONS
(Drug) is contraindicated in patients with:
1. Known hypersensitivity or allergy to the
drug.
2. Diabetic ketoacidosis, with or without
coma. Such patients should be treated with insulin.
PAGENO="0257"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13507
WARNINGS
SPECIAL WARNINGS ON CARDIOVASCULAR MORTALITY
(This subsection of labeling to be boxed set in
boldface type, andplaced at the beginning of~
~ectio~flab&~L~
The administration of oral hypoglycemic drug nay
be associated with increased cardiovascular mortality
as compared to treatment with diet alone or diet plus
insulin.
This warning is based on the study conducted by
the University Group Diabetes Program (TJGDP), a long
term prospective clinical trial designed to evaluate
the effectiveness of glucose-lowering drugs in prevent-
ing or delaying vascular complications in patients with
maturity-onset nonketotic diabetes. The study involved
1,027 patients who were randomly assigned to one of
five treatment groups (Diabetes, 19 (supp. 2):
747-830, 1970; Diabetes, 24 (supp. l):65-184,
1975).
PAGENO="0258"
13508 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The UGDP reported that patients treated for 5
to 8 years with diet plus a fixed dose of tolbut-
amide (1.5 grams per day) or diet plus a fixed dose of
phenformin (100 milligrams per day) had a rate of cardio-
vascular mortality approximately twice that of patients
treated with diet alone or diet plus insulin.
Total mortality was increased in both the tolbut-
amide- and phenformin-treated groups, but this
increase was statistically significant only for
phenformin. Despite controversy regarding the
interpretation of these results, the findings
of the,UGDP study provide adequate scientific
basis f or this warning.
Although only,one drug in the sulfonylurea
category (tolbutamide) and, one in the biguanide
category (phenformin) were included in this
study, it is prudent from a safety standpoint
to consider that this result may also apply to other
oral hypoglycemic drugs in these categories, in
view of the close similarities in mode of action
and chemical structure among the drugs in each
category.
PAGENO="0259"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13509
(Drug) should be used in preference to insulin
only in patients with maturity-onset diabetes
whose symptoms or blood glucose level cannot be con-
trolled by diet alone and only when the advantages
in the individual patient justify the potential
risk; see INDICATIONS. The patient should be
informed of the advantages and potential risks
of (drug) and of alternative modes of therapy
and should participate in the decision to use
this drug.
(Drug) is not effective in patients with juvenile
diabetes or insulin-dependent diabetes at any age. Such
patients should be treated with insulin. The concomitant
long term use of insulin and (drug) in an individual
patient is, in view of the potential risk of increased cardio-
vascular mortality siith (drug), less safe on a benefit-
risk basis than the use of insulin alone.
PAGENO="0260"
13510 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The effectiveness of any oral hypoglycemic drug,
including (drug), in lowering blood glucose to a desired
level decreases in a large number of patients as the
drug is administered over a period of months or years,
in part because the patient's blood glucose tends to
rise over time and in part because of diminished,
responsiveness to the drug. This phenomenon is known
as secondary failure to distinguish it from primary
failure in which the drug ~s ineffective in an
individual patient at the time of its initial adminis-
tration. See DOSAGE AND ADMINISTRATION.
Renal or hepatic insufficiency may cause
elevated blood levels of (drug) and increase the risk
of serious hypoglycemic reactions.
Pre,~~,n~: (Data and interpretation related to
reproduction and teratology studies to be supplied by
manufacturer).
Prolonged severe hypoglycemia (4 to 10 days) has
been reported in neonates born to mothers who were
receiving a sulfonylurea drug at the time of delivery.
Neonatal hypoglycemia has been reported more frequently
following use of the longer-acting agents. If (drug)
PAGENO="0261"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13511
is used during pregnancy, it should be discontinued
(tine period to be supplied by manufacturer) before
the expected delivery date.
PRECAUTIONS
~~ypog1ycernia: All sulfonylurea drugs are
capable of producing severe hypoglycemia.' Particularly
susceptible are elderly patients, patients with impaired
hepatic or renal function, patients who are debilitated or
malnourished, and patients with adrenal or pituitary
insufficiency. Hypoglycemia is more likely to occur
when caloric intake is deficient, after severe or
prolonged exercise, or when more than one glucose-
lowering drug is used.
(To be inserted for chlorpropamide only:) Because
of the long half-life of chlorpropamide, patients who
become hypoglycemic during therapy recjuire careful
supervision of the dose for at least 3 to 5 days,
during which time frequent feedings are essential. It
nay be necessary to hospitalize such patients and give
intravenous glucose.
Certain drugs may potentiate the hypoglycemiè
action of (drug), including plienylbutazone, oxyphen-
butazone, salicylates, sulfonamides, chloramphenicol,
probenecid, coumarin~, monoarni n~ oxidase inhibitors,
and beta-adrenergic blocking agents. See ACTIONS.
PAGENO="0262"
13512 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
When such drugs are administered to a patient receiving
(drug), the patient should be observed closely for
hypoglycemia.
Loss of Control of Blood Sujar When a
patient stabili7ed on any diabetic regimen is
exposed to stress such as fever trauma, infection
or surgery, a loss of control may occur. At
such times it may be necessary to d±scontinue
(drug) and administer insulin.
Certain drugs tend to produce hyperglycemia
and may lead to loss of control. These drugs
include the thiazides and other oral diuretics
corticosteriods and (to be supplied by manufacturer)
When such drugs are administered to a patient
receiving (drug), the patient should be carefully
observed for Loss of control
Pseudo-albumimuria(tolbutamideoni~y)
Urine containing a tolbi~tamide metabolite may
give a false positive reaction for albumin if
the acidification-after-boiling test is used
because this procedure causes the metabolite to
precipitate as flocculent particles. Use of the
sulfosalicylic acid test circumvents this problem.
PAGENO="0263"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13513
ADVERSE REACTIONS
Hypoglycemia: See PRECAUTIONS.
Gastrointestinal Reactions: Cholestatic
jaundice may occur rarely; (drug) should be
discontinued if this occurs.
Gastrointestinal disturbances, e.g.,
nausea, epigastric fullness, and heartburn are
the most common reactions, occurring in (manu-
facturer to supply estimate of incidence). They
tend to be dose related and may disappear when
dosage is reduced.
Dermatologic Reactions: Allergic skin reac-
tions, e.g., pruritus, erythema, urticaria, and
morbilliform or maculopapular eruptions occur
(manufacturer to provide estimate of incidence).
These may be transient and may disappear despite
continued use of (drug); if skin reactions
persist, the drug should be discontinued.
Porphyria cutanea tarda and photosensi-
tivity reactions have been reported.
Hematologic Reactions: Leukopenia,
agranulocytosis, thrombocytopenia, hemolytic
anemia, aplastic anemia, and pancytopenia have
been reported.
PAGENO="0264"
13514 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Metabolic Reactions~ Hepatic porphyria,
disulf iran-like reactions (manufacturer to
supply further details).
(To be inserted for chlorpropamide only:)
Endocrine Reactions: On rare occasions
(drug) has caused a reaction identical to the
syndrome of inappropriate antidiuretic hormone
(ADH) secretion. The features of this syndrome
result from excessive water retention and include
hyponatremia, low serum osinolality, and high
urine osmolality.
D9SACE AND &DMINI~~RATIpN
There is no fixed dosage regimen for the
management of diabetes nellitus with (drug) or
any other agent. In addition to the usual monitor-
ing of urinary glucose, the patient's blood glucose
must also be monitored periodically:
a, To determine the minimum drug dosage
that will lower the blood glucose adequately.
b. To detect primary failure, i.e.,
inadequate lowering of blood glucose when the
drug is first used, even though dose has been
raised to the maximum level recommended; and
PAGENO="0265"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13515
c. To detect secondary failure, i.e.,
loss of adequate blood-glucose-lowering response
after an initial period of effectiveness. (Drug)
should be discontinued, with careful monitoring
of blood glucose at least annually to be certain
that (drug) is continuing to lower the blood
glucose.
Short term administration of (drug) may be
sufficient during periods of transient loss of
control.
(Manufacturer to supply the following
details of dosage for each sulfonylurea:
1. Usual starting dose.
2. Maximum dose.
3. Dose beyond which a response is usually
not seen if patient has not already had some
response.
4. Usual maintenance dose.
5. Dosage interval, with reasons, e.g.,
avoid GI tolerance, short half-life of drug, etc.
6. Caution regarding dosage in elderly.)
HOW SUPPLIED
(To be supplied by manufacturer.)
56-592 0 - 75 - pt. 28 - 18
PAGENO="0266"
13516 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
(c) Labeling for oral hypoglycemic drugs of the biguanide
category shall be as follows:
DESCRIPTION
(Trade name, established name) is an oral
blood-glucose-lowering drug of the biguanide
category. It is a white, crystalline, water-
soluble compound, formulated as (to be supplied
by firm) for oral administration. (Manufacturer
to add structural formula and other appropriate
information).
ACTIONS
The mechanism of action of phenfornin is
not established but its ability to cause increased
peripheral glucose uptake in vitro appears to be
related to its inhibition of cellular oxidative
processes. It does not stimulate insulin production.
Many patients who at first demonstrate an adequate
glucose-lowering effect with (drug) subsequently
prove to be no longer satisfactorily responsive,
I.e., secondary failure may occur.
PAGENO="0267"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13517
(Manufacturer to supply information about:
1. Absorption.
2. Metabolism and excretion.
3. Plasma half-life and the effect of hepatic
or renal impairment on blood levels, metabolism,
and excretion.
4. Peak and duration of glucose-lowering
effect.
5. Mechanism of drug interaction with agents
that impair or potentiate drug effect.)
INDICATIONS
Identical to sulfonylurea label, except for
substitution of the following section relating
to concomitant therapy:
Concomitant Therapy:
Phenformin may be used in conjunction with
a sulfonylurea to control symptoms due to
hyperglycemia in patients with maturity-onset
nonketotic diabetes mellitus whose symptoms
cannot be controlled by diet and maximum
recommended doses of either drug alOne and
in whom insulin cannot be used for any of the
reasons cited above.
PAGENO="0268"
13518 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
In considering the use of concomitant
therapy, it should be noted that both phen-
formin and a sulfonylurea drug (tolbutamide)
have been reported to be associated with increased
cardiovascular mortality; see WARNLNCS. Thus
the use of phenformin in association with a
sulfonylurea may carry a greater risk than the
use of phenformin alone.
If a judgment is made that phenformin and
a sulfonylurea are to be used together in a
particular patient, it should be established
that the patient is responsive to both drugs.
This may be accomplished either by a trial
of each drug separately or by adding the second
drug and then tapering the dosage of the first,
observing for diminished control of blood
glucose. Once the need for both drugs is
established, the desired control of blood sugar
may be obtained by adjusting the dose of either
drug. The possibility of hypoglycemia should
be anticipated, and appropriate precautions
taken. See package insert for the appropriate
sulfonylurea for CONTRAINDICATIONS, WARNINGS,
PRECAUTIONS, ADVERSE REACTIONS, and DOSAGE AND
ADMINISTRATION.
PAGENO="0269"
COMPETITIVE PROBLEMS IN THE DRIJG INDIJSTRY 13519
CONTRAINDICATIONS
(Drug) is contraindicated in patients with:
1. Known hypersensitivity or allergy to the
drug.
2. A history of lactic acidosis.
3. Disease states associated with hypoxetnia
including cardiovascular collapse and acute inyocardial
infarction.
4. Severe renal disease.
5. Alcoholism,
6. Diabetic ketoacidosis with or without coma.
Such patients should be treated with insulin.
PAGENO="0270"
13520 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
WARNINGS
SPECIAL WARNING ON CARDIOVASCULAR MORTALITY
(Identical to boxed, boldface sulfonylurea labeling.)
(Drug) is not adequate therapy in patients
with juvenile diabetes or insulin-dependent dia-
betes at any age. Such patients should be treated
with diet and insulin. The concomitant long term
use of insulin and (drug) in an individual patient
is, in view of the risk of increased cardiovascular
mortality with (drug), less safe on a benefit-risk
basis than the use of insulin alone.
The effectiveness of any oral hypoglycemic
drug, including (drug), in lowering blood glucose
to a desired level decreases in a large number of
patients as the drug is administered over a period
of months or years, in part because the patient's
PAGENO="0271"
COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY 13521
blood glucose tends to rise over time and in part
because of diminished responsiveness to the drug.
This phenomenon is known as secondary failure, to
distinguish it from primary failure in which the
drug is ineffective in an individual patient at the
time of its initial administration. See DOSAGE AND
ADMINISTRATION.
Lactic Acidosis: There have been numerous
reports of lactic acidosis in patients receiving
phenformin. Lactic acidosis is an often fatal
metabolic acidosis characterized by elevated
blood lactate le~els, an increased lactate-to-
pyruvate ratio, and decreased blood pH. Azotemia
ranging from mild to severe is present in most
of the reported cases of lactic acidosis. Azotemia
can result from dehydration, and some patients
developing lactic acidosis associated with azotenia
have had normal serum creatinine levels when
properly hydrated. The following specific pre-
cautions should be observed when administering
phenformin:
PAGENO="0272"
13522 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
a. Impairment of renal function increases
the risk of lactic acidosis. Renal function
tests, such as serum creatinine, should be
performed prior to phenformin therapy and at
least annually thereafter. Phenformin should
not be used in patients with impaired renal
function, e.g., serum creatinine over 1.5 milli-
grams/lOO milliliters, except in extraordinary
circumstances.
b. Cardiovascular collapse (shock), con-
gestive heart failure, acute myocardial infarction
and other conditions characterized by hypoxemia
have been associated with lactic acidosis and
also may cause prerenal azotemia. Use of phenformin
in patients particularly prone to develop such
conditions must be carefully considered and the
risks weighed against possible benefits. When
such events occur in patients on phenformin
therapy, the drug should be discontinued promptly.
c. Gastrointestinal disturbances are the
most common adverse reactions to phenformin therapy.
These symptoms must be distinguished from the
symptoms of developing lactic acidosis. Anorexia
PAGENO="0273"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13523
and mild nausea are common side effects of phen-
formin, particularly upon initiation of therapy.
Nausea, vomiting, malaise, or abdominal pain
may herald the onset of lactic acidosis. The
patient should be instructed to notify the
physician immediately at the onset of any of
these gastrointestinal symptoms or of hyperven-
tilation. Phenformim should be withdrawn until
the situation is clarified by determination of
serum electrolytes and ketones, blood glucose,
and, if indicated, blood pH, lactate, and pyruvate
levels.
d. Lactic acidosis has a significant
mortality and, when suspected, must be treated
promptly by discontinuing phenformim and giving
bicarbonate infusions and other appropriate
therapy even before the results of lactate
determinations are available. Lactic acidosis
should be suspected in any diabetic patient
with metabolic acidosis in the absence of ketonuria
and ketonemia, uremia, and methanol or salicylate
poisoning.
PAGENO="0274"
13524 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
e, The physician should use special caution
after initiating phenformin therapy, after in~
creasing the drug dosage, and in circumstances
that nay cause dehydration leading to impaired
renal function.
f. Alcohol is known to potentiate the
effect of phenformin in elevating blood lactate
levels, and patients should be warned against
excessive alcoholic intake while receiving
phenformin.
g. Impaired hepatic function has been
associated with some cases of lactic acidosis.
Particular caution must be observed when admin-
istering (drug) to patients with hepatic disease.
~re~nancy: (Data and interpretation related
to reproduction and teratology studies to be
supplied by manufacturer).
PRECAUTIONS
~p~~ycemia: Hypoglycemia is unusual
in patients receiving (drug) alone, but may
occur when caloric intake is deficient, when
strenuous exercise is not compensated by caloric
supplementation, or when more than one hypogly-
cemic drug is used.
PAGENO="0275"
COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY 13525
(Manufacturer to supply paragraph on potentiating
drugs.)
Loss of Control of Blood Sugar: Identical
to sulfonylurea labeling.
~nClinica1 Status of Previous~
Controlled D~~~jc: A diabetic patient previously
well-controlled on phenformin who develops
laboratory abnormalities or clinical illness
(especially vague and poorly defined illness)
should be evaluated promptly for evidence of
ketoacidosis or lactic acidosis. Evaluation
should include serum electrolytes and ketones,
blood glucose, and, if indicated, blood pH,
lactate, and pyruvate levels. Acidosis of either
form necessitates withdrawing phenformin and
initiating other appropriate corrective measures.
Starvatio~jet~Q~~s This must be
differentiated from insulin-deficient ketosis
and is characterized by ketonuria with, little
or no glucosuria and relatively normal blood
glucose levels. This may result from excessive
dosage of phenfornin or insufficient carbohydrate
intake.
PAGENO="0276"
13526 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
ADVERSE REACTIONS
~yp~og1ycernia: See PRECAUTIONS.
Gastrointestinal Reactions: Gastrointestinal
distur1~ances such as anorexia, nausea, vomiting,
and diarrhea are the most common adverse reactions
(manufacturer to supply frequency) and are dose
related. These symptoms must be distinguished
from the prodromata of lactic acidosis. See
WARNINGS section for discussion of lactic acidosis.
They may also cause dehydration and prerenal
azotemia, which require discontinuation of the
drug until renal function is again normal.
Phenformin should be discontinued if vomiting
occurs, An unpleasant metallic taste is a
warning signal of impending gastrointestinal
disturbances.
Dermato1o~ic Reactions: (Manufacturer to
supply data, including estimate of incidence.)
Miscellaneous Reactions: Fatigue and
weakness. Anorexia, nausea, and vomiting may
occur in association with the intake of alcohol.
PAGENO="0277"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13527
DOSAGE AND ADMINISTRATION
There is no fixed dosage regimen for the
managementof diabetic mellitus with (drug)
or any other agent. In addition to the usual
monitoring of urinary glucose, the patient's
blood glucose must also be monitored periodically:
a. To determine the minimum drug dosage
that will lower the blood glucose adequately.
b. To detect primary failure, i.e.,
inadequate lowering of the blood glucose when
the drug is first used, even though dose has
been raised to the maximum level recommended.
c. To detect secondary failure, i.e.,
loss of adequate blood-glucose-lowering response
after an initial period of effectiveness.
Drug should be discontinued with careful monitoring
of blood glucose at least annually to be certain
that (drug) is continuing to lower the blood
glucose.
Short term administration of (drug) nay be
sufficient during periods of transient loss of
control.
PAGENO="0278"
13528 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
(Manufacturer to 5U~~iY the following
details of dosage:
1. Usual starting dose.
2. Maximum dose.
3. Dose beyond which a response is usually
not seen if patient has not already had some response.
4. Usual maintenance dose.
5. Dosage interval, with reasons, e.g.,
avoid CI intolerance, short half-life of drug, etc.
6. Caution regarding docage in elderly.)
HOW SUPPLIED
(To be supplied by manufacturer.)
(d) Each holder of an approved new drug application for an oral
hypoglycemic agent shall submit a supplement to his application under
the provisions of § 314.8(d) of this chapter to provide for labeling as
described in paragraphs (b) and (c) of this section. The labeling in
such supplement shall be identical in wording to the labeling in paragraphs
(b) or (c) of this section where precise wording is specified, shall
provide information on each of the points where wording is delegated to
the manufacturer, and shall contain no additional or extraneous information.
Such supplement shall be submitted within 10 days after (effective date
of the final regulation). Any oral hypoglycemic drug with labeling not
in compliance with this section and shipped into interstate commerce
after (60 days after effective date of the final regulation) shall be
subject to reguThtory action.
PAGENO="0279"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY 13529
Interested persons may, on or before (insert date 60 days after
d~teof publication in the FEDERAL REGISTER), submit to the Hearing
Clerk, Food and Drug Administration, Rn. 4-65, 5600 Fishers Lane,
Rockville, MD 20852, written comments regarding this proposal. Comments
shall be filed in quintuplicate and shall be identified with the Hearii~g
Clerk docket number found in the document heading. Received comments
may be seen in the above office during working hours, Monday through
Friday.
\ll /
Dated:____________________
PAGENO="0280"
13530 COMPETITIVE PROBLEMS IN THE DRuG INDUSTRY
jJubc
Cmzen
June 10,1975
Alexander M. Schmidt, M.D.
Commissioner,
Food and Drug Administration
Federal Building 8
200 "C" St. S.W.
Washington, D.C. 20204
Dear Dr. Schmidt:
This letter is to urge immediate publication of a warning to
patients and doctors about all oral drugs used to treat diabetes.
New evidence from four previously unreported studies combined with
previous evidence demands this action to prevent the unnecessary
death of thousands of diabetics and waste of more than 100 million -
dollars a year on these drugs.
It is now five years since the University Group Diabetes
Program (UGDP) reported that there was a significant excess cardio-
vascular mortality in patients taking an oral diabetes drug
(tolbutamide-ORINASE). Although ~romu1gation of an FDA-proposed
warning label for ~ such drugs was enjoined in 1972 by a Federal
District Court, this was reversed in July, 1973 by the U.S. Court
of Appeals. Thus, the FDA has delayed for almost two years the
1. "Although the specific sulfonylurea drug studied by UGDP was
tolbutamide, the conclusions apply equally to all sulfonylureas--
Diabinase, Orinase, and Tolinase--because of their close chemical
relationship." "...the conclusions apply to DBI and Meltrol as
well" (FDA Drug Bulletin, May 1972).
LEADING ORAL "ANTIDIABETES" DRUGS
TOTAL PRESCRIPTIONS PER YEAR FOR U.S.1
197Q ___
Diabinase (Pfizer) 5.845,000 6,201,000 +7.8%
Orinase (Upjohn) 5,290,000 4,998,000 -5.5%
DRI (Geigy) 4,035,000 4,282,000 +6.1%
Dyrnelor (Lilly) 1,553,000 1,462,000 -5.8%
Tolinase (Upjohn~~ ________ 1,975,000 j~4.6%
Total Prescriptions 18,369,000 19,381,000 +5.5%
(including all oral drugs)
(National Disease & Therapeutic Index, l972& 1973, LEA Inc., Ambler,PA.)
PAGENO="0281"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13531
publication of a warning label which would, `if properly worded,
cause most adult diabetics to be taken off these dangerous,
ineffective, and expensive drugs.
During this interval of irresponsible delay by the FDA1 approci-
mately 250 million dollars worth of these drugs have been consumed
in this country alone and, according to experts, 20,000-30,000
unnecessary deaths due to these drugs have probably oc~curred.2 Of
the 1.5 million people currently using these drugs, it is estimated
that 80% to 90%, at least, could be controlled as well or better-~
and certainly mare safely--by diet or, in a very small number of
cases, insulin.
Although the published medical literature contains ample
evidence of the increased risk of death and lack of efficacy of
these drugs, the FDA is aware of four additional studies the results
of which have not yet been made known to the public. A brief review
of these studies follows:
1. p~pub1ished Study on Mortai~t~L~ii) Josl in ~
Although many doctors who have been wedded tO the idea that
these pills benefit diabetics have attacked the findings of the
UGDP study, none have been as vocal as the doctors of the Joslin
Clinic in Boston. Aside from organizing the law suit against the
FDA to block promulgation of the warning label three years ago,
Dr. Robert Bradley (Joslin Director) and his cohort have devoted
considerable time to presentations at drug_company-financed meetings
around the country aimed at maintaining doctors "faith" in these
drugs. It is thus all the more striking to find that a study of
the Joslin Clinic's own patients yields results similar to those
found by the UGDP.
A Harvard Ph.D. thesis by Paula Kanarek entitled "Assessing
Survival in a Diabetic Population" and dated January 1973 was a
retrospective study of 6300 patients at the Joslin Clinic. Begun
after the UGDP study, its stated purpose was to study the effect
of various types of therapy on the survival of diabetic patients,
particularly to see "if individuals treated with tolbutamide
(ORINASE) are more likely to die from cardiovascular causes. * ."
The results of the study showed that:
a)~'For individuals who survive at least 5 years, however, the
relative survival experience for persons receiving tolbutarnide
is worse in all cases than that of those on diet. * ." (p. 111-16)
.certainly in all risk categories, the probability of
dying from cardiovascular causes in 5-10 years was consistently
higher in the tolbutamide c~roup than in the diet group (iII-l7)
2. Senate Hearings on Oral Hypoglycemic Drugs, Sept. 1974, p.10803.
3. ~ ~j, 854, May 26, 1975.
56-592 0 - 75 - pt. 28 - 19
PAGENO="0282"
13532 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
c)". ..the results of this observational study suggest that the
findings of the UGDP may be generalized to other diabetic
populations." (111-23)
2. FDA-Sponsored Study of the E ffect o~Lo uami~e on Developme~pJ~
~f çoron~ry ~rtQry Disease In Monkeys
- This study, FDA contract *72-l14,was done by researchers headed
by Dr. Robert Wissler, Professor of Pathology at the University of
Chicago. Begun in 1972, the final report was submitted to FDA in
February, 1975.
Monkeys were placed on an "Average American Diet" and one-half
were given tolbutamide at a dosage comparable to that taken by
diabetics. The study was continued for two years with the following
findings:
a) Coronary artery disease was found ~W~imes more often in
animals taking tolbutamide than in animals on diet alone.
b) The coronary artery disease caused by the drug was ~ii~
~~yer~ than that developing Spontaneously from the
diet alone.
Although the UGDP and Joslin studies showing increased death
in humans from cardiovascular `disease due to these drugs are clear
enough, the Wissler study is the first experimental confirmation of
the mechanism whereby the human deaths ahve occured. Since it has
been known for more than 30 years that most diabetics die from
cardiovascular complications, such animal studies should have
~ marketing.
3. A Study by Dr. C.R. Wu,, et al.,of New Jersey Medical School,
presented May 3, 1975 in Atlantic City at the American Federation
for Clinical Research meeting also looked at the effect of tolbut-
amide on the heart of diabetic dogs. At therapeutic doses (equivalent
to those taken by diabetic patients) the function and structure of
the heart was worsened by 1 year of treatment with tolbutamide.
4. IThpublished Study of Joslin Clinic Patients Showing Lack of
~~icacy of 4 Oral Diabetes Druq~.
In addition, we have just obtained a copy of a study by
researchers at the Joslin Clinic which clearly demonstrates that
the oral antidiabetic drugs fail to achieve their intended beneficial"
purpose, namely lowering of blood sugar.
365 adult-onset diabetics were placed on an individualized diet
and given either a placebo, tolbutamide (ORINASE), chlorpropamide
4. The Effects of Long Term Therapy with Oral Hypoglycemic Agents
on the Oral Glucose Tolerance Test Dynamics in Chemical Diabetics
(Tan, Graham, Bradley et.al.)
PAGENO="0283"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13533
(DIABINASE), phenformin (DBI) or acetohexarnide (DYMELOR). At the end
of four ynars on diet and either placebo or one of the four drugs,
~ no evic~ence that any of the four drug ~
tol~ar~c~~an diet and a plaç~.
This study, although completed more than 2 1/2 years ago and
presented in summary form at a meeting of the American Diabetes
Association June 23, 1973, has never been published. It is likely
that one or both of the following factors are responsible for its
non-publication:
1. The Joslin Clinic is the center of advocacy (aside from the
drug companies) for the use of these diabetes drugs.
2. The study was supported by Upjohn (makers of Orinase arid
Tolinase), Lilly (Dymelor) and Pfizer (Diabinase).
TheY degree to which the Joslin Clinic adheres to their "faith"
in these drugs despite all the evidence--including data from their
own patients--to the contrary can be seen in the testimony of Joslir~
Director Dr. Robert Bradley before the Senate hearings last fall:
"Data from the UGDP Study have thus far contributed nothing
to the controversy regarding the effectiveness of blood sugar
control and.. .the results cannot at present be extrapolated
to the diabetic population at large." Select Committee on
Small Business (p. 10986).
If the UGIDP study didn't convince Dr. Bradley, surely the combination
of increased mortality and lack of efficacy of these drugs in his ow~n
Clinic should.
WARNING LABEL AND MALPRACTICE
At the heart of opposition by both drug industry and doctor
opposition to the proposed warning label is the question of medical
malpractice. A strong label, warning against using the drugs unless
am adequate trial on diet therapy has been attempted and making
doctors (and patients) aware of the increased risk of cardiovascular
death,would likely become grounds for malpractice if these caveats
were not heeded. ~
on April 15, 197& we commented on the way FDA had already
weakened the warning label proposed for comment onJan.28, 1974
from the earliest draft published in the FDA Drug Bulletin, May 1972.
(See enclosed~.
In testimony before Senator Nelson last fall, Dr. Thomas Chalmers,
Dean of Mount Sinai Medical School discussed labelling:
"The new FDA labelling should be strong enough to warn all
physicians that they are distinctly putting their patient at risk
by using oral agents when diet or insulin will suffice. The wording
PAGENO="0284"
13534 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
should be such that doctors will be compelled to protect
themselves from a later malpractice suit by obtaining truly
informed consent from their patients, and by demonstrating
in the patient's record that the patient's symptoms could
not be controlled by diet and/or insulin." (Nelson Hearings,
- p. 10999).
Faced with such a warning label, most doctors, in the interests
of their patients (and themselves) would make a much greater effort
to avoid use of these dangerous drugs.
In summary, it is long past the time for FDA to have issued a
strong warning label for these drugs. Any further delay should
result in sanctions against all FDA officials who have been
responsible.
Sincerely,
Sidney M. Wolfe, M.D.
Anita Joh `on, esg.
PAGENO="0285"
COMPETITIVE PROBLEMS IN THE DRIJG INDIJSTRY 13535
1When Friends or Patients Ask About.
William H. Crosby, MD, Coordinator
John K. Davidson, MD, PhD
THE NELSON Committee hearings
in the US Senate (Sept 18 to 20, 1974)
explored the safety, effectiveness, and
use of hypoglycemic drugs. A dozen
witnesses, including the Food and
Drug Administration (FDA) Commis-
sioner ,~,L.Aj5xynder Schmidt, gave
testimony concerning the implica-
tions of the University Group Diabe-
tes Program (UGDP) articles for
medical practice. Concern was ex-
pressed because the FDA Bureau of
Drugs had not required appropriately
written package inserts warning phy-
sicians and the drug-consuming pub-
lic of the dangers inherent in the in-
discriminate use of sulfonylureas and
phenformin hydrochloride. There was
agreement that caloric restriction
with weight reduction is the pre-
len-ed method of treatment for the
maturity-onset diabetic who is above
deal body weight, and that such ther-
y is always safe and with intensive
~ducation and follow-up is usually ef-
fective.
The UGDP investigators ended the
tolbutamide (sulfonylurea) study in
1969 and the pheseformin study in
lii7l, when careful statistical analysis
showed a death rate more than twice
that expected for each drug when
compared to the death rates in the
three other groups in the study (pla-
cebo, a standard dose of insulin that
resulted in a moderately elevated
mean fasting blood glucose level, and
a variable dose of insulin that re-
suited in a near-normal mean fasting
blood glucose level). Death rates in the
placebo and insulin groups were sim-
ilar. There were no significant differ-
ences in the rate of development or
progression of chronic complications
(retinopathy, nephropathy, neuropa-
thy, arteriopathy) in the survivors of
any of the five treatment groups.
Are Oral Hypoglycemic Drugs
Safe and Effective?
In addition to the increased cardio-
vascular mortality with tolbutamide
therapy or with phenformin therapy
of approximately 1% per year, there
are many other adverse effects of oral
hypoglycemic drugs. These include
prolonged sulfonylurea-induced hy-
poglycemia and phenformin-induced
lactic acidosis, both of which have ter-
minated fatally in a number of pa-
tients.
Extrapancreatic effects of the sul-
fonylureas include hypothyroidism,
skin rashes and photosensitivity,
corneal opacities, blood dyscrasias, in-
creased fibrinolytic activity, hype-
natremia, water retention, jaundice,
liver-enzyme inhibition, heart ef-
fects (including inotropism, increased
oxygen need, and microgranulomas),
elevated blood pressure, altered elec-
troencephalogram in epilepsy, and in-
creased stomach-acid secretioti.
The sulfonylurea drugs compete for
carrier-protein binding sites with
many other drugs, including sqlfona-
mides, salicylates, phenylbutazone,
monoamine oxidase inhibitors, thia-
zides, and barbiturates. The pharma-
cological effect of all of these drugs,
including the sulfonylureas, may be
increased when they are displaced
from their combining sites; thus1 com-
bination drug therapy may have tiuany
unanticipated toxic consequences.
The difficulties in maintaining stable
anticoagulation therapy in the pres-
The FDA
and Hypoglycemic Drugs
The Author: John K. Davidson received
his B5 degree In 1943 and MD degree in
1945 at Emory University. Atlanta. He
was a research fellow of the American
Diabetes Association from 1961 to i965
at the University of Toronto, where he
won the Starr medal in 1963 for his work
on nonsuppressible insolin-like activity
and studies of beta cell fonction. He ret
ceived the PhD degree in Physiology in
1965 at the University of Toronto, and
joined the laculty with appointments in
physiology and medicine. He retorned to
Emory University in 1968, where he is
now director of the Diabetes Unit at
Grady Memorial Hospital and professor
of medicine Ha has been on the Buard
of Directors of the American Diabetes
Association since 1970, and directed its
sixth Allied Health Poslgraduate Course
in Atlanta in 1974. His research interests
ate in insulin immunity, immunologic in-
sulin resistance and its treatment with
sulfated insulin, and investigations of
various techniques of implementing ef-
fective diet therapy and weight reduction
in obese maturity-onset diubetics.
It you isish to suggest a topic or write an an-
seer for this texture, write to William H. Crosby,
ltD Scripps Ciinic and Research Fouvdaliort,
La Jolla, CA 92037
Reprint requesrs to Department of Medicine,
Emory University Ochool of Medicine, 69 Butler
Sr SE, Atlanta, GA 30303 lDr. Davidson).
JAMA. May 26. 1975 . Vol 232, No 6
FDA and Hypoglycemic Drugs-Davidson 853
PAGENO="0286"
13536 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
ence of sulfonylureas and the poten-
tiation of alcohol by sulfonylureas
with an occasional disulfiram-like re-
action are well substantiated.
Phenformin increases anaerobic
glycolysis and lactate production and
may produce lactic acidosis; it in-
creases blood pressure, heart rate,
and need for digitalis; and it de-
creases gluconeogenesis and glucose
absorption.
There has been a striking increase
in death rate and a decrease in life
expectancy in maturity-onset diabet-
cs in America, Europe, Asia, Africa,
and Australia during the last 20
years. These changes have paralleled
the inceas~~lyi~Fsnege1f
~e~hera ~
bridled enthusiasm amon man -
sician patients for the use of
~
treatments of choi~.
~ safe, it
should not shorten life expectancy
nor be associated with fatal side ef-
fects or with severe drug reactions.
The . increased number of cardio-
vascular deaths in the UGDP study,
the not uncommon occurrences of ir-
feversible hj~micjaind9~-
a~~and death with sulfonylureas and
fatal lactic acidosis with phenformin,
and the previously noted decreased
life expectancy of the maturity-onset
diabetic population throughout the
world are the dramatic cbnsequences
of unrestricted use of oral hypoglyce-
mic drugs. Thus, one is forced to the
conclusion that both sulfonylureas
and phenformin are unsafe.
It has been known for a long time
that ri and seconda drug
treatment failures are common wit
ti~ifon~li:ea1 and phenformin;
these failures have usually necessi-
tated insulin therapy. As early as
1957, it was noted that blood glucose
levels were frequently as low on pla-
cebo as on tolbutamide therapy. In a
1974 report, blood sugar levelsdid not
rise in 30 of 50 patients when single-
blind placebo was substituted for
chlorpropamide. In 20 patients, the
level rose modestly on a placebo reg-
men; however, in only four of the 20
patients did chiorpropamide help at-
tain fasting normoglycemia (blood
glucose level, <140 mg/100 ml).
If a drug is to be considered effec
tive, it should rave t or dela the ap- those above ideal body weight), (2)
pea ance of complications or pro ong the hyperglycemic maturity-onset di-
life, as well as lower the blood glucose abe~j~g~or below ideal bodywgjght,
l~~T'mn a ~ (3) the hyperglycemic pregnant din
sulfonylureas and phenformin appar- )~gtj.c., and (4) the hypogglycemiçjuve
~ The prime them-
prevent or delay complic~y)ons~~,~,,. peutic objective when insulin is used
d6~'~not 1owe~Eh~hlood lucose level is to attain normoglycemia as fre-
in an o tima as ion in the majority quently as possible, with hypogly-
o~patien s, neit er sulfo~y~,~2r cemia being as infrequent as possible.
~~~i~can be considered effec- The slight discomfort of insulin injec-
tiveJliiFipy despite the claim~"~? tions is a minor price to pay for
several rug companies and some the generally acknowledged physi-
iciansoTh7ontry~ ologic actions and life-sustaining
Safe and Effective Alternatives properties of the hormone. I~~L
therapy, optimally us~d, is safe, sac
Appropriate ~ (in- is almost universally eflectity in lo'~ -
cludingperiod~. of fasting ,for up to ering the blood glucose level.
one week) usually lowers the blood -
- - Responsibility for
glucose level dramatically in the
indlscnmlnate Use
obese diabetic. With ij~(pg~vei
s~~jan..follow-~p, a diet ad- At the present time, about 1,500,000
herence rate of 96% has been reported erans are beine treatecjm'it.j~
recently. Siieces&ratas in reducing oral hypoglycemic drug~s. At the Nel-
the weight of obese diabetic patients son Committee hearings, different
at Grady Memorial Hospital and witnesses estimated that from less
Emory University School of Medicine than 1% to 20% of the patients
in Atlanta have varied in different being ~`i~wiiTi these agents were
groups of patients from ~Q7e~to 90%.~ ~lhei~iJ by such therapy. Do-
with the most intensively instructed pending on which opinion one accepts
and monitored patients having the as valid, this means that from
greatest degree of success (unpub- 1,200,000 to 1,485,000 Americans are
lished data). To modify successfully a being inappropriately treated and
patient's long-established habits of should have their oral agent therapy
food intake, it is mandatory that the discontinued. The drugs are probably
physician be competent in calculating being used excessively because the
ideal body weight and in writing the physician wants to do something
correct dietary prescription. The pa- when he makes a diagnosis of diabe-
tient's adherence to diet is facilitated tea, and the patient wants something
by use of an easily understood diet done. The easiest thing to do is to
manual anTJhtënsive instruction write a prescription for a pill, because
over a long period of time by physi- 0it takes only a little time and it neces-
cian, nurse, and dietitian. Patientsat f~sitates no signifi~nt change in,
Grady Hospital are now given about I!~sient's life-style. Thus, the physi-
25 hours of individual and gro~pJ~ (~ cian writes a prescription for a medi-~
struction over a one-year period. Diet ~cation that he has been led to believe
therapy is universally saf~, and when is safe and effective, when it is almost
pursued aggressively it is effective in certainly unsafe and is frequently
a large majority of matiirity-oñiet di- ineffective.
abetics. The FDA has not required properly
In the diabetic above ideal body written package inserts for the sul-
weight who is not acutely decompen- fonylureas and phenformin because
sated (plasma glucose level, 500 of a legal barrier erected by litigation
mg/dl or higher, or moderate to large instituted by the Committee on the
acetonuria), only a one-week fast and Care of the Diabetic. For this reason,
intensive follow-up caloric restriction physicians and the dru -consuming
is needed. In addition to appropriate pj~j,ç have not been honestly an
diet and exercise ~ optimal j,~- fully informed of the dangers inher-
sulin therapy is neede~1 in (1) the ant in the continued widespread use
~i3Tdëiompensated diabetic (even of these drugs. ~
854 JAMA. May 26, 1975 * Vol 232. No 8
FDA and Hypoglycemic Drugs-Davidson
PAGENO="0287"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13537
has now been removed. Commissioner
Schmidt has indicated that when the
audit of the UGDP study by the Bio-
metrics Society is published [Feb 10
issue of J~4MA.-ED], the FDA will
move swiftly and forcefully to order
labeling of the sulfonylureas and
phenformin to reflect the results of
the UGDP study. He anticipates that
this action will discourage future in-
discriminate use of the drugs.
If the drugs are used in the future,
it would be prudent for the physician
to have the patient sign an informed
g~py,~ht ag~reement indicating that he
has been informed of the possible
hazards associated with such drugs
and that he accepts the full responsi-
bility for the effects of such therapy.
For physicians who are concerned
about the consequences of discon-
tinuing the use of oral hypoglycemic
agents by their patients, there is a
reassuring precedent for such action.
Four years ago, in the wake of the
UGDP report, oral agent therapy was
abandoned at Grady Hospital, involv-
ing 1~00 patients. Eighteen months
later, the blood glucose levels~,~Q_%
of these patients were satisfkctorily
controlled on di~Fthera alone.
Since 1972, t e emphasis on digt ther-
apy and weight reduction has been
intensified, and during the ladt year
insulin ~
in all patients who are above ideal
body weight.
Additional Readings
A study of the effects of hypoglycemic agents
on vascular complications in patients with adult-
onset diabetes, University Group Diabetes Pro-
gram. Diabetes l9lssppl 2):747-830, 1970.
Effects of hypoglycemic agents on vascular
complications in patients with adolt-onset diobe-
tow IV, A preliminary report on phenformin re
coIls, University Grnop Diabetes Program.
JAMA 217:777-784, 1971.
Proat TE: Adverse effects of the oral hypo-
glycemicdrugs. in Pro-eedsng.m,ffhc E'ighlh C'oo.
geese of the Jsternuli,,val Diabetes Federalioei.
Amsterdam, Excerpts Medics, 1974, pp 612-623.
Hurwitz D, McCuistion AC: Tolbutamide: A
double-blind study of its effect in diabetes. N
Engl J Med 257:931-933, 1957.
Len-Ran A: Trial of placebo in long-term
chlorpropaniide-treated diabetics. Diobetotogia
10:197-200, 1974.
Weinsier RL. Seeman A, Herrera MG, et al:
Diet therapy of diabetes: Description of a suc-
cessful methodologic approach to gaining diet
adherance. Diabetes 23.669-673, 1974.
Davidson 3K, Goldsmith MP: Diabetes (7oide-
hook: Diet Section. Colamhua, Ga, Litho-Krome
Co, 1971.
JAMA
75 YEARS AGO
May26, 1900
Carcinoma in Early Life
[Walter L. Bierring, MD, pp 1295-1298]
Carcinoma is usually regarded as the
malignant tumor of later life, having its le-
gitimate age limit at 35, though it is most
frequently met with after 40 years of age.
Wt'h increasing study of the malignant
neoplasms, age is gradually losing its hold
as a criterion of difference betseeen sar-
coma and carcinoma, the former being
found at almost any age, while carcinoma
:s beginning to he noticed more frequently
`bring tb: ":srlier de'-sdes of life.
i':iri'insma in early life often has a slow
"or-, mu 1km' th:mt si later sears may he
.:tt,.st or a consi,leral:ls: length of time.
ThaI `:ancvr in earle life is hei'oming more
iriolasn: :s very evident: whether it keeps
Pace siGh tb, general increase in carci-
noma will be difficult to determine. Cer-
tainly a collective study of the statistics re-
veals the unfortunate fact that the
mortality rate of carcinoma is on the in-
crease. Roowell Park calls attention to the
statistics of New York State, showing
2,363 deaths in 1887 against 4,456 in 1898.
In England and Wales, where careful
records are kept, the rate of occurrence of
cancer has increased from 1 in 5,046 Popu-
lation in 1840 to tin 1,306 in the year 1896,
an increaue of five times in fifty years.
This increase can hardly be ascribed as due
to improvements in methods of diagnosis,
for rather the reverse is the case, since
many cases that were formerly diagnosed
as cancer are now properly classified where
they belong in other lists.
While the etiology of cancer remains ob-
scure, there is but little hope that the fu-
ture a-ill offer any abatement is this pro-
gressive increase. As careful observation
continues to note the occurrence of carci-
noma in the earlier slecades of life, there is
removed from the list of predisposing
causes, the influence of age.
It has been an attrsm'tivm' explanation to
:setrliistm Is the lesseni':l physiologic resist-
ance in connective tissue the rOle of per-
mitting atyl)ical prisliferation of epithe-
hum beyond its normal limits. With
advancing age the submucous, subcutane-
ous and interstitial connective tissues Sn-
dergo atrophic changes, while the covering
or lining epithelial elements seem to retain
their usual vitality. When glands reach the
limit of their functional rOle in the orga-
nism, the danger from carcinoma is mOst
marked, while it diminishes again when
complete atrophy has taken place. When
cancer occurs thus in the developing dec-
ades of life, the above influences can
hardly be considered. . -.
In the entire field of pathology, no sub-
ject has proved more alluring to physician
and investigator than the etiology of carci-
noma. It was no wonder that the develop-
ment of bacteriologic methods gave a
mighty impetus to the search for an etio-
logic agent, the result of which is well
known. Numerous forms of cell inclu-
sions-cancer bodies-have been observed
and variously interpreted.
As yet it is evident that the real cause of
carcinoma has not been demonstrated.
Perhaps it will lie necessary to completely
modify our culture methods. Perhaps our
optical :sids are inaili'quate in magnifying
liss'er, hut unless tb, light conies soon, the
19th century svill close with the genesis sf
cancer as much a mystery as when the cen-
tory hogan.
JAMA, May 26, 1975 * Vol 232, No 8
FDA and Hypoglycemic Drugs-Davidson 855
PAGENO="0288"
13538 C.OMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
GOVERNMENT Rx's
FDA proposes
important revision
According to a recent FDA-proposed regula-
ti()n on Ial)elmg (package insert), a food, drug,
device, or cosmetic is "misbranded' if the
labeling does not reveal:
1. facts that appear in other statements
(et al) about the product;
2. consequences which may result from
using the product (either as prescribed
or as it is usually used).
However, the regulation's prohibitions are
more important than its requirements.
Material facts (10 not, according to the FDA,
permit or require a statement of differences
of opinion with respect to warnings (including
contraindications, precautions, adverse reac-
tions, and other information relating to possi-
ble product hazards) required in labeling for
food, drugs, devices or cosmetics under the
act." (Italics ours.) Furthermore, the proposed
regulation does not "permit or require a state-
ment of differences of opinion with respect to
the effectiveness of a drug unless each of the
Opinions expressed is supported i~y sul)stantial
evidence of effectiveness
In the regulation's preamble, the FDA
gives us significant insight into why they have
made the above proposal. The current regula-
hon states that ``the existence of a difference
of opinion aniong experts qualified by scien-
tific training and experience, as to the truth of
a representation made or suggested in the
labeling is a fact (among other facts) the
failure to reveal which may render the label-
ing misleading if there is a material weight of
opinion contrary to such representation. (In
other words, according to the regulation they
want to revise, differences of opinion are re
(Itlired.)
In 1971, the FDA published a notice that
the labeling for oral hypoglycemic drugs used
in the treatment of adult onset diabetes must
contain a warning against cardiovascular con-
sequences reportedly associated with the use
of these drugs. Later that year a group of
physicians petitioned the FDA to withdraw or
modify this requirement, because they
l)elieved that the warning was unjustifIed, aid
that if it were required, it should be acconi -
panied by a statement of the differences of
opinion among experts about the neci'ssit ~ IT
the warning. The FDA Commissioner denied
the petition, and manufacturers of oral hy-
poglycemic drugs had to include the warning
in their labeling. The FDA felt that an un-
diluted statement was justified, since there
was significant clinical evidence to su~)port the
warning, and that the physicians who pre-
sented the petition did not have enough cvi-
dence to dispute the need for the warning:
The physicians then brought suit to enjoin the
labeling change on the grounds that, i~y not in-
cluding the difference of opinion, the drug
would l)e misbranded according to the al)ove
regulation.
The,.EDA contended thatsince statements
~of drug effectiveness and tj:uthful labeling,
were supj,~osed to l)e ~)i~sed,,on `~substantial,~
eyjdeI)ce' rather. ~tb~piinedica~opiniqn,, the
labeling suggested by these physicians would
l)e misleading. Flowever, the district court en-
tered a preliminary injunction prohibiting the
Fl)A from requiring the warning because, the
court concluded, there was a r('asonable
likelihood that the labeling proposed by the
FDA did not com~)ly with current regulations.
\Vhen the F1)A appealed, the Unitel
~StatesCpur~ of' Appeal~icited th~ inconsisten~
cy l~veen the way~,the. curi:ei~t ~l)A~ggula;~
tion reads and the "substantial evidence~. re~_
(luireilleists that bad 1)een added to the Food
Drug Therapy / January 1975
PAGENO="0289"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13539
Prescribing Information
for
TUSSEND®~ and
TUSSEND®EXPECTOR~NT~j
CONTRAIN DICATIONS:
Hypersensitivity to any of the
formula ingredients.
INDICATIONS: Tussend and
Tussend Expectorant anti-
tussive-decongestants are
indicated when exhausting
cough spasm accompanies
upper respiratory tract con-
gestion. They are useful in the
symptomatic relief of upper
respiratory congestion due to
allergic conditions, the com-
mon cold, sinusitis and acute
bronchitis.
PRECAUTIONS: Tussend and
Tussend Expectorant should'
be used cautiously in individ-
uals with severe hypertension,
diabetes mellitus, hyper-
thyroidism or urinary reten-
tion. Continuous use over an
extended period is generally
contraindicated since hydro-
codone may cause addiction.
ADVERSE REACTIONS: As
with any narcotic analgesic,
hydrocodone may cause
gastrointestinal upset in
children and nausea in adults.
These reactions have been
reported, although rarely,
following the administration
of Tussend or Tussend
Expectorant.
DOSAGE AND ADMINISTRA-
TION: Adults, and children
over 90 lb., one tablet or one
teaspoonful; children 50 to
90 lb., 1/2 teaspoonful; children
25 to 50 lb., 1/4 teaspoonful.
May be given 3 or 4 times
a day as needed.
CAUTION: Federal law
prohibits dispensing without
prescription.
DOW PHARMACEUTICALS
The Dow chemical company
Indianapolis, nd. 46268
~i~d Drug Act by, Congress in .1962. The court
then voided the injunction and gave the case
back to the FDA for "further determination."
Now the FDA had to either allow inclusion
of conflicting opinions about the oral hy-
poglycemics drug warning or revise the
regulation. They decided to revise the regula-
tion.
According to the 1962 Food and Drug Act:
And:
"If an article is alleged to l)e ms~randed
l)ecause the labeling is misleading, then in de-
termining whether the labeling is mi~leading
there shall l)e taken into account (among
other things) not only representatior~s ma(le
or suggested by statement, word, design, de-
vice, or any combination thereof, but also the
extent to which the labeling fails to reveal
facts material in the light of such representa-
tions or material with respect to conse-
quences which may result from the use of the
article to which the labeling relates under the
conditions of use prescribed in the labeling
thereof or under such conditions of. use as are
customary or usual."
"Unless its labeling bears (1) adequate direc-
tions for use; and (2) such adequate warnings
against use in those pathological conditiOns or
l)y children where its use ~y be dangerous
to health, or against unsafe dosage or methods
or duration of administration or application, in
such manner and form, as are necessary for
the protection of users: Provided, That where
any requirement of clause (1) of this
paragraph, as applied to any drug or device, is
not necessary for the protection of the pUl)lic
health, the Secretary shall promulgate
regulations exempting such drug or device
from such requirement."
The law also (as everyone probably knøws
by now) states that a new drug may not be ap-
proved for marketing if it is unsafe or ineffec-
tive for the conditions for which it is supposed
to be used.
The FDA finds the use of the word "may"
(underlined in the quotation above) particu-
larly significant, since it implies lack of univer-
sal agreement among experts. The word
"may" thus is supposed to make it unnecessary
to include the specific medical controversy
surrounding the warning. The FDA comments
that warnings have been required on drug
(continued on p. 82)
Drug Therapy / January 1975
PAGENO="0290"
13540 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
lal)elS by the Food and Drug Administration
when there is significant medical evidence of a
poSSiI)Ie health hazard without waiting for a
causal relationship to be established by defini-
tive stu(lies, which in some instances may not
be feasible or would take many years. It states
that in this way, physicians are fully informed
of known potential dangers and the public is
better protected.
These warnings must be stated in clear
and~iini~~~[èrms without disc~or
~jii~iThici1Wons that the FDA says would under
ñiTh~or des~ [heir i anii~ij apd use mess,'
~ihat~ii1T~ii~h warnings are by their ver
nature, a )out ~OSSi) e anger on y. T ese
~
l~,utte~asne~mie~
labeling to reflect su~didebate, and, the sa
is e )a e an isa reemen s oul sr
in pro essiona journa s and S m osia 1)ut not
in rug a e ing. e prescription drug label
~ the guidance of a physi-
cian-it does not constitute a legal require-
ment and, the FDA adds, a physician does not
violate federal law when he prescribes a drug
contrary to a warning in its approved labeling.
In 1938, when the Federal Food and Drug
Cosmetic Act was passed, personal expert
judgment was the standard for determining
drug effectiveness. The only way to satisfy
that standard when the experts' opinions
differed was to state 1)0th sides of the issue in
the labeling for the drug. In the regulation
preamble, however, F1)A states that by 1960,
the scientific principles of modern drug test-
ing using statistically valid controlled studies
had become fully developed and accepted,
and claims of drug effectiveness (and thus the
truthfulness of' drug labeling) should be deter-
mined! by adequate and well-controlled in-
vestigations, including clinical investigations
by qualified experts.
The stated purpose of requiring scientific
studies was to eliminate individual clinical ob-
servation and opinion as the measure of a
drug's effectiveness, since the opinions of indi-
vidual physicians reflecting their personal ex-
perience, empirical observation, and tradi-
tional practice do not satisfy the requirements
of sound scientific investigation and thus do
not conform to the standards established by
the Drug Amendments of 1962. FDA notes
that the Drug Amendments of 1962 are
clearly intended to supersede tl~e earlier
reliance on medical opinion, and that
therefore regulations should reflect th~~ new
standard.
In commenting on the concept of "fair bal-
ance" argued by the physicians in attacking
the labeling for oral hypoglycemic drugs, FDA
states that a draft of proposed new regulations
will soon be published in the Federal Register.
These regulations will implement "fair bal-
ance" for prescription drug labeling in the
same way that present regulations do for
prescription drug advertising. In both adver-
tising and labeling, the required "balance"
refers to how information on drug safety and
effectiveness is presented and not to
differences of judgment al)out that informa-
tion.
The FDA concludes that the present
regulation is inconsistent with current legal
requirements and with contemporary medical
and scientific principles. Since Congress has
determined that t'he effectiveness of new
drugs must be established by substantial evi-
dence, the FDA thinks that a difference of
medical opinion about labeling claims of effec-
tiveness is not legally sufficient and is not a
material fact unless the opinion is supported
by evidence which meets the legal standard.
The FDA goes on to say that some controversy
exists in the medical profession about most
statements in prescription drug labeling, and
that permitting or requiring statements of
these opinions on all such matters would de-
str()y the present usefulness of prescription
drug Ial)eling.
The Commissioner thus concludes that
there is no justification for presenting
differences of' medical opinion in any warning
statements relating to possil)le product
hazards. The Fl)A feels that the law 1)resul)-
poses a difference of medical opinion, since
warnings of' drug dangers will be requiredl
even though the danger itself may not be es-
tablished l)ut merely potential. Furthermore,
they think there is O() reason to allow these
warnings to be discounted by including an
opinion that the warning is really not neces-
sary at all.
For these reasons the FDA proposes that
the standing regulations be revised!. (39
Federal Register 33229, Sept. 16, 1974) 1
Drug Therapy / January 1975
PAGENO="0291"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13541
Controversy in Internal Medicine II Ed. by Ingelfinger,
Ebert, Finland, and Relman. W.B. Saunders Co. Phila. 1974
Oral Antidiabetic Agents Have
a Limited Place in Management
and May Be Harmful*
ALBERT I. WINEGRAD, REX S. CIEMENTS, JR.,
and ANTHONY D. MORRISON
University of Pennsylvania School of Medicine
The manner in which most physicians now manage their patients with adult-
onset diabetes might best be termed "benign neglect." In part, this is a conse-
* quence of the continuing controversy over the aims and the efficacy of presently
available therapy. Just how benign current practice really is remains to be deter-
mined; however, it includes certain aspects that are difficult to justify, such as
the common misuse of the sulfonylurea compounds and the biguanide derivatives.
The indications for continued medical intervention are not so obvious in
adult-onset diabetics as in juvenile diabetics since, irrespective of their age at
diagnosis, patients with the adult-onset form of the disease exhibit little tendency
to ketoacidosis. There is general agreement that ~fforts to lower the blood glucose
are indicated in patients who are symptomatic as the result of hyperglycemia,
and that delay in the effective treatment of symptomatic hyperglycemia in adult-
onset diabetics may be a major factor in the development of hyperglycemic non-
ketotic coma. At the time of diagnosis, patients with adult-onset diabetes are a
heterogeneous group not only with regard to symptoms, but also with regard to
the presence of the complications of diabetes and the coexistence of other condi-
tions (e.g., hypertension and arteriosclerotic cardiovascular disease) that might
be expected to alter their prognosis. There is littlç~eement as to benefit of
that designed ~ symptoms directly related to
° This review was supported in part by Grant AM04722, National Institutes of Health,
U.S.P.H.S., and a grant from the John A. Hartford Foundation. Dr. \Vinegrs~l is the recipient
of Research Career Development Award GM06405 from the National Institute of General
Medical Sciences.
PAGENO="0292"
13542 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
What Is the Value of "Control"?
Physicians are divided on the value of attcmpts to achieve an approximation
of normal glucose metabolism in the hope that this will influence the develop
ment or progression of the late complications. (It is symptomatic of the state
the field that measures to modify other demonstrable abnormalities, such as
hyperlipoproteinernias, have received minimal consideration.) Much of the
present controversy over the value of "control" is nonsense. The players in this
philosophical sport are trapped by the definitions and unstated assumptions that
constitute the rules of this game. In the treatment of adult-onset diabetics, coo-
venience is the overriding consideration in the minds of most physicians. which
may be understandable since the value of treatment in asymptomatic patients is
disputed. The effort and expense required to achieve as~d to document an appro\i-
mation of normal blood glucose concentrations throughout the day preclude tLi~
type of treatment for most patients. None of the reported clinical trials desigm d
to evaluate the value of "control" has made a serious attempt to achieve normali-
zation of blood glucose fluctuations. The frequency with which this goal can
achieved in a large population of adult-onset diabetics by presently available
means has never been evaluated. In essence we have been subjected to a heat 1
dispute over the value of a form of therapy that is only rarely attempted and
even more rarely achieved.
Recent studies suggest that fluctuations in the blood glucose concentrati
in nondiabetic subjects on a normal diet are restricted to a relatively narr°\'
range. However, when evaluating blood glucose concentrations in adult-onset
diabetics, physicians almost invariably apply arbitrary standards that bear litti
relationship to the levels observed in nondiabetics. In addition, it is commoni~
assumed that if there were a relationship between the blood glucose level aod
the pathological processes responsible for complications, it should be appai
from comparisons of groups of patients who have had persistently diffei
degrees of abnormal blood glucose levels. Although the basis of this belief
difficult to perceive, it is rarely questioned. A blood glucose concentration tilL
fluctuates between 110 mg. per cent and 180 mg. per cent in a 39 year 01(1 100
throughout the day cannot be considered normal, even though most physicia~~
would consider this. good or excellent "control."
One of the undisputed aspects of the recent reports of the University Grout
Diabetes Program (UGDP)1' 2 is the record of the quality of "control" achie'
in adult-onset diabetics treated in university medical centers by a variet
therapeutic regimens, most of which are similar to those now cornnionly eu
ployed. None of the treatment groups in the UDPG study had a nomial fastin -
blood glucose at any time during the initial four plus years of the study. Diiiri
fluctuations in blood glucose on the patients' usual diet and with their tiLl'
physical activity were not assessed. However, at regular intervals the path it
were given a 50-gin, oral glucose load one half hour after their morning incdu
tion, and the blood glucose was determined one hour later. Although the 1071
mean blood glucose level at one hour was observed in the patients treate(l ~~it'
adjusted insulin dosages, even in this group the mean one-hour value
excess of 200 mg. per cent in more than half the tests during the initial four pilL-
PAGENO="0293"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13543
years. It is difficult to assemble totally appropriate age and sex matched control
data, but in a recent study of normal males with no family history of diabetes
(ages 22 to 28), the mean blood glucose one hour after a 50 gm. glucose load
was 90.6±5.7 iiig. per cent,3 and in the studies of Ned et al.~ the mean blood
glucose one hour after a 100 gm. glucose load in patients with no family history
of diabetes (ages 30 to 39) was 104.3±7.1 mg per cent in males and 93.7±6.~
mg. per cent in females.° The data available would suggest that most of the
patients in the UDPG study had persistently abnormal fluctuations in blood
glucose concentration irrespective of the treatment prescribed. Although the
UDPG study was a pioneer effort to improve the quality of clinical trials, it shares
one of the common deficiencies of such studies in that it included no group of
patients who had an essentially normal range of blood glucose fluctuation
throughout the day.
Another frivolous aspect of the current controversy is the simplistic fashion
in which the pathogenesis of the late complications of diabetes is viewed by
most physicians. Each of the diverse clinical syndromes lumped into this unfor-
tunate term has a very complex pathological basis, and in some instances a dis-
tinction must be made between processes responsible for acute symptoms and
those which operate in the production of predisposing pathological lesions. There
is little to support the widespread belief that each of the late complications is
primarily a reflection of diabetic microangiopathy. There is no convincing evi-
dence that alterations in capillary structure play a determining role in the
pathogenesis of the common symmetrical peripheral neuropathic syndromes
associated with diabetes mellitus,~ nor is there convincing evidence that this
process contributes to the increased frequency with which arterial occlusions
occur in specific portions of the vascular system. In patients dying of diabetic
nephropathy the pathological findings almost invariably include alterations other
than those unique to diabetes mellitus; thus, arteriolosclerosis and intimal fibrosis
of the renal arteries are often present.6 Therefore, in any organ system, the
pathological changes that are associated with diabetes mellitus may very well be
subject to modification by independently determined genetic and environmental
factors. However, this is rarely considered in efforts to evaluate the factors
responsible for the clinical course in diabetics. As an example, although the recent
studies of Hazzard et al.~ would suggest that inherited abnormalities in lipopro-
tein metabolism are commonly associated with myocardial infarction, the distri-
bution of patients with such abnormalities has not yet been assessed in clinical
trials of diabetic therapy.
Another reason why caution must be exercised in the evaluation of clinical
trials is the problem of asymptomatic but irreversible pathology in the patients
studied. It is notoriously difficult to (late the onset of an abnormality in glucose
tolerance in adult-onset diabetics, and this problem is magnified by the recent
realization that asymptomatie adult-onset diabetes may occur in children and
young adults and does not appear to progress to the ketosis-prone type of the
~ The effects of sex and increasing age on fasting blood glucose levels, and on the response
toa 100-gm. oral glucose tolerance test in patients with no family history of diabetes have been
well documented in patients up to age 39.
PAGENO="0294"
13544 cO~nETITIVE PROBLEMS IN THE DRUG INDUSTRY
disease.8 Thus, the presence of asymptomatic pathology in newly diagnosed
adult-onset diabetics cannot be used as a telling argument against a relationship
to the metabolic abnormalities associated with diabetes mellitus.
The failure to consistently observe improvement in specific diabetic compli-
cations following the institution of "stricter control" is frequently used as evidence
that the pathogcnesis of these conditions. is unrelated to the consequences of an
impaired insulin secretory mechanism. To cite hut one example of the limitations
of this reasoning, recent studies have demonstrated that the majority of a group
of newly diagnosed adult-onset diabetics w'ith normal standard neurologic
examinations and without symptoms of neuropathy could be shown to have
widespread fuiictional abnormalities in their peripheral nervous system when
suitably sensitive techniques were employed.9 Biopsies of peripheral nerves from
asymptomatic adult-onset diabetics without clinical evidence of neuropathy have
also revealed pathological changes similar to those found in the nerves of patients
with clinical neuropathy (although of a lesser degree) 10 Thus, the development
of the clinical manifestations of diabetic neuropathy may represent a late stage
in the pathological process and have irreversible elements. Whether or not
patients with clinically apparent peripheral neuropathy respond to efforts to
improve "control" can provide little information concerning its pathogenesis or
its prevention. The host of invasive techniques which would he required to
assemble suitably characteri~.ed subjects for study (e.g., coronary angiography.
renal and peripheral nerve biopsies, fluoroscein retinography, and so forth
almost precludes meaningful large scale clinical trials.
Many physicians also believe that the controversy over the relationship
between the metabolic derangements that result from an altered insulin secretory
mechanism and alterations in the capillary basement membrane in diabetics has
been resolved and that the two are clearly unrelated. This stems from the provoc-
ative studies of Siperstein and associates, who were the first to apply quantitative
techniques to the assessment of capillary basement membrane thickness
(CBMT) .~ They recognized the inherent technical problems in attempting to
measure CBMT in many organs and demonstrated that skeletal muscle biopsies
provide a means of obtaining suitable material for study from large numbers of
patients. Their data indicated that muscle CBMT was significantly greater in
diabetics and that the degree of thickening appeared to be unrelated to the
duration of known disease or to its "severity." ~\foreover, their studies suggested
that in patients genetically at high risk for the development of diabetes mellifus,
increased muscle CBMT was present prior to the development of a detectable
abnormality in glucose tolerance. Siperstein's work has been a major contribution
and stimulus. However, some of his observations and interpretations have been
seriously challenged by Kilo et al., who have concluded that muscle CBMT is
usually within normal limits at the outset of clinical diabetes mellitus and
increases with duration of disease.12 There are differences in methodology and in
patient selection in these studies, and the resulting controversy has not been
resolved. The studies of ~stcrby suggest that in the kidney the alterations in
glomerular capillary structure are not present at the outset of clinical diabetes in
young adults but progress with increased duration of the disease.13 Thus, it
would be premature to conclude that the demonstrable metabolic abnormalities
PAGENO="0295"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13545
in diabetes mellitus are totally unrelated to the development or progression of
alterations in capillary structure. A similar conclusion would have to be drawn
with regard to data available from studies of experimental diabetes in animals~
particularly in view of recent preliminary reports of the production of retina~
capillary microaneurysms in rats with chronic streptozotocin diabetes.14
From this cursory review it would appear that at the present time the physi-
cian treating adult-onset diabetics has a clear responsibility to prevent symp-
tomatic hyperglycemia; the advantages of maintaining a normal or nearly normal
range of blood glucose fluctuations remain to be evaluated hut cannot be
excluded. The choice of therapeutic aims must be a conscious decision by the
physician, and one in which an informed patient participates. The patient's age,
the presence of other serious medical problems, the presence or absence of
specific clinical complications, and the patient's motivation are practical
considerations..
The importance of Weight Control
Whether the physician believes that the prevention of symptomatic hyper-
glycemia is an adequate goal in a given patient, or whether a strenuous effort is
undertaken to achieve a normalization of glucose metabolism, the relationship
between increased body weight and hyperglycemia that so frequently exists in
adult-onset diabetics must be considered. In the UGDP study the adult-onset
diabetics averaged + 33 per cent of ideal body weight at the outset.'5 It has b~
observed repeatedly that many adult-onset diabetics will exhibit improvement in
~i~g~o~h T~emia if a significant reduction in bod~y ~ight c~
achieve . (Whether this is due to~~t reduction per seorto a decrease in
~iEoh~d~te~e r~ñ~ins to be deterrnined2°' ") The UGDP studies pi~~e
an in ication of the effectiveness o present efforts to achieve weight reduction
and the resultant improvement in diabetic "control." All the patient groups in
the UGDP study exhibited a fall in mean body weight at the time of the first
follow-up visit. However, the maximum decrease for any group was 4.2 lbs (less
than 3 per cent of initial mean weight), which was observed in the patients
treated with diet plus a placebo. In this group the early weight loss was not.
sustained, and the mean body weight remained close to 98 per cent of the initial
weight throughout the subsequent four plus years. The initial weight loss was
associated with a fall in both the mean fasting blood glucose level and in the
glucose level observed one hour after the ingestion of 50 gm. of glucose. Subse-
quently both of these parameters of "control" rose progressively throughout the
remainder of the study. Although the patients receiving tolbutamide or insulin
(in fixed or varying dosages) also received dietary instruction, their initial weight
loss was less than that observed in the placebo group, and no further reduction in
mean body weight was observed during the subsequent four years. These data
suggest that current efforts to achieve weight reduction in adult-onset diabetics
are ineffective, at least as currently practiced in medical outpatient clinics.
The difficulties encountered in achieving weight reduction in chronically
PAGENO="0296"
13546 COMPETITIVE PROBLEMS IN THE DRuG INDUSTRY
obese nondiabetic subjects have been well documented, and the psychiatric
implications of food to some patients are recognized; however, it is not certain
that the experience with these chronically obese patients provides a meaningful
index of the results that could be obtained in patients with adult-onset diabetes
Veight reduction r the safest ~
sufficient to prevent hyperglycemic s rn toms in the majority of adult-onset
diabetics. Moreover, in t e absence of si nificant wei~ t re uction, t e e cac Qf
6fber ai'~ts emp oye to lower blood glucose appears to be considerably
~
~ large percentage of patients. Diet instruction is confusing and reflects
the pseudoscience that presently surrounds the question of an appropriate diet
for adult-onset diabetics.
Dietary instructions should be as simple as possible. The justification for
modifications in dietary composition remains to be established: however, in the
face of evidence of persistent abnormalities in plasma lipoproteins, we suggest
modifications similar to those recently outlined by Fredrickson and his co-
workers.18 One of the factors that may contribute to the failure of diet therapy is
the long period between follow-up visits-three-month intervals being not
uncommon. The commercial weight reduction groups have demonstrated the
value of reinforcement provided by frequent weight checks, and brief visits in
which weight is checked by an office nurse can be helpful in many patients.
Increased physical activity is often dismissed as of little benefit; however, for
most 40 year old Americans a 15-minute daily walk would represent a significant
increase in physical activity. A conscientious effort to increase physical activit'
is in our hands a si"nificarii factor in those patients in w' om a, reduction in
~ight and improved glucose evels are achie~vd. It isO interest that a recent
~
insulin secretion observediuii~iaikedTy ~ by
oug t ere are many a u t-onset diabetics in whom lifelong patterns of
overeating and limited physical activity are practically immutable, the fact
remains that o encourage reduced caloric intake and ~rease
p ~ysical activity have been abandone man ysicians. There is little iitifi
ca iOn or t ~e use of any pharmacolo ie agent in asym tomatic adult-onset
~ possi ilty o attain~~i `e'cotrol"byweight_reduction
has been given an adequate trial. It is now eommon ~ begin
~ asymptomatic adult-
~
oreover, t e patient is instructed
that the oral hypoglycemic agents may provoke hypoglycemic symptoms w'hich
can be relieved by the ingestion of free carbohydrate. For many patients, who
are understandably anxious about their newly diagnosed disease, this becomes a
sanctioned invitation to unrestricted calorie intake. It is not surprising that sig-
nificant weight reduction and maintenance of normal body weight aie rarely
achieved under these conditions, fo~ bgth phvsiçi~n~p4. p~ept~ b~~c:pme to~
view the nature and amount of the drug iqgçs~cd ~
"controlling" adult onse diabetes
PAGENO="0297"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13547
Limited Effectiveness of Oral Hypoglycemic Agents
If the physician resorts to pharmacologic agents to prevent symptomatic
hyperglycemia in adult-onset diabetics before demonstrating the necessity for
their use, then one might hope that he would at least document the effectiveness
of the agent employed. Since a fixed dosage of 1.5 gm. per day of tolbutamide
was employed, the data derived from the UCDP study are not completely
applicable. ~
dos~g.~~nge,.a.nd the experience of this study is probably not too unrepresenta-
tive of the long-term efficacy of sulfonylureas in adult-onset diabetics in whom
significant weight reduction is not achieved. In the UDGP study there was an
initial improvement in fasting blood sugar and also in the blood glucose level
observed one hour after a glucose load. This improved "control" paralleled the
initial and transient weight loss which had occurred in these patients. Subse-
quently there was a progressive rise in mean fasting blood glucose and over the
remaining four plus years it rarely differed from that of the group receiving a
placebo by more than 15 mg. per cent (the mean values in 1)0th groups being
persistently abnormal). The blood glucose level after a glucose load also tended
to increase progressively in the patients treated with tolbutamide, and the mean
value was rarely more than 20 mg. per cent different from that of the placebo
group. At the end of four plus years the mean value in the group treated with
tolbutamide was 244 mg. per cent, whereas the value for the placebo group was
251 mg. per cent. Thus, one may question whether the use of tolbutamide under
these conditions ~
~
~
`T'ff~i~i~i~rcsponsible for the development of refractoriness to sul-
fonylurea therapy in patients in whom an initial response is observed are still
uncertain, and the data on its incidence are difficult to interpret.20 Values ranging
from 0.3 to 30 per cent have been reported, and there is a suggestion that the
incidence increases with known duration of diabetes. The uncertainty results, in
part, from differences in patient selection, in criteria for failure, and in the extent
to which other factors such as weight gain and dosage were considered. Unfor
tunately physicians exercise little selectivit in administerin sulfonylureaor
Jg~~i es ton u-onset ia ~ imposed by obesity2 arid
are slow to raise the question of efficac . One can only ness at the number of
patients present receivinl these drugs in w iom t eir withdrawal would not
signi cant v alter the dail' fluctuations in 00 lucose evels. Our own
ence wou ~ fraction of the patients who
Hive received the drug for a prolonged period iidln whom obesity persists. -
In gener~17~oth p~i~ians and patie are reuctant to stop ~i1Thypo-
glycemic therapy once it has been instituted. In many instances random blood
glucose values of 200 to 300 mg. per cent are observed over a period of months
to years before the physician reluctantly concludes that the drug has become
ineffective. The usual course under these circumstances is to resort to another
sulfonylurca in the hope of finding One that is "effective," and often to combine
this with a biguanide. ~ ~~rrent misuse of the sulfo-
56-592 0 - 75 - pt. 29 - 20
PAGENO="0298"
13548 COMPETITIVE PROBLEMS IN THE DRtTG INDUSTRY
nylureas and the bi uanides is that man atients are subjected to the expencc
an p0 entia hazard of long-term treatment to attain a relative imite coal-
free om rom ~ypergycemic symptoms-w i c tie agents are use&undercondi~
tions in which their efficacy is restricted and is only rarely adequately assessed,
~
of adult-onset diabetics in whom the use of the sulfonylureas can produce furth~r
improvement in the range of fluctuations of blood glucose over that resulting
from the correction of obesity. These agents may also be effective in some non.
obese adult-onset diabetics. We do not eschew the use of sulfon `lureas i
to prevent sym tomatic Ii' er cemi to achieve an a roximation of normal
uctuations in blood glucose. In both instances, however, the consequence f
~ n rited if obesit * esent. I
t ic a i on a p iarniacologic agent is ~
/ j ~ We do not believe that the
/ iguani es ave an established lace in the treatment of diabetics.
Many physicians are reluctant to accept t ic act t at t e su onylureas and
the biguanides are not appropriate agents in circumstances in which rapid cor-
rection of the metabolic abnormalities is required. An obvious instance is when
ketonemia develops either in association with an acute infection or following the
institution of therapy for an unrelated disease that requires the use of agents that
can impair endogenous insulin secretion and/or decrease its apparent effective-
ness. In these circumstances insulin is indicated, since neither the sulfonylureas
nor the biguanides provide significant protection against the development of
ketoacidosis. Even in the more common situation in which the previously
untreated patient presents with marked polydipsia, polyuria, and weight loss
without ketonemia, one cannot predict with any certainty whether the patient
will respond to sulfonylurea or biguanide therapy. Under these circumstances,
and particularly if there is an associated illness that predisposes to dehydration,
it is wiser to use insulin for the acute correction of hyperglycemia and the relief
of symptoms. Nonetheless, many physicians attempt trials at treatment with thq
sulfonylureas or the biguanides with a resulting delay in effective therapy that in
some instances may contribute to the development of hyperglycemic nonketotic
coma.
The use of insulin to achieve acute improvement does not imply that the
patients will necessarily require exogenous insulin once they have been stabilized
for a significant period. When there is a demonstrated requirement for the addition
of a pharmacologic agent to correct symptomatic hyperglycemia, there is a curious
reluctance to employ insulin in the management of adult-onset diabetics. This
stemi, in part, from the misconception that these patients are invariably insulin
hypersecretors and that insulin resistance is a major factor in the pathogenesis of
this syndrome. The bulk of the present evidence (which Kipnis has admirably
reviewed21) clearly indicates that most adult-onset diabetics exhibit an impaired
insulin secretory mechanism when compared with appropriately matched age, sex,
and weight groups. Moreover, although obesity in both the normal and diabetic
individual is associated with an apparent decrease in the biologic effectiveness of
insulin, the diabetic state per se does not appear to be associated with any signifi-
cant degree of insulin' antagonism (if one excludes specific. circumstances such as
I
PAGENO="0299"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13549
ketoacidosis). In nremains the preferred pharmacologic agentbeçauseofit~
assured and continued efficacy and its protective value in circumstances in which
~
Since 1972 there have been improvements in ~
preparations available, as the result of the application of methods developed to
separate insulin, pro-insulin, and related peptides. Preliminary data suggest that
these newer preparations may significantly reduce the immunologic responses to
the administration of porcine and bovine insulin and the associated clinical
problems.22 23 However, the present methods for the administration of insulin to
adult-onset diabetics are far from ideal. We have no practical method of repro-
ducing the timed relationship between food ingestion and insulin secretion that
Occurs in normals nor of selectively exposing the liver to the pulses of insulir~ that
normally occur in portal venous blood after eating. Therefore, there is little reason
to believe that present forms of insulin treatment reproduce normal physiology.
The primary advantage of the sulfonylureas is that they can be administered
orally. The resulting convenience is a real but not necessarily compelling con-
sideration when there is a manifest requirement for a pharmacologic hypo-
glycemic agent. The specific circumstances in a given patient should determine
the choice, and while our own preference is to employ insulin, there is no good
reason why a carefully controlled trial-preferably with one of the shorter acting
sulfonylureas-should not be undertaken. The physician should, however, con-
sider the possibility that eontraindications to the use of sulfonylureas may exist
(e.g., renal impairment or liver disease) and that sulfonylurea therapy may
influence the metabolism of other pharmacologic agents (e.g., bishydroxycumarin
and sulfonamides). The limitations of our knowledge concerning the role of
insulin deficiency in certain situations lead us to prefer the use of insulin in
patients with manifest symmetrical peripheral neuropathy with or without asso-
ciated chronic foot ulcers, and in individuals in whom repeated skin infections
or poorly healing surgical wounds are present. However, it must be admitted
that this is clearly a prejudice on our part.
Hazards of Treatment with Sulfonylureas
Insulin or sulfonylurea therapies are not without hazard. With insulin the
major concerns are hypoglycemia and problems resulting from the administration
of foreign protein( s). The safety of treatment with the sulfonylureas is, however,
the point most fiercely disputed at the moment. It should be apparent from thç
initial section of this discussion that we have serious reservations about the
conclusions that can be drawn from any of the published clinical studies, irre-
spective of the elegance of the statistical methods employed to deal with their
inherent deficiencies. A causal relationship between the treatment prescribed in
the UDPG study and the subsequent clinical course of these patients would be
difficult to establish on the basis of the information available. The excess cardiq
vascular moitaIity~obscrved in the groups treated with tolbutamide~~en-
formin cannot be dismissed, since the manner in which these agents were
PAGENO="0300"
13550 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
the misuse of ~hç~ç ag~pt~ jrj ~çç~nipipn practice. If subsequent
data provided by the UDPG group provide evidence of an unusual form 1
coronary artery disease in the patients treated with tolbutamide and phenformin.
or of an unusual propensity to cardiac arrliythmias, to pulmonary emboli, or ~
shock in the patients who died of cardiovascular disease, the case may be
strengthened. However, although we feel that the sulfonylureas have a limited
role in the treatment of adult-onset diabetes, we would not abandon their use en
the basis of the findings of the UDPG study. Nevertheless, this study has served
the useful purpose of reminding physicians that the use of sulfonylureas and
biguanides is not without hazard, and that as with any pharmacologic agent.
the possibility of unanticipated effects must be considered.
Hypoglycemia and hypersensitivity reactions have been the major problems
encountered in the use of the sulfonylureas. Hypoglycemia resulting froq~~3e
drugs can be prolonged and n~y ~
~ö~pob~di~ch aiThiorpropamide. Many of the reported instances havej~ltcd
~ im ait~
rena unc ion.-4 n a ition, it as ecn recognized that chlorpropamide can
induce a syndrome characterized by hyponatremia, impaired mental function,
and evidence of inappropriate antidiuretic hormone activity.25 The pernicious
aspect of this syndrome is that in elderly patients (in whom it has been most
frequently observed) the impaired mental function could easily be attributed to
other causes. There is little reason to believe that the effects of the sulfonylureas
are restricted to the pancreatic p-cells. Specific compounds have been shown to
alter thyroid function,24 to modify adenyl cyclase activity in cardiac muscle,27 to
augment antidiuresis,25 and to alter lipolytic activity in isolated adipose tissuc.~
However, with the exception of the UGDP study, the reported incidence of
serious problems resulting from the use of the shorter acting sulfonylureas has
been remarkably low. Nonetheless, the prolonged administration of sulfon lureas
to adult-onset ~ s with any pharmacologic
~ the indications for its use, and is not
acceptable unless Th ëfis~d~oiRtib~ted to b~e effe~tivë~~~ - -
Why Do We Use the Biguanides?
As we have previousl mentioned we find it difficult to *ustif ~
J biguanides in~trea~ei~ of diab ndwoul4discouraet~LJt is
doubtful that these agents would be employed at all if it were not for the fact
that they can be administered orally. While there are gaps in our knowledge of
the manner in which insulin and the sulfonylureas lower blood glucose levels in
diabetics, there is some reassurance that these effects arc mediated, in part, by
the correction of metabolic abnormalities resulting from an altered insuRn
secretory mechanism. Whether the long-term effects of the sulfonylureas result
primarily from an action on the pancreatic p-cells is a point that is difficult to
document, but the ability Of these agents to stimulate insulin secretion in adult-
PAGENO="0301"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13551
onset diabetics in whom they are effective is undisputed. In contrast, the manner
in which phenformin lo\vers blood glucose in human diabetics, or in animals with
alloxan diabetes, remains an enigma, ~ tar data a able are not
reassui~pg.An effect on insulin secretion has been excluded, although some
residual endogenous insulin secretion is necessary for phcnforrnin to be effective
in lowering blood glucosc.~~g~r et a!. have reported that phenformniinp~~~
~ peripheral
glucose utilization albeit with an increased rate of lactate production.3° The
applicability of data derived from other species must be questioned because of
the marked species variation in susceptibility to the hypoglycemic effects of
phenformin. Thus, Kreisberg and associates dispute any effect of phenformin on
decreasing hepatic glucose production or release in man.3° However, at high
concentrations phenformin does inhibit gluconeogenesis in isolated perfused rat
liver.30 In none of the isolated tissues in which the effects of phenformin have
been examined is there clear evidence that phenformin restores a pattern of
metabolism resembling that observed in the animal with a normal insulin
secretory mechanism.
More to the point, biguanides are ineffective in the prevention of keto-
acidosis, and there is no evidence that they have the assured efficacy of insulin
in the ~
suitable agent to use in those patients in whom a serious effort to achieve a
normal range of blood glucose fluctuation is undertaken with the aim of correct-
ing the underlying derangements in tissue metabolism that may contribute to
the pathogenesis of the late complications.,jii patients in whom efforts at weight
reduction have failed to remove the threat of symptomatic by er 1 cemia,
biguanides have no o vious advantage over insu in or tesulfonylureas (unless
one wish~ lewitseecs on gi~6~a sorptim~as an advan~g~). There is
~ of adult-onset
iabctes, un ess one will accept its use nj~flt to ~~~p~ent
with sulfonylureas. Sinesuitablealternativesaiiab1~th~ii&aiinnlQL
titio~lrisei~ailed in this practice esca es us. The UDPG study reported
a significant excess car sovascu ar mortality in the groups treated with phen-
formin, but again the causal nature of this relationship is difficult to establish.
However, there remains the distinct possibility that phenformin may represent a
pharmacologic hazard in a group of patients who tend to develop general or local
circulatory insufficiency. We believe this. may contribute to the association
between phenformin administration and the development of clinical lactic
acidosis.3'
Lactic Aciclosis
The pathogenesis of lactic acidosis in those instances in which there is no
obvious evidence of local tissue hypoxia is poorly understood. However, it is clear
that in many tissues the rate of lactate production under normal conditions is in
PAGENO="0302"
13552 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
a range that represents only a small fraction of the tissues' total capacity. In many
tissues thc ratcs of conversion of glucose to pyruvate and l'ictate ase kept it
small fraction of total capacity by chemical signals generated as a consequence
of the operation of the Krebs cycle and the electron transport system in which
oxygen is the final acceptor. The effect of anoxia on the rate of lactate production
in a tissue such as muscle results from a decrease in the signals which are usually
generated as a consequence of respiration and an increase in glycolysis. Under
these conditions the end product of gl~ colysis sppcsrs purnarily `is lactate sinci
a secondaiy consequence of imp'iiied iespn'stion is `in mci ease in the eytoplasrna
free NADH/NAD iatio which is one of the f'ictois determining the satso of
lactate/pyruvate in the cytoplasm There is no doubt that in lactic acidosis `is us
diabetic ketoacidosis the rate of pioduction of a norm'sl product can be incieasd
to le~ els th'it thi eaten the oi ganism s e'astence Lactate released into the cii eul'i
tion b) othei tissus is iemo~ ed in laige part by the liver where it is utill7ed to i
considerible degree foi the sesynthesis of glucose The oni) ~alue of this urn
phstic outline is to stiess the point that f'ictois that peirnit the e'tpiession of the
tiemendous htent cap'scitv for lact'ite production in many tissues or which
impair the capacity of the liver to dispose of lactate, may eventuate in lactic
acidosis It is well estiblished in idult onset diabetics thit chionic phenfoimin
administration i esults in sigmflcant elevation of blood lactate concenti `itions
While in most patients the levels obseii ed gii e little ciuse for concern the effect
does appear to be dose related 32 There is theiefore the possibility that at
sufficiently high concentiations phenfoirnin might induce lactic acidosis in
humans eithei by inereasmg peripher'il lactate pioduction or b) decieasin.
hepatic utilization oi both This would appeai to be the cisc since these arc
well documented instances in which phenformin was taken foi suicid il puiposes
w ith the subsequent development of lactic icidosis
The biguanides are unlikely to be the sole ciuse of the increased ficquencv
of lactic acidosis in diabetics since this syndrome has been observed in pitients
who are not receiving these diugs Adult onset dribetics are as a group mdi
viduals at inciessed sisk to the development of a number of acute conditions
which may produce local cireulitory changes conducive to the development of
either incieased lictate production or impaired disposition It seems difficult to
exclude the possibility that under these circumstsnces the pi esence of biquanides
may potentiate the dci eiopment of lactic icidosis The effoits to exclude this
possibility are not totally convincing thus the failure of bigusmdes to potentiate
markedly the rise in blood lactate in rats exposed to low oxygen tensions ignores
the relative insensitivity of this animal to the effects of these compounds. We find
ourselves in agreement with Oliva, who concluded: "It remains possible that the
association of phenformin and lactic acidosis is coincidental since lactic acidosis
may occur in diabetic subjects not taking phenfoi mm as is eli as in non di ibetic
subjects The v eight of the indirect evidence however strongly suggests thit
phenformin plays a causil or contubutoiy iole in the pioduction of lactic
acidosis 31 It is inteiestmg that the only death specifically ascribed to lactic
acidosis in the UDPG study occuried in a patient in the group receiving
phenfoirnin In sum since phenfoimin fills no unique iequirement in the trest
ment of adult~onset diabetes, since its mode of action is uncertain but unlikely
PAGENO="0303"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13553
to be corrective of derangements resulting from an impaired ilisulin secretory
mechanism, and since its relationship to lactic acidosis is unsettled, there is no
valid reason to employ it in these patients.
Summary
The present state of treatment for adult-onset diabetes is admittedly inade-
quate, except for the prevention of symptomatic hyperglycemia. The data derived
from clinical studies and from experimental work provide no basis for excluding
the possibility that normalization of blood glucose fluctuations may significantly
modify the development and progression of diabetic complications. However,
the value of this form of therapy has never been adequately tested, and its
immediate aims are difficult to achieve with present methods. It is an approach
that should be considered primarily in younger diabetics without evidence of
irreversible pathology who are capable of making an informed commitment to
this form of treatment.
In the majority of adult-onset diabetics the aim of therapy is of necessity
restricted to the prevention of symptomatic hyperglycemia, and irrespective of
the arbitrary assessments of "control" employed, most of these patients will have
blood glucose levels which persistently fluctuate in the abnormal range. The use
of any pharmacologic agent in this group of patients should be justified by
excluding the possibility that reduced caloric intake and increased exercise will
not remove the threat of symptomatic hyperglycemia.
In present practice the sulfonylureas and the biguanides are often used
without adequate indication and under circumstances in which they are unlikely
to be of any benefit. In addition, patients are exposed to the expense and potential
hazard of prolonged treatment with these agents without adequate concern for
their ef~cacy.
Insulin is the drug of choice when ketoacidosis threatens, or when an acute
improvement in symptomatic hyperglycemia is required. In asymptomatic pa-
tients with a demonstrated requirement for a pharmacologic hypoglycemic agent,
we believe insulin to be preferred, but a `veil controlled trial of a sulfonylurea is
not necessarily contraindicated. Bi~uanides have ~p ~ ~L
diabetes mellitus.
References
1. The University Group Diabetes Program: A study of the effects of hypoglycemic agents or'
vascular complications in patients with adult-onset diabetes. I: Design, methods and
baseline results. II: Mortality results. Diabetes 79 (Suppl. 2) :747-830, 1970.
2. Knatterud, C. L., Meinert, C. L., Klimt, C. R., et al: The University Group Diabetes Pro-
gram: A study of the effects of hypoglycemic agents on vascular complications in
patients with adult-onset diabetes. IV. A preliminary report on phenformin results.
J.A.M.A. 217:777-784, 1971.
3. Förster, H., Hasibeck, M., and Mehnert, H.: Metabolic studies following the oral ingestion
of different doses of glucose. Diabetes 21:1102-1108, 1972.
PAGENO="0304"
13554 COMPETITIVE PROBLEMS IN THE DRIJG INDIJSTRY
4. United States Department of Health, Education, and \Velfare, Public Health Service, Divi-
sion of Chronic Diseases: Genetics ansi the Epidemiology of Chronic Diseases. Wash-
ington, D.C., Government Printing Office, PHS Publication No. 1163, 1965, p. 113.
5. Winegrad, A. I.: Diabetic sseuropathy. (Editorial.) New Eng. J. Mcd. 286:1261-1262,
1972.
6. Warren, S., Le Compte, P. M., and Legg, H. A.: The Pathology of Diabetes Melhitu.s.
Philadelphia, Lea & Fcbiger, 1966, pp. 200-201.
7. Hazzard, W. R., Goldstein, j. L., Schrott, H. G., et al.: Ilyperlipidemia in coronary heart
disease: Lipoprotein characteristics of a classification based on genetic analysis. J.
Clin. Invest. 51:43a-44a, 1972.
8. Fajans, S. S.: What is diabetes? Definition, diagnosis, and course. Med. Clin. N. Amer.
55:793-805, 1971.
9. Chochinov, R. H., Ullyot, C. L. E., and Moorhouse, J. A.: Sensory perception thresholds
in patients with juvenile diabetes and their close relatives. New Eng. J. Med. 286:
1233-1237, 1972.
10. Chopra, J. S., Hurwitz, L. J., and Montgomery, D. A. D.: The pathogenesis of sural nerve
changes in diabetes mellitus. Brain 92:391-418, 1969.
11. Siperstein, M. D., Unger, R. H., and Madison, L. L.: Studies of muscle capillary basement
membranes in normal subjects, diabetic and prediahetic subjects. J. Clin. Invest.
47:1973-1999, 1968.
12. Kilo, C., Vogler, N., and Williamson, J. R.: Muscle capillary basement membrane changes
related to aging and to diabetes snellitus. Diabetes 21:881-905, 1972.
13. Østerby, II.: Diabetic glonserulopathy. A quantitative electron microscopic study of the
initial phases, Diabetes: Proceedings of the Seventh Congress of the International
Diabetes Federation, Buenos Aires, August 23-28, 1970 (International Congress
Series No. 231). Edited by R. R. Rodriguez and J. Valiance-Owen. Amsterdam,
Excerpta Medica, 1971, pp. 793-803.
14. Leuenberger, L., Cameron, D., Stauffacher, W., et al.: Ocular lesions in rats rendered
chronically diabetic with streptozotocin. Ophthal. Res. 2:189-204, 1971.
15. Coldner, M. C., Knatterud, C. L., assd Prout, T. E.: The University Group Diabetes Pro-
gram: Effects of hypoglycemic agents on vascular complications in patients with
adult-onset diabetes. III. Clinical implications of the UGDP results. J.A.M.A. 218:
1400-1410, 1971.
16. Grey, N., and Kipnis, D. M.: Effect of diet composition on the hyperinsulinemia of obesity.
New Eng. J. Med. 285:827-831, 1971.
17. Bruozell, J. D., Lerner, R. L., Hazzard, W. R., et al.: Improved glucose tolerance with
high carbohydrate feeding in mild diabetes. New Eng. J. Med. 284:521-524, 1971.
18. Levy, R. I., Fredrickson, 1). S., Shsilman, B., et al.: Dietary and drug treatment of primary
hyperlipoproteinemia. Ann. Intern, Med. 77:267-294, 1972.
19. Björntorp, P., Fahlén, M., Grimby, G., et al.: The effects of physical training and acute
physical work on plasma insulin in obesity. Europ. J. Clin. Invest. 2:274, 1972.
20. Krall, L. P.: The oral hypoglycemic agents. in Marble, A., White, P., Bradley, B. F., and
Krall, L. F. (eds.): Josbn's Diabetes Mellitus. Philadelphia, Lea & Febiger, 1971,
p. 315.
21. Kipnis, D. M.: Insulin secretion in normal and diabetic individuals. Advances Intern. Mcd.
16:103-134, 1970.
22. Schlichtkruli, J., Brange, J., Christianson, A. H., et al.: Clinical aspects of insulin-
antigenicity. Diabetes 21 ( Suppl. 2) :649-656, 1972.
23. Root, H. A., Chance, R. E., and Galloway, J. A.: Immunogenicity of insulin. Diabetes 21
(Suppl. 2):657-660, 1972.
24. Seltzer, H. S.: Drug-induced hypoglycemia: A review based on 473 cases. Diabetes 21:955-
968, 1972.
25. Weissman, P. N., Shenkman, L., and Gregerman, B. I.: Chlorpropamide hyponsstremia:
Drug-induced inappropriate antidiuretic-hormone activity. New Eng. J. Med. 284:
65-71, 1971.
26. Brown, J., and Solomon, D. H.: Effects of tolbutamide and carbutamide on thyroid func-
tion. Metabolism 5:813-819, 1956.
27. Lasseter, K. C., Levey, G. S., Palmer, R. F., et al: The effect of sulfonylurea drugs on
rabbit myocardial contractility, canine purkinjc fiber automaticity, and arlenyl cyclase
activity from rabbit and human hearts. J. Chin. Invest. 51 :2429-2434, 1972.
28. Stone, D. B., Brown, J. D., and Cox, C. P.: Effect of tolbutamide and phenformin on
hipolysis in adipose tissue in vitro. Amer. J. Physiol. 210:26-30, 1966.
PAGENO="0305"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13555
29. Kruger, F. A., Altschtild, B. A., Halloboiigh; S. L., et al.: Studies on the site and niecha-
nism of action on phenforniin. II. Phenformin inhibition of glucose transport by the
rat intestine. Diabetes 19:50-52, 1970.
30. Kreisberg, B. A., Owen, W. C., and Siegal, A. M.: Ilyperlacticacidemia in man: Ethanol-
phenformin synergism. j. Clin. Endocrinol. 34:29-35, 1972.
31. Oliva, P. B.: Lactic acidosis. Amer. J. Med. 48:209-225, 1970.
32. Craig, J. W,, Miller, M., Woodward, IL, et ii.: Influence of phenethylbiguanide on lactic,
pyruvic, and citric acids in diabetic patients. Diabetes 9:186-193, 1960.
33. Davidson, M. B., Bogarth, W. R., Challoner, D. R., et a!.: Phenformin hypoglycemia and
lactic acidosis. New Eng. J. Med. 275:886-888, 1966.
PAGENO="0306"
13556 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
IN VITRO EFFECTS OF TOLBUTAMIDE ON THE SYMPATHO-ADRENAL SYSTEM.
Chung-Yi Hsu, Gary Brooker, Michael J. Peach, and Thomas C.
Westfall, Charlottesville, Va. (Introduced by Joseph Larner,*
Charlottesville, Va.)
High concentrations of tolbutamide (6.6 mM and 10 mM)
provoked massive discharge of catecholamines from the perfused
cat adrenal gland. Differential estimation of secreted catecho-
lamines revealed preferential release of epinephrine by these
concentrations of tolbutamide. At concentrations within the
therapeutically effective range (0.1 to 1.0 mM), tolbutamide
depressed basal, nicotine-, KC1-, and glucagon-induced cate-
cholamine release from the cat adrenal glands and nicotine- and
KC1-,induced 3ti-nore~4inephrine release from perfused guinea pig
hearts in a dose dependent manner. This inhibitory action of
tolbutamide was quickly reversed following removal of the drug
from the perfusion solution. Prolonged perfusion for 20 mm
or longer with tolbutamide resulted in a sustained depression of
catecholamine secretion from the cat adrenal. Withdrawal of the
drug caused catecholamine secretion to reach a steady state
level which was higher than the level observed just before
tolbutamide perfusion was begun. This rebound phenomenon was
occasionally manifested by an outburst of catecholamine release.
Carboxytolbutamide, the major metabolite of tolbutamide, showed
no effect on basal as well as stimulated catecholamine secretion.
If tolbutamide acts similarly in vivo on the sjmpathoadrenal
system, then the detailed concentration versus time relationships
of tolbutamide could be of considerable importance in its pharma-
cological action and possible toxicological effects.
The Journal of the American Diabetes Association,DIABETES.
Volume 24 Supplement 2 , 35th Annual Meeting. June 15-17, 1975.
pp 414, #88
PAGENO="0307"
COMPETITIVE PROBLEMS IN THE DRIJG INDIJSTRY 13557
Reprinted from
21 March 1975, Volume 187, pp. 1086-1087
Inhibition of Catecholamine Release by
Tolbutamide and Other Sulfonylureas
Chung-Yi Hsu, Gary Brooker, Michael J. Peach and Thomas C. Westfall
Copyright@ 1975 by the American Association for the Advancement of Science
PAGENO="0308"
13558 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Sulfonylureas are oral hypoglycemic
agents which have been utilized exten-
sively in the treatment of adult-onset
diabetes mellitus. The primary action
of this group of agents has been at-
tributed to the direct stimulation of
pancreatic beta cells to release insulin
(1). However, the poor correlation be-
tween their insulin-releasing effect and
diabetic control has led to speculation
of extrapancreatic actions of sulfonyl-
ureas (2). Herein we report an effect
of tolbutamide and other sulfonylureas
on the spontaneous and nicotine-
induced release of catecholamines from
isolated cat adrenal glands (3) and on
amide amkte tolbutamide
((0.2 mM) (1mM) (1 mM)
Fig. 1. The effects of four sulfonylureas
on the release of [`HINE induced by nico-
tine in isolated guinea pig hearts, measured
in disintegrations per minute. Release of
[`H] was stimulated by a single injec-
tion of nicotine. Five minutes before and
after injection, one of the sulfonylureas or
the vehicle used to dissolve sulfosylureas
(control) was present. Asterigk denotes
difference from control is significant at
P < .001. NS., not significant.
nicotine-induced release of (3H]norepi-
nephrine ([3H]NE) from isolated guinea
pig hearts (4). These observations sug-
gest that sulfonylureas have a direct
inhibitory acfion on the sympatho-
adrenal system.
Isolated cat adrenal glands were retro-
gradely perfused with phosphate-buf-
fered Locke's solution, and the adrenal
catecholaminc secretion was monitored
by a modification of the procedure of
Robinson and Watts (5). This experi-
mental procedure has been reported in
detail previously (6). The effect of
tolbutamide on the spontaneous secre-
tion of catecholamines was observed by
perfusing the glands with Locke's solu-
tion containing tolbutamide (0.1, 0.3,
or 1 mM) for 5 minutes. To study the
effect of sulfonylureas on the nicotine-
induced release of catecholamines, each
gland was stimulated twice by per-
fusion with nicotine (10°M) for 1
minute. Five minutes before and 5
minutes after the first stimulation by
nicotine, the adrenal was perfused with
either tolbutamide (0.3 or 1 mM) or
tolazamide (0.2 mM). The adminis-
tration of nicotine was repeated 30
minutes after termination of the per-
fusion with sulfonylureas and served
as a control response.
Endogenous norepinephrine stores in
isolated guinea pig hearts were pre-
labeled with {3H]NE. Hearts were stim-
uated to release (3H]NE by single in-
jections of nicotine (4 x 10-i mole)
20 minutes after [°H]NE labeling. The
perfusate effluents from the hearts were
continuously collected and analyzed
for ~H]NE by liquid scintillation spec-
trometry (7). To study the effect of
sulfonylureas on nicotine-induced re-
lease of myocardial (~H1NE, sulfonyl-
ureas (tolbutamide, 1 mM; carboxytol-
butamide, 1 mM; tolazamide, 0.2 mM;
glybenclamide, 0.025 mM) were indi-
vidually added to the perfusion fluid
from 5 minutes before to 5 minutes
after the injection of nicotine.
The spontaneous output of catechola-
mines from the cat adrenal gland usu-
ally became stable 1 hour after the
initiation of perfusion with Locke's
solution. Tolbutamide (0.3 or 1 mM)
caused a decline in spontaneous cate-
cholamine output (see Table 1). This
inhibitory effect of tolbutamide was re-
versible. Nicotine-induced release of
adrenal catecholamines was also sup-
pressed by tolbutamide. When adrenal
glands were stimulated consecutively
at 30-minute intervals by the samedose
of nicotine (10-°M for 1 minute), the
response to the second stimulation was
Table I. Effect of tolbutamide and tolazasnide
on spontaneous and nicotine-induced release
of total catecholsmines from isolated cst
adrenal glands. Each gland served as its own
control. The data sre expressed as percent
of control. The number of glands studied is
indicated in parentheses after the mean
± S.E.M.
Catecholamines released
Drugs (% of control)
(mM) Nicotine-
Spontaneous* inducedt
Tolbutamide
0.1 95±2(3)
0.3 72±6(5)0 77± 5(4)0
1.0 62±8(5)0 56±10(7)0
Tolazamide
0.2 51 ± 7(3)0
IRate of spontaneous output is the presence of
aulfonylurea (na/mis) divided by the rate of
spontaneous output is the sbsence of sulfonylurea
(ag/mis)] X 100. Control spontaneous ostpot was
230 ± 30 ng/min. 0 INicotine-isduced release
is the presence of sulfonylurea (og) divided by
nicotine-induced release in the absence of sulfonyl
urea (sg) x 100. Control nicotine-Induced release
was 12,210 ± 1,330 ng. I siguiflant at P < .01
Inhibition of Catecholainine Release by
Tolbutamide and Other Sulfonylureas
Abstract. Tolbutamide and other suifonylureas inhibited spontaneous and
nicotine-induced release of catecholamines from the perfused cat adrenal gland
and nicotine-induced release of [°H]norepinephrine from isolated guinea pig
hearts. Of the sulfonylureas tested, the order of potency of this inhibitory effect
paralleled the hypoglycemic action. These results raise the possibility that the
inhibition of the sympathoadrenal system may contribute in part to the hypo-
glycemic action of sulfonylureas.
PAGENO="0309"
COMPETITIVE PROBLEMS IN THE DRIJG INDTJSTRY 13559
only 60 to 70 percent of the initial
response (data not shown). When
tolbutamide (0.3 or 1 mM) was pres-
ent only during the initial exposure to
nicotine, the second response to nico-
tine was then greater than the first.
Tolazanside (0.2 mM) exerted a simi-
lar inhibitory action (see Table 1).
Nicotine-induced release of [°H]NE
from the isolated guinea pig hearts was
also inhibited by several sulfonylureas
(tolbutamide, tolazamide, and glyben-
clamide). The order of potency in in-
hibiting catecholamine release by these
sulfonylureas was glybenclamide> to!-
azamide > tolbutamide (Fig. 1). Car-
boxytolbutamide, a major metabolite
of tolbutamide without hypoglycemic
action (8), showed no inhibitory effect
on induced 13H]NE release.
In vivo studies have led to the postu-
late that sulfonyluress may possess a
direct stimulatory action on the adrenal
medulla (9). Since epinephrine release
is increased by insulin-induced hypo-
glycemia (10. 11), the previous in
vivo observations with sulfonylureas
may be an indirect result of the hypo-
glycemia following insulin release. The
present in vitro studies demonstrate that
sulfonylureas act directly to inhibit the
release of catecholamine from the fe-
line adrenal medulla and from the adre-
nergic nerve terminals in guinea pig
hearts. This effect of sulfonyluress was
observed when catecholsmine secre-
tion was stimulated by nicotine or other
secretagogues such as glucsgon or KCI
(3, 4). Pittman snd Hszelwood selected
doses of insulin and tolbutamide which
caused similar degrees of hypoglycemia
in chickens and found elevated plasma
epinephrine levels only in those snimals
in which the hypoglycemia was induced
by insulin (11). This absence of epi-
nephrine release in response to tolbu-
tsmide-induced hypoglycemia suggests
that this drug probably inhibits adrenal
catecholamine release in intact animsis.
In general the metabolic effects of
cstecholsmines oppose the sctions of in-
sulin (12). Physiological concentra-
tions of catecholamines are known to
inhibit insulin release (13). It was in-
teresting to note that for the two sul-
fonyluress tested in cat adrenal glands
and for the four tested in guinea pig
hearts, the order of potency in in-
hibiting catecholamine release paral-
leled their hypoglycemic potency (1,
9).
Results from the present experiments
raise the possibility that this extra-
pancreatic action on the sympatho-
adrenal system may contribute in part
to the hypoglycemic effect of sulfonyl-
ureas.
CHSJNG-YI Hsu, GARY BROOKER
MICHAEL J. PEACH
THOMAS C. WESTFALL
Department of Pharmacology,
School of Medicine, University of
Virginia, Charlottesville 22903
Reforenees sad Notes
1, 5, F. Pfeiffer, K. Sclsoffllng, H. Ditschuneit,
IL Zeigler, is Oral Hylsoglyceeeic Agents,
G, D. Campbell, Ed. (Academic Press, New
York, 1969), p. 39.
2. J. M. Feldman aed H. 5. Lebovilz, Arch.
lssern. Med. 123, 314 (1969).
3. C.-Y. Hsu, Fed. Proc. 33, 524 (1974).
4. , T. C. Westtall, M. 3. Peach, Phor-
musologlst 16, 191 (1974).
5. R. L. Robinson and D. T. Walls, Cue. Chew,.
51, 986 (1965); R. L. Robinson, J. Pharmdcot.
Exp. The,. 156, 252 (1967).
6. M. J. Peach, F. 1,1. `Sumpus, P. A. Khairallah,
I. Pharmacol. Lop, The,. 116, 366 (1971).
7. T. C, Westfall and M. Brasled, Ibid. 182,
409 (1972).
5. 3. M. Feldman and H. 5. Lebovitz, Diabetes
18, 529 (1969).
9. A. BOnder, Aust. N.Z. I. Med. 5 (Suppl.
2), 22 (1971); W. E. Doling, E. H. Morley,
1, E. Nezamis, Proc. Soc. Lop. Blot. Med.
93, 132 (1956),
10. U. S. von Euler and R. Lutl, Metabolism 1,
528 (1952); A. Goldfien, M. S. Zilch, R. H.
Despointes, 3. 5. Brthsne, Endocrinology 62,
749 (1958).
II. R. P. Pitlman and R. L. Hazelwood, Comp.
Blochem. Physlol. 45A, 141 (1973).
12. 3. W. Enslsck and R. H. Williams in
Handbook nj Physiology, sect. 7, vol. 1,
Endocrine Pancreas, D. E. Steiner sod N.
Freinkel, Eds. (Williams & Wilkins, Balti-
mote, 1974), p. 665.
13. 0. Forte, Jr., and R. P. Robertson, Fed.
Proc. 32. 1792 (1973).
14. We thank Drs. W. 5. Doling and K. Gun.
deesen ot the Upjohn Company toe supply-
ing the sultonylareas used in these studies.
Supported by NIH.NS 10260, PHS NHLI
HL 12706, NIH P17 AM17042.Ol, und Hoff.
mann-LaRoche, Inc.
27 September 1974
PAGENO="0310"
13560 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
A STUDY OF
THE CHRONIC EFFECTS
OF TOLBUTAMIDE IN
THE RHESUS MONKEY
FDA # 72-114
By:
Jayme Borensztajn, M.D., Ph.D. Arthur Rubenstein, M.D.
Assistant Professor Professor
Department of Pathology Department of Medicine
University of Chicago University of Chicago
Godfrey S. Getz, M.D., Ph.D. Chung Hsin Ts'ao, Ph.D.
Professor Assistant Professor
Departments of Pathology Department of Pathology
and Biochemistry Northwestern University
University of Chicago
Seymour Glagov, M.D. Robert W. Wissler, Ph.D., M.D.
professor Donald N. Pritzker Professor
Department of Pathology Department of Pathology
University of Chicago University of Chicago
February 15, 1975
PAGENO="0311"
COMPETITIVE PROBLEMS IN THE DRuG INDIJSTRY 13561
A. INTRODUCTION
The aim of this project was to study the effects of chronic adminis-
tration of tolbutamide to normal Rhesus monkeys maintained on an atherogenic
diet. The rationale for carrying out such a project was derived from the
observations of the University Group Diabetes Project which suggested that
diabetic patients treated with t~lbutaznide had an increased incidence of fatal
events related to myocardial infarction when compared.to patients managed on
placebo or insulin (1). These studies raised the possibility that tolbutamide
administration might accelerate the development of atherogenic processes.
To test this possibility the Rhesus monkey model was chosen because
of ample evidence showing that atherosclerosis can be produced in this species
by simple dietary manipulations, and that the qualitative aspects of the
resulting lesions closely resemble atherosclerosis in man.
The "Average American Table Prepared Diet" was used for this study
because it was considered to be better to use a relatively mild (and slow)
atherogenic dietary regimen if one wanted maximum opportunity for a drug to
show either a positive or negative effect. Furthermore, it was thought beat
to work with dietary ingredients similar to those that people in `the U.S.A
commonly consume. Three chronic (2 year) experiments `conducted in this
laboratory over a ten year period have indicated that this ration `was well
accepted by the Rhesus monkey and consistently. resulted in mild to moderate
atheromatous change with blood lipid levels not too far above levels commonly
found in people consuming this type of diet.
PAGENO="0312"
13562 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Study of coronary artery lesions
The coronary artery sections were taken at nine predetermined sites in
the main coronary artery (Figure G.6.) and were sectioned on the freezing
~CO2) microtome and stained wjth Oil Red 0 for fat. The other nine adjoining
sections that were obtained from the sane sites were embedded in paraffin and
stained by hematoxylin and eosin and Gomori trichrome aldehyde fuchsin. They
revealed a very unexpected and disturbing trend; coronary artery lesions in
the treee main coronary arteries were almost two times more frequent end al-
most exactly three times more severe in the tolbutamide-treated animals than
in the control animals in relation to the amount of lipid deposited in the
intima and media (Table G.6.) ~* The intimal proliferation noted in the
coronary arteries of the tolbutaxside tested animals was also somewhat more
severe than in the controls but this difference was not statintically signi-
ficant. Figure G. 7. presents photcasicrographs of average types of coronary
lesions stained for lipids in paired animals. Lower and higher magnification
in the same lesioOs shows the intimal proliferation and lipid deposition in
sections of the two types of monkeys. Moderate intimal proliferation and
lipid deposition in the intimas and innermost media can be Seen in animals
treated with tolbutamide.
Pancreatic islet evaluation
The pancreatic islets of those animals were carefully examined by means
of two stains, the Gomori trichrcsne-aldehyde-fuchsin stain and the ~ciuori chrome
alum hematoxylin stain. These histological preparations were made from three
Bouin's fixed complete cross sections of the pancreas, taken from its tail,
body and head.
*The detailed microscopic evaluations for these coronary lesions as done
independently by each pathologist are presented pair-by-pair in Appendix IV.
PAGENO="0313"
TABLE 81.
PLAN OF THE ~PERIMEffl~ IN RHESUS MONKEYS FEB
AVERAGE ANERICAN DIET WITH OR WITHOUT TOLBUTNIIDE
G1~OuP
~
NUMEER OF
ANIMALS
DIET FED FOR A PERIOD OF 18 frONT-IS
.
j
9 (7)A
AVERAGE AMERICAN flIEr - 250 G/DAY
.
II
9(7)
AVERAGE AMERICAN Din- AND TOLBUTAMIDEB
~
20 MG/KG/DAY - PAIR FED
0
0
w
~J2
A. THE EXPERIMENT WAS STARTED WIT-i 9. ANIMALS PER GROUP1 BUT 2
ANIMALS DIED AT 1~4 MGNTHS AND THEIR CONTROLS WERE AUTOPSIED AT
ThE SAME TIME
B TOLBUTAMIDE - UPJOHN COMPANY
PAGENO="0314"
Ti\RLE G.6.
CD
0
0
02
L~i
02
MICROSCOPIC. FINDINGS IN CORONARY ARTERIES OF RHESUS MONKEYS
FED. AVERAGE kiERICAN DIET WITh OR WITHOUT TOLBUTANIDE
GROUP
No~ OF
PNIW~LS
. DIET .
. .
.FREQUENCY~
. .
SEVERI~~
.
.
.
II
..,
9(7)
.
9(7)
.
A\~ERAGEARERICA DIET
AVERAGE P~1ER!CAN DIET
... : .~
AND. TOLBIJTMIDE . .
31±6~3
.
: 65+~~5
. .. . . .
O~18±O~O3
.
O~56÷O~O6
.
p.
i. "Rarely, it may be necessary' to stop thl~Sidè therkpy before ~hyptdtalemla'
can be alleviated." S S `
j. "In states of pre-coma of hepatic origin, thiázide dltitètics' ma~~ trecitltate
coma~"
The promotional text of the advertisement contSin5s 1!~tulty "and `xhisleadii~g
representations as follows: ` `. `/ ~, `~ ` ~` 5'
a. The claims that the use of Salutensin will "get blo~ preoSu~e~dow5n so~u-
er" and that `use of Salutensin will yield "iliore ~ ilyper-
tension," than "just thia~Mde-rOserpine" have not been' approved foi~ labeling,
nor is there evidence in the New Drug Application to support theth. `
b. Referring to the combination of thiazide, resei~pine and protoveratriné~'A
the ad states: "Clinically, the advantages of such ,a combination have~ been
summed up as follows: `The concomitant use of résérpine and the' thia~idOs
and veratrum' has greatly Widened the therapeutic `and to~ic rahge, niaking
veratrum more effective and simple to administer *` * * `~tt is unfortunate that
more physicians `do not take advantage of veratrum,' thlazlde, tind reser~iné to'
lower the arterial, pressure simply, ~effectively and without ~ide effects," Tj~te
ad uses referenCe #2 (~`innerty) foi' the two-senténCe,statemhnt `quoted above.
Such statement Is objectiOnable for the following, reasOps / "
(1) The reader is led to believe that Finnerty iise5d ~ iti 13i5
study In fact, he prescribed Veriloid (alkavervir) `and' Unitensin (`eryptena~
mine)'. (l~rotoveratrine A iS a, purified single alkaloid `Isolated from' ~er~trnm
album. Alkavervir is a partially tiuriflOd extract epntainiug a thixture of alka-
loids, obtained from Vertitrum viride. Cryptetianhine is an alkalOid `derived from
an extract of Veratrum viride.) , ` `~ ` "
(2) The ad misleads by quoting Finherty out pf co~itext. The' i~e~der i~'led to
believe that the author waS advising th'e'reader `th~t t~ comblhia'tloti `of vera~
tr,um, reserpine and thiaziçle lOwers blOod ~ressiire~'ef~etiveiy aftd tt~ilhont~sitle
effect when a "simple" dosage' sch'edtile, `supposedly like tb~ öné `Bri~tOl `reconh-
mends, is followed. A review of Finnerty's artlcI~t' SlibW'ed thIs ~i~t "to be'~tbe
case, for in the `Several setitences lOCated betwehtC tIle two neiitØnces S tjuoted,
the autjiOr gives these detailed instrtictlons: the patffiuit `~h~thd ~kt ~t~the 5s~the
time every day the patient shotild hOt ett1~ fOi' tWO 1IOth~S attoS taking veM
ttuin'; a gool beginning `dosage of `vth t±unl ls~'2 mg~ t e~t1hiés, a day, é.g~,'
aftéy breakfast, mid~afternoon and at bed'Vihie~ and t~n `gi~aduaily~,tnci'ease~
over a~eriod ef `two to three' weeks-up tç~btit not e~ceediifg 4 mg. thteë times
a day.' ` / ` ` ` ` .//~ ~ , `, `: ~ :` ` /5 ,~`:` /
Also related to the problem of effectiveness and safet~r' ure the author S.
ments elsewhere in his article that the tLt~Zh4è~' (eblortlalhd&ne~-~whiCh
Incidenta'ly is not a thiazide) component of Ms eomttinntloti therapy was given
in a sIngle 50 mg daily dose (note thi[t the reeOmzuehde~l dosage range fGr the
hydrOllume~hiaStde in SalutetiSht"ih higher átid the' Scheduiè'perniit's"the' tloSb
56-592---Th------~22
PAGENO="0330"
13580 COMPETITIVE PROBLEMS ~N THE DIWG INDUSTRY
of the diuretic to be given up to four times a day) ; that it was necessary to
administer supplemental pott~ssium to those digitalized patients who had been
given ~ trial oi~ other thiazjc~es-ot1~r thai~ chiorthalidone-to avoid ai~rbyth~
mias , that a marked increase in toxic reactions occurs when more than ~ O~25
rug. of reserpiiie per day is~ administe~'ed (note that the dosage rai~ge recoin~
mended in the package insert for the reserpine component in Saluteusm may
be much higher than that recommended by Finnerty); that because mental
depression from reserpine has resulted in suicides, monthly checkup Is man-
datory; that the combinatlop of veratrum, thiazide and reserpine Is indicated
in moderately severe hypertension,. and that be does not advocate the combina-
tion in mild or severe hypertension.
21 UJ~.U. 352(f) (1) [Ileg. 1.106(b) (4)]: Salutensin ~s a prescription drug
within the meaning of 21 U.S.C. 353(b) (1) (0) in that it is limited to prescrip-
tion use by its approved new drug application. Accordingly, since adequate di-
rections for lay use cannot be written for a prescription drug, it is clear that
the labeling for Salutensin berç involved did not bear adequate directions for
lay use as required by 21 U S C 352(f) (1) Furthermore, the Saluteusin in
volved here was not Oxempt from 21 U.S.C, 352(f) (1) as provided for by regu-
latiOns 21 CFR 1.106(b) since it failed to comply with the condition for cx-
einption set forth in subparagraph 4(1) of such regulations. This condition
requires that any labeling of a prescriptiOn drug shall contain adequate Informa-
tion for use including lUdicatlons, effects, dosages, routes, methods and fre-
quency and duration of administration and any relevant hazards, contraindi-
cations, side etfect~ and precautions under which practitioners licensed by law
to administer the drug can use the drug safely and for the purpose for which
it is ipteuded and, if the drug is subject to 21 U.S.C. 355, the labeling providing
such information shall be substantially the same as the labeling authorized by
the approved new drug application for such drug.
In the case of Salutens~n, examination of the Mailing Piece designated as
"SH 3852 RV", which is labeling for the drug as described iii regulations ~1 CFR
1.105(1), has disclosed that such labeling Is not substantially the same as that
which is set forth in the approved new, drug application for the drug.
1. ThIs mailing piece deviates substantially from the approved package insert
labeling by failing to disclose the following side effects, warnings and precau-
tions:
a. Reduction of serum electrolyte concentrations may occur with Salutensin
"resulting in hypochloremia, hypochloremis alkalosis, hyponatemia . . ." The
warning in the mailing piece that "alterations In electrolyte balance . . . may
occur~' does not clearly inform the reader that the electrolytes, sodium and
chloride, can be depressed below normal levels to cause these potentially serious
cljnical states.
b If indicated potassium loss often may be easily replaced by including
potassium-rich foods in the diet (tomato juice or orange juice or other citric
juice, banana, etc.). Patients unable or unwilling to take fruIt juice may be
given potassium chloride 1 Ow 2 to 4 tImes daily by mouth. Rarely, it may be
necessary to stop thiazide therapy before hypokalepua can be alleviated
c. Some patients may have "insomnia" and "nightmares."
2. The promotional text of this mailing pi~ce contains these faults:.
a. The claims that the use of Salutensin will "get blood pressure down sooner",
and that the use of Salutensin will result In "more successful management of
hypertension , than lust thiazide reserpine lrrve not been approved for label
lug, nor Is there evidence In the NDA to support them.
b. This mailing piece also refers t~ the work by ~innerty (reference #2) In
such manner as to mislead the reader.
c The company's descri~tlon of the work by Smith (reference #3) exagger
ates a claim for etflc&cy ~et~ fails to mention s1d~ effects reported by the author.
(1) The reader is not told that the subjects in this study were hospitalized
and had an average age of T7. years. Without this Information he' Is prevented
from correetly ciraluating the. results ,Qf the author~ and from posing the ques~
t on of whether or pot these patleIits copid be expected to show similar blood
pressure reductions on just one of the Ingredients of Salutensin, i.e. the reserpine.
or the hydroflumethlazide. ` ` "
(2) Further, the mailing `piece falls to reveal' that the physicians attending
these patients were allowed to adjust the dosage of the drug within a range of
1 t~ 4 tablets, daily. Tbi~ ~tnformat1on Is clearly set out in the approved package
hlsert. The failure. to reveal this in the' mailing piece Is misleading because the
PAGENO="0331"
COMPETITIVE I~ROBLEMS IN THE DRUG INDUSTRY 13581,
headline implies that the results were obtained with "only one or two tablets t~
(lay." ~ . ~ ,~ ~ ~ ~ ~ ~
(3) The mailing piece does not show fair balance. It attempts to exaggerate
safety with a quotation from one paper (Finnerty), but, in an attep~pt to exag~
gerate efficacy claims, presents the report by Smith without revealing that twelve
of the author's 45 patients bad side effects such as nausea, vertigo, Weakness
and dizziness while on Sajutensin, and that dosage in these cases had to be
reduced in order to reduce the side effects. Further, the original Smith report~
In the NDA points out that several patients had side effects which were not
"minimized" by reduction in dose and that two additional patients had side
effects while on Salutensin; one with nausea, and another with abdominal pain.
The Smith report shows that some patients on Salutensin experienced side
effects on dosages (2 tablets daily) within the range recommended by the com-
pany.
Examination of the Salutensin Mailing Piece designated as "511 3919 RV-2"
disclosed the following:
1. The information in this. mailing piece is lacking in its treatment of the
drug's side effects, etc., as described above In the analysis of ma1~ing piece
N~, 511 3852 RV.
2. The promotional section of this mailing piece contains the same faults
as described above in relation to mailing piece No SIT 3852 RV plus additional
errors concerned with the use of two additional research papers.
a. The mailing piece misrepresents the work of Spiotta (reference #6, a~
report to Bristol Laboratories) and a graph "adapted from Splotta,"
* (1) The company seriously misleads by failing to tell the reader that the
study represented by the graph was on only one patient. As evidence, see use
of same graph in the enclosed October 1961 mailing piece which specified that
this is the result of a study of a single case. What is more, there were only
two patients In the Spiotta study who received serial additions of the ingredi'
ents of Salutensin In the order shown in the graph, namely Saluron, protovera-
trine A, and then reserpine.
(2) The company fails to inform the reader that when all the Ingredients
were added each of the 7 patients in this phase of this study was finally receiv-
big 200 mg. of ~iydrofiumetblazide, 0.5 mg. o~ reserpine and 0.8 mg. of protovera-
trine A per day (equivalent to 4 Salutensin tablets) and that not all of the
patients were receiving the three components in the same order shown in the
graph. The failure to disclose this is misleading because the headline in the
promotional section of the. mailing piece implies that the graphed results were
obtained with "only one or two tablets a day."
(3) The company also fails to tell the reader that Spiotta studied additional
patients who received only 2 of the 3 ingredients of Salutensin; that be found,
for Instance, results with hydrofiumethiazide plus reserpine which were similar
to those with Salutensin in some cases. Such information does not support the
idea in the promotion that Salutensin is better than a combination of only two
of its three components, and that protoveratrine A is needed for more successful
antihypertensive therapy. It is apparent that the Splotta study (and it Is the
only study in the NDA on the serial addition of each Salutensin component).
involved an insufficient number of patients to warrant making any ~.sound
judgment regarding the comparative efficacy claim which thu headline states
"instead of just thiazide-reserpine, use Salutensin."
b. The mailing piece presents the work of Thomas (reference #8, a report
to Bristol Laboratories) and `a graph purporting to show the effectiveness of~
S~lutensin'in longstanding hypertension of moderate severity. The reader is led
to believe that all 40 patIents received only Salutensin, and that the dose, as
claimed In the headline of the promotion, was "one or two tablets per day"
for all patients. On the contrary the approved package insert itself merely
states the following about results with Salutensin in the Thomas study:
* (1) "Thomas noted that In many patients It was possible to. eliminate' bydral-
azine which patients had been taking peevionsly." (Emphasis added)
(2)' "StatIstical analysis of the data indicated `that the supplementary dose
of reserpine could be considerably reduced `while they were on Salutensln lii
order to maintain satisfactory control of their hypertension." At least ten. of
the patients were taking supplementary doses of reserpine
(3) "A few of the patients required the addition of other antihypertensive
(reserpine, bydralazine, inversine, Singoserp, c~bk~rothlazide) agents' with Salu-'
tensin in `order to maintain satisfactory control ~of their hypertension." The
"few" patients referred to above numbered at least ten.
PAGENO="0332"
13582 COMPETITIVE PROBLEMS IN ~THE~ DRUG INDUSTRY
U) "Sonte received as Iittl~ ~as 1 tablet daily others as much as 4 tablets~
daily, the dose being adjusted according to individual patient response as the
study progressed." The headline in the mailing piece, however, ifivites the reader
to believe that Thomas also found that the daily dose required to keep blood
pressure down was"one or two tablets".
Examination of~ the Salutensin monograph appearing In the 1965 edition of
the PDIt reveal.s a third instance where. labeling is not substantially the same
as that which is set forth in the approved new drug application. In the same
way as described above at page 8, paragraph 1 in relation to Mailing Piece
"811 8852 RV" this monograph fails to provide full disclosure Concerning the
drug's side effects warnings and precautions.
II-~Prostaphlin:
21 U.S.C. 352(f) (1) (Reg. 1.106(5) (4) )-Prostaphlin i.s a prescription drug
and, as such, adequate directions for lay use cannot be written for it. It is
quite obvious then that the labeling of the `Prostaphlin here involved di~ not
bear adequate directions for lay use as required by 21 U.S~O. g5~(f) (1'). Fur-
thermore, the Prostaphlin was not exempt from 21 U.s.C. 352(f) (1) as provided
by regulations 21 CFR 1.106(b) ~ince it failed to comply with the conditions
for exemption set forth in subparagraph 4(i) of such regulations. This condi-
thai requires that any labaling. of a prescription drug shall contain adequate
information, for use including indications, `effects, dosages, routes, methods and
frequency and duration of athninistrati'ón and any relevant hazards, contrain-
dications, side effects `and precautloils under which `practitioners licen~ed by law
to administer the drug can use the drug safely and for the p1~frposes for which
it'ls Intended. Further if the drug `is subject `to 21 U.S.C. 857, the labeling
providing such info~mation `must be' substantially the ~ame as the labeling
required as a Condition fOr the certification Of alich drug (Sea regulatIons' 21
CFR 146.2(b) and 146.4(a) ~1)). In the ease Of Prostaphiin, examination of
the monograph for such' drug in the 1965 edition of the Physicians~ Desk Refer-
ence, which is labeling for the drug as deScribed in regulations 21 CFR 1.105(1)',
has disclosed that such labeling is not substantially the same as the labeling
(package insert) which was submitted for purposes of obtaining certification.
Examples of the omission of important items of informatiQn in the Prostaphlin
mohograph In the PDR which are Contained in the package `insert labeling sub-
mitted for purposes of certification are ~et forth below:
`1. That `safet~r for use of the `drug In pregnancy has not been established.
2. That periodic assessment of organ system functions, including renal, hepatic
nnd hemato~eVetic, should be made during leng-terni therapy with the drug.
8. A warning that anaphylactoid reactions to the drug' have been encountered.
4. A warning tkat hazards of anaphylaxis to patients with.' a history of peni-
d111h allergy xr4ust be `balanced against' the prognosis if the drug is withheld.'
The Prostaphlin monograph in the 1965 editIon of Physicians' Desk Refer-
ence also differs from the package Insert labeling submitted for purposes of
certification In that the monograph is `false and misleading because of the
following:
i The monograph implies that the drug Prostaphlin is safe and effective
when used in accordance with the information' contained' in the monograph,
when `jI~ fact thiS `monograph does not provide' all the information required
i~or sa~C aild effectt~e use of the' drug because it fails to include the warnings
referred to under the four points discussed in the' preceding paragraph.'
`2. `The mono~ráph includes `the statement "No anaphylactie reactions have
been encountered" `which statethCnt iscontrary tofact. ,
`3. The monograph tnclndeC th~' statement "Reactibns to' this penicillin ha~ve
1~een `infrequent ahd mild in' natiu~e". which l's misleading because of the `phrase
4'mIid in' nature." ` ` `
21 `U.S.C. 852(1)'.-It Is ftrther alleged that Prostaphlin when introduced'
lr~tO. interstate commerce wa~ `xnisbi~and~d within the' `meaning Of 21. U.S.C.
~52(1) in that it' was, represented as ~ drug composed ef a ~ertifiable antibiotic,
namely sodium oxadilhin, and it was ~not' from a; batch with reSpect id which
a' certificate or"release issued pursuant to Sectioi~ 857 wa~ in"e~ect *ith reSpect
to sñ~b drug since the cettiuIcatd WhiCh was issued on Február~' 4~ 1965~ for' such
`batch `had' bèeh obtained' throu~gh mfsrepreseiltation and Concealment of a ma-
terial fact. The application for' `c~rtiff~atloh constituted a mlsrepreSeñtatioñ
niid' was a concealment df a indterial' faci `in that it' purported' to be preCeded or
ac~~ónipani'ed by- sj~ecimeiis of all labeling te be `used for suCh drug when' `hi fact
PAGENO="0333"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1358*
it failed to include the ~thonograph of said drug which appeared in the 196~
Edition of the Physiciaus' Desk I~ofei~ence. (See~21 CFE 146.2(b) and 21 CF~
146.4(a) (1)).
EVIt~EN~E OF VIOLATIVE SHIPMENTS
On June 16, 1965, Food and Drug Inspector Joseph S. Slayton collected a
sample of Salutensin from a lot of 72 60-tablet bottles which were held at
McKesson and Robbins, Inc., Pittsb~irgh, Penilsylvania. A copy of an invoice
and shipping records in the. forii~ of a motor express bill of Lading and/or
freight bill showed that the lot had been received from the defendant's South
Hackensack, New Jersey warehouse on or about 5/13/65. Inspector Slayton
also obtained from Dr. James K. Spence of Pittsburgh, P~nnsylvania a state-
ment that he is a regular subscriber of the publication entitled "The Antietican
Journal of Cardiology." The doctor further identified pages 41 and 42 of the
Salutensin advertisement in this publication bearitig a Noveffiber 1964 issue
date.
On January 17, 1966, Food and Drug Inspector Alfred M Levy obtained a
statement from Donald J. Tie~rney, Production Manager of Fisher-Stevens, Inc.
(Mailing Service for Bristol Laboratories) atte~ing to the fact that a total of
9,693 four page mailing pieces identified as SH 3852 RY and Sli 3919 RV-~2 and
relating to Salutensin were mailed to ph~~rsicians in Pennsylvania; on May 26,
1964, and July 15, 1964. Actual samples ~of these mailing pieces were also iden-
tified and furnished to the inspector.
On October 29, 1965 Food and Drug Inspector Joseph P. Brochetti collected
a sathple of Protaphliri from a lot of 10 48-capsule vials which Were held at
McKesson & Robbins Drug Division, 445 Fort Pitt Boulevard, Pittsburg, Penn-
sylvania. A copy of an irifoice, the bill ~f lading, receiving record and dealer's
statement showed that the lot had been received from the defendant's Sotith
Jiackensack, New Jersey warehouse on~or about September 12, 1965.
* HEkEING Pt4ISUAET TO 21 TY.S~C. 355
Pursuant to the Notice of hearing regarding Salutensin dated August 9,
1965, Dr. Harold Frediani of Bristol Laboratories called the Food and Drug
Administration's Buffalo District dfflc~ on August 11, 1965 `aid w~s prot1ded
with d~tails concerning the ~peciflc charges involving representations in the
medical journal advertisement of Novethbe~ 1964 and in the labeling referred
to as mailing pieces S~I 8852 RV and SH 3919 RV-2. S S
The Hearing Was orlgffiaI1~r scheduled for August 24, 1965 but wa~ subse-
quently rescheduled at the request of the citee for September 7, 1965. The firhi's
answer was in the f~r'n1 of a 16~age letter dated: September 3, I96~. The letter
bore the ~ignatuie Of Hubert O~ Peltier, M.D., Medi(~al Direetdr Of IflristOl Lab-
oratorlea This Wi'ltten ahswer denied all dllegatidrIs. It ~il1aintained that the
advertisement' complained of complied With all pertinent requirements of the
law. Likewise, the answer maintains that the mailing pieces invOlvOd contfiitied
adequate directions and information for the practitiotier. The lirni contended
that the Salutensin aèWertisem'ent in the November 1964 issue of the American
Journal Of CardiolOg~T doti'stituted an "alerting device" when read by the physi-
cian, and further referred the physician to the official package circular. The
interence was that the advertisemefit "in its entirety" included package insert
information not present in said advertisement. The firm likewise denied that
the mailing pieces SH 3852 RV and Sil 3919 RV-2 failed to contain full dis-
closure relative to side effects and warnings, etc. They contended the pertinent
information given was all that a physician needed in order to use the drug;
and that Only unnecessary elaboration coiltained in the official'package circular
had been omitted. Throughout the written response, the firm contends and infers
that many of the cautionary and warning statemetits, (although present in ap-
proved labeling), are superfluous and would not be needed by a physician, who
would be aware of these omitted statements. The answer further indicated that
the respondents would appreciate meeting with appropriate persons in the
Washington headquarters o~ the Food and Drug Athflinistratlon to discuss the
matter with a view toward preventing such futuredifferences. S
Representatives of the firm met with Food and Drug Administration officials
in Washington on October 11, 1965. The majority of the violations were dis-
cussed and the firm promised corrections in `both future Salutensin advertise-
ments and mailing pieces.
PAGENO="0334"
13584 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Pursuant to the Notice of Hearing regarding Prostaphlin dated November 4,
1965, and setting the Hearing date for November 17, 1965, the firm submitted
an answer in the form of a four page letter dated November 16, 1965, signed
by Hubert C. Peltier, M.D., Medical Director of Bristol Laboratories. The firm's
answer, in essence, denied the charges made by the Government. However, it also
included a somewhat corrected galley proposed for the 1966 Physicians' Desk
Reference relative to the Prostaphlin monograph.
CONCLUSIONS
We believe that the factors which have 1~een set forth in some detail above
show that the labeling and advertisements used by this firm have not met the
standards required by law. The omissions and deceptive statements involved
were numerous and serious. Moreover~ the required warnings were already well
known to the company. As to Salutensin they were set down in the New Drug
Application labeling and as to Prostaphlin they were included in the labeling
submitted prior to antibiotic certification (package insert). In drug advertising
the law does not provide for product ~outing, or "puffing" when it entails a com-
promise in the requirement of full disclosure. The advertisement and mailing
pieces involved in the Salutensin charge are replete with half-truths designed
more to boost sales than to provide a physician with the information essential
to the proper and safe prescription of that drug. As to Prostapblin not only did
the company fail to submit the Physicians' Desk Reference labeling as part of
its request for batch certification as specifically required by regulations (21
CFR 146.2(b)) but the monograph which it did place in the Physicians' Desk
Reference was at serious variance with the existing approved package insert.
More oversight or carelessness cannot excuse the violations complained of here.
The manufacturers of potent drugs, better than others, know the potential haz-
ards of their products. Busy physicians should. and must be able to rely on
statements concerning a product without referring back to the original source
to look for inconsistencies and contraindications. We believe the prosecution
is fully warranted.
WITNuSSES
The principal witnesses in this case will include the government inspectors
who collected samples of the two drugs involved; witnesses to establish the
interstate shipment of the drugs, the Issuance of the PDIt,' mailing pieces. and
advertisement; medical officers from the Food and Drug Administration's
Bureau of Medicine who will testify concerning the approved New Drug Appli-
cation for Salutensin, the certification for this batch of Prostaphlin, the* ap-
proved labeling and the seriou~ nature of the alleged misbranding.
It Is requested that if the form of Information l~ amended, that the United
States Attorney furnish us with a copy thereof, and that we be kept advised
of the progress and disposition of the case. Upon request, we shall render every
further assistance.
Very truly yours,
WILLIAM W. GooDRICH,
A8si8tant Genei~al Counsel,
Food and Drug Division.
MARcH 26, 1969.
Re Bristol-Myers Co.-Alleged Violation of the Federal l~ood, Drug, and Cos-
metic Act, F.D.C. No. 52397
Mr. WILLIAM W. GooDRICH,
Assistant General Counsel, Food, Drug, and Environmental Health Division,
Department of HeaZth~ Education, and Welfare, Washington, D.C.
DEAR Mn. Goomucn: We have carefully considered the request for prosecu-
tion of the above-mentioned company for alleged violations of the Federal Food,
Drug, and Cosmetic Act stemming from its promotional activity in the field of
advertising and labeling for the drugs Salutensin and Prostaphhin in 1965.
Since this mattel- was first referred to us the triul of the Abbott case, while
unsuccessful, estabflshed a judicial precedent for the Government's contention
that the monographs appearIng in the Physicians Desk fleference are labeling.
We understand that the Bristol~M3rers Co. and the industry in general have
accepted this determination and have been more careful to make c~rtain that
PAGENO="0335"
COMPETITIV1i~ PRO13L]~M~ IN TIlE DR~YG INIILYSTEY 135~
the monographs closely correspond to the approved labeling. In addition, as a
result of the Court's interpretation of the phrase "~ubstan'tially the same" a~
used in the former regulation ~21 C.F.R. 1.105) the agency has c1iange~1 the
regulations under which the labeling violatiOn would be l~rought. This, o~
course, is also applicable in the case of the mailing piece, the use of which was
claimed to violate the labeling regul~tions~
Thus, the institution of the criminal action requested would not serve any
useful purpose in obtaining a judicial interpretation of the present labeling
regulations or as to what constitutes labeling,
Any prosecutions based upon the advertisement of Salutensin in the medical
publication for November, 1965, must be predicated upon the regulation which
was in force, at that time. This prosecution would necessarily involve certain
contentions relative to the meaning of the statute and regulations with refer-
ence to practices which are not distinctly set out therein. Inasmuch as the
agency Is in the process of issuing new regulations which will clearly and in
detail inform those subject thereto concerning what information is expected
to be included and what conduct is prohibited, it is our opinion that the agencys
interpretation of the statute can be best advanced by defending any judicial
challenge to the issuance of the regulation, or if none is forthcoming, by bring-
ing either a civil seizure or a criminal prosecution based on acts occurring after
the regulations have become effective.
The institution of previous criminal proceedings has made known to the In-
dustry that this Department intends to make use of criminal sanctions for
enforcement of the labeling and advertising provisions of the statute whenever
it is necessary to do so to prevent the use use of Improper promotional materi-
als. Industry publications contain accounts of instances In which dru~ manu-
facturers have, after consultation with agency rep~esentatlves, circularized the
medical profession calling attention to improper advertisements or mailing
pieces and correcting any misstatements therein.
Representatives of the subject company have stated that It has ma~de an
effort to comply ~vtth the'regulations and intends to do so in the future. They
have protested that a misunderstanding as to what practices constituted a
violation was possible under the regulations then In force and that Bristol-
Myers' violation was occasioned by such a mistake.
Accordingly, we do not' believe that the suggested criminal prosecution Is
required to obtain judicial detei~minatlon as to legal contentions of the agency
or as a necessary aid to enforcement of the Act and regulations. Therefore,
prosecution is declined.
While we stand ready to use any of the enforcement procedures prescribed
by the statute to Insure `that prescrl~tIQn ad~ertlsing flt~ the requirements, of
the statute and the regulatlOn~, candor requires us to observe that `In those
in~1ances where we have secured criminal convictions for thI~ type of violation,
the penalties imposed by the Courts have not been such a~ to provide effective
deterrents. On the other hand,' the technique reCently employed by the~ Com-
missioner of Food and Drugs, about which we have read in the trade media,
appears to, provide, more ass~irance of future compliance on the part of, those
Companies `subject `to the regulations. It seems obvious that an ethical pharma-
ceutical company will regard the so-called ,Dear Doctor iette~ as a more oppro-
brious experience than any other available enforcement measure, particularly
since it has not been the Commissioner's practice to recommend for prosecu-
tion the responsible individuals Of tile offending corporations. We urge that the
Commissioner of Food and Drugs continue to utilize this technique which he has~
stated has proved very effective? while resorting to civil or criminal l1ti~ation
only in respect to those, recalcitrant offenders, if any there be, as to whom
other measures have proved ineffCc,tIve~, If the Government can shOw that
prosecution is resorted to only in such circumstances, the Courts may well
adopt a different attitude in lmposingsentence.
We are closing our file on this matter.
Sincerely,
WILL WILsoN,
AssiBtGnt Attorney Ge~eraZ,
UriminaZ Division.
By ILu~oLn P. SHAPm0,
CMej' Adntin,istrative Reguk~tions Election.
PAGENO="0336"
13586 COMPEI'ITIyE PROBLEMS IN T~IE DRUG INDUSTRY
S JaNUARY 17, 1967.
~n re~y re~r t~, F.~.C. ~ `
~ ~T~Q~1~Y `~N~RAL, :
Dep~ti~nen~ `qf~ Jn$ti~e,~,L. S S
`Wi~~~on, D,U. , , S S S
DE~R MR. ATTORNEY GENERAL: We request the inStitution of criminal `pro-
ceedings in~der th~ `Federal Food, Drug,' aiid Cosmetk~ Act, against Rexar Phar-
~naca1 Corporation, `B~rookLyn, New york, Mr.' Armin Rosner and Mr. Martin
Benjamin. ` S S
The offenses complained of `occurred during the' period from about April 18,
i964, to about October 1, `1965, an~d invoTve the int~odiiction into' interstate corn-
meI~ce at Brook'yn, New YoI'k, for .deThrery' to Teaneck, Fort Lee and Paramus,
New jersey, of Oby-Rex tablets and'~ttme disintegration capsules and Obetrol
tàblet~. "~ `
There kre transmitted'heirewith a suggested form of criminal informatioi and
`the folltwing exhibits ` ` S ,
(1) Copies' of Notice `of Hearing. ` ` ` ` ` `
(2) Copies of bottle labeLs~fo~r Oby~-fl,ex and Obetrol tablets.
(3) Copies of package,' hisert labeling (the approved New Drug Application
hi'belixtg) for Obetrol5 tablets. ` ` ` S
(4) Copy of the advertisement for Obetrol tablets Which appeared in Modern
Medicine for September 13,?1965. S ` S
`SECTIONS ON THE ACT INVOLVED
`~he ~p~çrth ion thai~ges:, fpur violations of 21 U.S~C. 331(d) in that the
f~4ntt~ .eaused4~e iutrod*ction into interstate commerce of new' drugs, the
"Ob~~ ça~sules and t~b1ets,, wb4cb was in violation o~"2i U.S.C. ~55(a), `since
`ho apptoval' of an, application filed, pu-r~ant to 21' U.5C~ `3~5(b) was effective
with resp~ct tq the 4rugs.. ,~ , S , ` S S 5 5
* ~he `Information al~q~al1eges,jn one ~ount that t~efQ~etro~ tablets were mis-
branded within t~e'mehnipg of 21 IJ~S~Q., 352(n) in that the' defendants failed
to include in tl3e advertisement caused to be issued by tben~ with respect to',the
drug in ~Uie September 13, 1965,' issue of `the Mpdern Medicine~ a true statement
of information `in brief swnma~y relating' to shje effects and contraindications
and effectiveness of such drug~ as `require4 by regulations 21 SC~R 1~1O~(e.). and
(f)(2).' ` S ` ` ` S ` ` S
REASONS FOR INVESTIGATIONS S S
On July 24 1959 tb~ ft~m i~ec~ed an appiove~ New Drug Application for
`~ prodijic~ ~coown ,a~ O~etr~I ~aMets in .10 ai~id, 20'z~ffligra~, s1~rezigths. Obeti~ol
cous~sted of a combmat~on.pf sour 4ifferen1~ apipheta~ine s~lts in equ~l ~tr~ng~s
as'fd~bws ~etliampI~et~u~ije ~acc1lamte ~ethaIr1pb~ts~pine Hydrochloride
4fltpbelamine Sulfate DE~rQ ~mpbeta~nme Sulfate
Qbetrol, `was the trade name u~ed wher~ the drug was 5°id to whol~alers
while Oby Rex was the trade name used when the same drug was ~olcl to
doctors. The drug i~ill be referred to á~ "Obêts'ol" in the letter in the inte~'est of
clarit~'. ` `, ` 5 ,, 5
Ii~ 1962, the F9od and D,~ug Administration learned that tb~ firm was dis-
trib"u'ting a'. 30 milligram tab,lOt, and' caj~sule under the ,nam,ë of "Obetrol" `with-
out a, supp'ementaL New Drug Application. lowever, there was no eVidence
that `the firm was sbipp1~g this drug in interstate commerce at that time. On
February 6, 1963~ Food and Drug Administration Inspector Ernest ,Schn*lz
inspeèt~ed this firm and learned that it was still manufacturing the 30 milligram
,Obétrol cáp~ules and tablets, but apparently not shipping the drug in inter~tate
`commerce. Mr. Armin Uosner, tile President of the ~lrm, was asked by the in-
~pector about the 30 milligram dosage form and his i~easOns for not filing a
New Drug Application or a supplemental New Drug Application fo~ these prod-
ucts. Mr. Rosner replied that he was of the opinion that the New Drug Applica-
tion for Obetrol allowed dosage~ up to 60 milligrams per day. He pointed out
that the directions for use for fhe 30 milligram tablets provided that the daily
dosage fell within this range. He, therefore, questioned the need for submitting
a supplemental New Drug Application or a second New Drug Application for
the 30 milligram dosage form. He was advised by Inspector Schmalz to discuss
this matter with the Food and Drug Administration's New Drug Branch in
Washington, D.C.
PAGENO="0337"
CO~4PETITIVE PROBLEMS.. IN TH~ DRUG INDUSThY 13581
On Augusl~, 4 and 5, 1964, Fpq~an4 Drug Administration In~pectors ~Jarl ~,
Lorentzson and Cbar1e~ Thorne ii~speete'd the firmand learned that~ it' wa~ stl~1
znarn~facturing the 30 milligram Qbebroi ~tablet and capsuleand shIpping it'. i~
interstate commerce without an~apprO\red New Drug Apf1b~ation., ~ ,~a
up to this inspection, the Food and Drtig Adniinistration ~o1lected ~
30 milligram Obetrol time disintegration capsules and tablet~ In interState
commerce. (Counts I and U) `. S
Ln the spring of 1964, the firm submitted a Supp1e~en1i~'l New Drug A~pli~a#
tion. At that time, it came to the attention of the Bureau `of Medicine ~tthØ
Food and Drug Administration that the firm was advertising lt~ 10 and 20
milligram Obetrol tablets with the use of ~false and~ misleading claim ~Sbmé
of the objections raised by the Bureau of Medicine to the fiI'In"s adv~rtl~1ug
were as follows: . S S S
(1) It failed to list Obertol's side effects" `` ` `: S
(2) It implied that Obetrol was unique in that it was safer and more e,ffec-5
tive than other amphetamines. Thia `claim `was unsupportable âñd illogièal.
(3) It claimed that Obetrol was effective in "difficult cases," where~, `the
two papers referred to in the advertisement did not demon~trate~ this fact.' In
addition, these two papers contained identical cases, were writtei~ by' the same
authors, but were published in two separate journals. S
(4) It invited the misuse of the drug without proper r~gnrd' fOr patient-
safety by quoting, out of context, in such a way as to conceal the fact that some
patients could not tolerate the drug at all, and others found it necessary to
reduce the dosage, to avoid dangerous side effects. `
(5) It tampered with a direct quote thr~ugb the insertion of a phrase,, a
wrongful act aggravated (a) by' the fact `that `there was' no ififbrmation in
the author's article justifying the idea suggested' in the `inserted phrase; and
(b) by the fact that the tampering invited a dangerous over-çonfi~ience' In the
use of Obetrol in cardiovascular patients in whom the drug was contraindleated.
The quote from the author's article reads as follows, wilh. the Words the
defendants inserted being in brackets: "In the cOoperati~0 ~atithat (Obetro]]
was markedly befleficial in producing the desirable Weight lOss ~vjth mi4tmal
side effect5,'~ven In [the cdsO of a high j~rçentage of $tlebthj With cardthvas-
cular and other chronic ailments `which [r1orthally'1 make use of other a~phetá~
amines lmdesitable because of t1de,eftOcts.'~ `, S * ~, , , ` ~`
On August 4, `1l~64, Mt."Armin Rosner am,~ his átt~rne~ n~et With ye~tesenta~
tives of the 1~ood and Drug Administration and' were advised to discontinue
their' current advertising campaign with respect to Obëtrol tablets., The firm
then advised the representatives o~ the Food and Drug Administratiok that It
manufactured and sold a 30 milligram Obetrol tablet and capsule without an
approved New Drug Application, since it was of the opinion tb~tt the subx4ission
of a New Drug Application was not necessary, as it was selling this' drhg di-
rectly to physicians. S `
In a letter dated August 12, .1964, from Rexar Pharmacal Corporation'to the
~ood,and Drug Administration, the firm promised to discontinue all `advertis-
ing of the Obetrol 10 and 20 mg. tablets and to discontinue the manufa~turè.
of' Obetrol tabletS and capsules `in excess of 20~ milligrams. The firm then pro-
posed a revised labeling as part of, its supplemental New' Drug Application
for Obetrol tablets, The final l~beling for the drug was approve4 by the Food
and Drug Administration on `July 13, 1965.
Shortly after the labeling was approved, the Food and Drug Administration
learned that the firm had once more reinstituted an advertising and prthtto-
tional campaign for Obetrol. The firm placed an advertisement ja the Septem-
ber 13, 1965, issue of Modern Medicine, which did not state in brief summary,
oi at all, those precautions as set forth in the approved New Qrug'A,ppli&Ltion
labeling for the drug, which were, pertinent with respect to the use recommended
or suggested in the advertisement as required by. the regulations 21 CFR
1.105(e) and (f) (2) in that, the, brief s~mm~ry failed to state that the d~ug
should be uSed' with caution in indh'iduals `With anoxeria, insomnia, vasomotor
instability, asthenia, psychopathic perspnality, a ~i~tory of homicidal or nuicidal
tendencies, and Individuals *ho are ktiowii to be hypera~tive to' `sywpàtbom,i~
nieti~ agents or emotionajlyunstalle indivldual$ who are known to be suscept-
ible tO drug abuse,;"and that certalumOhoamilie oxMasq lnh~bitors may ~o~en-
tiate the action of Obetrol. Consequently, the~ rood and Drug Administration
collected a sample of Obetrol (Count V).
The defendants may claim that since the advertisement in Modern Medicine
contained a brief summary of side effects and contrainclications, as set forth
PAGENO="0338"
13588 COMPETITIVE PROBLEMS IN THE D1~UG INDUSThY
in the approved New Drug Application labeling, they have not violated Section
502(n) o~ th~ Act,~ no.r ~egu1ations 21 OFE 1.105(e) and (f) , as these sections
and regulations only require a brief summary of side effects and contraindica-
tions as set forth h~ the approved New Drug Application labeling, no mentloii
being made of precautions.
We. believe that "side effects" and "contraindications" certainly include "pre-
cautions". as this word Is used In the approved New Drug. Application labeling.
Congress clearly' intended that . prescription drtig manufacturers should provide
physicians with adequate warnings in prescription drug advertisements of
those conditions In which use of the drug entailed a high degree of risk. It is
pure sophistry to contend that Congress wanted physicians' to be warned of side
effeet~ gn'd ecintrajudications as `se~ forth in the approved New Drug Applica-
tion labeling, but not to be warned of the "precautions" ~s set forth in the
approved New Drug Application labeling.
On September 29, 1965, Food and Drug Administration Inspector Paul T.
Wiener inspected this firm and learned that It was still manufacturing and
shipping In interstate commerce 30 mg. Obetrol tablets without an effective
New Drug Application. As a follow-up to this inspection, the Food and Drug
Administration collected two samples of Obetrol 80 mg. tablets in interstate
commerce because the article was a new drug without an approved New Drug
Application (Counts III iind IV).
iiisiosr or FIRM ~ND INnIVIDUAL5
This firm flr~t came to th~ attention of the Food and Dr~ig Admiflistration in
early 1953 when the Connecticut State Division of Drugs, Devices, and Cos-
metics referred a sample of dextro amphetamine sulfate to the Food and Drug
Administration's Office because the label declared a fictitious name and address
of a manufacturer. The manufacturer and ~hipper were believed to have actually
been Rexa.r Pbarma~al Corp. As .a result of this complaint, an initial inspection
of the firm was made on March 11, 1953, which disclosed `that the firm was
operating with poor manufacturing contrái conditions. The original sample
which bore the fictitiOus name and address of the manufacturer was assayed
and, found to contain only 72% of labeled amount `of dextro-amphetamine sul-
fate. However, the sample was placed in perpianent abeyance because it could
not be dofjnjtely"ascertained that the sub$ct firm was the manufacturer. and
distributor of these amphetamine tablets.
The firm u as again inspected on December 15 1953 at which time according
to Mr Armin Rosner the firm ~ as doing a minimal business The firm was not
then shipping its drugsin interstate commerce.
The firm was inspected once more on February 15, 195~, at which time it was
learned that the firm was manufacturing Obetrol. a new drug, without an
eff cclii e New Drug Application and shipping the drug while using false and
misleading claims. As a follow-up to this inspection, a sample was collected
which resulted in a Hearing on September 28, 1956W The sample was placed in
permanent abeyance, however, because the firm agreed: (1) to revise its label-
ing in an eftort to bring the drug in compliance with the law, and (2) to submit
N~w Drug Application for the 10 and 20 mg. Obetrol tablets.
The fli~m was inspected on .~Tuly 25. 1955, at whjch time it was learned that
tia firm was shipping another new drpg namely Obertina tablets a combine
tion of amphetamine and rauwolfia serpentine without an effective New Drug
l~pplication Nowever the Food and Drug Administration could not obtain a
s~trnple of the product in interstate eon~mpr.ce and. no aetlon was bi~ought.
A follow-up Inspection, made qn July 8, 1958, showed that the firm was mark-
ing time while its New Drug. Application, for Obetrol was under study by the
Food and ~rug. Adminictration. `
The firm was again issued a Notice of He~r1ng in early 1963 because it had
shipped a new drug' consisting of thyroid and 30 mg.. of amphetamine salts
withou1~ an effective New Drug `Application, At the time of the' Hearing, the
respondent stated that a New Drug Application was unnecessary as the drug
was shipped under, what the firm termed. a physician-pharmacist relationship..
The firm had made this. drug to. order, for one, physician. The number was
placed rn permapent abeyance It was Its position that Rexar Pharmacal Corp
was simply asked to fill a prescriptI~n for the physici~~i~ No further action was
taken because' th.e firm agreed to discbhtlnue the interst~te distribution of this
product.. . ` ` *. . .. . ` .
PAGENO="0339"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13589
On February 6, 1963, the firm was again inspected and the Food and Drug
Administration learned that the firm was distributing Obetrol tablets iii viola-
tion of Its New Drug Application. A seizure was made of this drug in Los
Angeles, California.
EVIDENCE OF VIOLATIVE SHIPMENT
Counts I and II
This sample was collected on August 8, 1964, from J. Raymond McSplrit D.O.,
703 Cedar Lane, Teaneck, New Jersey, by Inspector Frederick T. Merola. The
sample was identified by Dr. MeSpirit, who furnished the Inspector with an
invoice showing that thO drugs were shipped by Rexar Pharmacal Corp. from
Brooklyn, New York on or about April 18, 1964.
Counts III and IV
This sample was collecled on October 5, 1965, from Dafliel Reider, D.O~, 2361
Lemoine Avenue, Fort Lee, New Jersey, by Inspector Paul T. Wiener. The
samph~ was identified by Dr. Reider, who furnished the Inspector with an
invoice showing that the drugs were shipped by Rexar Pharmacal Corp. from
Brooklyn, New York, on or about March 1, 1965.
Count V
rlhis sample was collected on December 15, 1965, from D. Katzm'an & Co., 670
Winters Avenue, Paramus, New Jersey, by Inspector Alfred M. Levy. The sample
was identified by Mr. Stanley Szewczyck, a buyer for the firm, who furnished
the Inspector with an invoice showing that the drug had been shipped by Rexar
Pharmacal Corp. from Brooklyn, New York, on or about October 1, 1965. On
December `29, 1965, Inspector Levy obtained from Richard P. Keating, M.D.,
130 Prospect Street, Ridgewood, New Jersey, the September 13, 1965, issue of
Modern Medicine, together with an appropriate affidavit signed by Dr. I~eating.
RESPONSIBILITY OF INDIVIDUALS
Both Mr. Armin Rosner,' President, and Mr. Martin Benjamin, Vice President,
share equally the responsibility for the conduct of the firm. During the Inspection
of February 6, 1963, the Inspector observed that both men were familiar with
the manufacturing and control procedures. They told the inspector they shared
equally In the responsibility for making major decisions in the firm's opera-
tions. During the inspection of August 3, 4 and 5, 1964, the Inspectors obtained
information and specimens of the advertisements of Obetrol tablets from Mr.
Martin Benjamin. At that time, Mr. Arinin Rosner told the inspectors that he
considered the responsibility for operations of the fitm, Including labeling, ship-
ping and sanitation, to be a joint onQ between himself and Mr. Benjamin. Both
men, the Inspector noted, knew exactly what was going on in the operatlob of
this business.
During the Inspection of September 29, 1965, Mr. Benjamin told the Inspector
that Mr. Rosner was in charge of plant sanitation, shipping and manufacturing
operations and that Mr. Benjamin was active In sales and promotion.
In the last three to four years, botl~ Mr. Rosner and Mr. Benjamin have dealt
with the Inspectors either singly or together during each of the inspections
made by the Food and Drug Administration. The firm Is small enough so that
each of the two men was aware of what was going' on in the firm. In addition,
they jointly set co~ipany policy and were equally responsible for the labeling
and the advertising for the drugs which were, shipped in interstate commerce.
HEARING HELD PURSUANT TO 21 U.S.C~ 335
A Notice `of Hearing was issued to Rexar Pharmacal Corp., Mr. Arwin Rosner
and Mr Martin Ben)amln on January 10 1966 charging the shipment in inter
state commerce of a New Drug, the 30 mg. dosage form of `Obetrol, without an
approved New Drug Application. It was also alleged that the firm had mis.~
branded its Obetrol `10 mg~ tablets because of medical journal advertising
which did not state Ill brief summary, or at all, certain side effects and con-
traindicatlons as set forth in its approved New Drug Application labeling for
the drug. Mr. Rosner and Mr. Benjamin, together with their attorney and con-
sultants, appeared at the Hearing.
The respondents first addressed themselves to the charge concerning the omis-
sion of a precautionary statement In the advertising for the drug, Obetrol. They
claimed that they `were guided by a press release issued by the Food and Drug
PAGENO="0340"
i!3590 COMPETmvE ?ROEL~M~ IN ~ DEUG INDUSTRY
-Administratiop on No'~ember 23, 19~4, pertaining' to ~he info~matiót~ required
in a brief stimmary. They also stated that the~ wei~e guided by other firms'
advertising for similar drug prodhcta In additio~, they f~lt that the' context
of the advertisements was geared for physicians, and since the doctors them-
selves received other material, such~ as brOchures, the doctors undoubtedly
v ould have some realization of the precautionary guides to be observed. How-
ever, they admitted that they were ~vrong in. omitting the. premutionary state-
`aents, and. they, assured' the Food a'nd Drug Admiinstratlon that this error
bad been corrected in the current advertising for the drug.
The firm knew full well that the gravamen of the regulations was pertaining
to prescription drug advertising. it had been warned In 1964 tha1~ its drug ad-
vertising for this `product was false and misleading. It had di~cu~sed Vith the
Food and Drug Administration the advertising drug regulations and bow to
correct its advertising. Despite this information, the `firm blatently advertised
this drug in the summer of 1965 without giving full inforinatioti as required
by the regulations. The precau'tionar~ statements, which the firm omitted,. con-
tained the moSt important information about the cautions to be observed in
patients with anorexia, insomnia, vasomotor instability, asthenla, psychopathic
personality, a history of homicidal or suicidal tendencies, and individuals who
are known to be hyperactive to sympathomimetic agents or emotionally un-
stable individuals who are known to be si~sceptible tp drug abuse; and that
certain monOamine oxidasé inhthlto,~s rn~y potentiate ~`the action of Obetrol.
With respect to the interstate shiptheñts of the Obetrol 30 mg. ,tab1et~. and
capsules, the firm stated that, this drug was sold by the firm siuc.e 19~. Hence,
the firm was `under the impression that the drug fell, within, the scope of the
`grandfather ~lause" and a New l5rug Application was `pot nece~sary.,
UnfortwiatelV for the de±endants, the "grandfather clause" (Pub. L. 87-781,
Section 107) requires `that the drug be generally recognised, as of October 9,
1962, as safe when used for the purposes intended. This drug was not so recog-
nized on that date.
The second point the firm made at this Hearing was that the' drugs were
sold directly .to p~ysici~ns and,not through regular commercial channels. This,
it said, was an ordinary `,`~hysi~ian-~har~naej5t" relationship, whereby the firm
was simply filling a, pr~,scription for the physicianS The defense' fails to explain
how .an ord?r i,nv~lvi~g some 3,QOQ tablets sold' to one physician and a 1,000
tablet order to a second physician is' the same as prescribing for an individual
patient as j~ u,~ually done In the. pbysicjan's~ daily. practice of medicine. Another
objection to the defendant's defense is that~ when~ there is a violatl~n of 21
U.S.C. 355(a), there is no exemption for any so-called "Physician-Pharmacist"
relationship, As you recall, the ~1rst four counts charge a violation of 21 U.S.C.
855(a)~ The firm had been previously advised that this conception of theirs
was wrong during the Hearing held in early 1963
The third point the firm made was that the 30 itig. tablets came within the
limits of the, 60 mg. daily dosage requirements under their New Drug Applica-
tion for the firm's 10 and 20 mg. Obetrol tablets. This argument had no merit
since the ,New Drug Application provided for the marketing of a `specific formu-
lation of the drug with specific labeling. The formulation and labeling for these
30 mg. preparations differed from that provided for by the New Drug Appli-
cation, hence, these preparations were not covered by the New Drug Applica.
tio~L. ` `
Lastly, the respondents stated that they were currently preparing to submit
~t New Drug Application for th'eir 80 mg. Obetrol tablets and capsules. This
Application is still pending. The firm was well aware of our position with re-
spect to the status of this drug. It was advised In August, 1964, that the Food
and Drug Adminfstratlon considered these preparations to be New Drugs with-
out an effective N~w Drug Application and was tOld that the Food and Drug
Administration could not condone the marketing of this drug in interstate
commerce without an approved New Drug Application. Yet, the firm chose to
continue the sale of this drug without an approved New t'rug Application. It
is noteworthy that, at the time of the Hearing, the firm indi~ated, thai it bad
ceased distributing this drug in Interstate commerce, but that it was ~till sell-
ing this drug in intrastate commOi~ce.
CONCLUSIONS
Warnings at Adrnini~trative Hearings and at meetings with the Bureau of
~tedicine of the Food and Drug Administration have gone unheeded by this
PAGENO="0341"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 135ki
firm. It had been advise~ as far back as 1958 that its 80 xr*g. Obetrol tablet
and capsules were NeW Drugs. t~espite this warning, the firm still faile~d to flu
a NeW t~rug Application for thes~ cI~ugs and continued td ship them in inter~
state commerce Without an apprb~ed NCw ~~rug Application. Sjmila~ly, the firn~
was advised in early i~64 that its advertising for its Obet~,ol 10 and 20 mg. tab-
lets was false aild misleading and that it could only advertise the drugs through
the use of claims set forth in the labelitig approve~1 in the firm's New Drug
Application. While It is true the firm discontinued advertising its 10 and 20
mg. Obetrol tablets until the new labeling for these drugs was approved in the
summer of `1965, the firm started to' il1~gal1y advertise again as soon as the
final printed labeling had been approved. T~h1s time the firm omitted the pre-
cautionary~ statements wl~ich relate to the side effects, contraiiidications, and
effectiveness required to be 1~air1y ~resente~l iii the advertising by the regulations
under the drug advertising sectioh of the law. It is onr opinion that prosecution
~f the firm and the two responsible individuals i~ fully warranted.
WITNESSES
rp~~ principal withesses in this case will be the Government inspectors who
collected the samp1e~; cooperating physicians who subscribe to Modern Mcdi-
cine and' medical officers of the Food and Drug Administration's Bureau of
Medicine who can testify as to the approved New-Drug appiicatious, approved
labeling, and the serious hature of the alleged medical journal advertising
misbranding.
It is requested that, if the InfOrmation is amended, the United States Attorney
Lurnish us with a copy thereof; also, that the United States Attorney keep us
advised of the progress of the case and its disposition. The New York Djstrict
of the Food and Drug `Administration will arrange for the presence of the
necessary' witnesses and assist in the presentation of the case. Upon request,
we shall render such further assistance a~, may be possible.
Very truly yours,
WILLIAM W. Goormicu,
Assistant General Counsel,
Food and Drug Division.
`U'N~TED STATES DEPARTMaNT~ OF JuSTICE,
Washington, D.C., April 24, 1967.
lIe Rexar Pharmacal Corp., Arthin Rosner, and Martin Benjamin, FDC No.
58053, Federal Food, Drug, and Cosmetic Act. ,
DEAR Mu. G00DRICII This Is in reply to your letter to the Attorney General
of January 17, 1967, in which you request the institution of criminal proceedings
against `the `above~captioned subjeets for viOlations of the 1~ederal FOod, Drug,
and Cosmetic Act committed' between Api~l1, 1964, and October, 1965.
We have carefully reviewed the staten~ients set out in your letter Concerning
the subjOcts' activities telative `to' the sale `of certain chug' products.
As to the fifth count, we are not disposed toward the `conclusion expressed in
`your letter that the words of the statute; also used in your regulations, r~quirjng
a statement `of ôontra4ndicatl'ons, side eiT~cts, and effectiveness `in prescription
drug advertisements are~ so dearly Inclusive of "precautions" a~ to~ `ive the `sub-
jects "fair warning" flint su~h items must be included. We have observed
that in the advertisement in Modern Medicine in September, 1965, the $~bjects
included,' nuder the statutory headings,' the' full and exact language f~ind
under those headings in the labelin~ approved `by the ~oód and' fl'rt~g Admial-
stratton. There Is nothing in the Act' or `the regulations to indicate that the
words `therein have a larger meaning than that of the approved labeling.
Accordingly, we' are of the opinion that the advertisement to wl~icb reference
was made in `your letter is not violative of the Act. MoreOver, we do not
`believe that the factual `situation here `Is' such' that the Government would be
* able to prevail in the event the theory suggested In your letter were `to be tried
out in a criminal prosecution of ~Re~ar and its officers. Therefore, prosecution
is declined as to the charges set out In Count V of your suggested informatJpi~i.
Inasmuch as the violations o~ April 18, 1964, and, March 1, 1965, were not
reported to u~ for criminal proseentien until `after the appearance of the adver-
tisement in Moder~.i Medicine in September, 1965, we `are uncertain as to whether
you are of the opinion that prosecution Is merited on the basis of those acts
PAGENO="0342"
13592 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
alone. Since your letter leaves the impression (p. 9) tl~at the subjects have
flied a supplemental new drug application covering the 30 mg dosage form for
"Obetrol" and you do not indicate whether or not the application has been
rejected, it would appear that In this respect compliance with the Act may be
deemed to have been achieved.
Therefore, we are withholding further action with regard to the violations
of April 18, 1964, and March 1, 1965, in order that you may inform us as to
whether you believe prosecution for these violations should be instituted. In
view of the factual situations outlined in your letter we are particularly inter-
ested in being informed as to the status of any supplemental application for
the 30 mg dosage form. You will undoubtedly appreciate the force of an argu.
merit to a jury that physicians could have prescribed two 80 nig doses per day
which would have been within the allowable lifitatlon of the approved labeling
and which wotild have been available to patients by `taking one 20 mg and one
10 mg tablet.
If there is any sound medical reason why such dosage should not be prescribed
and thus why 30 mg tablets should not be available to physicians, we would
be able to counteract any such defense with some force. Whether or not such
reason exists is therefore a factor to which we would attach considerable
Importance in the event prosecution is' requested as to the foregoing violations.
We also believe that it would be helpful to know `whether the subjects In any
way solicited the sale of the 30 mg tablets on April 18, 1964, and March 1,
1965, to Doctors McSpirit and Reider respectively, or whether such sales were
solely as a result of unsolicited orders by the purchasers.
Sincerely,
FRED M. VINS0N, Jr.,
As8l8tant Attorney General,
Criminal Divi~lon.
By ITAROLD P. SHAPIRO,
Chief, Admini8trative Regulatione Section.
DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE,
September 26, 1967.
Attention Harold P. Shapiro, Chief. Administrative Regulations Section.
Re Rexar Pharmacal Corp., Armin Rosner, and Martin Benjamin. Your ref:
FMV :JWK :mlh, 21-52-246. F.D.C. No. 58053.
Hon. F~no M. VIN5ON, Jr.,
Assistant Attorney General, CrinUnal Diiiision, Department of Justice, Wash-
ington, D.C.
DEAR Mn. VIN5ON: We have considered with the Food and t~rug Administra..
tion the questions `that you raised concerning this case.
The four counts which allege violations of 21 U.S.C. 331(d) are strbng
counts involving the distribution of iinapproved new drugs, and would by them-
selves, support criminal action.
The firm did submit a supplemental new drug application for Oby.Rex 80
mg., but it was iCcomplete and the firm has been advised of this. Therefore,
compliance has not been achieved. Moreover, the firm was advised In 1964 both
to discontinue their violative advertising and that a 80 mg. tablet would require
a new drug application.
It Is irrelevant that a physician could have prescribed a 20 mg~ and a' 10, mg.
tablet, thereby: giving his ,patient a total of 30 mg. at each dose. This does not
make It jeg~lfor the firm to market a 30 mg. dose without complying with
the new drug requirements. Had' a doctor ~o prescribed, be wOuld have ezceeded
th~ ,~Ihn1ts of the `sa~ety approval in the new drug application. While he may,
In `h:ls, discretion, prescribe an excessive dose for~ his oivn patient, he does so at
the rjsk of civil liability for exceeding the dosage that has been proved safe
as required by law The fact is that the consensus of medical opinion holds that
the dosage of ~0 mg. per tablet is riot generally recognized an safe. To say that
a person can take 20 5-grain aspirins at one time Is not to say that it is per-
missible to n~ak~ a 1.00-grain aspirin tablet,' * .~
The Administration's file does not reflect tliat the ~rm has detail men or uses
other means for, the direct solicitation of orders:from physicians.
PAGENO="0343"
COMPETITIVE PROBLEMS IN THE DRTYO INDUSTRY 13~8
In regard to the count based on the advertisement in Modern Medicine, Sep-
tember, 1965, we would ask you to reconsider yOur decision not to prosecute with
the following considerations,
We have previously written, yo~ our views as to the legal necessity f~r
including in the brief summary, information from the package insert titlØd
there "warnings" and "precautions". [United States v. Wyeth Laboratories,
F.D.C. 52673 Your ref: FMV :JNK :mfs. 21-62-326 22A-48-201 The law ré-
quires that every ad for a prescription drug shall present in brief sununar~'
form "such other information * * * relating t~ side effects, contraindications,
and effectiveness as shall be prescribed in regulations."
The regulations require that this summary shall fairly show the e~ect1ve-
ness of the drug in the conditions for which it is recommended, * * * togethe~
ss ith those side effects and contramdications that are pertinent with respect
to the uses suggested in the. advertisement and any other. use or uses for whicl~
the dosage form advertised is commonly prescribed. When the drug is an ap~
proved new drug this information shall be the information from the approve~I
new drug applicatIon.
Here, the information omitted, though headed precautions in the approved
labeling, was information relating to side effects and contraindications. It
warned against use of the drug by persons with anorexia~ insomnia, vasomotor
instability and other conditions. This was to tell the physician that such per-
sons might experience serious side effects-so serious that the drug should be
avoided. While the drug may not be compl~tely contraindicated for all persons
with these symptoms, it is contraindicated in some,, and in others, side effects
are to be expected.
So literally this information was related and pertinent to side effects and con-
traindications in the medical sense.
Far more important, however, is the fact that Congress used "side effects"
and "contraindications" in 502(n) to cover the relevant hazards-whether so
denominated in the labeling so-called precautions or warnings.
The legislative history on this point clearly shows that Congress intended
the physician to be fully informed by those ads, since they recognized that the
majority of doctors learn about drl~gs from such promotions. This has been
discussed in greater detail ~n our previous letter. For example, the provision,
that ads could be exempted from all of the requirements of informing the
physician, if they included a statement that full information could be. obtained
on request, was rejected on the fiqor of the House. Cong. Rec. House 87th. Cong.,
2d Sess., September 27, 1962, pp. 19928-9. The information to be supplied would
he the full disclosure inserts which were required even then to include "precan-
t~ons". The sponsor of this amendment, Representative Blatnik, specifically
stated that~ "There Is ample evidence to demonstrate that because of time limi-
tations physicians rely on drug advertisements a great deal in deciding which
they should prescribe for their patients. It is incumbent upon us t~ assure that
no false or misleading information is inadvertently relied upon.", . p. 19922, and
then, "I want to . see this bill so drawn as to make sure that doctors get all
available information * * *" [in the advertisement itself], p~ 19923. To leave
a doctor with the impression that no special consideration or care need be given
to a patient with anorexia, etc. is certainly misleading.
President Kennedy's proposed amendment on advertising to the $enate Bill
was offered on the ground that "Such advertisements should be required to
make fair disclosure to physicians of the information (good oI~ bad) needed to
permit them to do a better ~ob of selecting drugs for use in tbe~r practice
Letter from President I~ennedy to Senator Eastiand, August 4, 1992,. reported
Cong. Rec. Senate, August 6, 1962, p. 14682. This would, in our opinion, include
precautions. .
We think that a reading of `the Legislative History on this matter leaves no
doubt that the information lacking in this ad was of the sort that Congress
intended be'pmsented to the physician.
If we may be of any further assistance, please do not hesitate to call upon
us.
Sincerely yours,
WILLIAM W. Goonnxou,
48sistant Genera' Coun8eZ,
Food and Drug D~vlslon.
PAGENO="0344"
13594 `COMPET1TIV~ ~ROBLEMS I~ TIlE ;DRtG t~ThUSThY
OcroPE~ 25, 1967.
Re Re~arPharma~aj Corp., FDO!No.~580~3.
Mr. WII~rAM W. GOODRICH,
Aesi$tant Genera~ Counsel, DepartmetU of Heait1~, `Education, anI Welfare,
~Wa~1vinØon, D.C.
DRAR 1~ln, GOODRICH: This iS in response to your letter of September 28, 1967,
`in which you answered certain questions which we ,raised in our letter to you
dated `Apr~1 24, 1967. The views ~Ou express concerning the subject's advertise-
ment of the drug Oby-Rex in' Modern Medicine for September, 1965, have been
carefully Considered. However, s~e have observed that, although the statute
~nd the regulations ~pertaihing to advertising refer only to "side effects" and
~ontraindications", other re~i1ations pertainitig to the labeling of ~prescrip-
tion `drugs. refer to "i~eIevant `baz~ttds,' cOhtralndicatlôhs, side effects, and pre-
cai~tlOns'~ (foi' examiile, see; `21 C.F.R. ~ectiOn 1.106(b) (~) (i); 1.10~6(b) (4) (i);
130.4(c) ;~i30~9(d) (`I); 130.11(á) (1) and (3)),: Thus, ,`whCre the' agency `has
`desired. tc1~req~iire inclusion of informatiOn hi the labeling of a prescription
drug of "hazards" or "precautions", it has done so by using thoSe specific wOrds
even though in' the same sentenCe it has required the inclusion of "side effects"
and "contraindicatlOns." Accordingly, It appears that the agency has recognized
that ~acIi ~of these `words has a separate definite meaning ahd we could not
,s~s~iit a eo~tentiOn that "j~lrecautions" are inCibded by implication in the
statutory language, particularly In an instance in which the labeling of the
drug as approved by `the `agency distinguished between what constitutes "side
effects and contraindlcations" and "p~eeautions."
Therefore, prosecution on the jasi~ of th'e failure of the subjects to include
in the advertisement a statement of the "precautions" is declined.
Insofar g~ the counts pertaining to the shipment; on April 18, 1964, and
March 1, 1965, to Doctors ~tcSpjrit and Reider, according to their request, of
~o mg strength' tablets of' Oby-Rev and Oby-Rev M tablatC a~e coffcerned, we
are not persu~~ded `that such `drastic action as ctiminal prosecution is required,
~r that'if it were instituted a successful rOsttlt could be obtained.
`The offenses took place between two and one half and three and ` one half
ye~ra~agt~bi~'t were not reported until January of this year. Despite your state-
ment that compliance has not been achieved, no additional violations have been
repQrted and the stthjects ha~e filed a supplemental new `drug application
requestli~g approval for the 30 nig dosage form. Although this application has
bcCh cbatacte~ized us "incomplete" it has not been with'drawn nor has the
ag4~ne~'táken á~iy steps lOoking toward a refusal to' approve it. App~rently the
subjeè,ts `âa~e attempting ,to com~1et'é the ap~lication. `Thus, `despite your sthte~
p~ebt aS to~h'e' cOfisensus of ni~!iieai' Opinion, the reCord befOre the agency is
~n~b `a~te:leave op~n `the question `as to, whether suCh dosage' form is proper:
I~i addition, the faCts that under the' approv~d'.labelir~g a, dosage totaling 60
mg pcr~ day~may ev~rItuaily `b~ ~eached in lhe n~mlnistràtton o~ this drñg and
that `the dosage form of `the `shipment was ~~~ciilcafly Oi~dered by' licensed
phyatcialis wilo are legally entitled and ethically required to prescribe for their
p~ttOi~t~ `the dosa*e dictated b~ their `own judgment, based fipOn `a knowledge
of `thCfr patients' requirethei~ `Will' te~id to weaken any ctiminal prosecution.
Accordingly, it is ~çur view that `the time' which ha~ eiap~ed `since the COm-
mission' of the ~1olati'on and' the factual' Situation `make thi~ an unattractive
c~i~ for i~xdt~al action. Moreover, shice the subjects have not sold `this dosage
India hiulua1~éiy~in `fñte±state cOmmeree'and `have filed 4 `snp~lemental applica-
tioii ~ovcr~g~it~use,' It would ~ee~ prosecution fs net tueeessaz.~r to `enforce corn-
~lignee wtth~tbe re~nlrepaetit ef t1~e, law; ` `, `
thus, prosecution is de~l1ned~ "
Sincerely, , , " ~` `` , P~ED ~f. VzNso~, Jr.,'
4ssistánt~AttOr~iey' General,
- , Crim*iai Dioi8ion.
By IIA1iOLD P. SItAPIRO,
Chief, Administrative Regulatjons Section.
PAGENO="0345"
COMPETITIVE PROBLEMS IN T~E I)RUG INDUSTRY 13~9~5
DEPARTMr~T O1~ HRALVI, EDtCATION, ~ND WELrARu,
4prU 18,1968.
Attention JIarold P~ Shapiro, Chief, Administrative Regulations Section.
Re Rexar Pbarmacal~ ,Corp~, Your Ref. ~MV':JWK :adj 2J--52-~246, J?DC ~ó.
53053.
lIon. FRED M. VIN50N, Jr.,
Assistant Attorney General, Criminal Division, TJ.$. Department 01 Justice,
Washington, D.C.
DEAR Sin: You will recall that Mr. Barrett and I met with Mr. Shapiro an~
Mr. Murphy last November to discuss your reasons for declining prosecution~
set forth in your letter of October 25, 1967.
We went over the file together and we pointed out that the new drug charge
was a serious One because (1) the product was labeled with no warning Infor-
mation at all (no package insert was included) and with dose recommendation
of "one table once to twice daily", (2) the strength o~ this tablet (30 mg.)
which contains methamphetamine (speed) along with amphetamine makes It' a
dangerous product, (3) our medical advisors supplied a memorandum describ-
ing adverse reactions to 30 mg. dose~ of ai~iphetamine, and (~l) the tw~ sales
involved were not isolated transactions.
We asked our New York Office for further information. They have sent us an
Inspection repOrt covering an inspection in September 1967 and a list of inter-
state shipments of this product made iii June 1965 (the month the violation
occurred). There were four additional shipments made that month, two ~to con-
signees in Ohio, one to a purchaser in Mas~achusetts and One to Long Beach,
California. Each shipment involved a bottle or bottles of 100 tablets or capsules.
Our letter of September 28, 1967, said that the firm does not have detailmen.
The file shows that Mr. Benjamin personally delivered one shipment as a sales-
man,
You asked whether the fii~m had in fact discontinued the 30 mg. dosage. The
1967 report shows that the firm claims to have discontinued interstate ship.
meats but continues to do about $20,000 per year in intrastate business in prod-
ucts without approved new drug applications, including Oby Rex iti a 3O mg.,
Timed Disintegration Capsule and Tablet' fOrm. The firm was also using a
"Dear Doctor" promotional letter with physician samples to elevate `mood and
and help relieve despondency, claims not approved in the new drug'labeling, for
those products and not included in the prescribing Information.
It seems, therefore, that the firm despite several warnings has not yet decided
to comply with th~ new drug requirements of the law. The pro~osed defendants
seem to have little concern' for p~itient safety or little appreciation of the hazL
ards in the use,of'high doses of amphetamines. `
"As to the advertising charge; we think there can be no doubt whatever that
the information the Company headed "precautions" jn' its labeling Was informa-
tion "i'eiated to~ side efeecta and contraindications". Congi'ess uSed differetit
language' in Section `502(n~~ than the Agency usOd in its ftill disclosure labeling
reg'ulatious promulgate~1' several years befote, but the `intent of' `the Congress
was plain, as we have' shown `ib our' discussion of these ádvertisin~ cases' with
sour people. ` ` ` ` . `
We simp1~ cannot agree that any ~nateri~tl n~w included in' product brochureS
apprOved' o~er the past 28 years in' new drug clearances that is headed "precau-
tions" or "warnings", `or indeed' anything other than material `headed' ~`side
effects" or "contrai'ndicatiOns",' is not' information related to "side `effects" t~nd
"contralndications"~ as `tbos~ terms `were `used in the 1962 Drug Amendments.
Not only is this directlrcdntlary to the Congressional `intent and' to' the plain
statutory language, it is untrue as a factual matter. In this case, the `precau-
tionary information left' out of the ad was the most important information
needed for safe `use of the drug.
Your dOclination of the new drug charge was based on your appraisal of
the l~kellhood of success, the drastic nature of the proposed action, the time
lag between the' violations and repOrt~ng them to you, your understanding of
the "incomplete status" of the NDA, and on the startling statement that a
physician is ethically and legally bound to treat patients with drug dosages
beyond the limits of safety proven through the new drug procedures.
This action is not drastic when one considers that the' drug ft~volved is a
potent amphetamine mixture, with well recognized hazards, which 4espite some
years of clinical use in medicine has not beèti shown or recognized' to be safe
in the 30 mg: dosage. To place suck an unp~Oved dosage on the market In utter
disregard fo1~ the new drug procedure is a substantial violation of law and a
distinct disservice to physicians who prescribe and patients who are expected
56-592--75-23
PAGENO="0346"
13596 COMPETITIVE PROBLEMS IN TIlE DRVG INDUSTRY
to use prescription drugs. The Incomplete application was rejected by the incom-
plete letter.' Our regulations, upheld on this point by the Sixth Circuit `in the
Turkel case, call fOr filing over protest if the Cornpan~ wishes to pursue its
application. This Gompany `has not done so `and lt& application 15 a closed one.
Exceeding approved dosage of drugs Is one of the surest ways to patient
injury. Physicians have neither the legal nor the ethical right to experiment
with their patients with excessive dosages without the patients informed con~
sent, and when drugs of interstate origin are used, without complianCe with the
IWD regulations.
Very truly yours,
WILLIAM W, GOODRICH,
Assistant General Counsel,
Food and Drug Division.
~ 29, 1968.
Re Rexar Pharmacal Cprp., FDC No. 58058.
Mr. WILLIAM W. GOODRICH,
Assistant General Counsel,
Department of Health, Education, and Welfare,
Washington, D.C.
DEAR Mn. GOODRICH.: This is to acknowledge receipt of your letter of AprIl 18,
1968, requesting us to reconsider our decision* to dedlln6 prosecution in this
matter.
Your comments relative to the alleged violation of the advertising provisions
o1 the Federal Food, Drug, and Cosmetic Act appear to be the same as those
made in at least one other similar matter. (See Syntex Laboratories, Inc., FDC
No. 53222.) Our views regarding the narrow issue presented by similar facts
were fully set forth in our letters to you pertaining to that matter and in our
letter of October 25, 1967, concerning the present submission. We adhere to them
and to our decision not to prosecute this matter Insofar as the advertisement is
concerned.
Neither does the Information contained in your letter warrant any change In
our decision not to prosecute on the basis of the" shipments to `Dr. ~1cSpirit In
AprIl 1964, and Dr. Reider in March of 1965. Despite the statement that there
were four additional shipments in ,June 1965, no evidence has been submitted
proving that, the sales were to persons other than duly qualified physicians
who ordered the larger dosage `form for the purpose of dispensing them to bona
fide patient~ being treated by them. The sales of the drug In New York State
do not constitute violations of the Federal law and cannot be construed as
exhibiting an intent to violate the. Federal Food, Drug, and Cosmetic Act.
Accordingly, we are still of the opinion that the age of the alleged violations
and the factual situation present a very poor basis for criminal prosecution.
Ti~e facts are such as to raise grave doubts as to whether they constitute a
viQlatlon In the legal sense.. Moreover, It Is our belief that the circumstances
upon which the Government *ouid be compelled to rely are so unappealing to
a judge a~id a jury as to render a cOnviction highly unlikely.'
We desire to correct the statement attributed to us In your letter that `a
physician is ethically' apd, legally bOund to treat patients, wib drug dosages
beyond the limits of safety proven thtongh new drug procedures" That Is not
our position. The Correct statement of Our position, to ~Vhich we adhere~ appears
In the next to the last paragraph of our letter of October' 25, 1967.
For the reasons `Indicated abOve, we ~continue . to `be of the view that prOsecu-
tion Qf this ease Is not feasible. We are therefore closing Our file.
Sincerely, ` , `
Fnnn, M.. Vixsox, ,Tr.,
A8sistant Attorney' General,
Criminal Division.
`By HAROLD P. `SHAPIRO,
Chsef AdnHntstratsve RegulatiQna ~ectsOn
MAuCE 16, 1967.:
~n reply refer to P\D~C.~No. ~3222; ` `~ ,, ,.,
The I~ouoràb1e'Arvonxxv GENERAt~,. ` ,~ ,` $
~epartment of ~1ustice, WasbingtQn, D.C. $ `,., / , ,
D~AR Mn. ATTORNEY OENERA~: `We `request the IñstltÜtlon of, criminal, pro-
ceeth~gs in the District of New Jersey under the Federal Food Drug and
PAGENO="0347"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRt 13597
Cosmetic Act, against ~yntex Laboratories, Inc., 1401 Hiliview Drive, t'tilo Altq,
`California, and 45 Walnut AV me, Clark; New Jer~ey. The offenses complained
of occurred on or about September 29, 1965, November 18, 1965, and February ~8,
1966, and involve the introduction Into interstate commerce at Clark, N~w
,Ter5ey, for delivery to Berkeley, California, of quantities of Norinyl, a prescr~p-
tion drug.
There are transmitted herewith a suggested form of criminal Information a)id
the following exhibits:
A. Copy of Notice of Hearing.
B. Carton and bottle labels.
C. Package insert (approved new drug application labeling).
D. Copy of monograph in 1965 Edition of Physicians' Desk Reference.
* B. Copy of advertisements in the November 1, 1965, November 15, 1965, ai~d
February 1, 1966, Editions of The American Journal of Obstetrics and Gyne-
cology; the November, 1965, and February, 1966, Editions of Obstetrics and
Gynecology; and the February 14, 1966, Edition of Modern Medicine.
SECTIONS OF ACT INVOLVED
The Information charges violation of 21 U.S.C. 331(a) in that the defendant
caused the Introduction into interstate commerce of quantities of Norinyi
which were misbranded as hereinafter described.
Count 1.-Time drug was misbranded within the meaning of 21 U.S.C. 352(f)~
(1) in that its labeling failed to bear adequate directions for use and It was
not exempt from such requirement since its labeling, a monograph in the 196~
Edition of the Physicians' Desk Reference, failed to comply with the require~
ments of regulations 21 CFR 1.106(b) (4) (i).
Counts II an~1 111.-The drug was misbranded Within the meaning of 21 U.S.C.
352(n) in that the defendant failed to include In advertisements caused to be
issued by it in the (Count II) November 1 and 15, 1965, Edition of The Amer-
ican Journal of Obstetrics and Gynecology and the November, 1965, Edition of
Obstetrics and Gynecology, and in the (Count III) February 1, 1966, EditIon
of The American Journal of Obstetrics and Gynecology, the February, `1966,
Edition of Obstetrics and Gynecology, and in the February 14, 1966, EditIon of
Modern Medicine, a true statement of information in brief summary relating
to the side effects and contraindicatiOns of the drug as required by regulations
21 CFR 1.105(e) and (f).
BACRGEOUND INFORMATION
"Norinyl" is a registered trade name used by the defendant fo~ a drug com-
posed of 2 mg. of Norethin4rone and 0.1 mg. of Mestranol. At the time of the
alleged violations, Norinyl was commonly prescribed as an oral cobtraceptive.
Syiltex submitted to the FOod and Drug Administration a new drug applica-
tion for Norinyl which was approved on March 5, 1964. At this time, the Com-
missioner of the Food and Dr~1g Administration sent a letter to Syntex Labora
tories Inc In which be said that the claims made In the labeling were limited
by the representations made in the new drug application The Commissioner
also said that `the approval of the new drug, application In no' way relieved
Syntex Laboratories from complying with all of the provisions of the Federal
Food, Drug, and Cosmetic Act. *
The 1965 edition of the Physicians Desk Reference published by Medical
Economics, Inc., bore a monograph for Norinyl which appeared on pages 962
and 963. The Physicians' Desk fleference is published annually in cooperation
wIth the subscribing manufacturers, Th~, purpose of the `Physiciabs' Desk Refer-
ence is to make avallabie to physicians Information on ma~or pharmaceutical
specialties.. Information appearing In the Physicians' Desk Reference is solely
that furnished by the manufacti~rers. This is explained by the, foreward in the
19th EditIon, which contains the following ~tatenient: `~.`The function of the
publisher is the compilation o~ga~ization, and distritution ~f the Informm~tion
I~acb product description has been prepared by the manufaçturer~ and edited
and approved by the manufacturer s Medical Department Medical Dire~tor ot
Medical Consultant.", , ` ` *
EVIDENCE OF MISBRANDING * ` ` ,, * *
Count 1.-The drug, Norinyl, which is a prescription' drug and which is a new
drug subject to 21 U.S.C. 355, was not exethpt from the requirements of 21
U.S.C. 852(f) (1)' that adequate directions ~or use appear In its labeling since
PAGENO="0348"
1359S COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
it failed to comply with the condition ~or e~em~tf on set forth in the regulatIons
21 C~R 1.106(b) (4).. Such condition requires that, i~ case of a drug subject, to
21 tLS.C. 355, the labelji~g be substaut~aUy the same as the labeling ai4horized
by the approved new drug application for such drug. Th1~ w~s not t'~ie with
respeót to Norinyl because it~ labeling, the 1965 Edition of the Physiclans' Drug
Reference, failed to include the follOwing information which was in the new
drug application labeling:
(1) A statement concerning the length of experience with the drug, an~1 that
although no deleterious effect of the drug on pituitary, ovarian, adrenal or
uterine function has been noted, the long-range effect on thesç~ and other organs
must await more pro1onge~ ob~ei~vation.
(2) The information regardjn~ the, possibility of pregnancy if the treatment
schedule is not adhered to; and that, if the regular menses fail to appear and
the treatment schedule has. n~dt been adhered to, or, it the patient misses two
regular menstrual periods, the possibility of pregnancy should be resolved
before resuming Nori~y1. If pre ~tncy is estfibli~shed, Norinyl should be discon-
tinued during the period of gestatioli, on the basis that virlllzátion of the
female fetus has been reported with oral use of progestational agents or
estrogen.
(3) The Important information . regarding the possible causal relationship
between progestational agents and intravascular clotting. In addition, the Pliysl-
clans' Desk Reference labeling contained the statement "it provides maximum
protection against unplanned pregnancy and minimizes `undesirable side effects
resulting in fewer patient dropouts" which statement was not supported by
the approved new drug application labeling.
Counts II and IlL-In regard' to the acWertisements upon which the mis-
branding charges are based, examination has shown that they omitted certain
information which was iti the new drug application labeling and that they
Included certain information which was net in such labeling, ~flie nature of such
information is alleged in' ` the criminal fliformation. Because of such omission
and of such inclusion, the advertisements f~til~d: to include' true statements
relating to the side effects and contraindications of Norinyl as required by
regulations. S
EVIDSNCE OF VIOL4~TIVF SRIPM]ENTS
Count 1.-A sample of tw~ 1QO-tablet bottles of ~orinyl was taken at random
from a lot of ten such bottles in the shelf stock of the Permanente Services
Pharmacy, South San Francisco, California, by Food and Drug Inspector Frank
W. Scholl. The lot was identified by Donald B. Mtirray, Manager and Pharma-
cisf. in Charge of the Dapite, Division of Perma~ente ~ervices, The, ~erkeley,
California, Who said that the lot'l~ad be~n. ~`çceh~eU ,fron~ Sytitex Labor~t~ries,
Inc., Clark, New' Jer~ey7 on Octqber ~; , 1965' and subsequently shipped to the
above mentioned phai~macy Mr ~1irray ~npplie~ the Ipspec~or with Supporting
docuix~ents. , 5 5 5' ` `
Count II -$To ph~slcal sttm~le was dblalr~ed Howeve1' Donald F' Murray of
the Dapite Division of ~ermahent~ Serflees S~i'd that his t~rm had' recei'vod 86
twenty-tablet packages and' 48' 100-tabfet bottles of Norl,nyi fr~n Syntex Lab~
Qratories, Inc., clark, New `Jersey, on November 26, 1965. Mr. Murray `supplied
Inspector Frank D. Korun with stipporting documents `relating to the shipment.
Mr. lflchard IVickel, Assistant Administrator fOr the Samuel Merritt flospital;
Oakland, California, provided Inspector Korun with copies Of advertisements'
fOr `Norin~l Which appeared in the, l'~ovember 1~ 196~, ariçl Koveniber 15, 1965,
Editions' o~ The ,Amèric~n Journal Of Obsteti~ics àiid Gynecology, and the' No-
vember, 1965, Edition Of Ohptetrfo~ pxid ~yne~ology.
`Count III,-~-A sample of three 100-táb1~t .bqltle~ wa~ taken by T~od aXid. Drug"
In~pecto~ ~`ránk I(orun ~t ràndOnl ~rbm a lot of 48 such bdttles at the Dapite
Division of'?~frihanen'le~ Ser~tices, ~e,,~erkeie~;, %iiforuia. .`2~b~'1Ot Was ideñti.
fled b~ Mr Dçu~ald Mtirra~ who said that thO'flrm receivecj the lot frotu ~
tex Lab'oratoi~ies,' The., Clark, ` New Jersey, on February ~8, ~ COpies ~
docutneil'ts relating tO~'the ship~neñt were provided by ~rs. ~lbrence Thirns; ~
employee of Dapite. Mr, flichard Wlcke'l, Assistant Athii~nistrator for `the
Samuel Merritt `Hospital, Oakland,' California, proVided Inspector Koran with
copies of advertisements for ,Norinyl which appeared in the' FObruary 14, 1966,
Edition of Modern Medicine, the February,' `1966,' Edition of Obstetrics an4I'
PAGENO="0349"
COMPETITIVE PEOBLEMS IN THE Dl~UG INDUSTRY 135~9
Gynecology, and the February 1, 1966, EdItion of The American Journal of
Obstetrics and Gynecol9gy,
The headquarters of Syntex ~abora~Qr~qs, Inc., a~e ~esently located In
Alto, California, but `nianufaCturiug a1 the time of the allege~ violatiQas was
carried out elsewhere. The drug i~i questiQu was manufactured by Synte~ Lab-
oratories emplQyees' i~ing eq1i~puiOnt~iôchte~;Qn the premises of Warner-Cbilcott
Laboratories, Inc., Morris PIai~s, New J'érsey, This ~a~ufacturlng opêratio~i
was under the control of Syntex, J~aboratories, Inc. The shipments of the drug
were made by the Syntex Laboratories Er~ñcb In Clark, New Jersey.
HEARING HELD PURSUANt TO 21 U.5.C. 335
A Notice of Ilearing issued pursuant to 21 U.S.C. 335 on April 8, 1966, ad1
lressed to Syntex Laboratories, Inc., 1401 Hiliview Drive, Palo Alto, California.
A personal response was made by Mr. Vincent Kleinfeld, Attorney for the
firm, on April 21, 1966. Mr. Kleinfeld did not question the interstate nature of
the shipments and did not deny that the Physicians' Desk Reference was ac-
companying labeling as is charged in Count I. Neither did be deny the firm's
responsibility for the shipments, for the placing of the monograph in the Physi-
cians' Desk Reference, nor for the placement of the advertisements In question
in the November issues of the various medical journals.
Mr. Kleinfeld said that the firm had no intention of deliberately omitting
information from the Physicians' Desk Reference or from medical journal
advertising that they believe would have been desired by the Food and Drug
Administration. I-Ie said that honest men can differ in opinions as to what is
required under the regulations and that Syntex's doctors had believed their
medical journal advertising and the monograph in the Physicians' Desk Refer-
ciace were in full compliance with the law. Mr. Kleinfeld stated that he himself
did not realize the full extent, of the information desired by the Food and Drug
Administration to appear i~i these publicatioi~s.
Mr. Kleinfeld said be was helping the firm to set up the procedure. for the
development of advertising and labeling that will assure full compliance with
the requirements of the Food and Drug AdmUnstrat.ion. He had prepared and.
submitted to the Food and Drug Admini~tratjon a new propQsed monograph
for the Physicians' .Desl~ Reference to be published in the next quarterly supple-
ment after it is approved by the l?ood and Drug 4dmjnistration. He had also
prepared ~ statement for use in medical jpurnal advertising, as soon as he
obtained approval from the Food and Drug Administration. He stated that he
was of the opinion that. these proposals would meet all the points raised in
the Notice of Hearing but that they Would be changed if they were still not
&itisfacto~y to the Food and Drug. Administration,
On June 21, 1966, Mr. Kleinfeid, and other representatives of Syntex, Labpra-
tories, Inc., met with representatives of the Food and Drug Administration In
Washingtofl, D.C. A proposed monograph f~r the I~hysicians' Desk Reference
was submitted and discussed as was propo~ed material for journal advertisi~ig.
The firm indicated an intention to comply with the~ recommendatioils and re-
quirements of the Food and Drug Administration.
SEIZURES
No seizures were made of the shipments which are the subject of the proposed
Information.
CONCLUSIONS
We have considered the representatioils which wOre made on behalf of the
firm at the above hearing. However, we cannot ignore the facts which show
that Syiltex Laboratories, Inc., has had ,an ~ttënsive experience In the develop-
ment of new drugs and their subsequent distribution in accordance with the
flew drug regulations; and that such firm was fully aware of the regulations.
which required that the contraindications `and side effects set forth in the ap-
proved new drug application labeling be presented in the Physicians' Desk
`Reference monograph and in the medical journal advertisements.
In these circumstances and in the Interest of protecting the health of the
consuhdng public, it is our opinion that criminal prosecution is' necessary to
impress upon Syntex Laboratories, Inc., Its responsibility fOr compliance' `with
the law, aüd to deter other firms from similar violations of the law.
PAGENO="0350"
13600 COMPETITIVE PROBLEMS IN THE, DRUG INDUSTRY
WXT~ES$ES
The principal witness In this case will,' be the Government Inspectors WhO
collected the samples and made the inspections, the Government Analyst who
analyzed the tamples, witnesses to establish the inte~state,origin of the samples
and the issuance of the advertisements, ~iid medical officers of the Food and
Drug Administration's Bureau of Medicine who can testify as to the approred
new drug application, the approved labeling, and the serious nature of the
alleged medical journal advertising misbranding.
It is requested that,' if the form' of Information l~ amended, the United
States Attorney furnish us with a copy thereof; also, that be keeps us advised
as to the progress of the case and its dieposition.
The l'~~ew York District Office of the Food and Drug Administration located
at 850 Third Avenue (at 30th Street), Brooklyn, New York 11232, Telephone:
798-1300, will arrange for the presence of the necessary witnesses and assist
in the presentation of the case. Upon request, we will render such further
assistance as may be possible.
Very truly yours,
WILLIAM W. GooDRIcH,
Assistant General Counsel,
Food and Drug Division~
APRIL 22, 1968.
R Syntex Laboratories, Inc.,' F.D.O. No. 53222, Federal Food, Drug, and Cos-
metic Act
Mr. WILLIAM W. Goo~nxcn,
AsSistant General Counsel, Department of Health, Education, and Welfare,
Washington, D.C.
DEAR Mn. GooDRICH: In view of the expressions contained In your letter of
March 1, 1968, we have re-examined our previous determination relative to
prosecution of this matter. The reasons set out In your letter tend only to rein-
force our opinion that this matter does not present a case for prosecution and
i'o not a proper settifig In Which to attempt to sustain a judicial Interpretation
of the regulations issued' pursuant to Section 352(n) of the Act.
We believe that any attempt to secure a judicial interpretation which will
enlarge the meaning of the `statutory terms on the `basis of the meaning of the
word "relating" as suggested by you would be frustrated by the factual situation.
The difficulty inherent in any such attempt is that the labeling which was
approved by the Food and Drug Administration contains the same words as are
found in the statute and regula~tions,' i.e., "side effects" and "contraindleations."
Under each heading, the specific' items or' conditions are listed. Many ~f the
conditions which you now contend are side effects or contraiiidlcatlons are not
listed under those headings In the labeling but are set out under other headings
lii the labeling. In a criminal prosecution, such a factual situation creates an
impossible barrier to success. In `all' likelihood, the only result would be to
obtain a judicial expression contrary to your desire. In passing, We consider
the possibility that the court might find an analogy between the present' situa-'
tion and that of Haynes v. United States, -~-~ U.S. -, decided January 29,
1968, wherein the Supreme Court commented that "so' much could not be de-
rived from so little.," , , ` `
Moreover, the argument presented relative to the meanings of certain lan-
guage used in the advertisement as compared to that of the labeling presents
so fine and tenuous a distinction as to render conviction most unlikely. In other
instances the suggested violation appears to consist of a failure to furnish
information which does not appear In the approved labeling.
Specifically, the argument that the statement in the advertisement that the'
drug is contraindicated in pregnancy Is not satisfactory because the doctor
should have been told to discontinue the drug "at the earliest possible sign of
pregn~ney" is untenable. Obviously, the physician is not going to use', the drug'
to prevept pregnancy if the patient is pregnant. Neither would' it seem logical
to erpect that a `physician would continue its use after the patient became
pregnant Moreover under the circumstances, any physician would be aware
that contraindicatlons o~ the drug In pregnancy can only mean that it should
be discontinued if `the patient becomes pregnant.
As to the nec~ssity tot `discontinuance, at the `earliest' possible sign, It will
be observed `that the approved labeling under the heading of "Side effects" notes
that symptoms "resembling early pregnancy" as well as changes In the `men-
PAGENO="0351"
COMPETITIVE PROBLEMS IN THE DRUG INDTJSTRY 13601
~trua1 cycle, including occasional inter~menstrna1 bleeding and spotting arn~
sometimes with the period being missed entirely, may occur. Thus, the ap~
pioved labeling certainly indicates tJ~at some of the e~trliest pOssible ` sign~
of pregnancy may be false one~ which do not justify discontinuancç of the
medication Accordingly, a warning to the physician to discontinue the drug
at, the "earliest possible" signs of the condition the drug is meant to prevent
is not consistent with the approved labeling. ,
Under these circumstances, we see no justification for any~ attempt to predi-
cate a criminal prosecution in the factual situation above described.
Your letter states that "contraindication for psychic depression in the ap-
proved labeling is directed to patients with a history of psychic depression
and not just to those with presently observable psychic depression." However,
the language used in the labeling approved by the Food and Drug Administra-
tion is not that clear or unatnbiguous. The labeling stated under "contraindi~
~ations that patients with a history of psychic depression should be carefully
observed, and the drug is discontinued if depression recurs to a marked degree.'~
The advertisement stated that the drug was contraindicated in "severe depres-
sion.,,
Thus, the labeling does not say the drug should not be used in patients with
a history of psychic depression or even that it must be discontinued if that
condition should appear. Its use is contraindicated only if the condition reap-
pears to a marked degree. The advertising statement contraindicating the drug
in `the event of severe depression, therefore, does not significantly differ from
that of the labeling. Its terseness is not a vice, indeed, a "brief summary" not
a verbatim quote of the labeling is all that the Act requires.
It will be observed that it is also arguable that the flat contraindication of
the advertisement may be considered to be more restrictive than the permissive
and rather ambiguous wording of the labeling.
While the labeling of the drug might well have included under the heading
side effects a warning about the blood clotting possibilities of the drug the
fact is that it did not. The Food and. Drug Administration approved labeling
which treated it in another fashion, and many courts and juries would con-
sider it manifestly unfair for the Government to attempt to impose criminal
sanctions for the failure of the subject~to set this condition out as a side effect
without having first been advised of the necessity so to do.
Additionally, the agency's position is considerably weakened by the fact
that in the labeling the coupling of the reference to blood clotting with the favor-
able comment of the Ad Hoc Advisory Committee has the effect of minimizing
the potential threat to the patient. This circumstance would also tend to dimi-
nish the seriousness of the failure to list the condition as a "side effect."
Much of the same objection applies to the contention relative to the failure
to set forth as a side effect a statement found in the labeling under a `different
heading relative to the effect of estrogens on calcium and' phosphorus. In addi-
tion to the questionable fairness of a prosecution in these circumstances, the
labeling statement does not refer to a definIte knowledge' that the drug has a
deleterious effect on calcium and phosphorus metabolism but only to a general
medical learning that estrogens are `known to have such an influence. The
complaint, therefore, seems to relate to a failure to list as a side effect of th'e
drug" a matter of~medical learning which is applicable to, estrogens. Thus, the
information does not seem to relate to a side effect of the drug advertised, and
for that reason a prosecution based upon this omission is, not sound.
Nor do we believe ,that the failure of the advertisement to include the specifl~
details relative to the length of evperience with the drug forms a basis for
criminal action. Such Information `wa~ not required as a part of the statement
in the tabeling concerning side effects, and the advertisement, did, by inference
at least, inform the `physician that "prolonged observation~" bad not taken
place.
That portion of the advertisement which speaks Of the low Incidence of side
effects does not appear to constitute a violation of the statute. In our view,
the advertiselnelit does not falsely represent the incidence' of side effects which
attends the use of the drug. Although w~ think it is `at least debatable whether
the advertisement claims a superiority over other products in connection with
the incidence' of side effects, such statements are within the area of `allowable
promotion of the drug The statute is in our opinion oriented toward requiring
truthfulness in the' area pertaining to safety and effectiveness. We do not read
the legislative history as ,indicating that the statute is designe4 to prevent the
usual' effort to convince consumers that the advertiser's product is superior to
other almost' identical products so long as the' factual statements are true.
PAGENO="0352"
13602 COMPETITIVE PROflLEMS I~T' ~TItE DRUG' INDUSTE~
Yoñr letter ieads us to the conclusion that yoiido not contend that `the adver-
titement fOr Norin~rl claimed any use not included in the approved Iabelh~g
bitt `only that `it did' itot set forth all the side `effecte with `the specificity ~w~tich
the agency finds desirable. ~`Our' letter also `states that it `i~ your position that
advertisements of newdrngs suh~ect'to prescription use for which applications
were approved after October 10, 1962, cannot contain any statement that i~ not
in the approved New Drug Application.
We' have observed that hadthe advertisement set forth as "side `e~e~ts" some
of the statements which' you contend should' have been Were lnchZded such act
would be' a direct' violation of the position stated above because the statements
of the conditions allegedly `omitted are' not found under that héa~,ling in the
approved' labeling. YoUr contentions,' therefore, seem to be diametrically opposed
to your argument as' to the necessity that tb~ ad~~ertisëment conform to the
labeling. ,
Although we are declining prosecution based upou the alleged violatioi~s of
the `advertising provision of the Act, we do not have the same reservatiOns
concerning the violation alleged in Cottnt I of th~ suggested form of informatiOn
which relates to the use of improper labeling in the monograph in the Pltysi-
clans' Desk If eferenee.
We are engagecl in the process of ,redrafting that Count in accordance with
the ideas expressed in our rast communication., A copy of our con~mitnication
to the United States Attorney will be forwarded to `you `when it is mailed.
Sincerely,
FRED M~ ~VINSON, Jr.,
Assistant Attorney General,
Criminal Division.
By HAROLD P. SHAPIRo,
Chief, AdmiflietrGtive Re9uiations $ection.
UNXTEI' STATtS DEPARpaENT O~' JuSTIpE,
T~T~'~cEi~ SrA~rEs ATTORNEY,
FOfl T'HR Dls~ittc~r `ov Nuw JERSEt,
Newark, N.J., `May 21, 1968.
Attention Ifarold' P. Shapird, ~hie~, Aciministr~tive R~itlations Section.
Re Syntex Laboratories, InC., ~t~O No. 53222, Federal `Food, Drug, and Cosmetic
Act. Your ~f: FMV :JWK :nic, 21-48-353.
DEPARTI~rENT o1~ JitsTIcE,
Washington, D.C.
DEAR Mn. SHAPIRO: We have, reviewed the file ir~ the above-referenced matter
and have concluded that the case lacks prosecutive merit.
We recognize that discrepancies between, the mqnograpji; and the' approved
labeling do exists However,, the ~~gulati~ns themselves r,equi~e, only' that' the
labeling be "substantially" the same, as tliel~~bejing authorized' by the approved
new drug ~pplication. The word "~ubstant1aUy", certainly: gave the company
some license to synopsize the in~or~uat1on contained in t'he, approved new drug
applIcation. A reasonable doubt evists in `Qur minds as to whether the labeling
in the monograph is, substantially different fr~m the labeling in the approved
new drug application, and we believe that a lay jury would also harbor such a
doubt.
Moreover, the discrepancies between the monograph and the approved labeling
do not appear to have been the result of a purposeful evasion of the regulatory
requirements of the Administration but of an honest difference of opinion as to
what was required under the regulations. In fact, at the hearing held on April
8, 1966, the firm offered to obviate all future difficulties by submitting sample
advertising and' monographs to the Food and Drug Administration before
publication. `
In closing, we note that ,the c~ffe~se complained, o~, occu~re4 in Septeinbei~,
1905, almost three years ago. ~o indications of violations of the Food and
Drug law subsequent,to ,that date are ~eporied. ,Prosecut'ion at this rather late
date would' not be in, the best interests of the Federal Government.
Very truly yours,
DAVID M. SATE, Jr,~
U.S. Attorney.
By MARLENE GRoss,
Assistant U.S. Attorney.
PAGENO="0353"
COMPETITIVE PROBLEMS IN TIlE DRUG INDUSTRY 13603
DEPARTMENT OF HEALTII, EDUCATION, AND WELFARE,
May 28, 1968.
Attention Fred M. Vinson, Jr~, Assistant Attorney General.
Re ~yxLtex. Laboratories, Inc. ,!ou~ ref: FMV :JWK :mch 21-48-353, FDC No.
53222.
Hon. RAMSEY CLARK,
Attorney Genera';
Department of Justice,
Washington, D.C.
DJ~AR Sin: This is with reference to your letter of April 22, 1968, and the
United States Attorney's leti~ér of May 21, 1968, detlining prosecution on both
the advertising and the labeling violations we have reported to you.
The United States Attorney declines for the reasons that (1) there Is a
reasonable doubt whether the labeling (FDA monograph) is substantially dif~
ferent from the approved labeling, (2) the violations appear tO be the result
of honest differences of opinloil as to what is required, and (3) the o~ense is
almost three years old and ther care no indications of violations since 1965.
(1) There can be no reasonable doubt that the PDR differed from the ap-
proved labeling in a substantial way. It was too brief, it was promotionally
slanted, and it omitted important information for safe prescribing.
(a) .A comparison of the approved labelibg and the PD~I moilograph plainly
shows that some of the most vital safety information w~ts omitted from the
`monography. A copy of the 1965 omnograph is enclosed. It exaggerates the
effectiveness of the drug, and it minimizes the side effects, precatttlohs and
contraindicatiOns. We simply cannot understand `how it can b~ said that this
,abbreviated material', and ~particularly the seven lines devoted to side effe~t5,
cAn possibly be said to' bO ~i,ibstautial1~ the same as the approved directiOns for
safe use of this drug. The substanc~ of the monOgraph, and th'e, ~ub~tanrce of
the approved labeling are nbt'the same. S
rundamentally, what the Company did was to present a reassuring write-up
for the physician's desk which omitted the most important information he needed
to' have' to prescribe this drug for his patient~ with safety. As the criminal
`information which your office drafted `shows, the physician ~ivas not alerted
to the possibility `of effects' on the fetus if a pregnant' woman should take the
drug not knowing she was pregnant, he was not alerted to the limited `experi-
ence with the dri~ig and Its possible loflg range effects upon a vai~I~ti~ of organ
and endocrine systems, includThg~ `pituitary,. ovarlati `adrenal,' uterine, livei~ â~d
thyroid be w~ not alerted to the possibility of intravasci~ilar elbtting (a risk
nOw' k~own to be `even moi~e' ~iOns tMn In 1965), ~he' *va~ n~t Warned `kbo~it
effe~ts un patients ~rith any condition involving calcium Or ~hosphOrus' metabo-
`lism, an~ he was not told to, use tlie drug with care in any patient who had a
history of psychic depression; ` `
We feel coñfldeitt that we can prove by acceptgblë medical eviden~e that the
pr~scribing Information in PDR-1965 was not substantially the'same as the ap-
`proved labeling, as the regulations require. And we are equally confident that
we can prove the omissions and changes were significant `from the standpoh~t of
patient Safety. Our Advisory Comithttèe on Obstetrics and Gynecology," quoted
in the Company's lab~lin,g, ~aid that the physiCian must deCide for hi~'patient
whether to dccept the risk involved `in the use of oral contracepti~es, small
though it may be, but that hA~"~can 4° this WIsely onl~ when there is ptesentad
to him dispassionate scientific knowledge of the a'v~ailable data," The' PDR
monograph did not serve this purpose, and the violative ads we will discuss
later compounded the hazard. ,,
`(2) There was no' basis fo~ eor~sid'ering this violation a result of an honest
difference of opinion as to "What is required. The `applicable regulations' bad
been in effect since 1961, and the Company's performance shows that it did not
comply with what was clearly required-it did not pre~ent substantially the
same information in PDR a~ in the apprQved 1a~eling. It had the labeling at
hand and it chose to omit some of the informatIon from the PDR monograph.
(3) Syntex has not been in compliance sincd'1965. The ad Charges you elimi-
nated would have shown ftie United State~ AttQruey thiut the same failures to
inform continued after that date, Aad ~itbin the `past few, months, on January
22, 1968. we required' the Company to mail a letter t~ all physicians in the
United States calling attention to its failure tO iñeludç~' ii~i its then `current
ads appropriate w~rnings abo,i~j use of the drug in ~sychic depression and
about ,the pqs~ibi1ity p~ t1lro~uiboembolic episodes. A copy of the letter is
enclqsed,' ,
PAGENO="0354"
13604 COMPETITIVE PROBLE&tS IN THE DRUG INDUSTRY
Turning to the reasons given in your letter of April 22, 1968, for declining
to ~rosecute the advertising violations, we must reiterate what we have fre-
quently said to your representatives, that the brief summary is required to
disclose true information related to side effects, contraindications, and effec-
tiveness as specified in our regulations, not merely the information headed
"side effects" and "contraindications" in the approved labeling.
We are confident that any qualified expert would regard the information
about clotting, calcium and phosphorus metabolism, effect on organ and endo-
crine functions, etc., as information related to "side effects" and "contraindi-
cation" as those terms were used in the 1962 Drug Amendments. The fact that
the information in certain instances was headed "Thyroid Gland", etc., does not
make it less information related to side effects than the material in the labeling
under the heading "Side Effects~" A mere reading of the package insert estab-
lishes this.
While criminal prosecutions do sometimes involve strict interpretatIons of
regulations, there is "no canon against using common sense" in applying a
criminal law.
As to the point you make that our understandings of what was said in the
advertising are based upon tenuous constructions, we think we should have an
opportunity to present these points to a court or `jury with an explanation of
their important medical significance.
Surely a jury would understand the need to warn a physician, and through
him .the patient, that a missed dose or a miscalculation in taking the oral con-
traceptive drug may result in pregnancy, and that if that occurs there is danger
to the unborn child from continuing to take the drug. Telling the physician
that the drug. is contraindicated in pregnancy does not tell him what he really
needs to know for his patient's safety in bearing a child. It is the need to alert
the user to the hazard to the unborn child which may result from a missed
dose or a miscalculation that calls for this early pregnancy warning.
On psychic depression, we have noted above that as late as January 1968,
Syntex was still not properly presenting the message about this hazard to the
profession.: A corrective letter was necessary. What is missing in the ads is the
warning that any patient with a history of psychic depression should be care-
fully observed (this means the patient should be followed much more closely
than the routine patient) if a decision is made by the physician to prescribe the
oral contraceptive. A statement that the drug is contraindicated in "severe
depression" does not adequately advise the physician of the risk he runs with a
patient who has bad depression in the past but now has it under control.
Blood clotting has been a problem with these drugs for a long time. When
these ads ran, the labeling was required to say that such episodes had, occurred,
that a causal relationship had not been established, and that the prob1e~ was
under investigation. The ad said nothing about this. Instead, it said the drug
was contraindicated in. thrombophlehitls and pulmonary embolism (current or
past). This is quite different from telling the physician that the cause and, effect
relationship of the oral contraceptive drugs to these serous complications, partic-
ularly in the older age groups, was then under scientific ipve~tigation.
As you know, this problem has caused a progressive strengthening of the
warning as more experience has been gained.
Finally, we must express our disagreement with the idea that a drug com-
pany can use a literally true statement to convince prescribers that its products
aye superior to other identical products. A half-truth is still. sometimes a great
* lie. Promotional practices with this class of products have been characterized
by the use of such half-truths to gain marketing advantages. We have had to
require several companies to discontinpe such ads and to send letters of correc-
tion to the profession. For from beIng harmless puffing-which in any event is
not tolerated in prescription drug advertr~ements-this kind of promotion affirm
atively misrepresents both the `effectiveness and the `safety of the oral contra-
ceptives.
We have written you at this length because of the serious consequences to the
profession and to patients of improper `prescription drug ndvertlsing~
Six million women are considered to be on oral contraceptive drugs at this
time in the United States. Their safe use depends upon proper promotion of the
drugs to `the profession. Syntex ha~ seriously failed in the instances cited in our
recommended prosecution case.
Yours very truly,
WILtIAM W. Goomnicri.
A88istant Genera1 Uoun~?eZ,
Food and Drug Divi3ion.
PAGENO="0355"
COMPETITIVE PROBLEMS I~ THE DRUG I~DUSThY 136~5
MAY 27, 1968.
Hon. RAMSEY CLARK,
Attorney General,
Department of Justice,
Washington, D.C.
Attention Fred M. Vinson, Jr., Assistant Attorney General.
Re Syntex Laboratories, Inc.,Your ref: PNV :JWK,,;meh 21-48-353, ?? No. 5822~.
DEAR Sin: This Is with reference to your letter of April 22~ 1968, and the Urdted
States Attorney's letter of May 21, 1968, decliningprosecutlOfl on both the adver-
tising and the labeling violations we have reported to you.
The United States Attorney declines for the reasons that (1) there Is a rea~
sonable doubt whether the lableing (`1?? monograph) Is substantially different
from the approved labeling, (2) the violations appear to be the result of honest
differences of opinion as to what is required, and (3) the offense is almost three
years old and there are no indi~ations of violations since 1965.
(1) There can be no reasonable doubt that the ??? differed from the approved
labeling in a substantial `way. It was too brief, it was prom~tionaUy slanted, and
It omitted important information for safe prescribing,
(a) A comparison of the approved `labeling and the PBR monograph plainly
shows *th~t some `of `the most vital safety information was~ omitted from the
monograph. A copy of the 1965 monograph is enclosed. It exaggerates the effec-
tiveness of the drug, and it minimizes the side effects, precautiotis and contra-
indications. We simply cannot understand `how it can be said that thjs abbrevi-
ated material, and particularly the lines deVoted to side effects, can possi-
bly be said to be substantially the same as the approved directions for safe
use of this drug. The substance of the monograph and the substance of the
approved labeling are not the same.
Fundamentally, what the Company did was to present a reassuring write-up
for the physician's desk' which omitted the most important information he needed
to have to prescribe this drug for his patients' with safety. As the ci~iminal infor-
mation which your office drafted sbow~, the physician was not alerted to the
possibility of effects on the fetus if a pregnant woman should thke the dru~ not
knowing she was pregnant, he was not alerted to the limited experience with the
drug and' its possible long range effects upon a variety of organ tind endocrine
systems, including pituitary,' ovarian, adrenal, uterine, liver, anil thyroid,' he
was not alerted to the possibility of Intravascular ~1ottipg (a risk now known to
be even more serious than In 1965); he was not warned about effects on patients
with any condition lnvolving'cfllClUm or phosphorus metabolism, and he was `not
told to use the drug with care In any patient who bad a history of psychic
depression.
We feel confident that we can prove by acceptable medical evidence that the
prescribing Information in PDR-1965 was not substantially the same as the
approved labeling, as the `regulations require. And we are equally confident that
we can prove the omissions and changes mor significant front the standpoint of
`patient safety. Onr Advisory Cotamittee on ObstetricS and Gynecology, quoted in
the Company's labeling, said that `the physician must de~lde for his patient
whether to accept the risk involved In the use of oral contraCeptives, ~ma1l
though it may be, but that he "can do this wisely only when there is presented
to hint dispassionate scientific knowledge of the available data." The PDR mono-
graph did not serve this purpose, and the' violative ads we will discuss later
compounded the hazard.
(2) There was no basis for considering this Violation a result of an honest
difference of opinion as to what is reqnired~ The applicable regulations bad been
in effect since 1961, `and the Company's performance shows that it did not comply
with what was clearly required-it did not present substantially the same Infor-
mation in PDR as In the approved labeling. It had the labeling at hand and It
chose to omit some of the information from the PDR monograph.
(3) Syntex has not been in compliance sl~c~ 19~5. The ad charges you elirrii-
nated would have shown the United States Attorney that the same failures ~o
inform continued after that date. And within the past few months, oh Janu-
ary 22, 1968, we required the Company to mail a letter to all physicians in the
United States calling attention to it~ failure to include in Its then current ads
appropriate warnings about use of the drug in psychic depression and `about the
possibility of thromboenbolic episodes. A copy of the letter is enclosed.
PAGENO="0356"
13606 COMPETITIVE I~RQBLEMS IN THE DRUG' INDUSTRY
Turning to the reasons given In your letter of April 22, 1968, for declining to
prosecute the advertising violations, we must reiterate what we have frequently
said to your representatives, that the brief summary is required to disclose true
information related to side effec~ts, comtraindjcatj'ons, and effectiveness as sped-
fled in our regulations, not merely the information headed "side effects" and
"contraindications" in tl~teapp~oved labeling.,
We are confident, that any qualified expert would regard the information about
clotting, calcium and phosphorus metabolism, effect on organ and endocrine func-
tions, ete~, as information related to "side effects" and "contrtxlndicatjons" as
those terms were used in the ~962 Drug Amep~lments. The fact that the informa-
tion in certain instances was headed "Thyrold' Gland", etc,, does not make it
less information related to side effeets than. `the material in the labeling under
the heading "Side Effects." A mere reading of the package Insert establishes this.
While criminal prosecutions do sometimes Involve strict interpretations of
regulations, there Is "no canon against using common sense" in applying a
criminal law.
As to the point you irn~ke that our understandings of what was said in the
advertising are based upon tenuous constructions, we think we should have an
opportunity to present these p~lnts to a court or jury with an explanation of
their important medical signif1ca~ce.
Surely a jury would understand the need to wai~n a physician, and through
him the patient, that a, missed dose or a miscalculation in takir~g the oral contra-
ceptive drug may result in pregnancy, and that. if that occurS there is danger
to the unborn child from eoutii~u~ug~to take tliedrug~ Telling the physician that
the drug is contrain~flcnted In *pt~egnancy does not tell him what lie really needs
to know for his patient's safety in ~bear1ng a child. It is theY need to alert the
user to the hazard to the `unhornehild which may result from a missed dose or q~
miscalculation that calls for this early pregnancy warning.
On psychic depression, we. have noted above that as late as January 1968,
Syntex was still not properly presenting the message about this hazard to the
profession. A corrective letter was necessary, What iS missing in the ads Is the
warning that any patient with a history of psychic depression should be care-
fully ~observed `(this means the patient shoi,ild hie followed much more closely
than the~ routine patienty if a decision iS made by the physician `to prescribe
the oral con'tra'ceptive~A statement that the drug is contraindicated in. "severe
depression" does not adeqnatel'yady}~~~e ~pbysician `of tile risk he man with `a
patient whojias bad de~resnion~in the paM bat now h'as it under contrtyl.
Blood clotting has'been~aproblem~with these drugs for a long time. When these
ads ran, the labeling was required to `nay that Such epiSode~ had eccurred, that
a causal relationship had not been astablishe~, and that the problem was under
investigation. The ad said nothing about this. Instead, It said the' drug was
contraftidjeated in th'r'onibop1il~hjti~, and pulmonary embolism (current or past).
This is quite different from telling the `physician tl~at the nause and effect rela-
tiapshjp of the oral contraceptive drugs to `these serious complications; particu-
larly in the older age groups, was then' under~ scientific investi'g~tion.
As you know, this problem has caused a progressive Strengthening of the
warning as moreoxperience has been gained.
Finally; we must express our disa~reeraent with the idea that a drug company
can use a literally true statement to convince pres~ribers' that Its products are
superior to other identical products: A balf~truth `IS still sometimes a great lie.
Promotional practices with this class ef products have been characterized by the
use of such half-truths to gain marketing advantages. We have had `to require
several companies to~ discontinue such ads and to send letters of correction to
~he profession. Far from being harmless pufling-which in any event is not
tolerated in prescription drug advertisemeflts-_this kind of promotion affirma-
tively misrepresents' both the effectiveness and the safety of the oral contra-
ceptives.
We have writtep you at this length because of the serious consequences to the
profession and to patient's `of Improper prescription drug advertising.
Six million women are estimated to be on oral contraceptive drugs at this
time in the United States. Their safe use dcpends upon proper promotion of
PAGENO="0357"
COMPETITIVE PROBLEMS IN T~IE DRUG INDUSTRY 13607
the drugs to the profession. Syntex ha~ seriously failed in t~ie instances cite~j
ill Oi~r reeOmmertded prosec~ition case.
Yours ~rery truly,
WILLIAM W. GOoDRIc~iI,
Assistant General Counsel,
Food and Drug Division.
SEPTEMnER 20, 1968.
Be Syntex Laboratories, Inc., F.D.C. No. 58222-Federal Food, Drug, an~1
Cosmetic Act.
Mr. WILLIAM W. Goon1~IcH,
Assistant General Counsel,
Department of.Hcalth, Education, and Welfare,
Washington, D.C.
DEAR Mn. Goonnicu: Enclosed I~ a copy of our letter of eeen date to the United
States Attorney at Newark, New Jersey, r~qnesting~that his offi~e rOconsider the
above matter in light of your letter of May 21~ 1~~8. YoUr kttention is invited
to our suggestion on page two that the tinited States Attorney interview the
Food and Drug Administration experts who would be called as witnesses and
review the expected testimony of non-Government experts who may also be
called. We believe that such aCtion is necessary to ascertain whether or not
the differences between the ~ and the approved labeling are substantial.
It ts ~luggested that you arrange with the United States Attorney for him
to interview the Food and Drug Administration experts and furnish him with
a statement of the testimony w~ich may be expected from any other experts
who will appear for the Government.
Sincerely.
FRED M. VINS0N, Jr.,
Assistant Attorney General,
Criminal Division.
By HAROLD P. SHAPIRo,
Chief, Administrative Regulations Section.
SEPTEMEER 20, 1968.
Be Seyntex Laboratories, Inc., F.D.C. No. 53222-Federal Food, Drug, and
Cosmetic Act.
Mr. DAVID M. SATZ, Jr.,
U.S.. Attorney,
Newark, N.J.
DEAR Mn. SATZ: After receiving yoUr letter1 Of `May 21, 1D68, recommending
that the abo~re-eapti'oned matter be, c'lo~ëd' ~i't1iout `~proseentlon, we received a
letter from Mr. Goodrich, `Assistant General Cotti~isè1, Food and Drpg Division,.
taking exceptioi~i to your concluslops arid ~questir~g, ~t further conslderatio~.
In view of the atrott~ differeflces `of' OpiDi~n held by `Mr. Goodrich, your Office,
and ourselves, we suggOsted that ~ ~heeting be árrknged' whereby alil parties
could discuss their views. ,
However, our efforts tq a~range ~uch a fl~eeting l~ave been to' no avail, and
sirice the itiatter `is growing~ stale, ~we belIeve `it' is appropriate to f~rward a
copy of Mr. Goodrich's 1etter~ We understand that Mr. Good~ich's áss!staiit,
Mr. ~otUieb has previoUsly sbowfi `Miss ~i~oss a rough draft Of this letter So
that she is familiar with r~s contents
You ~vi1l observe that Mr `~kx~4ricb no~ o~nly disagrees with tour concitision
that the alleged viotations of the Act which re1,~ite to the monograph on th~
drug Syfl1~~x, which' `appeared inthe P1~y~ici&n~ J3~s~ Ref ere~ice for, 196~, do not
justify criminal prosecfltion, bitt aiso With' our decision that tb~aljege~ viola:'
`tions of the, ~dvertistng provisions of the Act do not provide a baSis fo~ prose-
cution. Insofar as our decision on the latter problem is concerned, the views set
forth in Mr. Goodrich's letter have long been known to us and were thoroughly
considei~ed at the timC we reached our conclusion. They, therefore, are not, in
our view, so persuasive as to result in a change of our views.
PAGENO="0358"
13608 COMPETITIVE PROBLEMS IN' THE DRUG INDUSTRY
We have noted that Mr. Goodrich strongly disagrees with your analysis of.
the discrepancies between the monograph and the approved labeling, In view
of (the decision of Judge Lepoid in the Abbott ease, we `think that your point may
be well taken but the question Is impossible to evaluate In the absence of ex-
pected testimony. Itt would appear that the nature of these differences is a
matter that e~ujd heat be decided after discussion witl~ medical experts of the
Food and Drug Administration, who would be called at witnesses, as well as
a combination of the statements of any outside experts who would be expected
to testify on behalf `of the Goyernment. It is, therefore, suggested that you should
interview the appropriate Food and Drug Administration experts and examine
the statements of expected witnesses before making a final decision.
While it appears that the subject has seen the error of its ways and has
voluntarily complied with the desires of the Agency both with respect to its
P.D.R. labeling and advertising practices, we believe you should balance these
factors with the Agency's opinion as to the seriousness of the offense and the
necessity for criminal action. We suggest that the Court's attitude toward this
kind of an offense may be ascertained at the time it Imposes sentence `in the
Ciba matter, which, we undei~stand, is expected to take place shortly and may
be considered by you in evaluating this matter,
Accordingly, your reconsideration of this matter in the ~ig~t Of Mr. Good-
rich's comments will be appreciated. We trust that you will advise us In your
final decision. ,
Sincerely,
FRED M. Vixsox, Jr.,
4ssistant Attorney General,
Criminal Division.
By HAROLD P. SHAPIRO,
Chief, Administrative Regulations Section.
UNFIEO STATES DEPARTMENT OF JUSTICE,
UNITED STATES ATTORNEY,
rOR THE DISTRICT OF NEW JERSEY,
Newark, N.J., October 30, 1968.
Attention Harold P. ShapIro, Chief, Administrative Regulations Section.
Re Syiltex Laboratories, Inc., F.D.C. No. 53222-Federal Food, Drug and Cos-
nietic Act. Your Ref: FMV :JWK :mc, 21-48-35g.
DEPARTMENT OF JUSTICE,
Washington, D.C.
DEAR Mn. SHAPIrO: At your suggestion we have again reconsidered our file
in the above-referenced matter and have again concluded that the case lacks
prosecutive merit for substantially the reasons set forth In our letter dated
May 21, 1968.
We do not believe that interviews with medical personnel concerning the
discrepancies between the monograph and . approved labeling would affect our
opinion regarding the prosecutive merits of this action The regulations them
selves give the company sowe license to synopsize the material of the approved
labeling in the monograph. Even If medical personnel could convince us that
there were substantial differences and that we could in fact carry our burden of
proof, we feel that prosecution of this offense that occurred over three years
ago would r~ot be in the best interests of the Federal. Government, especially In
light of the fact that revised and current monographs do comply with the
regulations. *`
On October 2~, 1968, the Ciba Pharmaceutical Company was fined $200.00
on each count of a two count Information charging violations similar to those
here, In view of the time expended on this c~e by this o~ee. and by the Food
dnd Drug AdminIstration we found this resi4t most discouraging In view of
the, factors outlined above and in our previous c~rrespondence, we doubt that
we could achieve a mOre staisfactory result In this rnatter.
Very truly yonrs
DAVm M SATE Jr
TJS. Attorney.
By `M'4Rt.EN$ GROSS,
Msistant U.S. Attorney.
PAGENO="0359"
COMPETITIV1~ PROI3LEMS IN THE DRUG INDUSTRY 136Q9
UNITED STATES DEPARTMENT OF JUSTICE,
Washington, D.C., March 25, 1969.
Re Syntex Laboratories, Inc., F.D.C. No. 53222-Federal Food, Drug, `and Cos~
metic Act. Your Ref: MG :ch 748447.
Mr. WILLIAM W. GOODRICH,
Assistant General Counsel, Food, Drug, and Environmental Health DlvlsioiI,
Departmen~t of Health, Education, and Welfare, Washington, D.C.
DEAR Mn. GOODRICH: We note that you have been furnished with a COPY o± th~
United, States Attorney's letter to us declining prosecution in the above matter.
We have carefully considered whether this office should persevere in Its effort
to persuade the United States Attorney to agree to prosecution of the above~
captioned matter. Mr. Sats has filed criminal informatlons in two `similar case~
on pi~evious occasions and Is, therefore, familiar with such prosecutions am~
sympathetic to the enforcement efforts of the Food and Drug Administration.
It has been noted that in the Abbott case the court seemed to feel that some edi-
torializing was permissible under the regulations then In effect, and in the Ciba
and Armour cases very small fines were imposed upon the defendants for viola-
tions similar to those reported in the instant matter. Also We have observed
that new and more comprehensive regulations are about to be adopted in the
field of advertising and that the industry seems' to have accepted the' ruling
of the Abbott case that the Physician's Desk Reference constitutes labeling.
For the above reasons, we do not believe that, any useful purpose will be
served by directing the United States Attorney to institute a prosecution which
is contrary to his best judgment. Prosecution is, therefore, declined.
Sincerely,
WILL WILSON,
Assistant Attorney General,
Criminal Division.
By HAROLD P. ShAPIRo,
Chief, Administrative Regulations Section.
UNITED STATES DEPARTMENT OF JUSTICE,
UNITED STATES ATTORNEY,
FOR THE DISTRICT OF NEW JERSEY,
N&ieark, N.J., March 28, 1969.
Attention Alvin L. Gottlieb.
Re Syntex Laboratories, Inc.,' Federal Food, Drug, and Cosmetic Act. Your Ref:
58222.
U.S. DEPARTMENT OF EEALTH, EDUCATION, AND WELFARE,
Office' of the General Counsel,
Washington, D.C.
Dw~ MR. GOTTLIEB: In view of the fact that prosecution has been declined
in the above-captioned matter, we are returning herewith all exhibits furnished
to this office.
Please be advised that we are closing our file in this matter.
Very truly yours,
DAVID M. SATZ, Jr.,
U.S. Attorney.
By MARLENE GRoss,
Assistant U.S. Attorney.
MAY `25, `1967.
In Reply Refer to P.D.C,. No.53548.
E~on. RAM5nY, CLARN,. ` `
Attorney General, `. ` ` `
Dep~rtment of Justice, . ` `
Washington, D.C. `
DEAR Mn. CLARI~: We request the institution of criminal proceedings under
the Federal Food, Drug, and Cosmetic Act, against Warner-Lambert Pharma.
ceutical Company, a corporation trading as Warner-Chiieott ~aboratories, at,
Morris Plains, New Jersey.
The offenses complained of occurred during the period from about February 4,
1966, to about AprIl 22, 1966; and involved the introduction, into interstate
PAGENO="0360"
13610 COMPETITIVE PEOI3LEMS IN THE DRUG INDUSTRY
commerce at Morris Plains, New Jersey, for delivery to Glendale, New York,
and New Ysrk, New York, Quantities of the drug, Peritrate SA (Sustained Ac-
tion Pentaerythritol tetranitrate) tablets; wiljcl~ was misbranded, and. a new
drug for which no approval of a New Drug Application was effective for all
the purposes for which it was offered.
TIi~re are transmitted herewith *a suggested form of criminal lnformatiou
and the following ~xbibits:
1. CopIes of Notices of Hearing.
2. A copy of the approved New Drug Application labeling.
3. A copy of the mailing piece "490T718 Octol~er 1965" which included a re-
print of `Pentaerythritol ~etranitrate As Adjunct Therapy In the Immediate
Postinfarction Period" and "The U~e of Coronary Vasodilators In Acute Coro-
nary Occlusion."
4. A copy of the adverti~ement from the JQurnal of the American Medical
Association ot January 3, 1966, and rebruary 7, 1966, i4entmed PE-GP-527-4C.
5. A copy of the advertisement from the April 15, 1966, and prll 22, 1966,
issues of Me4icc~l World News.
6. The package insert for Peritrate S4~.
SECTIONS OF THE ACT INVOLVED
Count I of the Information charges violation of 21 U.S.C. 331(a) in that tile
d~fendánt caused the introduction into interstate commerce of an article of
drug which was misbranded within the meaning of the following sections of
21 U.S.C.
352(a) in that its labeli~ig, namely, the reprints contained In the mailing
pieces qontained false and misleading statements and representations concern-
ing the drug;
352(f) (1) in that ~aid labeling failed to bear adequate directions for use,
and the drug was not exempt, since tile labeling was not substantially the same
as the labeling in the ~pprbved New Drug Ap~1ication; and
352(n) in that the defendant failed to include in the advertisement caused
to be issued by it wth respect to the drug in the January 3, 1966, issue of the
Journal of the American Medical Association, ~t true statement of information
in brief summary relating to the effectiveness of such drug as required by 21
CFR 1.105(e). Since these charges are set forth at length in the information
and wOre also alleged in the seizure of this drug, they will not be repeated
here.
Count II alleges violation of 21 U.S.C ~31(d) in that the drug may not be
introduced into interstate `commerce in `that it was a new drug within the
meaning of 21 U.S.C. 321(p) since it was not generally recog~iized as safe, and
effective for all of the uses fOr which it was offered in the labeling and approval
of a New Drug Application with respect to such uses was not effective.
Count III alleges that a s~e~or~d shipment of the drug wasmisbranded within
the meaning of 21 ILS.C. 35~(n) in, that the defendant `failed to include in the.
adverti~4~nent caused to be issued itt ~the April 15, 1966, and April 22; 1966 issues
of Medical World News, a true, statement of information in brie~ summary'
relating to the effectiveness of ~uchL drug as required by 21 CPR 1.105(e).
BACEGROIJND INFORMATION
Peritrate `SA is a tradetnark held by the defendant for the drug, pentaerythri-
tol tetranitrate (PETN) in a time release dosage form Warner-Chilcott Labora-
tories submitted to the Food and Drug Administration, a New Drug Application
for Peritrate SA which was approved on Nq~eniber. 23, 1959. This permitted
the thug to be offered for use In angina pectorts' ~iatients. ` ,
The mailing piece referred to in this case consisted of reprh~ts of ~rtieles' in
the medical literature concerning experimental use of PETN.. The mailing piece
was actually addressed and mailed by the firm of Clark-O~eill, Itic.; Pah~le~,~
New Jersey. The list of doctors to whom the literature was to be sent `wad
suppli'e~ by' Warner-OhllcOtt Labotatories `b~ means Ol' `stiOker labOls which
were applied by the flrn~i and it has no record of names. However, Mr. ~obn
Owens, Executive Vlce~?reSldent of Clark-O'Neill informed FDA Ii~speetor~
that such a mailing usually includes the New York metro~ol1tan area A total
of 91,000 pieces were sent to medical people' throughout the cotintry.
The advertisement In the Journal of the American Medical Assoeia4ion
(JAMAl issue of January .3, 1966 is a five page color ad. The copy of the ad
PAGENO="0361"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13611
enclosed is in black and white, and the graphs do not reproduce clearly because
of the contrasting colors used in the magazine. However, the printing can all
be read. If this case goes to trial the FDA will obtain back copies of the mag~-
zines from the publishers for use. There are no extra copies available at this
time. The ad highlights the study made by Alexander Oscharoff, M.D. and re-
ported in the article "Pentaerythritol Tetranitrate as adjunct Therapy In tht~
Immediate Post Infarction Period." The same ad was rerun in the February 7,
1966 issue of JAMA.
The advertisement for the drug which was run in the April 15, 1966 and
April 22, 1966 issues of Medical World News is a two page ad which reports a
study conducted by B. L. Brofman, M.D. using Peritrate SA.
On February 28, 1966 in the Eastern District of New York, a libel was filed
against the shipment represented in Counts I and II and the drug was seized
on the same day. The libel alleged that the drug was misbranded in the same
manner as Count I does, and also alleged its shipment in violation of the new
drug provisions of the Act as now alleged in Count II. Warner-Lambert Pharma-
ceutical Company filed a claim of ownership and on May 12, 1966 consented to
the entry of a decree of condemnation without admitting the allegations of the
libel. The rest was then destroyed.
Hearings pursuant to 21 U.s.C. 335
Counts I and 11.-On April 18, 1966, a hearing was held in the Food and
Drug Administration's offices in New York, New York, with the following people
representing Warner-Lambert: Mr. Robert Clark, President, and Dr. Frank
DiTraglia, Medical Director, Warner-Chilcott Laboratories, Mr. James Hoge
and Mr. William F. Weigel of the law firm of Rogers, Hoge & Hills.
Mr. Weigel stated that the firm recognized the seriousness of the charges,
based on the JAMA advertisement but denied any violation of the Food, Drug,
and Cosmetic Act. He also submitted a written statement referring to the
charges set forth in the Notice of Hearing.
Mr. Weigel said that the claims with respect to myocardial blood flow, in-
creased oxygen, and collateral circulation were made prior to October 10, 1962
and were therefore in his opinion grandfathered. This would include the claim
concerning collateral circulation in the mailing piece allegedly supported by
study of Lumb and Hardy.
The written statement asserts that the firm never intended to make any
claim for the use of the drug in the postinfarction period as found in the mail-
ing piece but felt it was necessary to set forth in detail the manner in which
the drug had been used in the Oscharoff study.
The statement continued that the Oscharoff studies, which claimed that 22
percent more patients treated with Paritrate SA at a time closely following
myocardial infarction remained alive after two years than patients treated
with a placebo, were used in good faith and the firm had no reason to question
the results of the study. In using these results as part of their labeling and
advertising, the firm maintained that it was unaware of the fact that technically
it might be making a new claim for the drug. Mr. Weigel indicated that since
no case law was established as yet on interpretation of the advertising provi-
sions of the Act, it was his opinion that it should be established first in civil
actions before criminal prosecution should be instituted.
Count 111.-On July 22, 1966 a second hearing was held with the following
people present, representing Warner Lambert Pharmaceutical Company: Wil-
liam F. Weigel, Esq., Attorney with Rogers, Hoge and Hills, Frank J. DiTrag-
ha, M.D., Corporated Medical Control Director, Mr. Frank Markoe, Jr., Senior
Vice President, Secretary and General Counsel, Warner Lambert Pharmaceut-
ical Company. This was held to discuss the deficiencies of the advertisement in
Medical World News.
Again Mr. Weigel stated that the firm admitted that the shipment had been
made and that the firm issued the advertising which was the subject of the
hearing, but denied any violation of the law. He emphasized that the ad offered
the drug only for use in the treatment of angina pectoris, in accordance with
the policy of the Food and Drug Administration that nitrate-containing drugs
are considered as useful only in the management of angina pectoris. Warner
Lambert denied the charges contending that Dr. Brofman's findings were clin-
ically significant and that the manner in which the subjects were chosen, the
methodology employed, and the evaluation of the results were in accordance
with generally accepted procedures for such studies. Dr. DiTraglia dictated a
prepared statement of his interpretation of the study substantially as follows.
56-592 0 - 75 - pt. 28 - 24
PAGENO="0362"
13612 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Study A as done on 20 patients demonstrated equal therapeutic results for
Peritrate and the placebo, but this was not statistically significant because the
manner in which the study was run left doubt as to whether the effectiveness
of the placebo was not a carry over from the drug which had been administered
just prior to the placebo period. In Study B the drug was discontinued for a
period prior to use of the placebo thus ruling out any carry-over effect from
the drug. Dr. DiTraglia stated that the "Company did not believe that it was
justified in portraying this * * * finding concerning methodology (the differ-
ence between `Study A' and `Study B') and accordingly limited its references
to the reports of those patients who received the active medication in the
approved manner as demonstrated by `Study B' ".
CONCLUSIONS
We believe that the evidence indicates that prosecution is warranted in this
case. The advertisements in question did not meet the requirements of the Act
and would clearly be misleading to physicians who read them. The purpose
of these ads is to provide information to the physician who does not have the
time to read all of the reports in the medical literature. The law provides that
the ad be factual and provide full information on the product, both good and
bad, so that the physician can rely on the representations at to the effectiveness
and use of drugs without going behind the ad into the studies referred to.
Clearly, these ads will not give this service. It is necessary to read beyond
what is reported to be fully informed on the studies.
In addition, the labeling for Peritrate SA issued by the firm, contained
seriously misleading representations with respect to the drug. Mr. Wiegel is in-
correct in his interpretation of the grandfather clause. All claims made before
passage of the Kefauver-arris Amendments are not automatically protected.
Claims made on October 9, 1962 are grandfathered only if the drug was not
a new drug at that time.
Drug companies have the responsibility to follow the law in all respects.
When one company has not met its responsibility, it is open to criminal prose-
cution and we believe that it is warranted in this instance.
WITNESSES
The principal witnesses in this case will be the Government inspectors who
collected the samples and other witnesses to prove interstate commerce; and
medical officers of the Food and Drug Administration's Bureau of Medicine
who can testify as to the approved new drug application, approved labeling,
and with other medical experts the serious nature of the alleged labeling and
journal advertising misbrandings.
It is requested that, if any form of information is filed other than the one
enclosed, the United States Attorney furnish us with a copy thereof. The New
York District of the Food and Drug Administration will provide assistance to
the United States Attorney. Upon request, this office shall render such further
assistance as may be possible.
Very truly yours,
WILLIAM W. GOODRICH,
Assistant General Counsel.
APRIL 23, 1968.
Mr. WILLIAM W. GOODRICH,
A ssistant General Counsel, Department of Health, Education, and Welfare,
Washint,ton, D.C.
Re Warner Lambert Pharmaceutical Co., FDC No. 53548, Federal Food, Drug,
and Cosmetic Act.
DEAR MR. GOODRICH: This is in response to your letter of February 13, 1968,
submitting for our consideration a suggested criminal action against Warner
Lambert Pharmaceutical Company arising out of the introduction into inter-
state commerce on April 27, 1966, of the drug "Proloid" which was alleged to
have been misbranded because of an advertisement which appeared in certain
medical journals during February, April, and May of 1966.
We have carefully examined the allegations set forth in the suggested form
of information, your comments relative to the nature of the violations, and the
four exhibits enclosed in your letter. The information set forth in your letter
and its enclosures is not sufficient to enable us to ascertain the exact manner
PAGENO="0363"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1361,3
in which the advertisement is violative of the statute and regulations. It Is
observed that the narrative account in your letter which endeavors to explain
the offense does so in almost the same langu~ge as the allegations of the Infor-
mation, and provides little or no additional explanation of the reasons why the
advertisement violates the law.
Inasmuch as the advertisement did not on its face appear to create, at leaSt
in the minds of lay persons, the impression and implications seemingly set out
in your letter and suggested form of information, we have discussed this matter
personally with Dr. Robert S. McCleery, of the Bureau of Medicine, to make
certain the position of the FDA and to ascertain the nature of the evidence
available to prove the accusation. We have considered the views of Dr. M~-
Oleery and have studied the materials submitted by him.
It appears that one of the principal reasons prosecution has been suggested
is the contention that the first paragraph of the advertisement creates the im-
pression, in the mind of the reader, that in his paper Greer was comparing
liothyronine with Proloid, when in fact he was comparing it with desiccated
thyroid. Our own realing of the advertisement does not create that impression.
The wording of the advertisement seems to compare liotbyronine, not with
any particular thyroid product, but with more slowly acting thyroids.
In addition, we understand from Dr. McOleery that liothyronine has a more
abrupt effect in raising the metabolic rate than does Proloid, and that as
between Proloid and desiccated thyroid there is practically no difference in the
rate at which the metabolism is raised. Accordingly, the fact that Greer may
have used desiccated thyroid in his studies is of little Importance since in this
area of comparison there is little or no difference between the action of Pro-
bid and that of desiccated thyroid.
We understand that another of the agency's objections is founded upon the
contention that the comparison between liothyronine and Proloid fails to dis-
close that there Is a potential danger of precipitating cardiac complications
from the use of Proloid. However, Dr. McOleery has advised that the omission
refers not to cardiac complications which arise by reason of any abrupt action
on the part of Proloid, but such cardiac reactions as might arise generally from
the use of all thyroid preparations. Since the advertisement is comparing the
dangers arising from an abrupt action of a drug with the effect of a slower
acting product, we fail to see how the omitted information could result in any
misrepresentation.
Another of the agency's contentions relates to the failure of the advertise-
ment to state that Proloid is less rapidly metabolized than liothyronine, and,
therefore, cannot be withdrawn as rapidly when toxic manifestations appear.
Since the gist of the comparison between liothyronine and Proloid is that the
latter is slower in raising the metabolism rate, we inquired whether it would
not be obvious to a physician that Proloid is less rapidly metabolized and were
informed that the rate at which the drug is metabolized is partly dependent
upon and partly independent of the general metabolic rate of the body, and, thus,
it would be to some extent obvious and to some extent not obvious.
In view of this explanation, and the lack of any statement in the approved
labeling which indicates that there is a possibility of such toxic manifestations
appearing as would require immediate lessening of the metabolic rate, this
ground clearly forms no sound basis for criminal action.
The agency contends that the advertisement lacked fair balance because it
failed to include three statements by Greer in two of which he indicated that
other products were not superior to desiccated thyroid; the third deals with
the caution to be exercised in treating elderly patients or those with known
cardiac complications. But since the quotation from Greer is not such as to
create an impression that Greer is of the opinion that Proloid is, in general,
better than other thyroid products, there would seem to be no necessity that
Greer's views as to the relative over-all superiority of the various products or
his caveat concerning treatment of certain types of patients be expressed.
Nor is the statement, apparently true, that Proloid is a more precise prepa-
ration whose standards exceed USP requirements one which would require the
further statement for which the agency contends. The statement in the ad-
vertisement is not a quotation and is referenced to a paper by H. S. Kupperman.
The language quoted as being omitted is taken from Greer. It seems that the
statement in the advertisement is one of fact from which qualified physicians
can form their own conclusions. Under these circumstances the opinions of
PAGENO="0364"
13614 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
others as to the medical conclusions to be drawn from the facts are not
required.
The last quotation, the omission of which the agency argues creates a crimi-
nal offense, does not appear to be related to the use of the drug advertised but
seems to refer to a general medical procedure recommended in the use of all
thyroid products by elderly patients. Such an omission again seems to be only
a failure to set forth an opinion or conclusion of one medical researcher and is
not necessary to complete a factual statement about the drug advertised.
There is apparently some basis for questioning the failure of the ad to disclose
the chemical identity of some of the components of Proloid with those of the
drugs with which the ad undertook to make some comparisons, although we
also understand there may be some difference of opinion as to whether there
is complete identity, as reflected by the statement of the Company. But assuming
the Agency's contention in this regard is supportable, we do not understand
you would contend for prosecution on this factor alone.
Accordingly, it is our view that the matter is not one which justifies the
bringing of a criminal action. Even if the agency contentions are technically
correct, a fact as to which there appears to be at least an arguable basis for a
different view, they involve technical refinement of the nuances which flow
from the statements used as in our judgment to preclude a successful prosecu-
tion. We strongly feel that this is not the kind of a matter in which to under-
take criminal action.
Accordingly, prosecution is declined.
Sincerely,
FRED M. VIN50N, Jr.,
Assistant Attorney General,
Criminal Division.
By HAROLD P. SHAPIRO,
Chier, Administrative Regulations section.
DECEMBER 5, 1966.
The Honorable ATTORNEY GENERAL,
Department of Justice,
Washington, D.C.
Attention Harold P. Shapiro, Chief, Administrative Regulations Section.
Re Wyeth Laboratories, Inc., F.D.C. No. 52677, Federal Food, Drug, and Cos-
metic Act. Your reference: FMV :JWK :ik, 21-62-326.
DEAR MR. ATTORNEY GENERAL: This is in reply to your letter of October 5,
1966. We enclose a revised form of Information which charges the violative
shipment of Serax capsules on July 21, 1965 [in lieu of June 17, 1965], as well
as on April 25, 1966.
At your request, the evidence of the July 21, 1965 shipment has been obtained
In order to avoid charging a shipment which occurred before the advertise-
ment's publication dates of June 28 and July 5, 1965.
The July 21, 1965 shipment charged in Count I of the revised form of Infor-
mation involves the introduction into interstate commerce by Wyeth Labora-
tories, Inc., at Paoli, Pennsylvania, on July 21, 1965, via the Pennsylvania Rail-
road, of 1248 100-capsule bottles of 15 mg. Serax capsules for delivery to Wyeth
Laboratories, Division of American Home Products Corporation, Baltimore,
Maryland.
The April 25, 1966 shipment charged in Count II involves the introduction
into interstate commerce by Wyeth Laboratories, Inc., at Paoll, Pennsylvania,
on April 25, 1966, via Needham's Motor Service, Inc., of 180 500-capsule bottles
of 10 mg. Serax capsules, for delivery to Wyeth Laboratories, Division of Amer-
ican Home Products Corporation, Secaucus, New Jersey.
~The violative shipments were made pursuant to Wyeth Finished Stock Trans-
fer Orders No. 51481 and 12226, as confirmed by signed statements made by the
respective branch managers at the branches where the lots were received after
their shipment in interstate commerce.
We have noted your request concerning copies of the advertisements which
appeared in the April 25, 1966 issue of the Journal of the American Medical
Association, the April 18, 1966 issue of Medical Economics, and in the March
1966 issue of the American Journal of Psychiatry. Because those advertise-
ments do not photocopy well, one set of the actual advertisements has been ob-
tained for your use and is hereby transmitted together with photocopies of the
advertisements.
PAGENO="0365"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13615
The United States Attorney will be supplied with a set of the actual adver-
tisements by the Philadelphia District of the Food and Drug Administration. We
should be happy to render any further assistance as may be possible.
Sincerely yours,
William W. Goodrich,
Assistant General Counsel,
Food and Drug Division.
IN TEE UNITED STATES DISTRICT COUNT FOR THE EASTERN DISTRICT
OF PENNSYLVANIA
United states of America v. Wyeth Laboratories. Inc., a corporation, No. 21
U.S.C. 331 and 333.
Information
COUNT I
The United States Attorney charges:
That Wyeth Laboratories. Inc., a corporation, organized and existing under
the laws of the State of New York, and trading and doing business at Philadel-
phia and Paoli, Pennsylvania, the defendent herein, did, on or about July 21,
1965, within the Eastern District of Pennsylvania, in violation of the Federal
Food, Drug, and Cosmetic Act, [21 U.S.C. 331(a)], unlawfully cause to be intro-
duced and delivered for introduction into interstate commerce at Paoli, Penn-
sylvania, for delivery to Baltimore, Maryland, a number of bottles containing
a drug, namely, Serax;
That displayed upon said bottles was certain labeling which consisted, among
other things, of the following printed and graphic matter:
100 capsules
SERAX (Oxaxepam) 15 mg.
Caution: Federal law prohibits dispensing without prescription.
Wyeth Laboratories Inc., Philadelphia, Pa.
That said drug, when caused to be introduced and delivered for introduction
Into interstate commerce as aforesaid, was misbranded within the meaning of
21 U.S.C. 352(a) in that said drug was a prescripition drug which was a new
drug subject to 21 U.S.C. 355 and said defendant, the manufacturer of said
drug, failed to include in the advertisements caused to be issued by said defend-
ant with respect to said drug in the June 28, 1965 and July 5, 1965 issues of the
Journal of the American Medical Association, a true statement of information
in brief summary relating to the side effects, contraindications and effectiveness
of said drug as required by regulations, 21 CFR 1.103(e) and (f), to wit, (1)
the aforesaid advertisements did not present a brief summary which fairly
showed the effectiveness of said drug in the conditions for which it was recom-
mended in the advertisements, together with a showing of all side effects and
contraindications of said drug that were pertinent with respect to the uses
recommended and suggested in the advertisements, including the information
from the approved new drug application labeling for said drug concerning said
side effects and contraindications, and (2) the aforesaid advertisements lacked
fair balance in presenting with respect to said drug information on effectiveness
and information on side effects and contraindications.
COIYNT II
The United States Attorney further charges:
That Wyeth Laboratories, Inc., a corporation organized and existing under
the laws of the State of New York and trading and doing business at Philadel-
phia and Paoli, Pennsylvania, the defendant, herein, did, on or about April 23,
1966, within the Eastern District of Pennsylvania, in violation of the Federal
Food, Drug, and Cosmetic Act, [21 U.S.C. 331(e)], unlawfully cause to be intro-
duced and delivered for introduction into interstate commerce at Paoli, Penn-
sylvania, for delivery to Secaucus, New Jersey, a number of bottles containing
a drug, namely, Serax;
That displayed upon said bottles was certain labeling which consisted, among
other things, of the following printed and graphic matter:
500 capsules
SERAX (Oxaxepam) 10 mg.
Caution: Federal law prohibits dispensing without prescription.
Wyeth Laboratories Inc., Philadelphia, Pa.
PAGENO="0366"
13616 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
That said drug, when caused to be introduced and delivered for introduction
into interstate commerce as aforesaid was misbranded within the meaning of
21 U.S.C. 353(n) in that said drug was a prescription drug which was a new
drug subject to 21 U.S.C. 355 and said defendant, the manufacturer of said
drug, failed to include in the advertisements caused to be issued by said defend-
ant with respect to said drug in the April 25, 1966, issue of the Journal of the
American Medical Association, in the March 1966 issue of the American Journal
of Psychiatry and in the April 18, 1966 issue of Medical Economics a true state-
ment of information in brief summary relating to the effectiveness of said drug
as required by regulations, 21 CFR 1.105(e) and (f), to wit, (1) the aforesaid
advertisements did not present a brief summary which fairly showed the
effectiveness of said drug in the conditions for which it was recommended in
the advertisements, (2) the aforesaid advertisements lacked fair balance in
presenting information on the effectiveness of said drug, and (3) the aforesaid
advertisements recommended and suggested said drug for uses which were not
set forth in the approved new drug application labeling for said drug.
UNITED STATES ATTORNEY FOR THE
EASTERN DISTRICT OF PENNSYLVANIA.
DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE,
May 15, 1967.
The Honorable ATTORNEY GENERAL,
Department of Jn8tice,
Washington, D.C.
Attention Harold P. Shapiro, Chief, Administrative Regulations Section.
Re Wyeth Laboratories, Inc., FDC No. 52677, Federal Food, Drug, and Cosmetic
Act. Your ref: FMV :JWK :mfs, 21-62-326, 22A-48-20.
DEAR MR. ATTORNEY GENERAL: This is in reply to your letter of February 15,
1967, advising of your tentative conclusion that no violation of 21 U.S.C.
352(n) (3) can be proved against the above named company.
We believe that a violation of that provision has occurred, since the "message"
portion of the advertisements contain false and misleading statements of effec-
tiveness of the drug and those misstatements cannot be corrected by a "true
statement" of such properties elsewhere in the advertisement under the heading
"prescribing information". This conclusion is supported by the language of the
section and by the legislative history of 21 U.S.C. 352(n). We think both show
that an advertisement cannot meet the "true statement" requirement when it
is inconsistent within itself and when it fails to achieve truth in its central
theme.
The legislative history shows that Congress intended 21 U.S.C. 352(n) to deal
completely with the problems of false and misleading advertising. The require-
ment that the advertisement include "a true statement of . . . such other in-
formation In brief summary relating to side effects, contraindications, and effec-
tiveness as shall be required in regulations which shall be issued by the Secre-
tary" was meant to give the Secretary authority to promote truthful informa-
tion in the entire advertisement.
In sponsoring the amendment which restored the section on prescription drug
advertising to the Senate bill, Senator Kefauver pointed out that "eminent
medical authorities testified that much of drug advertising is misleading and
some is false. They further emphasized that when a doctor is misled, his pa-
tient's health is endangered." Senate Report No. 1744, 87th Cong., 2nd Sess., p.
39 (1962) [Emphasis added]. Representative Dingell quoted a study by the
American Medical Association, during debate on this measure, stating "the
most heinous and despicable practice that doctors found was this-'a company
knowingly misleading a doctor through bad advertising' ". 103 Cong. Rec. 21086
(September 27, 1962).
This provision was intended to provide for "truth in drug advertising", ac-
cording to Representative Celler, who likened the desired effect of this amend-
ment to the restrictive regulation of information on securities, stating:
"Our securities laws provide that stock may not be sold to the public against
false or misleading information. I fail to see why we should not be equally
stringent in the sale of lifesaving-but potentially dangerous-drugs.
PAGENO="0367"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13617
"While I am well aware of the many medical miracles wrought by our modern
wonder drugs, and while the drug companies have many splendid accompli~h-
ments of which they may justly be proud, I am not convinced that false or mis-
leading drug advertising is the price we must pay for pharmaceutical progress."
108 Cong. Rec. p. 21086 (September 27, 1962).
Representative Multer referred to the function of the regulations:
"The Secretary's regulations necessarily would be limited to a fair summari-
zation of side effects, contraindications, as well as effectiveness so that a bal-
anced story appropriate to an advertising page could be presented.
* * * * * * *
"There can be no justification for half truths either in an advertising piece Or
in labeling copy. We are dealing with drugs that are lifesaving and which are
used in serious, debilitating diseases. Fairness alone requires that the message
to physicians be a balanced one giving both the good and the bad of the drug,
within the limits of advertising copy, and that the patient's interest not be
obscured by a glowing picture of the effectiveness of the drug with no mention
at all of its possible side effects and contraindications." 108 Cong. Rec. 21091
(September 27, 1962).
These statements illustrate the Congressional concern that the advertisement
not mislead the physicians in any way. Certainly, the most obvious way in which
advertisements mislead is by omission of side effects and contraindications, and
Congress, naturally, focussed a great deal of attention on these facets. HoW-
ever, the frequent use of the term "misleading" and the context in which it
was used, clearly demonstrate that Congress was concerned with the truth of the
whole advertisement.
This concern is evident in a number of examples of drug advertisements
which were presented on the House floor. Representative Blatnik described for
his colleagues "two examples of how ads can mislead". 108 Cong. Rec. 21084
(September 27, 1962). The first was a 3-page advertisement for the drug
"Singoserp". He pointed out that the headline for the ad states: "For the first
time, the drug that lowered this patient's blood pressure without side effects."
Yet elsewhere the ad states that the drug "has infrequent side effects". He
criticized, as well, the use of meaningless illustrations:
"Obviously they had quite a bit of space because you notice at the bottom
one-third of the lefthand page is a picture of a scale. Now that does not provide
any fundamental basic knowledge to a doctor because he knows what a weigh-
ing scale looks like, but it is in the picture."
The ad was printed in several medical journals, although the screening
comittee for the Journal of the American Medical Association rejected it be-
cause: They felt this advertisement, and I quote, "carries a misleading implica-
tion of a broad, too all-inclusive nature."
That criticism goes far beyond the mere omission of a "brief summary" of
side effects, contraindications, and effectiveness. It is directed to the misleading
nature of whole advertisement, and, we submit, reflects the opinion of Repre-
sentative Blatnik, the sponsor of the floor amendment that was adopted, as to
the abuses which the prescription drug advertising provisions were intended to
correct.
The second example was fin ad for the drug "Tao". Again, Representative
Blatnik criticized the waste of space devoted to a picture of a slide rule:
Would you not think they would have used that space to tell what the bad
effects, the side effects might be? In short, to tell the whole story. They do not
need a slide rule to illustrate what the drug does.
He then points out the misleading nature of the ad: Listen to what it says:
On the first page it is implied that Tao is a superior antibiotic. And on the
other page it says:
Tao encompasses even strains of common pathogens (notably straphylo-
cocci) [sicj resistent to penicillin and erythromycin.
They say it is superior to penicillin. What does the American Medical Associa-
tion say about it? They said that this was "altogether too strong and mislead-
ing" a statement. This is a quote about this particular advertisement. This
particular advertisement claimed that the drug has a greater range of efficacy
than penicillin. The American Medical Association committee reported to the
advertising agency that this claim "seems completely false since clinically it is
quite inferior to penicillin in the treatment of most infections."
* * * * * * *
PAGENO="0368"
13618 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The American Medical Association rejected this as being completely .false
and misleading, yet 6 months later the same ad appeared in four widely used
medical journals * * *
Here again, the criticism is directed to the false and misleading nature of the
advertisement as a whole. The statements in the "message" portion of the ad-
vertisement were false and misleading. The addition of a "ture statement" in a
"brief summary" of side effects or contraindications or effectiveness might
afford the physician other information on which to judge the drug, but could
not itself correct the misleading effect of the advertisement. Certainly, the ad-
vertisement would be misleading if it asserts that "Tao" is soperior to penicillin
in the "message" portion even if the "brief summary" said that it was not.
The total effect would be something less that a "true statement" of effectiveness
of the drug.
Similarly, Representative Dingell described an advertisement for the drug
"Medrol", as follows:
"I was surprised to find in my study of material involving one medical prac-
titioner that an ad appeared in one of the standard publications on the subject
showing the photographs of a patient who allegedly suffered from colitis. I was
further surprised to find in my additional study that this so-called reputable
manufacturer who was advertising had actually gone so far as to use X-ray
photos of persons who had not received the drug.These two different photo-
graphs of two different patients appeared in the ad with statements in the ad-
vertisement purporting to indicate that it was the same patient on a before
and after treatment basis, before and after he had received the drug Medrol."
"This is the kind of advertisement that the amendment to the drug adver-
tisement section tries to correct, to assure that the doctor shall receive to the
fullest e~rtent possible the fairest and most complete statement of side effects,
contraindications, and efficacy of the drug in simple form." 108 Cong. Rec. 21064
(Sept. 27, 1962). [Emphasis added]
The false illustrations in that advertisement occurred in the "message" por-
tion, and would not be affected by a true statement of side effects, contraindi-
cations, or effectiveness, "in brief summary", or otherwise. From this example,
the conclusion is inescapable that Representative Dingell believed that section
502(n) was intended to eliminate false and misleading statements throughout
the advertisement.
The drug advertising examples given in debate demonstrate that Congress
Intended to promote truthful drug advertising in all aspects. The concept of
truth in drug advertising would be meaningless if it were interpreted to apply
to only a portion of an advertisement. Such an interpretation would illogically
limit the meaning of the phrase "true statement" and would do violence to the
understanding of the provision by Congress. Rather than preventing half-truths,
the provision would enable advertisements to mislead, as has occurred in the
instant case, by permitting truth to be mixed with fiction.
Certainly, Congress did not want this amendment to promote confusion to
physicians. To limit regulation of the advertisement to examination of a portion
of It-whether labeled "brief summary" or otherwise-could only have that
result. Only if the entire advertisement be subject to the requirement of "a
true statement" can 502(n) have the enforcement effect sought by Congress.
This very Issue was taken up in the 1963 hearing to establish regulations on
prescription drug advertising. The Commissioner stated and the Pharmaceutical
Industry agreed that the "true statement" concept applied to the whole ad
message. On October 1, 1963, Mr. Larrick wrote to Mr. Gesell, counsel for the
Industry, as follows:
I. Fair Balance and Prominence. It seems clear to us from the legislative his-
tory of section 502(n) that Congress intended this new section to deal com-
pletely, and not partially, with the problems of false and misleading advertising
which had been called to its attention. The legislative history clearly shows
that Congress intended the administering agency to have jurisdiction over the
entire advertisement and that the phrase "brief summary" was introduced
only to authorize use of a stripped-down statement of the drug's effectiveness,
side effects, and contraindications when the sponsor wished to limit the size of
his ad.
Our regulations are not intended to prohibit use of graphic presentations,
headlines, or similar advertising techniques. Our basic purpose is to provide
assurance that the advertisement will fairly present the message to the physi-
cian of what the drug will do, what its limitations are, and what side effects
PAGENO="0369"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13619
and contraindications may attend its use. Somewhat different size type may
be used in presenting information with respect to side effects, contraindica-
tions, and warnings than that used in the eye-catching headlines. But the regu-
lations would not permit the concealment, subordination, or deemphasis of this
essential side effect and contraindication information to minimize its disclosure
as a part of the total message the advertisement conveys. With this background,
we will answer the first four questions.
This was placed in the record and was taken to be the understanding of th~
provision and regulations by all parties.
Moreover, this interpretation is consistent with other regulation in this field.
The terms "false" and "misleading" have been uniformly interpreted to apply
to an entire piece of labeling or advertising. Labeling is false and misleading if
it contains ambiguities, exaggerations, half-truths, and over-emphasis in an.V
respect. United States v. 95 Barrels * * * Vinegar, 265 U.S. 438, 442-443 (1924)
V. B. Irons, Inc. v. United States, 224 F.2d 34 (CA. 1, 1957), cert. den. 354 U.S.
923; United States v. 46 Cartons * * * Fairfaj~ Cigarettes, 113 F. Supp. 336
(D.N.J., 1953). Nor do disclaimers in the piece cure false and misleading state-
ments. V. B. Irons, Inc. v. United States, supra; Research Laboratories v. United
States, 167 F.2d 410 (CA. 9, 1948), cert. den. 334 U.S. 843.
The very same considerations apply to advertisements. See Rhodes Pharmacal
Co., Inc. v. Federal Trade Commission, 208 F.2d 382, 387 (CA. 7, 1953) and
Murray Space Shoe Corporation v. Federal Trade Commission, 304 F.2d 270,
272 (CA. 2, 1962) where the Court cites the Food and Drug case United States
v. 95 Barrels of Vinegar, supra, for the proposition that the advertisement as a
whole must be examined, and that if the advertisement contains statements
capable of both a misleading and a truthful interpretation, the advertisement
is misleading. P. Lorillard, Inc. v. Federal Trade Commission, 166 F.2d 44 (C.A.
4, 1950) ; Beckenstette v. Federal Trade Commission, 134 F.2d 369, 371 (CA. 10,
1943). The entire context is examined to determine whether or not the overall
impression is misleading and deceptive to the audience. Federal Trade Commis-
sion v. Standard Education Society, 303 U.S. 112 (1937); P. Lorillard, Inc. v.
Federal Trade Commission, supra. Even though words and sentences of an
advertisement may be literally and tecnically true, when they are framed in a
setting which may mislead or deceive, the advertisement is violative. Becken-
8tette v. Federal Trade Commission, supra.
We enclose a redrafted information for your consideration which charges
that the advertisements in their entirety do not include true statements as to
effectiveness, side effects, and contraindications. We believe that this informa-
tion is sufficient to withstand a motion for a directed verdict.
Very truly yours,
WILLIAM W. GooDRICH,
4ssistent General Counsel,
Food and Drug Division.
PAGENO="0370"
13620 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
PREPARED STATEMENTS
Statement by
Neil L. Chayet, Esquire
as Counsel
on behalf of the
Committee for the Care of the Diabetic
before the
Subcommittee on Monopoly
of the
Select Committee on Small Business
U.S. Senate
January 31, 1975
accompanied by
Dr. Robert Bradley
Chairman
Committee for the Care
of the Diabetic
PAGENO="0371"
COMPETITIVE PROBLEMS IN THE DRTJG INDTJSTRY 13621
Mr. Chairman:
On behalf of the Committee for the Care of the Diabetic,
Dr. Robert Bradley as Chairman and I as Counsel are pleased to testify
today before this Committee on a matter of extreme importance to the
medical and scientific community: the continuing controversy relating to
the UGDP study.
The Committee on the Care of the Diabetic is a group of 180
physicians who represent leading diabetologists and it includes clinicians,
researchers and academicians throughout the United States. The group
was formed in November 1970 shortly after the UGDP results were
released. Its chairman, as I mentioned, is Dr. Robert Bradley, who
is here with the today, and its coordinating, committee consists of Drs.
Henry Dolger. Peter Forsham, Holbrooke Seltzer, and John B. O'Sullivan.
For the last five years. the Committee on the Care of the
Diabetic has been engaged in the continuing and often bitter controversy
which has surrounded the University Group Diabetes Program (UGDP).
This controversy has involved the Congress, the Courts, the National
Institutes of Health, the Food and Drug Administration, the manufacturers,
and scientists, physicians and patients throughout the United States and,
indeed, throughout the world.
PAGENO="0372"
13622 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
At the onset, we wish to state, for the record, the position of
the Committee on the Care of the Diabetic with regard to the TJGDP study.
First it is important to stress what we are not seeking to bring about.
We are not asking that the FDA ignore the UGDF findings completely
or that these findings not be reflected in the labeling accompanying these
drugs. In fact we find it difficult to understand why, some 19 months
after this matter was decided by the Court of Appeals for the First
Circuit, the FDA has not acted definitively to bring about new labeling.
We are seeking new labeling to insure that this study and other
relevant studiks will be made known to physicians and their patients.
However, such labeling must indicate the fact that there is great controversy
surrounding this subject and that a material weight of scientific and medical
opinion does not accept the results of the UGDP study. In short, what we are
seeking is fair balance in the governmental handling of this matter. I realize
that nearly eight million dollars is a great deal of money and t en years is a
very long time, and criticism of such an effort is neither to be given nor
taken lightly; nevertheless, costly mistakes in research, as in all human
endeavors, have been made before, and will be made again, and criticism
and controversy will not disappear merely because those who seelc to justify
this study pretend that it eit her does not exist or is of no account.
We have set out in this testimony some of the fundamental prob-
.terns confronting analysis of the UGDP as a study and the Biometric
Society's report on that study.
PAGENO="0373"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13623
I. The need for science tobe free from governmental in-
tervention where there exists validsç nti controverS~y.
The power of government is awesome. It must be
wielded with discretion as will as tact, with equity as we.l.l
as understanding. Partisanship by government throws what
can be overwhelming weight into the balances of scientific ex-
change. Such partisanship fosters positions which are both
dangerous and incomprehensible.
After repeatedly seeking the raw UGDP data since
1971, our Committee most recently received a letter from
Dr. Whedon of the National Institutes of Health, dated January
27th, stating: `To my knowledge, no one in the Department
of Health, Education and Welfare has ever had any of the raw
data of the UGDP study." In effect, this data is not available
to the experts of our Committee.
Despite our inability to obtain TJGDP raw data, we did
obtain the Biometric Society report this past Tuesday which
states:
"Dr. C. Klimt, the Director of the Coo rdinatiCet
and his staff provided extensIve tabulations and original data
of the UGDP trial."
PAGENO="0374"
13624 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The Committee on the Care of the Diabetic has re-
peated.ly protested the partisanship of government agencies
in this controversy. The release to the press of the UGDP
study before its scientific presentation triggered a storm of
controversy. Our criticism of multiple flagrant defects in
the UGDP study was and still is, apparent.ly, considered ~
challenge to the National Institutes of Health which expended
over $7. 7 million for a study which neither tested oral hypo-
glycemic therapy as it is practiced nor provided any insights
into insulin therapy.
* For years patients were carried in a study monitored
by NIH personnel- -a study which has become the subject of
severe cxiticism by many of the leading diabetologists in the
United States.
* Otter agencies in HEW, such as the FDA, have enter-
ed the scene and, despite the clear existence of valid scientific
controversy, the government has to this day sought to impose
its will by excluding from ora.l hypoglycemic package inserts
the very "fair balance" required by its own regulations. When
challenged in the courts, it seeks by regulatory means to ex-
empt itself from the preexisting requirement of fairness.
PAGENO="0375"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13625
* In a manner not clear to us, a principle investigator
of the UGDP was engaged by the FDA an d appears from doc-
uments available to us to have participated in the preparation
of FDA material relating to the legal aspects of this contro-
versy.
* When chaflenges to the validity of the study cou.ld not
be silenced, the NIH arranged for Dr. Thomas C. Chalmers
to have the Biometric Society undertake an ana.lysis primarily
of the statistical aspects of the study, despite the fact that
many if not most of the central criticism was clinical in na-
ture. The same Dr. Chalmers now appears as the author of
an editorial in the S. A. M. A. in which he modestly omits his
personal role as a government official in arranging for the
Biometric Society study as he proceeds to hail its findings,
to support it with selections of references relating to experi-
mental and human studies, while ~
reports in the literature.
I do not know what this is called in science. In politics
it has been called a "cover-up". There is need for further in-
quiry into the whole area of the treatment of diabetes and also
the role of government agencies and individuals in the UGDP
affair.
PAGENO="0376"
13626 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The government has continually been a party at
interest in the UGDP controversy. Government officials
should have been sensitive to American judicial and sci-
entific processes and should have delegated both the selec-
tion of the review body and its mandate to an uninvolved,
nonpartisan group. Instead, its actions, in effect, were
equivalent to being judge, jury and prosecutor, publicist
and apologist.
PAGENO="0377"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13627
Ii. The rights of patie nts tp be ull~free1yn properly
info rmed
Diabetes is strongly affected by emotional factors.
Confidence in the physician-patient relationship is essential
for optimal management of diabetes.
As has happened so often in the past with the UGDP
study, patients have been exposed to frightening headlines
and unsubstantiated charges before the actual report has either
appeared in the scientific press or been presented to a jury
of its peers. This week, newspapers, TV and radio were
flooded with statements unsupported by data before even the
experts and diabetologists of our Committee could obtain any
re.levant information. It does not serve the interests of pa-
tients nor of the physician-patient relationship, nor of science
itself to repeatedly confront such situations in which there is
neither time for suitable study and deliberation nor for cor-
rective action on the part of the medical profession.
This week, the worried diabetic who coflsults his physi-
cian can get little comfort from his doctor who has no data on
which he can base an informed opinion on the Biometric Report.
Both Dr. Chalmers editorial and the Biometric Report will not
appear in J. A. M. A. until February. How is the diabetologist
56-592 0 - 75 - pt. 28 - 25
PAGENO="0378"
13628 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
or the practicing physician to know upon what Dr. Chalmers
(in editorial and press release) has based his estimate of
10, 000 to 15, 000 deaths; how can they determine, as we
have, the fact that some of the very references Dr. Chal-
mers puts forth in support of the validity of the UGDP study
are in reality in contravention of the UGDP findings? Trial
of medical treatment by headline and press release is as bad
for science as it is for law, and results, medically, in dis-
ruption of patient treatment as it raises unsupported doubts.
PAGENO="0379"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13~29
III. Observations on the Biometric Report and the
Chalmers Editoria.L
At the outset, it should be noted that the authors of
the Biometric report do not claim tO be qualified to appraise
the clinical aspects of the TJGDP study. The Biometric re-
port states: `The choice of specific selections criteria
adopted by the UGDP was a responsibility that was shared
with medical experts and is not a topic on which this com-
mittee as a whole claims primary competence." This dis-
avowal of clinical ~ompetence was well taken, as will be
seen below.
The Biometric report recognizes that "the result
~f [the UGDP] decision to terminate the study [is toj leave
us with some residual uncertainty as to the meaning of the
findings."
The report further states, "We discovered a puzzl-
ing anomaly concerning the distribution of the two sexes in
the four treatment groups within clinics... The discrepancy
in Seattle alone would represent an unusual event in random
allocation... A more important point is whether these find-
ings provide evidence of a breakdown of the randomization
PAGENO="0380"
13630 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
procedure- -a contingency which might have grave impli-
cations for the credibility of the whole study... We were
not able to find an assignable cause for the surprising a Ilo-
cation of the sexes to treatments but have no reason to think
the study has been compromised by a breakdown of the ran-
domization to the treatment groups." *
The Biometric Society could have considered other
eviden cc of a possible breakdown of the randomization pro-
cedure if they examined the two-fold and even more than
three-fold difference in total time at risk as between male
and female patients, differences in compliance, autopsy
rates, etc. Among the defects in the UGDP study, the Bio-
metri~ report states, `The omission of a history of smoking
was a blunder. U This is praising with a faint damn as there
were other blundering omissions such as the failure to iden-
tify patients on thiazides, as well as the failure to identify
those patienfs with family histories of heart disease.
The Biometric Society mildly observed that the UGDP
report "made use of some relevantly unfamiliar and explora-
tory techniques'. These very techniques have, for several
years, caused critics to raise the question of statistical man-
ipulation of data.
* It is interesting to note that the calculation of age at death of the data presented
by the UGDP shows that alt tolbutamide patients, males and females, had a mean
age of 65. 2 and all placebo patients 61. 5 and the difference at death between
placebo females and toibutamide females was six years, 59. 8 for placebo and
65. 8 for tolbutamide group.
PAGENO="0381"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13631
The Biometric Society report notes that "On the
question of cardiovascular mortality due to toibutamide
and phenformin we consider that the UGDP trial h~s raised
suspicions which cannot be dismissed on the basis of other
evidence presently available." However, we believe the
Biometric report incorrectly minimizes some of the evi-
dence which it analyzes, as in the case of the Paasikivi study
which is inconsistent with their thesis of the toxicity of tol-
butamide. They omit a recent report of a prospective study
which has become a landmark in the field of epidemiology,
that of the Framingham group. We have included a copy of
this report, published in the American Journal of Cardiology,
as an appendix to our testimony.
The Biometric group then focuses upon a critical issue:.
"There remains the gus stion of generalization of thes~
findin~. As has been frequently point out, t conditions
of drug use in this study were, to some extnt abnormaL
ud~eiS ~ pentu~-
able to obtain inadequate control on tolbutamide could be
shifted to insulin."
PAGENO="0382"
13632 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
In the conclusions of the UGDP group there is a brief
warning against extrapolation of UGDP findings to other dos-
ages and other types of patients.
One could never judge from the press releases, the
news stories, the public actions of government officials that
both reports related to a hypothetical test situation and not to
the actu!tp~actice of oral hypoglycemic treatment. Under
certain circumstances, the Biometric report finds that
"There remains the question whether tolbutamide, although
ineffective in a fixed dose regimen, might be an effective
therapy as ordinarily used." I am sorry that I cannot supply
the page citation but wi.ll do so when it is published next month.
In their analysis, the Biometric Report apparently ac-
cepts the study of Keen et al as statistically valid and atten-
tion is drawn to the fact that those authors concluded "a sig-
nificant degree of primary protection against cardiovascular
events can be conferred by tolbutaniide in moderately and
mildly hyperglycemic people."
In concluding their analysis of the Paasikivi study
they state, "This study neither confirms nor contradicts the
UGDP findings as the population under consideration was not
PAGENO="0383"
COMPETITIVE PROBLEMS IN THE DRtTG INDUSTRY 13633
one of maturity onset diabetes, and the patients taking tot-
butamide had been exposed to a relatively small dose for a
shorter time than applied in the UGDP study." H owever,
the report here is internally inconsistent for the Biometric
group elsewhere holds that `The concern about possible tol-
butamide toxicity would not really be lessened if it could be
shown that the study group contained some nondiabetics. A
drug found toxic in such subjects would not likely be counted
safe for well documented diabetics either."
If the authors of the Biometric Report are suggesting
that cardiotoxicity ~~se is a central issue, then it becomes
difficult to comprehend their failure to accept the Paasikivi
report which in any consideration of toxicity qua toxicity is
most pertinent. This study was carried out on 178 survivors
from a first myocardial infarction. The sensitivity of a dam-
aged heart to cardiotoxic substances is reflected in the nar-
rowing margin between effective and toxic dose of the cardio-
glycosides when used in such situations. It would be remark-
able that a cardiotoxic agent given in the period of greatest
danger "during the first 12 months [after myocardial infarc-
tion] would show a significant difference in favor of tolbuta-
mide." Fifteen controls died and only six tolbutamide pa-
tients--2-1/2 times more controls than the treated patients.
PAGENO="0384"
13634 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
Then 19% of the control group died as compared with 14%
of the tolbutamide group with no significant difference in
survival after five years. Such a study would weigh heavily
against any conclusion that cardiotoxicity was involved in
the UGDP tolbutamide patients.
It is also disturbing to note that neither the Biometric
Report nor the Chalmers editorial saw fit to call attention to
the recent report on the Role of Diabetes in Congestive Heart
Fai.lure, The Framingham Study" (Am. Jrnl. Cardiology 34:
29-34, July 1974). The findings are pertinent to any consid-
eration of cardiotoxic effects of oral hypoglycernics. May we
quote from the paper:
`Role of Insulin;
Further examination of the ~group of patients with dia
betes who had congestive_heart failure revealed that more than
half were takin~jnsuiin. Treatment of diabetes was therefore
su~j~cted to analysis revealing that the on.~y~subgroup of dia
betic subjects that sustained a substantial and statistically
g~ficant increased risk of co~gq~tive failure was the gr~
treated with insulin. This was demonstrated_~yabivariate
regression ana~ysis accounti~gj~r age (Table 9)."
PAGENO="0385"
COMPETITIVE PROBLEMS IN THE DRIJG INDIJSTRY 13635
The table reveals that insulin increased the risk of
congestive failure over oral hypoglycemic agents and other
treatments in both men and women aged 45 to 74 years.
This finding, from this cias sic non-controversial prospec-
tive study of 18 years which reviewed 5, 209 men and wo-
men, is in contradiction to the findings of the briefer UGDP
study reporting on 823 patients. The Framingharn study is,
however, consistent with the findings of Paasikivi study.
We find it exceedingly disturbing that the great
Framingham study, acknowledged as a true landmark, was
neither referred to by Chalmers in the editorial or in the
Biometric Report. To the best of our knowledge it a.lso has
been conspicuous by its absence from press releases.
Our cursory review of the UGDP study and other
studies such as the Framingharn study points out that valid
scientific controvers~ exists with respect to any conclusions
to be drawn in this area at this time. There exists important
scientific studies challenging some of the fundamental assump-
tions and conclusions drawn from the UGDP study. In the light
not only of the UGDP findings but of the controverting studies
there is need for new, truly unbiased and well controlled objec-
PAGENO="0386"
13636 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
five prospective studies. Action by this Committee can
make a significant contribution to the future of American
medicine and science by recommending that U. S. govern-
mental agencies divorce themselves from partisan partici-
pation in the presence of scientific controversy and that
government follow a doctrine of equal opportunity for re-
search and studies by qualified individuals holding different
views.
PAGENO="0387"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13637
American Journal of Caraiology 3~i: 29-3l~. (July 197lt).
Role of Diabetes in Congesfive Heart Failure:
The Framingham Study
WILLIAM B. KANNEL, MD, FACC
MARTHANA HJORTLAND, PhD'
WILLIAM P. CASTELLI, MD
Framin,gham, Massachusetts
Bethesda, Maryland
The Incidence of congestive heart failure was determined in relation to
prior diabetic status in 5,209 men and women aged 30 10 62 years fol-
lowed up for 18 years in the Framingham study. Men aged 45 to 74
years had more than twice the frequency of congestive failure as their
nondiabetic cohorts, and diabetic women had a fivefold increased risk.
This excessive risk appears to be caused by factors other than pccel-
erated atherogenesis and coronary heart disease. Even when patients
with prior coronary or rheumatic heart disease were excluded, the dia-
betic subjects had a four- to fivefold Increased risk of congestive heart
failure. In women (but not men) with prior coronary disease, diabetes
also imposed a threefold increased risk of congestive failure. Fudher-
more, the Increased risk of heart failure in the diabetic patients persist-
ed after taking into account age, blood pressure, weight and cholester-
ol values as well as coronary heart disease. Women with diabetes ap-
peared to be especially vulnerable and, irrespective of coronary dis-
ease status, had twice the frequency of congestive heart failu~e as
men. The excessive risk of heart failure among diabetic subjects was
confined to those treated with insulin. The data suggest that diabetes is
another discrete cause of congestive heart failure and that some form
of cardiomyopathy is associated with diabetes, as a result of eñher
small vessel disease or metabolic disorders.
Congestive heart failure is a common end stage of heart disease due
to a variety of causes. The incidence is far from trivial. The annual
rate is 2.3/1,000 men and 1.4/1,000 women aged 30 years and over.'
Despite the availability of potent glycosicles and diuretic agents, Con-
gestive heart failure continues to be a lethal process, and half of the
patients die within 5 years of onset.1 Previous study' revealed that
hypertension and coronary heart disease were the dominant cau~ies,
but 14 percent of men and 26 percent of women with congestive flail-
ure also had diabetes, an apparent excess. The purpose of this report
is to explore the role of diabetes in the development of congestive
heart failure and to assess its contribution taking into account the
presence of coronary heart disease and atherogenic factors such as
hypertension, high serum cholesterol levels, overweight and increased
age.
Methods
From the Framingham Heart Disease Epidemiol-
ogy Study, Framingham, Mass., and the Nation-
al Heart and Lung Institute, National Institutes
of Health, Bethesda, Md. Manuscript accepted
February 27, 1974.
Address for reprints: William B. Kannel, MD.
Framingham Heart Disease Epidemiology Study,
123 Lincoln Sl., Framingham, Mass. 01701.
The Framingham study was initiated in 1949 to explore the epidemiology
of cardiovascular disease in a general population sample of 5,209 men aud
women aged 30 to 62 years. These subjects have been followed up for the de-
velopment of cardiovascular disease including congestive heart failure. At
every biennial examination each participant has had, in addition to a histocy
and physical evaluation, a 13 lead electrocardiogram, a chest X-ray film, tests
of vital capacity, urinalysis, measurements of blood sugar, uric acid and cho-
lesterol levels and deterr~inations of Framingham relative body weight,
Detailed descriptions of the sampling procedure, response rate, methods q(
examination and laboratory procedures and the criteria for the outcome of
disease have been reported previously.2
July 1974 The American Journal of CARDIOLOGY Volamo 34 29
PAGENO="0388"
13638 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE I
Criteria for Congestive Heart Failure
Major criteria
Paroxysmal nocturnal dyspnea or orthopnea
Neck vein distension
Rates
Cardiomegaly
Acute pulmonary edema
S, gallop
Increased venous pressure »=l6cm H,O
Circulation time »=25 second
Hepatojugular reflux
Minor criteria
Ankle edema
Night cough
Dyspnea on exertion
Hepatomegaly
Pleural effusion
Vital capacity decreased 1/3 frOm maximum
Tachycardia (rate »=120/min)
Major or minor criterion,
Weight loss »=4.5 kg in 5 days in response to treatment
Patients were considered to have congestive heart failure
if two major or one major and two minor criteria were present
concurrently.
TABLE II
Annual Incidence of Congestive Heart Failure by Age, Sex and
Presence of Prior Coronary or Rheumatic Heart Disease:
Year Follow-Up Study
it
Total Pxpulatino
3D xi ROD
*
Inci-
Subjects wi
lb Prinr C
Persnn
New
denco
Persno
New
Yearn at
CIIF
per
Years at
COP
tncidence
Age (ye)
Risk
Events
l0OOt
Risk
Events
per 10003
Men
45-54
14.100
28
20
1,074
17
158
55-64
10,414
43
41
1,564
28,
179
65-74
3,700
26
70
762
14
184
Women
45-54
17.598
ii
6
678
5
74
55-64
13,688
45
30
1,202
24
200
65-74
5,232
34
65
766
20
265
CHD = coronary heart disease; CHF = congestive heart
failure; RHD = rheumatic heartdisease.
Criteria
At each biennial examination a diagnosis of c~ngestive
heart failure was entertained on clirsical grounds and the
opinion of a second examiner obtained. All suspected cases
thsis uncovered from biennial clinic examinations or from
interim ivfvsrmation olutained from hospital records and
physician's office repvsrts were reviesved by a panel of issves-
tigalors vising unifusrns criteria. The diagnosis of congestive
heart failure was accepted only in persouss with at least tuvo
major or one major and tsvo minsr criteria present concur-
rently as indicated in Table I. Minor criteria that could be
attributed to some other medical conditisan were rejected.
Only about half the persons diagnosed as having congestive
heart failure at the time of the clinical examination on the
basis of the examination or interim hospital or physician's
office information were included in this study.
Persons who had congestive failure at the time of the ini-
tial examination were excluded, leaving a population of
5,192 men and women aged 30 to 62 years at risk. Follow-
up study during the ensuing 18 years was reasonably coin-
plete, with 80 percent of subjects receiving every possible
biennial examination. The remaining 20 percent were seen
at less frequent intervals, and admissions to the only gener-
al hospital in town were monitored daily, Only 2 percent of
the sample were completely lost to follow-up study.
Glucose intolerance was assessed from casual blood sugar
determinations, urinary sugar values, or a history of clinical
"diabetes." Blood sugar levels were determined by the
Somogy-Nelson method. A diagnosis of diabetes was made
in subjects who (1) had an abnormal glucose tolerance test
during hospitalization or their physician's laboratory eval-
uation, (2) were taking insulin or oral hypoglycemic agsnts,
or (3) had casual blood sugar values> 160 mg/100 ml.
Statistical Techniques
Incidence rates for congestive failure were ascertained
according to diabetic status, age, sex and coronar3i (or rheu-
matic) heart disease status. Subjects were reclassified by
age at each examination; a case svas defined as a subject
free of congestive heart failure at a given examination but
having failure at the time of the next biennial examination.
The method of Mantel.Hasnzsl3 was used to construct
summary chi squares to assess differences in the frequency
of congestive heart failure in diabetic and nondiabstic
subjects and to estimate the rink of heart failure in the
presence of diabetes. To assess the joint and net effect of
diabetes taking into ~ccosmnt other related atherogenic vari-
ables, musitivariate coefficients were computed and com-
pared with bivariate coefficients taking only age into ac-
count. Analysis was confined to subjects aged 45 to 74 years
since too fesv cases occurred before age 45 for meaningful
analysis. The bivariate function to assess the regression of
incidence of congestive heart failure on diabetic status and
age svas estimated by the nuethod of Walker and Duncan.4
The mnultivariate function wan estimated by the Walker
and Duncan maximal likelihood method using, in addition
to the variables of age and diabetic status, systolfc blood
pressure, serum cholesterol and relative sveight.
Results
Frequency of congestive heart failure: During
the 18 year follow-up period, congestive heart failure,
as defined by the specific criteria, developed in 97
men and 86 women aged 45 to 74 years. The inci-
dence of heart failure increased sharply with age. As
expected, the incidence was considerably greater in
the subjects with prior coronary or rheumatic heart
disease. There was a male predominance at all ages
(Table II), but this predominance appeared to wane
with advancing age. Among subjects with prior coro-
nary or rheumatic heart disease, there wax a male
predominance only under age 55 years. About half of
the subjects with a diagnosis of congestive heart fail-
ure had coronary disease; more than three-fourths
had hypertension. About 16 percent had antecedent
diabetes, an apparent excess over the expected rate.'
30 July 1974 The American Journal of CARDIOLOGY Volume 34
PAGENO="0389"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13639
TABLE Ill
Annual Incidence of Congestive Heart Failure According to Age Sex and Diabetic Status at Each Biennial Examination:
18 Year Follow-Up Study
Diabetic States
Age 4
Person Years
at Risk
5 to 54 Years
New CIII
Cases
Incidence
CHF
perJlO,000
Age
55 to 64 Years
*
Age
65 to 74 Years
Person Years
at Risk
New CIII
Cases
Incidence
CIII
per/lO,000
Person Years
at Risk.
New 6SF
Cases
ln~idence
.CHF
per/10,000
Men°
.
Nondtabetic
Diabetic
Total
13,696
404
14,100
26
2
28
19
50
20
9,864
550
10,414
40
3
43
41
55
41
3,428
272
3,700
20
6
26
58
221
70
.
Women'
.
Nondiabetic
Diabetic
Total
17.268
330
17,598
8
3
11
5
91
6
13,156
532
13,688
36
5
41
27
94
30
4,904
328
5,232
25
9
34
57
274
~5
* Meo: For alt ages combined difference in incidence is statistically sigsificant at(P <0.05). (chi square 6.50); relative risk 2.4. Women:
For all ages combined difference in incidence is statistically sigsificant (P <0.01) (chi square 42.54); relative risk 5.3. Method of Mantel-
Haeozel.
Diabetes and congestive heart failure: During
the 18 year study, diabetes was present or developed
in 141 men and 151 women, allowing 1,226 person
years of follow-up experience for men, 1,190 for
women, Almost 40 percent of those labeled "diabet-
ic" were treated with insulin. In most of these pa-
tients, diabetes was already evident at entry into the
etudy. Another 40 percent were treated with oral hy-
poglycemic agents and about 20 percent were fol-
lowing a diet only or were receiving no treatment. Ex-
amination of the incidence of congestive heart failure
according to diabetic status at each biennial exami-
nation' revealed a distinct excess risk in diabetic
subjects of both sexes and at all ages (Table III). The
small numbers do not allow accurate age-specific es-
timates of risk, but for all ages combined, men had a
2.4-fold increased risk and women a 5-fold increased
risk (Table IV).
Role of factors leading to atherogenesis and
coronary heart disease: Diabetic subjects are gen-
erally believed to have higher, than average blood
pressures, lipid values and relative weights and to
manifest coronary heart disease at an accelerated
rate. The excess frequency of congestive heart failure
in diabetic subjects could derive from these abnor-
malities, particularly the hypertension° and coronary
heart disease. However, an examination of the preva-
lence of antecedent coronary or rheumatic heart dis-
ease in all subjects with congestive heart failure re-
vealed the same proportion of subjects with prior
heart disease in the diabetic and nondiabetic groups
(Table V). This finding suggests that the increased
risk of congestive heart failure in the subjects with
diabetes has another cause than accelerated athero-
genesis and coronary heart disease although age and
severity of coronary disease are not taken into ac-
Risk of Congestive Heart Failure According to Sex an
Status at Each Biennial Examination: 18 Year Follow
d Diabetic
-Up Study
Incidence
Credo Age-
Perssn Years Annual Adjusted'
Diabetic Status At Risk per 10,000 per 10,000
Relative
Risk
Men Aged 45 to 74 years
Nondiabetic 26,988 31,87 32.14
Diabetic 1,226 89.72 75.98
2 36
`
Women Aged 45 to 74 years
Nondiabetic 35,322 19.53 19.75
Diabetic 1,190 142.85 101.60
5 14
`
o Indirect method,
Significartt at P <0.05 (chi square = 6.50).
Significant at P <0.01 (chi square = 12.53).
TABLE V
Prevalence of Antecedent Coronary (or Rheumatic) Heart
Disease (CHD) Among Patients Aged 45 to 74 Years with
Congestive Heart Failure (Diabetic vs. Nondiabetic
Subjects): 18 Year Fellow-Up Study
Diabetic Subjqcts Nendiabelic Subjects
With CIII Without CHD With CHD Without CHD
Men
b
6
54
32
Women
10
7
39
30
Total
15
13
93
62
TABLE IV
July 1974 The American Journal of CARDIOLOGY Volume 34 31
PAGENO="0390"
13640 COMPETITIVE PROBLEMS IN
TABLE VI
Annual Incidence of Congestive Heart Faiture (CHF) According
to Sex and Diabetic Status at Each Biennial Examination
Excluding Subjects with Coronary (or Rheumatic) Heart
Disease Before the Development of Failure: 18 Year
Follow-Up Study
Annual Incidence
Diabetic Status
Person New Age~
Years At CHF Crude Adjusted'
Risk Cases per 10,000 per 10,000
Relative
Risk
Men aged 45 to 74 Years
Nondiabetic
Diabetic
Total
23,844 32 13.42 13.53
970 6 61.86 51.41
24,814 38
*
Women Aged 45 to 74 Years
Nondiabetic
Diabetic
Total
32,892 30 9.12 9.23.
980 7 71.43 50.54
33,872 37
5.48t
Indirect method.
Significant at P <0.51 level (chi square = 0.15 and 17.27 for
men and women, respectively).
TABLE VII
Annual Incidence of Congestive Heart Failure (CHF) According
to Sen and Diabetic Status at Each Biennial Enamination in
Subjects with Coronary (or Rheumatic) Heart Disease Before
the Development of Failure: 18 Year Follow-Up Study
Incidence
Diabetic Status
Persnn New Age-
Years CHF Crude per Adjusted'
At Risk Cases 10000 per 10,000
Relative
Risk
Men Aged 45 to 74 Years
Nondiabetic
Diabetic
Total
3,144 54 171.76 172.07
256 5 195.31 190.99
3,390 59 ... ...
1 11
`
Wnmen Aged 45 to 74 Years
Nondiabetic
Diabetic
Total
2,436 39 160.10 147.39
210 10 476.19 465.17
2,646 49 ... ...
16
~
Indirect method.
Significant at p <0.01 level (chi sqnare = 8.51).
count in this tabulation. This hypothesis is confirmed
by the finding that even when patients with prior
coronary or rheumatic heart disease were excluded,
patients with diabetes still had a four- to fivefold in-
creased rink of congestive heart failure (Table VI).
Also, among pernons with prior coronary or rheumat-
ic heart disease,.diabetic women had an excess rate of
THE DRUG INDUSTRY
congestive heart failure. This excess could not b
demonstrated for men in the small number of case
available (Table VII).
Furthermore, a comparison of the regression coef
ficients in the bivariate (taking only age into ac
count) and multivariate case (taking into accoun
blood pressure, cholesterol and relative weight a
well) indicates that the effect of diabetes is not me
diated through these atherogenic traits (Table VIII).
An examination of the regression of the incidence of
congestive failure on diabetic status in men with and
without coronary heart disease also reveals substan-
tial and significant coefficients only for patients
without coronary heart disease. The regression coeffi-
cients in the multivariate case are only slightly re-
duced compared with those in the univariate case
(Table VIII).
Taking all these facts into consideration it appears
unlikely that diabetes promotes congestive failure by
accelerating coronary atherogenesis. Nor does it ap-
pear that hypertension accounta for the increaned
risk. In women, significant regressions are noted in
both those with and those without prior coronary
heart disease although the coefficients are somewhat
larger in the latter group. Also, the coefficients are
substantially larger in women than in men.
Role of insulin: Further examination of the group
of patients with diabetes who had congestive heart
failure revealed that more than half were taking insu-
lin. Treatment of diabetes was therefore subjected to
analysis, revealing that the only subgroup of diabetic
subjects that sustained a substantial and statistically
significant increased risk of congestive failure was the
group treated with insulin. This was demonstrated by
a bivariate regression analysis accounting for age
(Table IX).
Discussion
The finding of an increased rink of congestive heart
failure in diabetic subjects is not unexpected in view
of the frequent association of diabetes with hyperten-
sion, hyperlipidemia, obesity and coronary heart dis-
ease. The strength of the relation, particularly in
women, and the demonstration that the excess risk of
myocardial decompensation in the diabetic subject is
independent of all these atherogenic traits are unex-
pected and indicate some other mechanism. It has
been suggested that some form of cardiomyopathy is
associated with diabetes.6 The findings reported
herein lend some substance to this claim.
Metabolic causes of diabetic cardiomyopathy:
Such diabetic cardiomyopathy could result from dig.
betic microangiopathy or an abnormal myocardial
metabolism. The heart normally derives most of its
energy for contraction from free fatty acids, although
glucose, lactate and pyruvate are also used.7~ Any
reduction in oxygen tension immediately produces
changes in the electrical and contractile performance
of the heart. Efficient metabolism of fatty acid and
pyruvate is hampered by hypoxia, leaving only the
glycolytic pathway to generate the high energy phos.
32 July 1974 The American Journal of CARDIOLOGY Volume 34
PAGENO="0391"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13641
TABLE VIII
Regression of Incidence of Congestive Heart Failure on Diabetes: 18 Year Follow-Up Study
Regression Coefficient
Men Aged 45 to 74 Years Women Aged 45 to 74 Years
Bivariate* Moltioaniatef Bisariate' Moftivaniatet
With Prior CHD or RHDt
.
Coefficient
0.10 006 1.13
ttestl
1.15
0.21 0.12 3.05 2.97
(no. cases/no, at risk)
(59/1,700) (58/1,641) (49/1323)
Without Prior CHD or RHDJJ
(47/1,275)
Coefficient
1.33 1.01 1.67
itest
2.94 2.17
(no. cases/no, at risk)
3.88 3.15
(38/12,407) (36/11,899) (37/16,936) (37/16,16ô)
0 The bivariate function was estimated bythe method of Walker and Duncan' using the variables diabetes and age.
The multivariafe function was estimated by the method of Walker and Duncan' using the variables diabetes, age, systolic blood
pressure, serum cholesterol and Framiogham relative weight.
To be included in this group at a given examination (n) the subject most have had coronary or rheumati~ heart disease on or be-
fore examination (n + 1).
I At test value of 1.96 is significant at the 0.05 level.
II To be included in this group at a given examination (n), the subiect must not have had coronary or rheumatic heart disease on or
before examination (n + 1).
phate bonds to fuel the heart'o work. It thuo appearo TABLE IX
that glucose and inoulin fuel the failing hypoxic
heart. ~ This hypotheois would explain the difficulty
of the diabetic iochemic heart but does not account
for myocardial decompensation in the absence of ________________________________________________
hypoxia.
However, in the diabetic heart there is come evi-
dence to suggest that free fatty acids are also not
used efficiently. Hearts of rats with alloxan-induced
diabetes have been found to accumulate increased
myocarçlial triglyceride in lipid droplets.8 Human di-
abetic hearts extract more fatty acid and ketonen
than the nondiabetic heart, but in the light of the
foregoing this phenomenon could indicate that more
free fatty acid is being shunted into structural lipid
as less is metabolized for energy. It may be that the
diabetic heart, because of faulty utilization of fatty
acid must, as in the case of ischemia, fall back on gly-
colytic metabolism for energy. Thin places it in jeop-
ardy because the utilization of glucose, which is insu-
lin-dependent, is also faulty. Thin energy crisis may
be further compounded in the diabetic subject by a
block between pyruvate and the tricarboxylic acid or
Krebs cycle.
These major metabolic disturbances could provide
a metabolic basin for eventual myocardial failure.
The fact that only the insulin-dependent diabetic pa-
tient appears peculiarly susceptible to congestive
heart failure supports the foregoing in that only these
diabetic patients appear to have difficulty with ex-
cessive and faulty acid metabolism leading to ketosis
as well as impaired glycolysis. The fact that subjects
with diabetes of adult onset treated by diet or orally
administered drugs have no increased risk of conges-
tive heart failure suggests that the central metabolic
Regression of the Incidence of Congestive Heart Failure
(Without Prior Coronary or Rheumatic Heart Disease')
on Treatment of Diabetes: 18 Year Follow.Up Study
Bieariate
Treatment Cssfficientf t Valse~
Men Aged 45 to 74 Years
Insulin 1.76 3.26
Oral hypoglycemic agent 0.68 0.66
Other 0.92 0.90
Women Aged 45 to 74 Years
Insulin 2.21 4,47
Oral hypoglycemic agent 0.56 0.54
Other 1.49 1.45
0 To be included in this group at a given examination (n), the
subject must not have had coronary or rheumatic heart disease
on or before examination (n + 1).
The bivariate function was estimated by the method of
Walker and Dnncanl using the variables type of diabetic treØt-
ment and age.
Al value of 1.96 is significant at the 0.05 level.
defect promoting failure may be ketosis and insulin
deficiency. Once heart failure ensues, the process ap-
pears to be self-perpetuating since it has been shown
that failure further suppresses insulin release.'°
Role of insulin-treated diabetes: The finding
that the excess risk of congestive heart failure is con-
fined to the insulin-treated diabetic subjects raise~
July 1974 The American Journal of CARDIOLOGY Volume 34 33
PAGENO="0392"
13642 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
three possibilities: (1) Only severe, long-standing dia-
betes leads to congestive failure. (2) Only insulin-
dependent diabetes promotes cardiomyopathy. (3)
The insulin treatment itself is damaging to the myo-
cardium. The lack of any discernible increased risk of
congestive heart failure in the group treated with tol-
butamide or diet tends to exclude severity or dura-
tion of diabetes as the sole mechanism since an inter-
mediate risk would be expected in this group under
this hypothesis. Insulin in itself would not seem to be
the culprit since in the patient with keto-resistant di-
abetes of adult onset endogenous insulin levels are
often high either spontaneously or as a result of stim-
ulation. by orally administered hypoglycemic agents
Thus, the most tenable hypothesis is the difference in
the kind of diabetes, implicating insulin-dependent,
ketotic insulinopenic diabetes of early onset as the
promoter of cardiac decompensation.
It is hard to exclude duration of diabetes or its se-
verity from consideration. Congestive heart failure
could be primarily a function of either factor and
hence exhibit a particular relation to insulin-treated
diabetes. Data are too scarce in this cohort to assess
the effect of the type of treatment required or used
versus the duration or severity of diabetes, and these
aspects are difficult to disentangle without conduct-
ing a controlled experiment. It would be expected
that the insulin-treated group would have more small
vessel disease such as nephropathy and retinopathy
(and perhaps in the heart as well). We cannot tell
from our data.
Role of large and small vessel coronary dis-
ease: The excess occurrence of heart failure in pa-
tients with diabetes could be a result of either large
or small vessel disease in the coronary arterial circu-
lation. Such diseases particularly of the small vessels,
1. McKee PA, Castet5 WP, McNamara PM, et at: The natural his-
tory of congestive heart failure: the Framingham study. N Engt
Med 285:1441-1446, 1971
2. Gordon T, Kaneel WB: The Framingham study: an epidemiologi-
cat investigation of cardiovascular disease. Sections 1-27, U. S.
Govt Printing Office, NHLI, Bethesda, Md. 1968-1972
3. Mantel N, Haenzel W: Statistical aspects of the analysis of data
from retrospective studies of disease. J Natt Cancer Inst 22:
719-748, 1959
4. Watker SN, Duncan DB: Estimation of the probability of an
event as a function of several independent variables. Biometrika
54:167-179, 1967
5. Kannel WB, Castettl WP, McNamara PM, et at: Rote of blood
pressure in the development of congestive heart failure: the
Framingham study. N Engt J Med 287:781-787. 1972
6. Rubler 5, Dtugash J, Yuceogtu YZ, et at: New type of cardio-
myopathy associated with diabetic gtomerulosclerosis. Am
Cardiol 30:595-602, 1972
7. Taylor SH: lnsu5n and heart failure. Br Heart J 33:329-333,
is more apt to be severe in the insulin-dependent dia-
betic patient than in the patient not treated with in-
sulin. The ischemic myocardium, which is more de-
pendent on glucose and insulin for energy, would be
especially vulnerable. All diabetic subjects should
have difficulty in coping with an ischemic myocardial
episode in view of the dependence of the hypoxic
heart for energy on the glycolytic metabolic pathway,
which is impaired regardless of the type of diabetes.
And, indeed, once coronary disease develops, the dia-
betic subject fares worse than the nondiabetic subject -
in relation not only to congestive failure, but also to
recurrence of infarction, myocardial rupture and sur-
vivaL"
Accelerated coronary atherosclerosis has been
noted in diabetic subjects, and these patients seem to
have more myocardial infarctions, especially silent
infarctions.'3 The latter observation suggests some
difference in pathogenetic mechanism from that of
the nondiabetic infarction. Myocardial and small ves-
sel abnormalities have been studied less extensively
than large vessel disease in the diabetic patient.6
Myocardial hypertrophy and diffuse, patchy fibrosis
have been reported more frequently than macroscop-
ic myocardial infarction.6 Microangiopathy has been
well described in the skin, kidney, retina and skeletal
muscle, but has not been well documented in the
heart.6 More systematic studies of the diabetic myo-
cardium and its small vessels such as those of Blu.
menthal and co-workers'3 are urgently needed. These
investigators reported more proliferative lesions of
arterial branches of all sizes and of venules as welL
They also found arteriosclerotic-appearing lesions in
the small arteries and arterioles at least twice as fre-
quently in diabetic as in notldiabetic patients with
coronary disease.'4
1971
8. tJngar I, Gilbert M, Siegel A, et at: Studies on myocardiat me-
tabolism in diabetes, Am J Med 18:385-396, 1955
9. Opte LM: Metabolism of the heart in health and disease. Parts
1-3. Am Heart J 76:685-598. 1968; 77:100-122, 383-410,
1969
10. MaJid PA, Sharma B, Meeran MKM, et at: tvsulin and glucose in
the treatment of heart failure. Lascet 2:937-931, 1972
11. Editorial: glucose and the heart. Lancet 2:1295-1296, 1972
12. Goldenberg 5, Ales M, Blumenthal HT: Sequelae of arterioscle-
rosis of the oorta and coronary arteries: a statistical study in di-
abetes mellitus. Diabetes 7:98-108, 1958
13. Kennel WB, McNamara PM, Felnlelb M, et at: The unrecog-
nized myocardiat infarction: 14 year follow-up esperience in the
Framingharn study. Geriatrics 25:75-87, 1970
14. Blumenthal HT, Alex M, Goldenbert S: A study of lesions of the
intramural coronary artery braeches In diabetes meltitus. Arch
Pathot 70:27-42, 1960
References
34 July 1974 The AmerIcan Journal of CARDIOLOGY Volume 34
PAGENO="0393"
COMPETITIVE PROBLEMS IN THE DR'IJG INDUSTRY 13643
STATEMENT OF
EDWARD M. CHESTER, M.D.
PROFESSOR OF MEDICINE
CASE WESTERN RESERVE UNIVERSITY
DIRECTOR OF THE AMBULATORY MEDICINE TEACHING CLINIC
CLEVELAND METROPOLITAN GENERAL HOSPITAL
3395 SCRANTON ROAD, CLEVELAND, OHIO 44109
BEFORE THE
SUBCOMMITTEE ON MONOPOLY
SENA~E SMALL BUSINESS COMMITTEE
U.S. SENATE
JULY 10, 1975
56-592 0 - 75 - pt. 28 - 26
PAGENO="0394"
13644 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
I am pleased to respond to the invitation to testify concerning the use of oral
hypoglycemic agents in the treatment of diabetes mellitus.
I am a Professor of Medicine at Case Western Reserve University and Director
of the Ambulatory Medicine Teaching Clinic at Cleveland Metropolitan General
Hospital. A large segment of my time is devoted to teaching 3rd and 4th year
medical students during their clinical clerkships. My research efforts have been
directed toward an understanding of the eye changes which are associated with
diabetes mellitus. For the past 16 years I have been active in the' direction of
the Diabetes Clinic at Cleveland Metropolitan General Hospital. Prior to 1959,
when I joined the full time faculty of Case Western Reserve University. at Cleveland
Metropolitan General Hospital, I had been engaged in the practice of internal
medicine for 18 years in a suburban area of Cleveland. My practice dealt chiefly
with patients who suffered from cardiovascular disease and/or diabetes mellitus.
During the years of practice I served as a part time teacher of Case Western
Reserve University at Cleveland Metropolitan General Hospital.
The Diabetes Clinic at Cleveland Metropolitan General Hospital provides care
for approximately 500 patients with' diabetes mellitus per year. Approximately 80%
of the patients have the maturity onset form of the disease. Prior to 1973 the
majority of this group of patients were treated with oral hypoglycemic agents with
a limited degree of success. Although dietary instruction was provided for the
patients, there was little compli4nce.
PAGENO="0395"
COMPETITIVE PROBLEMS IN THE DRTJG INDUSTRY 13645
When the results of the UGDP study were issued in 1970, we became concerned
with our use of the oral hypoglycemic agents. We urged the physicians who cared
far patients with diabetes in the clinic and hospital to pay heed to the results of the
above study and to re-evaluate their treatment of the maturity onset group of
patients. In an attempt to learn the extent of the use of oral hypoglycemic agents
and their cost, the amounts of these medications dispensed by our staff were
recorded from 1968 to early 1972 (See Table I). Review of these data disclosed
an alarming increase in the use of these agents from 1968 through 1970. Response
to the recommendations of the UGDP study was reflected by a modest decrease in
the use of the oral agents during 1971 and 1972. Because we believed that the use
of these agents was still excessive, the following letter was dispatched to the
Pharmacy Committee of the hospital early in 1973.
May 24, 1973
Emanuel Wolinsky, M.D.
Chairman of the Pharmacy Committee
Dear Doctor Wolinsky:
The results of the University Group Diabetes Program (UGDP)
(Diabetes, 19, Supplement 2, 747-830, 1970) allows one to develop the
following conclusions concerning the safety and effectiveness of the
oral hypoglycemic drugs, specifically the sulonylurea group (Tolbutamide
and Chlorpropamide). (1) In the group treated with Tolbutamide there
was a significant increase in deaths from cardiovascular disease, as
compared with those treated with either insulin or strict adherence
to a calculated diet. (2) That Tolbutamide was not as effective as
either; jesulin or strict adherence to an isocaloric diet in the control
of levels of blood sugar
The UGDP study subsequently reported comparable results with
the use of Phenformin (J.A.M.A., 217, #6, 777-784, 1971).
It is only fair to point out that there are skeptics who do not accept
the results of the above study.
PAGENO="0396"
13646 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
I accept the results of the study and believe that the use of
Sulfonylureas (Tolbutamide and Chlorpr opamide) and the ~iguanides
(Phenformin) should be restricted because they appear to be hazardous
to health and are far less effective and more expensive than insulin.
I suggest that we implement a form of control which would
restrict the use of Suifonylurea drugs (Tolbutamide and Chiorpropamide)
and Phenformin with the following exceptions:
1. Patients who cannot administer insulin to themselves because
of severe visual impairment or other physical handicaps such as
neurologic disorders which impair use of arms and hands.
2. Patients who refuse to use insulin.
In order to accomplish such control the department of medicine
would provide a list of physicians who could authorize the use of the
drugs under discussion. Other services may wish to provide a
similar mechanism.
In 1972, $30, 000 were expended for Tolbutamide, Phenformin,
and Chlorpropamide. Substitution of insulin would be less costly.
Sincerely yours,
Edward M. Chester, M.D.
The results of this educational reminder and form of control produced the
results noted in Table I. Table II indirectly indicates that many of the patients
previously receiving oral agents were started on insulin therapy. Continuous
review of the use of the oral agents is in progress with the intent of further
decreasing their use except under the circumstances noted in the letter of
May 24, 1973. It is apparent that restriction of the use of these medications in a
hospital can be accomplished by education of patients and physicians and by providing
a method of control. The problem is unfortunately not as simple for a variety of
reasons when one attempts to achieve similar results with patients who are under
the care of private physicians. Among these reasons are:
PAGENO="0397"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13647
1. That the conclusions of the UGDP study are not accepted by some eminent
authorities on diabetes mellitus.
2. That education of physicians lags well behind the knowledge developed
through research. Unfortunately drug company literature provides the major
source of information for many physicians.
3. The lack of adequate patient education in their understanding of diabetes
and the hazards of oral hypoglycemic agents.
4. The failure adequately to impress the patients with sufficient understanding
of the importance of a calculated isocaloric diet and their failure to comply in
this respect.
5. The ease of using oral medication compared with the injection of insulin.
The UGDP study clearly demonstrated that standard doses of oral hypoglycemid
agents did not effectively reduce levels of blood sugar over a 5 year period. These
data confirmed previous studies which disclosed that the success rate in managing
diabetes with tolbutamide at the end of 5 years was only 13% (DeLawter and Moss,
J.A.M.A., 181, 1962, 156). Relapse or secondary failure is recorded as 22%
within 5 years (Camerine -Davalos and Marble, J. A. M. A., 181, 1962, 1). Six
to seven years after therapy with oral hypoglycemic agents only 6-12% remain well
controlled. (A.M.A. Drug Evaluations, Publishing Sciences Group Inc., 2nd Edition,
1973, p. 130, Alton, Massachusetts). Yet despite the ineffectiveness of these
hypoglycemic agents and their demonstrated relationship to increased mortality from
cardiovascular disease, these drugs are still widely used in the treatment of
maturity onset diabetes. I have noted previously the reasons for the failure-.onJhe:.
part of physicians and patients to heed the warning of the hazards and ineffectiveness
of this group of drugs.
PAGENO="0398"
136~t8 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
There appear to be 3 approaches to this problem. These include an
immediate restriction of their use through firm wai ning and labelling via the
F D A a long term educational process and the development of more rigid drug
testing requirements
Recommendations
I. Immediate warning to all physicians of the hazards by a bulletin from the F. D. A.
stating the following:
A A suitably calculated isocaloric diet selves as the cornerstone for the
treatment of maturity onset diabetes
B A su table tr al of dietary management for seveiai weeks should be
instituted first
C. If adequate levels of blood sugar cannot be obtained with this regimen,
even in the absence of symptoms associated with diabetes insulin
therapy should be instituted.
D The oral hypoglycemic agents are a potential hazard to health and should
be used after the patient has been advised of this fact with caution
only under the following circumstances
(1) If the patient is handicapped by serious visual loss or other
physically incapacitating disorders
(2) If the patient refuses to use insulin.
The drug companies should be required to include the above facts on the
package Inserts of medication despite the fact that physicians infrequently read
them Some method of identifying these medications as ha&ardous must be
developed for patient protection
PAGENO="0399"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13649
Ii. Long term educational effort. Medical school educators, clinicians who
care for patients in university and community hospitals must emphasize the
facts known about these drugs to medical students, house officers, and
physicians in practice. Efforts should be made to reach the last mentioned
through post graduate courses and through the development of self educational
units in an attempt to provide more reliable and scientifically based information
to counteract the biased and often inaccurate statements issued by pharmaceutical
companies and the "throw away" psuedomedical periodicals.
A vital step in the educational process is the need to encourage and
support the young investigator. Greater availability of research and training
grants through the National Institutes of Health or other government agencies
should be encouraged. For it is through the development of Such investigator~
and teachers that the many problems related to diabetes may be resolved.
lit. Adequate long term trials befote drugs are released for use. Most drugs,
and this applies to the oral hypoglycemic agents, were initially tested for their
ability to lower levels of blood sugar in animals. Search f~r toxicity was made
as well. These studies were short in duration. After short term trials in ma~
were made by able investigators and clinicians the drugs were released.
Subsequent long term studies of these drugs were retrospective and dealt only
with tbeir.ability to alter levels of blood sugar and lipids. The UGDP study
was the first well controlled prospective study and was designed to determine
the role of these drugs in the development of vascular disease. Thus many
years elapsed before medications, which were commonly used, were found to
be hazardous to health and to possess very limited effectiveness. Standards for
PAGENO="0400"
13650 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
long term studies must be developed by the F.D.A. to insure adequate clinical
trials of drugs before their release.
The steps indicated above are likely to be met with severe outcry and resistance
by pharmaceutical companies and scientists and clinicians who do not accept the
conclusions of the UGDP study. Support of the medical societies, particularly the
American Diabetes Association would be essential.
Restriction in the use of the oral hypoglycemic agents would significantly alter
modes of care for the patient with diabetes. To begin, it would needfully provide
a great emphasis on the importance of dietary management. In many instances with
adherence to diet adequate reduction of blood sugar and removal of symptoms would
follow. Physicians or their assistants would have to instruct patients in the use of
insulin when diet alone did not suffice. Thus more teaching would be needed for each
patient. Perhaps more teaching related to mechanisms involved in the production
of the disease, the need for preventing infection, manifestations of hypoglycemia,
and other measures would be taught as well. Since the cost of insulin is considerably
less per patient than oral hypoglycemic agents there would be a decrease in total
cost.
The issue of the clinical use of research information is exemplified by the mixed
reception of the results and recommendations of the UGDP study. Why, one may
ask, are there delays in the transmission of research data to its clinical
applications? These are several reasons:
1. Early research data may be presented initially to select groups in research
societies and published in journals which are read by only highly trained specia~ist:s.
In addition most articles are not published for at least 6 months after they have been
submitted.
PAGENO="0401"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13651
2. Further delay occurs because of the need for clinical testing.
3. When the information is finally released there are varying degrees of
receptivity and understanding. Here we deal with a variety of variables which
Include initial training and continuing education of physicians. Medical educators,
both basic scientists and clinicians, and medical societies must play an important
role in narrowing the gap between delivery of research information and its clinical
application.
Corrective measures in this regard are mostly likely to be effective if medical
students, fellows, and house officers in training are adequately prepared to receive
and evaluate research data. This requires improvement in the teaching of basic
science, biostatistics, and clinical pharmacology during medical school and post
graduate training programs. As a teacher of students and physicians in training
during their formative years, one is aware of the need to stimulate them to share in
the joy of learning. Such an effect develops and fosters intellectual curiosity,
critical thinking, and the self discipline required for continued intellectual development
throughout their careers.
During their period of formal training they will recognize the need to continue
their education once they embark upon their careers as practitioners. Reading
current literature, attending medical meetings, utilization of self educational
material, and attending specific post graduate coerces ar~ effective approaches.
Physicians should be urged, if practicing in groups, to exchange information and
ideas with peers. Journal clubs and conferences could be developed. As a former
practitioner, I found that becoming a part time teacher at a university affiliated
hospital was an excellent learning experience and a considerable stimulus to
encourage my own intellectual development. Medical schools should encourage
suitably trained physicians to participate in clinical teaching.
PAGENO="0402"
13652 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The task of communicating with the well established practitioner is more
difficult. Those who are well trained in various major specialities generally keep
abreast of new developments in their area of interest and expertise through many of
the educational methods previously mentioned. Unfortunately, there is another
group of physicians, who because they are either overworked or inadequately
trained, find or take little time to read or attend educational meetings, and rely
upon ill informed pharmaceutical company representatives and "medical throw
aways" for their sources of information. Many of them observe that because of
their lack of scientific background and the tremendous burst of new information that
they cannot understand and profit from current medical literature. They are thus
poorly prepared to accept new research data which are clinically applicable. As a
result they are not equipped to be critical of some of the claims by drug comp3nies
of the effectiveness of various forms of therapy.
* It is difficult for me to envision major corrective measures for this group.
Obviously they should be urged to attend post graduate courses in which efforts would
be made to bring them abreast of current understanding of disease and therapy.
The Academy of General Practice has made efforts to promote such courses.
Medical schools, medical societies at local and national levels must share in this
educational process.
PAGENO="0403"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13653
TABLE I
COMPARATIVE USAGE OF ORAL HYPOGLYCEMIC AGENTS
CLEVELAND METROPOLITAN GENERAL HOSPITAL
The use of Tolbutaniide, Chlorpropamide, and Phenforroin for the years which
there are accessible records are as follows:
CHLORPROPAMIDE
Tablets Used
TOLBU TAMIDE
Year
1968
1969
1970
1971
* 1972
1973
1974
1975 (To date)
1975 (Projected)
PHENFORMIN
* Year
1968
1969
1970
1971
1972
1973
1974
1975 (To date)
1975 (Projected)
270,800
525, 300
522, 200
320,600
286,600
.144, 200
83, 500
46, 100
92,200
C~p~ules Used
74, 300
93, 200
97, 000
95,700
61, 300
42, 800
15, 600
1,000
2,000
$62. 00/Thousand
$62. 00/Thousand
$62. 00/Thousand
$62. 00/Thousand
$62. 00/Thousand
$64. 00/Thousand
$69. 62/Thousand
$75. 56/Thousand
$75. 56/Thousand
Cost
$73. SO/Thousand
$73. 50/Thousand
$81. 00/Thousand
$80. 00/Thousand
$80. 00/Thousand
$69. 10/Thousand
$69. 00/Thousand
$69. 00/Thousand
$69. 00/Thousand
Total Cost
$16, 789k 00
$32, 562. 00
$32,376.00
$19, 877. 00
$17,769.00
$ 9,228.00
$ 5,813.00
$ 6,966.00
Total Cost
$ 5,461.00
$ 6,850.00
$ 7,857.00
$ 7,656.00
$ 4,904.00
$ 2,957.00
$ 1,076.00
$ 138.00
Total Cost
$ 5,140.00
$ 7,440.00
$ 8,294.00
$ 7,598.00
$ 7,377.00
$ 8,517.00
$ 2,787.00
$ 3,371.00
Tablets Used Coit
Year
1968
1969
1970
1971
1972
1973
1974
1975 (To date)
1975 (Projected)
76, 000
110, 000
127,600
116, 900
113, 500
104, 500
34, 200
18, 800
37,600
$67. 64/Thousand
$67. 64/Thousand
$65. 00/Thousand
$65. 00/Thousand
$65. 00/Thousand
$81. 51/Thousand
$81. 51/Thousand
$89.66/Thousand
$89.66, Thousand
June 25, 1975
PAGENO="0404"
13654 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
TABLE U
Yearly Use of Insulin Stocked in Pharmacy Cleveland Metropolitan General Rospital
Based on Accessible Records
Year 10 ml. Vials Total Cost
1968 7,250 $ 6,587.lS~'
1969 6,594 $ 8,119.90
1.970 6,522 $ 7,675.00
1971 7,125 $ 8,405.45
1972 8,065 $ 9,927.00
1973 9,803 $12,130.96
1974 9,814 $15,232.00
1975 (Projected) 9,192 $12,437.00
PAGENO="0405"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13655
Philip Pelig, M.D.
Professor and Vice Chairman
Department of Internal Medicine
Yale University School of Medicine
333 Cedar Street
New Haven, Connecticut 06510
Statement Before Subcommittee on Monopoly
Senate Small Business Committee
July 10, 1975
PAGENO="0406"
13656 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. Chairman and Members of the Subcommittee
I am pleased to have this opportunity to participate in these hearings on the
oral hypoglycenic drugs. Over 5 years have now elapsed since the initial presentatio
of the findings of the University Group Diabetes Program indicating an increased risk
of death from cardiovascular disease in patients treated with tolbutamide or
phenformin. Since that tine there has been considerable debate and controversy in
the medical profession as to the validity of these findings and their implications
with respect to the treatment of diabetic patients. My discussion will focus on
the following areas: (1) Those aspects of the pharmacology and clinical applications
of the oral hypoglycemic agents in which there is fairly uniform agreement among
proponents as well as opponents of the UGDP study. (2) The impact which the findings
of the UGDP study have had on medical practice. (3) The mechanisms by which the
prescribing habits of physicians may be altered.
Virtually all experts in the field of diabetes agree that the oral hypoglycemic
agents are drugs of convenience. They are convenient because they may be taken
orally as opposed to the injections of insulin. More importantly, they are convenient
because they do not require the self discipline and compliance inherent in a weight-
~
-reducing dietary regimen. In contrast to the effects of insulin in the patient
with diabetic coma, the oral hypoglycemic agents are not life saving drugs.
Furthermore, no convincing evidence is available which indicates that ~ ~
blood sugar by oral agents retards or pr~y~s~he ~
cations of diabetes which may affect the eyes, kidney or nervous system. It is thus
clear that these drugs are useful in a very limited number of patients with adult-onse
diabetes: namely, those with symptoms due to an elevated blood sugar in whom dietary
measures have failed and in whom insulin is impractical or refused by the patient.
While some experts would include patients with an elevated blood sugar who are
asymptomatic, there is universal agreement that these drugs are overprescribed in
the United States.
PAGENO="0407"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13657
All of the above was in fact well recognized before the UGDP study was reported.
The effect of the UGDP has been to add evidence of a relationship between oral agents
and increased cardiovascular mortality. This relationship has been considered
conclusive by some, persuasive by others, and at the least possible by all, including
the most severe critics of the UGDP. Given the fact that 1) these agents are drugs of
convenience, 2) they are widely overprescribed, 3) they may increase cardiov4scular
mortality, and 4) that the practice of medicine is usually governed by the axiom
"Primum non nocere" - "above, all do no harm," one may question whether the ftndings
of the UGDP study have resulted in a change in the clinical treatment of diabetes.
Unfortunately the answer is very definitely no~ The most recently available data
reveal that the total prescriptions for oral hypoglycemic agents increased 5.5%
between 1972 and 1973. This represents a total of over 19 million prescriptions
costing over $100,000,000 and involving over 1½ million patients.
Since all agree that these agents are overprescribed and at the least possibly
toxic, it is apparent that the experts in the field of diabetes have failed tc~
appropriately influence the clinical management of this disorder. To rectify this
situation I would propose the following:
1. Leading proponents as well as criticsNof the UGDP study should meet ~or
the purpose of issuing a joint statement in which the primacy of diet and
the obvious need for restriction in the use of oral hypoglycemic agents
is clearly spelled out. Such a statement can be divorced entirely from the
UGUP study. It should be noted in this regard that the critics of the UGDP
study often emphasize the limited indications and rarity with which they
employ oral agents in their own practices. However)so long as such
statements are immediately followed by a statement attacking or discrediting
the UCDP, the end result is a perpetuation or exaggeration of the abuCe of
these agents which characterizes current medical practice.
PAGENO="0408"
13658 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
2. Emphasis should be placed on dietary management rather than oral agents
in the instruction of medical students and in postgraduate medical courses
on the treatment of diabetes.
3. Research should be undertaken on developing improved methods of assuring
patient conpliance and success in adhering to weight-reducing diets.
4. Most importantly, the labeling of oral hypoglycemics should be changed to
include:
a. A warning that evidence has been reported that these agents may increas
cardiovascular mortality; and
b. That use of these agents should be restricted to adult-onset diabetics
in whom dietary measures have failed and insulin is refused or
impractical.
Mr. Chairman, there has been much discussion in the lay press and medical
journals of the need to maintain the physician's freedom of choice in the treatment
of his or her patients. I believe that our overriding concern as physicians is to
do no harm. As experts in the field of diabetes our primary obligation should be
to improve the lot of our patients by influencing current treatment practices
rather than perpetuating a situation which is at the least wasteful and at worst
causing an unnecessary shortening of life span in adult-onset diabetics.
PAGENO="0409"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 18659
(STATEMENT BY
(JOSEPH LARNER M.D., PhD., PROFFESSOR AND CHAIRMAN
(DEPARTMENT OF PHARMACOLOGY AN]) DIRECTOR OF THE DIABETES AND
(ENDOCRINOLOGY CENTER, UNIVERSITY OF VIRGINIA SCHOOL OF MEDICINE,
(CHARLOTTESVILLE, VIRGINIA
(BEFORE SUBCOMMITTEE ON SMALL BUSINESS
(JULY 10, 1975)
Honorable Gaylord Nelson and members of the Subcommittee:
Senator Nelson:
My name is Jbseph Lamer, I am a scientist (pharmacologi~t,
biochemist) and physician who has been interested in problems
of diabetes for nany years, and who has been working on the
mechanism of insulin action for 15 years. I was called and as)çed
to testify before this committee, and am pleased to do so. In
connection with the five points raised in your letter of June 19,
1975, I respond as follows:
1. The proper labeling of the oral hypoglycemic drugs in
the light of the studies recently conducted with these drugs.
Having reviewed the literature, I have come to the following
conclusion which is quoted from Chapter 71, written by myself ar~d
R.C. Haynes, Jr. of a standard textbook in Pharmacology (Goodman
and Gilman' s textbook, 5th edition to appear in September, 1975).
"The sulfonylureas should be used only in subjects with
diabetes of the maturity-onset type who cannot be treated with
diet alone or who are unwilling or unable to take insulin if
weight reduction and dietary control fail. The physician must
realize that he is using these agents only to control symptoms
associated with hyperglycemia and that dietary control with or
without insulin is more effective for this purpose."
56-592 0 - 75 - pt. 28 - 27
PAGENO="0410"
13660 COMPETITIVE PROBLEMS IN THE DRuG INDUSTRY
The major complications and life-threatening disorders associated
with diabetes are heart disease, kidney disease, blindness, and
limb gangrene. There is no evidence that sulfonylureas ameliorate
or prevent these disorders. While in many instances in medicine
the physician must prescribe for overt symptoms, we all prefer
to correct the underlying problem if possible. Unfortunately,
this is not presently possible with diabetes without additional
basic and clinical investigation. There is no evidence that
sulfonylureas will assist in the underlying problem. Since these
agents would appear to relieve primarily the symptoms of hypo-
glycemia, one should restrict their use until less costly and
perhaps safer measures have been used (diet with or without
insulin).
For this reason, I feel that there should be stronger labeling
of the oral hypoglycemic drugs in the package insert. With regard
to the nature of the labeling, I feel that the stronger 1972 FDA
draft is preferable to the weaker 1974- draft for the reasons just
discussed.
2. The effect of these studies on medical practice.
To my knowledge these studies have had a variety of effects
on medical practice. The total utilization of this group of
agents, however, has not seemed to change much. For example,
when the results of the studies were initially announced, some
physicians changed their patients to other sulfonylurea analogues
not realizing that the fundamental pharmacology should be quite
similar to the drugs studied. This obviously demonstrates the
PAGENO="0411"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13661
need for additional postgraduate training and education of some
of the medical community. Some physicians accepted the results
of the study and some questioned the design and control nature
of the experiment. This controversy has undoubtedly been apparent
to this committee. On the whole, these studies indicate that
the use of oral hypoglycemic agents should be limited to the
small percentage of patients with diabetes for whom other
therapies have proven impossible to carry out.
3. The availability of scientific evidence, if any, which
demonstrates the benefits of oral hypoglycemics.
I know of no evidence that directly demonstrates that the
oral hypoglycenics are life-saving or life-prolonging in the
therapy of diabetic patients.
The major therapeutic problem in diabetes is no longer the
acute ketoacidosis which used to be the cause of death before
the introduction of insulin. Rather, it is the long term or
chronic vascular complications of the disease. In other words,
the major problem, now, is the well recognized thickening and
other damage to the blood vessels throughout the body leading
to kidney disease, heart disease, blindness, and gangrene in
the limbs. We still do not know the answer to the following
fundamental question, "If the blood glucose level in the
diabetic patient could be controlled as precisely as that of
a non-diabetic through the use of an insulin delivery system
y~~to be devel9~p~, would there still be vascular complications?"
Or, alternatively, is there some factor or factors involved other
PAGENO="0412"
13662 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
than proper insulin delivery which leads to these harmful effects
on the blood vessels? Basic and clinical investigators are
working on this question with respect to insulin at present.
Since no answer exists for insulin, which is a direct hormone
replacement therapy, obviously no answer exists for the oral
hypoglycemic agents. For this reason, I feel there is no direct
evidence that these 9ral agents are beneficial, i.e., life-saving
or life-preserving.
4. The problems of translating the results of basic
research developed by medical scientists to the practice of medicine.
This is a very broad question, and we could spend a great
deal of time discussing it. Briefly, I am of the opinion that
scientists today are more aware than ever before of the importance
of applying their fundamental studies to the practice of medicine.
For example, in my own field, Pharmacology, there has been a
strong development in the area of Clinical Pharmacology which
addresses itself to this problem: namely, the application of
fundamental laboratory findings to the patient in order to
understand and treat the disease process. For example, the
sulfonylureas have been used clinically for about 20 years, yet
a great deal of information regarding these compounds is still
lacking. The metabolism of these compounds in patients and their
precise mechanism of action are still unknown. These have been
complicated problems and require additional studies in both
animal systems as well as patients. Scientists are very interested
in coordinating such diverse efforts and studies. I feel that
PAGENO="0413"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13663
clinical research work in this area should be further nurtured,
but that it must be balanced by a broad base in fundamental ani~nal
research as well.
5. Any other aspects of the subject which you think might
be helpful to the Subcommittee.
I feel strongly~that the time has cone in terms of the oral
hypoglycemic agents to restudy their effects in animals and
patients. It is my feeling that since recent animal studies
are proving of considerable interest in terms of the actions
of these drugs on organs such as the heart, adrenal glands, and
liver, it would be wise to restudy these compounds in animal
systems during the time their clinical use is reevaluated in
order to see whether we can gain an understanding of the
mechanism of the cardiovascular deaths or even reproduce them
in animals. Here I note with particular interest two recent
pieces of data in animals: 1) the summary statement of the
work of Wissler et al. which states that in rhesus monkeys fed
an average American diet for 74 weeks containing 20 mg/kg
tolbutamide, there were present in the coronary arteries two
times more frequent and three times more severe atheromatous
changes than in the coronary arteries of control monkeys.
2) the work of Hsu etal. from our Department of Pharmacology
at Virginia which demonstrates that in heart, adrenal medulla
and other organs, sulfonylureas inhibited catecholamine release
from the nerve endings of the autonomic nerves. Thus the
function of the autonomic nervous system, which provides the
PAGENO="0414"
13664 COMPETITIVE PROBLEMS IN THE DRIJG INDUSTRY
involuntary control for many of the organs of the body, is
significantly influenced by these, drugs. Therefore, I feel
that it is time to caution physicians about the use of these
drugs, and to restudy them in the clinical and basic laboratory
much more extensively.
PAGENO="0415"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13665
STATEMENT BY
PAUL MEIER, PH.D.
PROFESSOR OF STATISTICS
UNIVERSITY OF CHICAGO
CHICAGO, ILLINOIS 60637
BEFORE MONOPOLY SUBCOMMITTEE
SENATE SMALL BUSINESS COMMITTEE
31 JANUARY 1975
Mr. Chairman. I speak as a member of the Biometric Society Committee
on Biometric Aspects of Controlled Clinical Trials whose report is under
discussion today.
Professor White has outlined our problem and our findings, and Professor
Zelen will speak about some of the particular criticisms made of the U.G.D.P.
report. I shall speak a little more generally about the role that I see for
clinical trials in guiding our decisions about modes of therapy.
It happens that in March of 1970 I testified before this committee o~t the
subject of risks of Thronbo-embolism due to the use of oral contraceptives.
spoke then of the deplorable lack of prospective controlled clinical stud~Les
on the effects of oral contraceptives. I discussed possible reasons for that
lack. Let me quote a few lines from that earlier testimony. My explanat~.on
for the lack of substantial research on this important problem was as follows:
"Frankly, the required research, although important, is not
especially appealing to scientists. It is not fundamental
and it is not exciting. It is difficult, it is expensive,
and it is fraught with the risk of attack from all sides.. Who
would willingly prepare himself for such a study, make an
application to be weighed competitively with others on scien-
tific merit, and risk the loss of support halfway through the
study when a review committee with different views or priorities
comes to consider renewal of support, when he stands to gain so
little in scientific recognition or otherwise?
Evidently, for whatever reasons, there is no sound body of
scientific studies concerning these possible effects available
today, a situation which I regard as scandalous. If we proceed
in the future as we have in the past, we will continue to stumble
from one tentative and inadequately supported conclusion to another,
always relying on data which cone to hand, and which were not
designed for the purpose. The planning of better studies is
difficult, and the recruitment of investigators willing to commit
their efforts to these purposes nay be more difficult still. I
believe both are possible and essential to the public welfare."
At the time those words were written, I had no knowledge of the U.G.D.P.,
but they could scarcely have been more apt.
PAGENO="0416"
13666 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
It is true that the U.G.D.P. had defects. It~is true, also, that it
falls short of proving the case against Tolbutainide. Nonetheless, as
Professor Cor.tfield remarked in testimony here last September, the U.G.D.P.
today provides the best available information on the possible toxicity of
Tolbutamide.
As to defects, there are no studies which are entirely free of them,
and it was the judgment of our committee that this study was weLl conceived
and executed and that those defects we could identify did not give reason
to doubt the findings.
As to it being inconclusive, that was inevitable in the nature of the
case. Once the investigators became convinced that there was substantial
evidence of toxicity, and not of corresponding benefit, they had no choice
but to withdraw the drug.
Thus we are left with an ominous yet inconclusive result, and I believe
that this is a typical outcome which we nay expect to see repeated in many
other instances. It may be, in such a case, that the community of physicians
will decide that, although not conclusive, the evidence is sufficient to
abandon the drug. Or, on the contrary, as in the U. G.D.P * case, they may
conclude that the evidence does not require them to give it up.
In the latter case, however, I can see no alternative to the initiation
of a new clinical trial, conducted by physicians unconvinced by the first one.
I should expect, in any event, that both physicians and patients should be
made as fully informed about the evidence as is feasible.
I go so far as to hope that the experience to date. with oral hypoglycemic
drugs nay convince us that clinical trials should be a continuing component of
drug surveillance for any drug, from the first day of its release, and so long
as substantial doubt about the balance of risks and benefits remains.
PAGENO="0417"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13667
Statement by
P. J. Palumbo, M.D.
Certified Internist and Endocrinologist
Clinician and Clinical Investigator with
special interest in diabetes and its
complications and hyperlipidemia.
Assistant Professor of Medicine
Mayo Medical School
Rochester, Minnesota 55901
Before Subcommittee on Monopoly
Senate Small Business Committee
January 31, 1975
PAGENO="0418"
13668 COMPETITIVE PROBLEMS IN THE DRUG INDuSTRY
DIABETIC TREATMENT AND SURVIVORSHIP
P. J. Palumbo, M.D.
The comparison of treatment for a disorder can only be
evaluated through controlled randomized, clinical trials.
Hints and leads from retrospective studies can be
extremely valuable in leading to a new hypothesis and may be the
basis of justification of a randomized trial. However standing
alone they cannot form the basis of any firm conclusions concerning
treatment effects.
The preliminary analysis of our data of the incidence,
prevalence and mortality of diabetes mellitus in Rochester, Minnesota
between 1945 to 1970 contains some hints that survivorship may be
lower in diabetics on oral antidiabetic agents, but group differences
preclude any firm conclusions regarding this observation. Such an
observation would point to the need for controlled randomized
clinical trials to study the possible adverse effect of various
treatments on survivorship in the diabetic.
The University Group Diabetes Program was a randomized
trial study to evaluate the influence of treatment on diabetic
complications. A statistically significant, adverse effect on
survivorship was noted after patients had been on tolbutamide and
phenformin for five or more years. These data have been reviewed
and the conclusions have been found to be sound.
In my opinion, as a diabetologist, another randomized
tria\l study of treatment in diabetes is not ethically justified,
PAGENO="0419"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13669
as the data from the University Group Diabetes Program clearly
indicate an adverse effect of the oral antidiabetic agents on
survivorship in the diabetic. The use of these oral agents,
therefore, should be curtailed.
PAGENO="0420"
13670 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
STATEMENT BY
HENRY T. RICKETTS, MD
PROFESSOR OF MEDICINE EMERITUS,
UNIVERSITY OF CHICAGO,
950 EAST 59TH STREET,
CHICAGO, ILLINOIS 60637.
BEFORE THE SUBCOMMITTEE
OF THE SELECT COMMITTEE,
ON SENATE SMALL BUSINESS,
JANUARY 31, 1975
I have studied diabetes mellitus, cared for diabetic patients,
and conducted researches in this specialty for 34 years. I have
been president of the American Diabetes Association and co-founder
and president of the Chicago Diabetes Association. I have served
on the Study Section of Endocrinology and Metabolism, Grants Division,
National Institute of Arthritis and Metabolic Diseases, and have
served as a contributor and an assoc~ateeditorof the journal, Diabetes.
My connection with the Committee of the Biometric Society was
that of a consultant diabetologist, and I attended most of their
meetings. I was struck by the thoroughness with which the members
of the Committee made their investigation. I detected no bias for
or against the UGDP study. The Committee listened to more who
criticized the study than to those were were less opposed or favorable.
The Committee did not hesitate to ask the Coordinating Center in
Baltimore for raw data when a point was in doubt, and members made
trips to the Center and to several participating clinics to check
methods, procedures, and results. No uncertainty was too small to
leave unresolved.
PAGENO="0421"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13671
The UGDP was set up to determine whether various treatments
for diabetes would minimize the mainly vascular complications
that notoriously accompany that disease. It is ironic that a
full report dealing with complications has not yet been published
because, in the third and fourth years of the study, an alarming
preponderance of deaths had accumulated in the tolbutamide group~
The investigators, then, perforce, had to turn their attention to
mortality and survival.
I was not a participant of the UCDP study, but I followed it
closely. Despite some imperfections, I think that the results arid
conclusion of the UGDP have shown tolbutamide and phenformin, and
probably their cousins, to be dangerous drugs, especially when taken
for extended periods of time. I stand by my opinion of four years
ago, expressed with the help of a committee of the American
Diabetes Association in the editorial statement accompanying the
first report of the UGDP (Diabete~, Supplement No. 2, Vol.l9:74783O,
1970). I quote:
"...The UGDP mortality study shows that death rates
were essentially the same in the IVAR group, which
maintained more nearly normal fasting blood glucose
levels, as in the more poorly controlled groups of
PLBO and ISTD. This would appear to mean that efforts
PAGENO="0422"
13672 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
to establish `good'control of hyperglycemia in
the kind of population studied (had]no effect on
mortality...
"The real lesson of the data is that if
diet plus insulin does not reduce mortality below
that experienced with diet alone, it is highly
improbable that oral hypoglycemic agents will do so.
There is indeed no doubt about the reality of
the greater number of cardiovascular deaths observed
in the TOLB group as compared with all other treat-
ment groups. Inquiry into the reasons for this has
been both intensive and extensive. Aside from the
most proximate explanation, that tolbutamide may have
been directly and solely responsible, the possibility
that the TOLB population, by chance and despite random-
ization, entered the study with more or greater risk
factors than the other populations had to be scrupulous-
ly investigated. Although this possibility has, in
the opinion of the ADA Ad lIoc Editorial and Advisory
Conrnittee, not been exiuded, the weight of statistical
analysis makes it probable that the excess cardiovas-
cular mortality in TOLB is attributable either to the
PAGENO="0423"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13673
drug itself or to unconsidered and unknown factors.
In the absence of evidence for the latter, suspicion
would naturally attach to tolbutamide.
`The mortality study is at least suggestive
enough to put a damper on what appears to be the
indiscriminate use of all oral hypoglycemic agents
in the treatment of mild or moderate, adult-onset
diabetes. Although tolbutamide, for practical
reasons, has been the only sulfonylurea drug inves-
tigated by UGDP, the chance that other compounds of
this family may be similarly involved cannot be
dismissed despite differences in molecular structure.
It would not be justifiable at this point, however,
to prohibit the manufacture and use of sulfonyurea
drugs, for they will probably continue to fill a need
in special circumstances."
If these drugs are dangerous, what course should we take? You
have just heard that their manufacture should not be forbidden,
and for reason. For example, how do we treat a diabetic patient
who ought to be taking insulin but is living alone with a broken,
or amputated, or paralyzed arm that prevents him from using a syrir~ge
and needle? One who is blind and cannot measure his dose of insulin?
PAGENO="0424"
13674 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
One who is old and tremulous? One who is mentally disturbed?
And finally, one who refuses to take insulin? In another vein,
there are diabetics who are engaged in hazardous occupations and
ought not to take insulin for fear of reactions. We ought to make
allowance for these patients even though the oral agents are not
very effective and, I believe, in the long run, may be harmful.
But if we continue to make these agents available, as I think
we must, how do we protect other diabetics who would like to use
them but should not?
Insulin comes to the patient with a package insert that carries
a great deal of information, including certain warnings. The oral
agents come to the patient in silence because they have been regarded
as innocuous, needing no instructions except the doctor's directions
for dosage and timing. This must change.
But it is the physician who should lead the way, and I hope
that the report of the Biometric Society will in time convert the
many current unbelievers. Meanwhile, it might not be too radical
to ask the FDA, under proper authority, to transfer the oral hypo-
glycemic agents to the circumscribed Schedule II of dangerous drugs
along with barbiturates, amphetamines, and certain narcotics.
Physicians might learn that the oral agents are not exactly safe,
and the requirements of BNDD prescriptions, if for dubious need,
PAGENO="0425"
COMPETITIVE PEOBLE1~~tS IN TEE DRUG ThThUSTEY 13675
might become a salutary nuisance. This arrangement, of course,
would have holes in it (and I can see some), but it might~ have
the effect of helping to reduce the use of a product that too
many patients could well do without.
56-592 0 - 75 - pt. 28 - 28
PAGENO="0426"
13676 COMPETITIVE PROB~JEMS IN T~E DEtG INDtTSTR~
STATEMENT B,Y
ETHAN A. H. SIMS, M. D~
PROFESSOR OF MEDICINE, COLL~GE OF MEDICINE, UNIVERSITY OF VERMONT
BURLIftGTON, VERMONT 05401
(on sabbatical leave at Endocrine Dlv.., Tufts N. *E. Medical Center
Boston, MA 02111
BEFORE MONOPOLY SUBCOMMITTEE
SENATE SMALL BUSINESS COMMITTEE
JULY 10, 1975
Mr. Chairman, Members of the Committee: I ~am glad to have
the Opportunity to testify before this Committee relative to the proper
labeling of oral hypoglycemic agents in th~ light of recent studies,
and on the problems of translating the results of research to the
practice of medicine. I plan to center my remarks on the optimal man-
agement of the overweight, non-insulin-dependent diabetic as this
relates to' the recommendations included in the proposed labeling. My
qualifications include care of diabetic patients at Yale-New Haven
Hospital and in Vermont over the past 30 years and former direction
of a Metabolic Unit and of an N'IAMD training program in diabetes. 1 have
served on special study, sections of the NIH for review of proposals
for Diabetes and for Obesity Centers. I am a member of the advisory
and editorial group for the Fogarty International Center Conferences
on obesity and of the workshop on Obesity of the National Diabetes
Commission.
SUMMARY
I. Obesity is now recognized as a factor predisposimg to non-insulin-
dependent diabetes in a genetically susceptible person. Untreated
obesity may pose a greater long-term risk in relation to cardio~vascular
and other disease than the use of oral agents.
II. Insulin, In addition to Its well known function of lowering blood
glucose, is a storage hormone. The use of insulin, or of oral agents
PAGENO="0427"
COMPETITIVE PROBLEMS IN THE DT~tYG INDUSTRY 13677
which stimulate release of insulin promote fatness.
III. Intense preoccupation with one aspect of the UGDP study has~ blurred
our perceptions of other vitally important data in the study. These
data indicate adverse effects of therapy both with sulfonylureaS and with
insulin in promoting further development of obesity. This is a recog-
nized risk factor for cardiovascular disease.
IV. One should not go on Writing package labeling recommendatiGflS
based on data comparing the five treatment modes chosen by the invest-
igators of the UGDP in the l960s, when these options are now out of date.
They do not include preventive and rehabilitative measures available
in 1975.
V. Lifestyle changes in eating and in physical activity are essent-
ial components of the management of non-insulin-dependent diabetes, as
well as of cardiovascular disease and are often sufficient in them-
selves to restore near normal function. Initiated early they may
provide effective prevention.
VI. The proposed FDA package labeling for oral agents should reflect
these considerations. They should also be written in a form conducive
to the education of the patient.
VII. The schedule of peer review and publication of the results of
randomized prospective clinical trials requires modification, if the
results of future studies are to be accepted by interested parties.
VIII Efficacy of a drug must be considered in the light of other
available options for management. On this basis there is little evidence
of acceptable efficacy of the sulfonylureas, except under special
circumstances which preclude other therapeutic options. In the case
of phenformin the efficiency:efficacY ratio is not acceptable.
PAGENO="0428"
1S~7S cO ~TIT O~tbEMS ~ ~RVG ~T~ET
I. PROPER LABELING ~F THE ORA YP~GGt~Y~MIC DRUGS tND~E~L'~GHT:
OF RECENT STUDIES
I believe that the Intense preoccupation with the incr'ease'in
córd1ova~cu1ar flt~rta11ty from use of t1~e oral ~agents h'a's drawn attent~
Ion away fr~om matters o~L even greater importance in the nTana~ement
of the non insulin &epei~kent diabetic These fnclude the facts that
4
1) ObesIty is now recognized to~ bea predlspas'lng factor for diabetes
In a permon ge~ne-tlcal1y su pt44le~. 2) ~~flhS1j~f~, In `addttlon to
its well-known futiction of Towering blood glucose, is a storage
hor~ori~e and i-ts use and the `use of oral agents which prOmote release
of InSulin promote obesIty. 3) Lifestyle changes in eating and in
physical activity are essential components of good management of
non-insUliri-'depen~Ient dtebet~s and ~of cardfovarcular:disease as well.
These changes are often sufficient in Iftieniselves to restore near normal+.
function.
We should not classify diabetes into juvenile and maturity onset,
since the correlation of types with age is poor; Diabetes occurs in
two forms, each requiring entirely different management: 1. e. insulin
dependent, usually lean and hungry, and non-insulin-deper,dent, usually
obese and also insulin resistant, Our studies in Vermont have
shown £hat normal volunteers who deliberately gain 20 to 30 per cent
above their basal weight develop Insulin resistance
i~ similar to that seen In the spontaneously obese.
They also had a diminished ability of muscle and fat cells to utilize
glucose. Their fat cells enlarged, but did not increase in number, again
simfiar to those who spontaneously develop obesity in adult years. These
results suggest that in the naturally obese patient at least a part.of
the Insulin resistan~e that places astress upon the pencreatlc reserve
of Insulin Is secondary to weight gain and is reversible.
The UGOP investigators chose their five treatment regimens because
PAGENO="0429"
COM?~flTI~t ~OB~tJEMS I~ TH~ DRUG I1~iISTRY 1&~79
they rep11cat~e~ th~ cc~mmoniy. a eptedbpt~ofls~of thi ea~1y ~0s~ but
they. db not ~`epresent the best optIons available now. Epidenhlol-
ogists and cardiologists have defined the risk factors forcardio-
vascular disease, which include more than just elevated blood sug~r
and fats, I. e. obesity, smoking, and physical inactivity. Of the
1,500,000 patients now taking oral agents, probably 5b per cent are
more than 25% over weight (54% of those entering 6 of the 12 clinics
of the UGOP study). When these patients are ti4eated with "diet a1on~",
It is generally little more than a token gesture in the direction of
a low-caloric diet. The success rate for Weight loss is notoriously
poor. There is important data in the UGDP study the significance
of which is overlooked. With qualified dieticians available) th2re wka~
an initial drop of slightly under 3 %~~rL all groups. Only those
taking placebo or phenformin maintained their weight ttiráughout the
16 folloWup periods, and there was no differdnce between placebo and
phenformin. On the otherhand, those taking either ~tolbutamide or
insulin lost less weight initially and' regained ~weigIrt above their
baseline values s that In all there was an 8 per cent difference
between the mean rate of the placebo group. This was to be expectE~d
since when the patient receives either a sulfonylurea drug or Insulin
plasma insulin is increased, thus increasing the tendency to store
fat. The Indications listed in the latest FDA recommendation for
package labeling calling fOr use of insulin if dtdt~fal'ls t~u,iSiouht'er
to our ~ürrent concepts'of the pathopt1y~sio1ogy of ndn-1v~sulin'~dePendén~
diabetes when it is associated with overweight ahdwhen there are
other valid options.
OTHER OPTIONS FOR TREATMENT. There are other o~tlons in'
addition to the five of the UGDP study. These differ from the 000P
treatment modes in that they have the potential for reversing, the
patient's diabetic state. They Include vigorous andcomj~reheoSive
PAGENO="0430"
13680 COMPETITIVE PROBLEMS i~ THE DRUG INDUSTRY
regimens which Include education and Increase in physical activity.
Now adjuncts Include behavioral self-modlficatløn and proteln.sparlng~
starvation.
a) Withdra~l ~ Q~.) Aqe~n~s ~ Intensive We~qht Rcitl9~*.
An Important result of the JGDP study was the decision of Dr. John
Davidson in Atlanta to discontinue oral agents for 1500 patients at
Grady Hospital in Atlanta, as reported on Sept. 18, 1974 to this
Committee. I suggest that his recent paper (JAMA 232:853, `75~ on
his experience be incorporated In the record of this hearing.
60% of these patients have been control1ed~ without drugs or insulin
by an intensive regimen of dietary treatment and exercise, which
includes 25 hours of Instruction and the use of special manuals. 50~
90 % have lost significant amounts of weight, and it has been
possible to discontinue insulin therapy in all patients who were
Initially above ideal body weigiLt. The short term costs for educ~
atlon of the patient are considerable, but so are the long-term
costs to the patient ~of oral agents, which are also susceptible to
secondary failure. Dr. Davidson's importa~t e~perimen~, a pilot study
which Still requires reporting in full detail, demonstrates not
only that patients may do as well with no medication as with oral
agents, but also that their overt diabetic state may be at least
temporarily reversed.
- b). Proteip-Spgrlng~Starvation. We are becoming more experienced
In tbe use of.brief-- and occasionally long per-i~ds of semi-starvation
to lnltlateweight reduction without consuming essential body protein
s well as fat stores. This can be accomplished not by by providing
carbohydrate to spare protein, which was the dictum in the past, but
by providing optimal protein and minimal carbohydrate to minimIze
Insulin stimulation. Sucb~a temporary Spartan aid to acute weight
loss is Often surpr~sin~1y ~ve11 tolerated by patients, who may report
PAGENO="0431"
COMPETIPIVE PROBLEM~S IN THE 1~ETJ~ INDt$TRY i3~81
loss of appetite plus a feeling of well being.. (Genuth JAMA 231: `74).
But It obviously cannot represent a panacea and must be cart4led out
under clo~e supervision To sustain weight loss other measures are
needed.
c) Inci~ease j~ Phy~icai Ac~ivity~ I have 1~eft'to the last
what may weilbe one of the most important) even if the least
tried and the most poorly documented adjuncts for the prevention
and treatment of non-insulin-dependent diabetes, namely increase
in physical activity. The only reference to exerciSe to date in
these hearings was on Sept. 18, 1974 (p 10880). Mr. Gordon quest-
ioned Dr. Schmidt regarding the suggestion of Dr. Jesse Roth of the
NIAMD that "vigorous exercise lowers blood sugar and that there seems
to be a persistent beneficial effect In additiOn to the immediate
effect". In answer Dr. Schmidt gave short shrift to exercise asL
a major factor in the management of diabetes. I disa~ree with him
in this, and have support In the recommendations of the White House
Conference on, Food, Nutrition, and Health (Panel 11-3 p 51 Adults
in an Affluent Society: The Degenerative Diseases of Middle Age). A
relevant study Is that of Dr. Per Bjorntorp in Sweden, who has measured
the insulin response of obese middle-aged men before and after a course
of physical training. tven though he encouraged them to maLntain
their excess weight, their resting insulin and insulin response to
glucose was strikingly reduced (Metabolism 19:631 `70). This is not
surprising, as any insulin-dependent diabetic learns that exercise
lowers the requirement for insulin. Some formerly obes~ persons find
that sustelnett increase in physical activ~ty is the only way they can
malntain.'weig'ht loss.
A lifestyle lncomp~tible with good health often lies be'hind
many patients with non-insuj1~n-dependent diabetes. A change' In life'
Style with respect to composition of the diet and level of activity
PAGENO="0432"
18682 co~n~xi~ o~i~s IN T~E DRUGI U~T~7S~RY
is the first and somet+mes the only measure requ1redfor~ many ~uch.
diabetics. For those patients who definitely are lnsul1n~depenaent
insulin is a necessary and logical adjunQt tO diet therapy. These
include 1) the acutely decompensated diabetic (even those above ideal
weight) 2) the hyperglycemic maturity onset diabetic at or below
ideal weight 3) the hyperglycemic pregnant diabetic and of course
4) the young *insulin~~dependent dlabettc.~ However, for the o~'er~
weight diabetic, use of oral ag~e;nts or Insulin ~s ~both~illogicai
and unnecessary, If the measures iridicatec above can restore metàbolic~
balance.
Many patients enter thehealth care system toolate in the course
of their diabetes or of their lives to modify their lifestyle effect-
ively, and the therapeutic options are limited. Our profession
should ma~k.e maximum efforts to reach younger members of high~rlsk
families with educa'ticrn and programs~ for effective prevention,
If `all the above Is true, one might Indulge In an extrapolation
regarding the results of usfn,g oral agents in the non-Insulin depend-
ent diabetic. Such an. extrapolation is obviously highly conjectural
since solid data is not yet available, but I believe that lt.poin'ts
In the direction of important truth. If we grant that there are
approximately 1,500,000 pat1en~s reptttedly t~aking oral agents and that
501 are gro~s1y ov~erwelght, we~ have 750,000 patIents wlth'::diabetes
and obesity who are prob'~bly also less physically act~tve than tJiey~
should be. If we assume that~ 90% of them are not exposed to any
vigorous and comprehensive regimen such ~as that at the Grady Hospital,
675,000 are left with their obesity essentially untreated, and 4 out of
5 are taking an agent which increases their obesity. The taking' of
oral medication lulls both physician and patient into believing that
something worthwhile is~ being accomplished, while the op~tIons which
could make a fundamental difference in a patients ,llfe and~, survival
PAGENO="0433"
~T~EMS 1* THE P~iYG~ INDV~TRT 13~8'3
are~be~tng ne te~d~. This to i~iyTmir~d, i~~an ifl~por~t~nt con~id~a~t3o'r~ ;.
whfcft ~1wa~rs-'~ev~en tire ~e~rl~us ~oi~cev~ about tU~toxl~1ty of th~
a'ftnts Even If t1~e~ b~al aflrits we~ pr~ove~~ ~to~ have j~o ~to~xI~ ~aat$n,
their d~tr~imental~ rQ1ea~$ ~a~s~u'bst4tute-'~ `for ~th~r's~a'f~e ~nd potsn~t~V~'
rehab~14t~&~ti~s .~lir~es wou1~i remain. `~
To change the way v~e are doiiig ~thlngs~ today is no~t a simple
matter, and It Is inappropriate to blame the physician or the.
patient for the result of large forces at work In our e~onomy. A
considerable reallocation of resources over a period of titne.woul~d
be required to bring about a shift from ~a syNptomatic to. a reh~ab~il.~
itative form of therapy.' Dr.. Leon White, Commissioner of Heaith
and He~pit&ls in Boston~ recently listed the destructive lifestyle
habits in this country and the diseases and disorders they~produce.
Social excess of alcohol, overeating, and lack of exercise contribute
to obesity and cardiovascular disease. These habits also contribute
to diabetes In the genetically predisposed. As reported in the
Harvard Medical Newsletter (l;no 39, June `75) Dr. White sta'ted that
If the battle to modify lifestyle is to be fought, a major enemy is
advertising. But advertising is not the only enemy. Lifestyle
modification, if it is to be successful, will adversely affect the
pharmaceutical ~industry, the tobacco industry, the alcohol products
industry, the food products industry, and the auto Industry. The
real challence is to improve health without wrecking the economy.
THE RESPONSIBILITY OF TH~ FDA REGARDING LABELING
The question remains as to whether the Commissioner of the
FDA should indlca'te priorities or treatment or treatment options
and whether these represent a constraint on the freedom of the'
responsible physician or mak him liable to suit if he doe~s not
follow such priorities. Since there is a stated responsibility
PAGENO="0434"
13684 00 trzvE ~ ~
under tite Federal Food., trrug, and~Cosmet~c ~
FDA to assure effica y~as~~Ieila~ safety of E%trl*, ~~bai4eve~that~
the~ PD~ haw ~`respoiislbi11ty to maka sucb~ i~ndlcatiOns, since they
~a~fe~t~the ruiat1vee~ffic~cy of a~drug4. ~
might be rated as effective,. but if a ~upeYlór~al~erna~1ve b~e~eines
av~aflable,~its .rela~tin,e ~ SiPté tire~'types~
and stages. of diabvtes vav~y, ft teems'to me appro~priat~ 1ó~t~ the
FDA to suggest priorities for the use or non~use of' d~rugs at these
various stages. The physician retains the ultimate responsibilitY
of deciding how his particular `patient relates to the general `guide-
lines. Ha shouTd not be nted~colegally'vulnerable for electing any
particular' option provided that he can justify his decision an~J a-l~so
makes his pati~ent an informed partner in the choice', whenever pOssible.
SUGGESTED CHANGE IN PROPOSED LABELING FOR SULFONYLUREA DRUGS
In line with the above considerations I suggest that the
section on Indicatjon~ for the use of sulfonylurea drugs submitted
for the Federal Register (page 40 of the copy available for the hearings)
be modified as follows. (Changed wording is underlined).
~j~JetiC patients, with. n i~ul.1nr,depend,ent, diabetes who are
ovej~wei.9ht commonly exhlbi,t' insulin, resistance and have elevated
fasting Insulin concentrations a,nd ~ncr,eased,,,t'hou,gh,relat1YelY i,nad
eqQate Insulin response to~lucose,,. Such patients can, fregueptly
,ehabi,lit,ated,,,and,th.eir overt diabet~ reversed bya vigorous and
comprehensive regimen otdietary restrlction.~,in,creased physlca1'ac,~
ivlty. anthweiqh't loss. Th~us neither treatment with~(4rUg) nor
i'ns,ulin ,i s 1r~dl,cate4~ unless japplisationof such meuu-res,ist~ptal iz.
impractical. (prug).ls Indicated in maturity onset' non-ketotic
diabetics `of normal ~or' subnormal weight whose ~)~p~gl,Y~cejpi!*cannot `be
controlled by carbohydrate restriction a~,d Increased, actlvj,,,ty and
PAGENO="0435"
COMPEPIPIVE PROBL1~MS IN TH~ DRtYG INIY~JSTRY 136 5
In whom Insulin cannot be used because of .,,,.slmllar,factorS,
When,. e&l~i asy~p~t~rna ~ eJ,~vatI~nfo~o4
glo~e. ,canno~tbe .ç r~ll.e&by~,the~ ahgve measures and in whom 1nSulir~
..unanswered scientific question."
* I have.substituted the word hyperglycemia for asymptomatic
here because I do not consider that the UGDPstu~y, with Its limited
range of population and limited measurements l~ d~clsiv.e in determin-
ing whether hyperglycemia Is related to microangiopathy' Also with
a high renal threshold for glucose prolonged hyperglycemia of major,
degree may be present without syisptomatic glucosuria.
II THE EFFECT OF THE STUDIES OF THE USE OF ORAL AGENTS ON MEDICAL
PRACTICE AND PROBLEMS OF TRANSLATING RESULTS OF RESEARCH TO PRACTICE
Judging by th,e prescription rates for the various oral agents,
the Inipact of the UGDP study on general practice has been negligible.
It Is ironic that the sale of phenformin, the agent which has been
most clearly shown to produce tachycardia, hypertension, and a sl~nif-
icant Increase in both cardiovascular and overall mortality has
ncreased In sales (up 5% from `72 to `73). There are suggestions,
however, that the use in the vicinity of large medical centers has
decreased. The study has provoked a great deal of inquiry into
* the use and limitations of prospective randomized clinical trials.,
The Reasons for the minimal impact have been discussed by others In
* detail before this Committee, and I will mention only a:few:
1) Both physicians and patients look for quick solutions in the
form of pills or injections. This is understandable because correct-
ion of our many health problems today calls for alterations in life
style which are not easy to accept and which often call for resOurces
which most physicians do not have available. It l's the large problem of
PAGENO="0436"
13686 COMPETITIVE PROBLEMS IN THE DRUG INDTJSTEY
symptomatic vs. preventive and rehabilitative medicine.
2) The disagreement, wrangling, and finally name calling (scandal,
coveru~, drugh~use whore,. etc.) that hat develoced out of the controversy over.
one single, though important, aspect of the LJGDP study has tended to weaken
public confidence in the medibal~ pr~fession and has obscured ou~ ability to
perc~eve the factors of most critical importance to the patient.
The Plahning of future randomized, controlled Clinical Trials.
To aváid some of the features which have detracted from the impact of th~
sug~est the following;
That direct finan~cial support of particular studies by vested interests
UGDP I
1)
be avoided.
2) That agreement be obtained in advance regarding both the conduct of
the study and the release of data.
3) That the final data and that at any critical point in the study be
reviewed by a third party unaware of the identity of the experimental groups.
4) The customary piecemeal process for review . and publication of medical
work is inappropriate and impractical in the case of a randomized clinical
trial which may involve ingrained traditions of clinical practide and academic
theory on the one hand and multimillion dollar commercial Interests on the
other: It seem to me that the first release should be of a full report which
vitally interested parties have had an opportunity to review prior to
definitive issue.
5) We are entering a period when patients will take a more active role In
deciding their own options for treatment of chronic health problems. In
parallel with this I hope that we will see informed, cooperative patients in
the early stages of their disease or disorder taking part in clinical studies.
tends
6) Medical advertising often A to neutralize the impact of randomized
clinical trials, if the results from a particular product are unfavorable.
Our profession has a responsibility to see that advertising accepted for our
professional publications does not undermine or weaken the Impact of research
PAGENO="0437"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13687
and government regulations. As an example, the medical journal which published
the full report of the experience of the UGDP with phenformin has continued to
print advertisements for phenformin which give no warning regarding hypertension,
tachycardia, and cardiovascular death as adverse reactions.
7) Teaching in medical schools and in post-graduate courses should emphasize
methods of evaluation of therapeutic agents to a greater extent. At the
University of Vermont during several years we have used the UGDP study and its
problems as the basis for an interdepartmental teaching exercise. The
excellent short book by A. L, Cochrane in England, Effectiveness and
Efficiency (Nuffield Provincial Hospitals Trust 1972) should be available in this
country as required reading for all health professionals.
III. AVAILABILITY OF SCIENTIFIC EVIDENCE DEHONSTRATING BENEFIT OF ORAL AGENTS
The UGDP demonstrated no increased effectiveness of tolbutamide as opposed
to placebo with respect to mortality or gross evidence of vascular disease in
the groups of maturity onset diabetics of the type now receiving oral agents
in today's practice. As in other studies there was also evidence of secondary
failure. Thus,on the issue of efficacy) were the drug up for initial review
today, it doesn't seem likely that it would be approved. The UGDP study also
showed that, in this group of largely overweight non-insulin dependent diabetics,
lowering the blood glucose concentrations alone gave no objective benefit, altho
the techniques for evaluating microangiopathy were insensitive. Therefore, the
one justification for continued approval of the sulfonylureas appears to be that
there is a very small percent of patients who can't take or will not take insulin,
do not respond to other measures, and are severely symptomatic and hyperglycemic.
However, to give blanket approval for use of the drug to accomodate this
relatively minute fraction of the diabetic population leaves the door wide open
for continuation of widespread inappropriate use, with all the disadvantages
indicated above. It Is unrealistic to believe that package labeling with
suitabl~ warning and suggested priorities will alter the patterns of prescription.
56-592 0 - 75 - pt. 28 - 29
PAGENO="0438"
13688 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
It therefore seems to me to be justifiable and appropriate to classify the sulfonyl-
urea compounds, along with other drugs with circumscribed uses and significant
adverse effects, to sharply limited cliihical situations by placing them in
a restricted category and requiring written justification for their use in a
particular situation. This would not curtail their use where there is strong
indication, but would limit much uninformed or ill-considered use.
The justification that is advanced for continued use of phenformin is that
it does not stimulate insulin release while lowering blood gluco~o and therefore
is ideal, as the advertisements say, for releasing patients from entrapment
in their fat cells. The UGDP study showed that the patients who took phenformin
maintained their initial weight loss, and had approximately 8 % lower body weight
at the close of the study than those taking tolbutamide or insulin. There
was, however, no difference from those taking the placebo and the clearcut
evidence of tachycardia, hypertension and increased total and cardiovascular
mortality seen in the UGDP study in addition to the potential of producing
lactic acidosis indicates a price in toxicity too great to pay for any relative
advantage over oral agents with respect to weight loss. Therefore I believe that
it is past time that this agent should be disapproved.
PAGENO="0439"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13689
Statement of
Cohn White
Professor of Public Health
Yale Universj~y School of Jedicinc
New Haven, Connecticut ()(5l~)
Before the
Sub-committ~ee on lonopolv of the
Small Business Committee
U. S. Senate
January 31, 1975.
Mister Chairman and Members of the Sub-committee:
I am the chairman of a committee which was appointed by the
Biometric Society and funded by the National Institutes of Health to
carry out the following mission:
i. to make an in-depth assessment of the scientific quality
of the UGDP study and in particular of the biometric
aspects of the design, conduct, and analysis of the trial;
ii. to make a similar assessment of other controlled trials of
oral hypoglycemic agents.
The committee consisted of six members:
John P. Gilbert, Harvard University
Paul Meier, University of Chicago
Chris L. Rt~nke, Free University, Amsterdam
Rodolfo Saracci, Pisa, Italy
Marvin Zelen, State University of New York at Buffalo
Cohn White, Yale University
PAGENO="0440"
13690 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Two officers of the Biometric Society attended several of the meetings
as observers: Peter Armitage, London School of Hygiene and Tronical
Medicine; and Berthold Schneider, Ilannover, l~est German~.
The research associate for the committee was Theodore lolford, and
the consultant diabetologist was Henry 1. Ricketts.
The full committee met on six occasions over a two ~`ear neriod and
has completed a report which will be published on February l() in the
Journal of the American ~1edical Association.
The work of the U.G.1).P. is still in progress and I think it is
fair to say that diabetologists in general await with interest the findings
on the treatment by insulin. There has never been a study of comparable
scope and thoroughness on the long-term effects of this agent in subjects
with maturity-onset diabetes. In the meanwhile, however, controversy has
arisen about the data concerning tolbutamide, and the committee saw as
its main task the investigation of the reported excess cardiovascular
mortality in the subjects receiving this drug. It is interesting to note
that the U.G.D.P. presented results on phenformin which are quite com-
parable to those on tolbutamide: the death rate from cardiovascular causes
was approximately the same in the two cases. The findings on phenformin,
if one can judge from the absence of criticism, appear to have been accepted
by medical scientists, even if they have not so far been translated effect-
ively into medical practice. Yet these findings also were made by the U.G.D
using the methods that have come under heavy criticism when applied to
tolbutainide.
Because of the many factors which influence survivorship in a chronic
disease such as maturity-onset diabetes, careful methods of investigation
PAGENO="0441"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13691
are needed, and, in particular, control groups are essential. ConsequentlY
we reviewed only such trials as were controlled. It then became clear that
the major study to consider, other than the U.G.D.P., was the study in
Bedford, England, organised by Dr. H. Keen and Dr. 1. .1. *Jarrctt. It
should be said at once, however, that the Bedford study, based on 125
patients in each of two treatment groups was not comparable in size or th
detail to the U.G.D.P. in which approximately 201) patients were followed on
each of five treatments.
The work of the committee appointed by the Biometric Society fell ibto
four sections:
1. Visits were made to the U.G.D.P. co-ordinating Center and to two
of the co-operating clinical centers to study methods used in the trial.
2. The methods and findings of the U.G.D.P. study were discussed
with several authors who had written about them, and the Bedford study was
discussed with Dr. Keen and Dr. Jarrett.
3. The published criticisms of the (J.G.D.P. were reviewed in detail.
Comparable criticisms of the Bedford study do not exist, though several of the
major criticisms made about the U.G.D.P. would apply a fortiori to the Bedford
study.
4. New analyses were made of the data from the U.G.I).P. and Bedford
studies, the data being kindly made available by the directors concerned.
Critics have pointed out that in the U.G.D.P. study the total mortality
was not significantly higher in the tolbutamide group than in the placebo
group, even though there was a significant difference in the case of deaths
from cardiovascular causes. We consider that this cirticism has some weight
but is not convincing. Criticisms that have been commonly made but which,
in our view, are not correct are:
PAGENO="0442"
13692 COMPETITIVE PROBLEMS I~ THE DRUG INDUSTRY
(1) the excess mortality in the tolbutamide group was due to the
data obtained from just a few clinics:
(2) the studies of Keen et al and of Paasikivi contradict the J.G.I).P.
(3) the baseline differences among the treatment groims account for
the finding of the adverse effects from tolbutamide. bn this point I might
remark that none of the critics, to my knowledge, has given serious con-
sideration to the multiple logistic method that was used by the tl.(.D.P.
to take the effect of baseline risk factors into account. Until they do
they have not carried out an adequate review of the IJ.G.U.P. analysis.
(4) the findings on the effect of tolbutamide are flawed by the
failure to adapt dosage to individual need.
(5) the evidence was not adequate to justify the discontinuation of
the oral drugs.
In our analysis of the U.G.D.P. data we have used the same multiple
logistic model as was employed by the U.G.L).P. investigators, hut have taken
additional variables into account to allow for the time each subject was under
study and for differences between clinics. We confirm the principal find-
ing from the simpler study of failure rates, namely that the cardiovascular
death rate was higher in patients receiving tolbutarnide than in those receiving
placebo. This difference remains after adjustment for the effect of baseline
variables and cardiovascular risk factors.
We have also made an analysis in which the extent of adherence to assigned
treatment was taken into account. The highest death rate was found in the
tolbutamide group who adhered 100% to their treatment and who did not modify
the dose.
In an analysis of the data from the Bedford trial we found no difference
in death rate between the placebo and the tolbutainide group. As indicated
PAGENO="0443"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13693
above, we do not interpret this failure to find a difference as a con-
tradiction of the more thorough U.G.D.P. study.
The conclusion of the committee is that it remains with the proponents
of the oral agents to conduct scientifically adequate studies to justify
the continued use of such agents.
PAGENO="0444"
13694 COMPETITIVE PROBLEMS IN THE DRtTG INDUSTRY
Statement of Dr. Marvin Zelen, Director
Statistical Laboratory, State University of
New York at Buffalo
Mr. Chairman and Members of the Committee. Thank you for
this opportunity to appear before this Committee. My comments
will be divided into two parts.
How is it that able and respected clinicians can disagree
with the interpretation of the UGDP data? The tolbutamide
cardiovascular death rate is more than double compared to other
treatments. Yet many clinicians who treat adult onset diabetes
find it difficult to accept such a figure. For many of them,
this elevated cardiovascular death rate does not appear to have
been perceived in the clinic.
Let us examine other factors which may lead to elevated
cardiovascular mortality. According to the UGDP data, the
cardiovascular death rate for individuals above the age of 53
is approximately five times that of individuals 53 or younger;
people with arterial calcification at time of diagnosis have
four, times the cardiovascular death rate compared to those
without arterial calcification; the initial glucose tolerance
test (GTT), as used by the UGDP investigators, shows that. those
with a GTT above 723 (the median value) have double tho rate
of cardiovascular deaths compared to those who have a GTT below
the median; men have a doubled cardiovascular death rate compared
PAGENO="0445"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13695
to women. Although the numbers quoted are "rounded" for
simplicity, it is clear that in the clinic there are many
factors simultaneously influencing cardiovascular deaths.
Several of these have greater or equal effect on the cardio-
vascular death rate compared to the effect of tolbutamide.
As a result, it would be difficult for a clinician to perceive
an elevated cardiovascular death rate associated with tolbutamide.
Such an effect would be almost completely obscured by these
other important factors. Only if there is careful and structured
record keeping on a large number of patients would a changed
cardiovascular death rate of 2-3 be detected. The analysis of
such multi-faceted data requires more sophisticated data
analytic methods than those in common usage by clinicians.
Next, I wish to discuss some features of the Biometrics
Society report. A criticism .of the original UGDP analysis
is that it failed to explore the effects of several factors
acting simultaneously on the cardiovascular mortality. Our
Committee did in fact consider this matter very carefully.
We found that when one examines the group of older women (age
greater than 53) the tolbutamide cardiovascular death rate is
almost five times that of the placebo group. It is in this
group of older women where the tolbutamide excess cardiovascular
mortality is most dramatically shown.
Finally, I wish to comment on the problem of planning
and analyzing clinical investigations in which patients are
PAGENO="0446"
13696 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
expected to be on chronic medications for a period of many
years. It is important in planning these long term studies
to allow the clinician to change the medication if it is in
the best interests of the patient. This can result in an
altered dose or even a change in the medication. The tJGDP
protocol did allow the clinician this freedom. .A protocol
which does not allow this flexibilIty may not be in the best
interests of the patients under study. In addition to
modified or changed medications, patients may, on occasion,
not take their medication at all. In the Biometrics Report,
these problems were examined in considerable detail. It is
our conclusion that the greatest statistically significant
difference between tolbutamide and placebo occurs in the group
who have taken their prescribed medication in exactly the
manner specified in the protocol for the entire period of
follow-up.
To conclude, I wish to state that the interpretation of
the data is difficult due to the small number of deaths relative
to the total number of patients. In our endeavors we have
analyzed the data in many other ways which have not been put
into our final report. Our conclusion is that the weight of
evidence points to tolbutamide as being responsible for excess
cardiovascular mortality.
PAGENO="0447"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13697
EXHIBITS PROVIDED BY THE FOOD AND DRUG ADMINISTRATION
STATEMENT
BY
ALEXANDER M. SCHMIDT, M.D.
COMMI SSIONER
FOOD AND DRUG ADMINISTRATION
PUBLIC HEALTH SERVICE
DEPARThENT OF HEALTH, EDUCATION, AND WELFARE
BEFORE THE
SUBCOMMITTEE ON MONOPOLY
SELECT COMMITTEE ON SMALL BUSINESS
UNITED STATES SENATE
JULY 9, 1975
PAGENO="0448"
13698 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
Mr. Chairman:
I am pleased to appear today before this subcommittee to discuss
current Food and Drug Administration (FDA) actions relating to oral
hypoglycemic drugs. As you are well aware, labeling for this class
of drugs has been the subject of extended public controversy and legal
challenge. The Food and Drug Administration (FDA) has now published
a proposed regulation providing new labeling for this class of drugs,
which appeared in the Federal Register of July 7, 1975, and has invited
comments on this labeling. A public hearing will be held on August 20,
1975 to afford interested persons a further opportunity to comment.
On September 20, 1974, I summarized before this subcommittee the actions
of the FDA that followed the report of the results of the University
Group Diabetes Program (UGDP) study. Today I will review the events
that have taken place since my previous testimony and I will discuss in
some detail important aspects of our proposed labeling changes.
REVIEW OF BIOSTATISTICAL ISSUES BY
THE C~MMITTEE Of THE BIOMETRIC ~CIETY
Because of the controversy concerning the UGDP study and the oral
hypoglycemic labeling previously proposed by FDA based on the findings
of that study, we decided that publication of proposed labeling should
await completion of a detailed review of the UGDP study by the Biometric
Society, a distinguished international organization of biostatisticians.
The Society's report was published in the February 10, 1975 issue of the
PAGENO="0449"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13699
Journal of the American Medical Associatiofl. Testimony before this
subcommittee on January 31, 1975, by the members of the Biometric
Society who had conducted the review, had provided a preview of their
conclusions.
The Biometric Society Committee assessed the scientific quality of
the UGDP study, particularly the design, conduct, and analysis of the
trial, and evaluated other controlled trials involving oral hypoglycemic
agents. The committeediscussed in detail the published criticisms of
the UGDP study and found that "most of the criticisms unpersuasive."
Specifically, the committee concluded that:
1. The criticism that patient selection had been
inappropriate was "largely irrelevant" to the validity
of the evidence for the toxicity of the oral agents.
2. The criticism that total mortality in the tolbutamide
group was not significantly different from that in the
placebo group had some weight and "the toxic effect of the
oral hypoglycemics cannot be affirmed with the certainty
that would be present if total mortality were significantly
different."
3. Excess mortality in tolbutamide-treated patients was
not confined to a few clinics, as critics had claimed.
PAGENO="0450"
13700 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
4. The committee particularly analyzed the criticism that
there were important differences in base-line cardiovascular
variables among the groups and concluded that there was
"~ * * no evidence that the base-line differences arising
from the randomization contributed in any important way to
the finding of adverse effect from tolbutamide."
5. The criticism that oral hypoglycemic drugs were given
in fixed dosage was not relevant to the question of whether
the drugs were toxic.
6. Although the committee acknowledged that, "It would have
been easier to interpret the findings if there were more
data on mortality." [that is, if the study had been carried
on longer], it did "* * * not criticize the UGDP investigators
for having made the decision when they did." The committee
further said: "Nevertheless., the result of that decision is
to leave us with some residual uncertainty about the meaning
of the findings, a point that is well understood by the UGDP
investigators themselves."
7. Other studies said to contraJict the findings of the
UGDP study do not in fact do so.
PAGENO="0451"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13701
In addition to evaluating criticisms of the UGDP study, the
Biometric Society Committee conducted extensive new analyses of
the UGDP data, taking into account the effect of various base-line
variables and cardiovascular risk factors. The committee's a~,alyses
confirmed that cardiovascular mortality was increased in the tolbutamide
group. The increase was statistically significant for the patient
population taken as a whole and in the subgroup of females, especially
in women over the age of 53, but not in the male subgroup. This does
not mean that the study demonstrated that the drug carries less risk
in males. On this point the committee concluded: "The data do not
support the same conclusions for men, but one possible reason is that
the smaller number of patients in the male group results in lack of
sensitivity to detect differences of moderate magnitude."
An important finding was that the highest death rate occurred in the
group of patients who adhered most closely to the tolbutamide regimen
and did not have their dose modified. Also, when the analysis was
conducted according to the survival modeling method, which takes into
account the proportion of time each patient received the assigned
medication, women in the tolbutamide group showed statistically
significant increase in both cardiovascular and total mortality.
The Biometric Society Committee summarized conclusions in the final
section of its report as follows:
PAGENO="0452"
13702 COMPE:TITIVE PROBLEMS IN THE DRUG INDUSTRY
`On the question of cardiovascular mortality due to tolbutamide
and phenformin, we consider that the UGDP trial has raised
suspicions that cannot be dismissed on the basis of other
evidence presently available.
"We find most of the criticisms levelled against the UGDP
findings on this point unpersuasive. The possibility that
deaths may have been allocated to cardiovascular causes
preferentially in the groups receiving oral therapy exists,
and, in view of the `nonsignificance' of differences in total
mortality, some reservations about the conclusion that the
oral hypoglycemics are toxic must remain. Nonetheless, we
consider the evidence of harmfulness moderately strong.. The
risk is clearly seen in the group of older women * *
Whether it affects all subgroups of patients cannot be decided
on the basis of the available data, owing to the small number
of deaths involved in these subgroups.
In conclusion, we consider that in the light of the UGDP findings,
it remains with the proponents of the oral hypoglycemics to conduct
scientifically adequate studies to justify the continued use of such
agents.
ADDITIONAL INFORMATION PERTAINING TO ORAL HYPOGLYCEMIC DRUGS
In addition to the Biometric Society report, other information has
become available recently.
PAGENO="0453"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13703
1. The UGDP recently published its detailed report of the results
of the phenforriin study (Diabetes, 24 (suppl 1): 65.-184, 1975). In
addition to reporting that cardiovascular mortality and total
mortality were greater in the phenformin-treated group than in
the other treatment groups, the report presented evidence that
phenformin therapy resulted in increased blood pressure and heart
rate, thus suggesting possible mechanisms by which this drug might
influence cardiovascular mortality.
2. At hearings before this Subcommittee on January 31, 1975,
Dr. P. J. Palumbo reported that a retrospective study of diabetic
patients treated at the Mayo Clinic suggests that survival was lower
in those patients treated with oral hypoglycemic agents, compared
to those patients treated with insulin or diet. Dr. Palumbo's
full study has not yet been published.
3. A retrospective study of diabetic patients treated at the Joslin
Clinic, reported in a doctoral thesis by P. Kanarek, can be interpreted
as providing results that are consistent with those of the UGDP.
Although we have seen this study, it has not yet been subjected to
full review by statistical and epidemiologic experts. At this point,
we can say that certain subgroups of insulin-treated patients appear
to have better survival rates than tolbutamide-treated patients with
comparable glucose abnormalities. Studies of this type, however,
always present difficulties in interpretation because of doubts
56-~92 0 - 75 - pt. 28 - 30
PAGENO="0454"
13704 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
regarding comparability of the treatment groups and because
treatments are not rai~domly allocated. Thus, although the
retrospective studies of Palumbo and Kanarek may or may not, when
fully analyzed, add support to the UGDP findings, the prospective
UGDP study must be accorded far greater weight and is alone a
sufficient basis for our proposed actions.
4. Drs. Tan, Bradley, Gleason, and Soeldner reported on the long-
term (4 years) effects of hypoglycemic agents on the oral glucose
tolerance test and blood lipids in chemical diabetics at the annual
meeting of the American Diabetes Association in 1973 (abstract in
Diabetes 22 (suppl 1): 290, 1973). The investigators' abstract
indicated that there was no significant difference in the improvement
in glucose tolerance in patients receiving oral hypoglycemic agents
compared with patients receiving placebo. The full report of this
study has not yet been published, but it appears that the investigators
studied glucose tolerance on the day following discontinuation of the
drugs. Their findings thus would indicate only that the oral agents
do not lead to improved glucose tolerance in the absence of continued
use of the drug.
5. Dr. R. W. Wissler, et al., in an FDA-supported study, examined
the chronic effects of tolbutamide in the rhesus monkey. He found
that coronary artery lesions were almost two times more frequent
and three times more severe in the tolbutamide-treated animals than
PAGENO="0455"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13705
in the control animals. The FDA recently received the final report
on this study, which has not yet been published in the literature.
At Dr. Wissler's request, FDA is supporting a further review of
the pathologic findings by several independent pathologists.
6. Dr. D. F. Wu, et al., reported on the effects of tolbutamide on
heart function in dogs with chemically induced diabetes at the
meeting of the American Federation for Clinical Research on May ~,
1975 (abstract in Clinical Research 22:2l5A, 1975). The investi~ators
found that, after one year of treatment with tolbutamide, left
ventricular function was reduced and cardiac morphology altered
compared to the control groups.
The animal studies do not necessarily bear directly on the excess
cardiovascular mortality seen in tolbutamide-treated patients In
the UGDP study, but they do suggest several mechanisms by which this
might have occurred.
PROPOSED LABELING FOR ORAL HYPOGLYCEMIC DRUGS
Mr. Chairman, as you know, it has been the position of the FDA
since 1970 that the findings of the UGDP study should be reflected
in a warning in the labeling for oral hypoglycemic drugs and,
in turn, in the use by
PAGENO="0456"
13706 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
physicians of these drugs. Let me emphasize that this view does not
require that we conclude the study provides absolute proof of hazard.
The UGDP study is an adequate and well-controlled study -- by far the
most extensive and bast examination of the long term effects of Oral
hypoglycemic agents yet undertaken -- and the finding of an increased
cardiovascular mortality in tolbutamide and in phenformin~treated patients
cannot be attributed to any shortcomings of study design or execution.
This finding, despite any residual uncertainty that may remain, requires
a clear warning to physicians. Prudence dictates that a warning be
issued whenever there is sufficient evidence to believe that a drug may
be hazardous or carry a risk and that such warning is necessary to assure
the safe and effective use of the drug by physicians.
Enough time has now passed for interested persons to have studied the
Biometric Society report and the recent detailed UGDP report on
phenformin. The Agency has, therefore, published for comment a regulation
proposing new labeling for the oral hypoglycemic labeling. Interested
persons may comment on the proposal by September 5, 1975, and a public
hearing will be held on August 20, 1975. Final labeling regulations
will not be published until after all comments and materials have been
considered.
The proposed labeling contains two sections of particular importance:
a Boxed -Warning stating that there may be an increased risk of cardio-
vascular death associated with the use of oral hypoglycemic drugs and a
new indications section that limits use of these drug to patients
PAGENO="0457"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13707
whose symptoms or blood glucose abnormalities cannot be controlled by
diet alone and who cannot take insulin for one of a number of specified
reasons. Let me discuss each of these sections in greater detail; they
are reproduced in full in an attachment.
The Warning describes the UGDP study and its findings. It has been
contended that certain studies said not to support the findings Of the
UGDP study should be mentioned in the warning section to provide `fair
balance." We have concluded, however, and made clear in revised
section 1.3 of our regulations (also published July 7, 1975), that if
scientific data exist to support a warning, the warning must be presented
in unambiguous terms without disclaimers or qualifications that would
undermine or destroy its usefulness. There is, therefore, no mention in
the proposed warning of other studies involving the oral hypoglycemic
drugs. The mention of studies in which increased cardiovascular mortality
was not found, would serve only to encumber the warning.
The Warning also points out that although only one sulfonylurea and one
biguanide were included in the UGDP study, it is prudent from a safety
standpoint, in view of the similarities in chemical structure and mode of
action of drugs within each of these two categories, to consider that the
UGDP findings may apply to all other products in each category. The
Warning is thus identical for all the sulfonylurea drugs; only one
biguanide drug is marketed in this country.
PAGENO="0458"
13708 COMPETITIVE PROBLEMS TN THE DRUG INDUSTRY
Finally, the warnings section makes explicit the clear implication of the
finding that tolbutamide and phenformin may carry a risk not associated
with insulin: `(Drug) should be used in preference to insulin only in
patients with maturity onset diabetes whose symptoms or blood-glucose
level cannot be controlled by diet alone and only when the advantages in
the individual patient justify the potential risk; see Indications.
The patient should be informed of the advantages and potential risks of
(drug) and of alternative modes of therapy and should participate in the
decision to use this drug."
`4e have concluded that a patient population exists for which these drugs,
properly labeled, can be considered as safe and effective. We have also
concluded, however, that this patient population is a limited one. The
proposal to limit the treatment population to patients in whom insulin
cannot be used has been opposed in the past on the ground that it
interfered with the practice of medicine. We recognize that drug labeling
has an impact on the practice of medicine. For this reason the Food and
Drug Administration has an obligation to ensure that labeling is as correct
and accurate as possible. It must, however, meet the statutory standard
of describing the conditions of use under which a drug may be considered
safe and effective. If a known hazard or potential risk leads to the
conclusion that a drug may be used safety only in certain patients, this
limitation on use must be expressed in labeling.
The indications section, in addition to describing the population in
whom these drugs are indicated, points out that "in considering the
use of (drug) in asymptomatic patients, it should be recognized that
PAGENO="0459"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13709
whether or not controlling the blood glucose is effective in preventina
the long-term cardiovascular or neural complications of diabetes i~ an
unanswered scientific question.' This emphasizes the different benefit-
risk considerations that obtain in the symptomatic patient who needs
alternative treatment if insulin cannot be used, and the asymptoma1~ic
patient, whose need for alternative treatment is problematical.
Mr. Chairman, you asked that I comment on the promotion of these drugs.
Advertising for these products has not violated general legal require-
nents, but it has been based upon labeling that has been in need of
modification. It is clear that promotional materials must change
radically to reflect the new warning and the restricted indications.
You can be assured that we will be monitoring the promotion of these
products closely after the new labeling becomes final, to see that they
do so.
It is important to recognize that the use of the oral hypoglycemics
remains widespread despite the UGDP findings and despite the rather
limited capability of the drugs, after a few years of use, even to
lower the blood sugar. Total prescriptions for this class, according
to the National Prescription Audit (a survey of IMS America) have been
stable between 19 million and 21 million since 1967 (except for an
apparent dip in 1969). There is thus a great deal of common practice
to overcome before use of the oral agents will recede to its proper
lebels.
It is anticipated that publicity attendant upon our publication of
proposed labeling and announcement of the upcoming public hearing
PAGENO="0460"
13710 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
as well as public interest in todays hearing will bring the new
labeling to the attention of physicians and help begin to persuade
them that the UGDP findings should change the way they treat diabetic
patients. In addition, the FDA plans to issue a Drug Bulletin when
the labeling for these drugs is made final. We will monitor the
use of these drugs and will take additional measures as necessary to
publicize the final labeling.
I will be pleased, Mr. Chairman, to answer any questions.
PAGENO="0461"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13711
PROPOSED LABELING INDICATIONSICONTRAINDICATIONS AND
WARNINGS SECTIONS FOR ORAL HYPOGLYCEMIC DRUGS OF
THE SULFONYLUREA CATEGORY.
INDICATIONS
* (Drug) is indicated to control symptoms due
to hyperglycemia in patients with maturity-onset
nonketotic diabetes mellitus whose symptoms
cannot be controlled by diet alone and in
whom insulin cannot be used because of patient
unwillingness, erratic adherence to the injection
regimen, poor vision, physical or mental handicap,
insulin allergy, employment requirements, or
other similar factors.
(Drug) may also be used to lower blood glucose
in asymptomatic patients whose blood glucose
elevation cannot be controlled by diet alone and
in whom insulin cannot be used for any of the
above reasons. In considering the use of
(drug) in asymptomatic patients, it should
be recognized that whether or not controlling
the blood glucose is effective in preventing the
long term cardiovascular or neural complications
of diabetes is an unanswered scientific question.
The use of (drug) may be associated with an
increased risk of cardiovascular mortality as
compared to diet alone or diet plus insulin; see
WARNINGS. For this reason, it should be used only
PAGENO="0462"
13712 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
when the advantages in the individual patient
justify the potential risk. The patient should
be informed of the advantages and potential
risks of (drug) and of alternative modes of
therapy and should participate in the decision
to use this drug.
The foundation of therapy in the obese
maturity-onset diabetic is caloric restriction
and weight loss. Proper dietary management
alone is often effective in controlling the
blood glucose and eliminating symptoms of
polydipsia and polyuria. Use of (drug) must be
considered by both the physician and patient as
a treatment in addition to diet and not as a
substitute for diet or as a convenient mechanism
for avoiding dietary restraint.
Many patients who are initially responsive
to oral hypoglycemic drugs become unresponsive
or poorly responsive over a period of time,
usually 1 to 5 years. (Drug) should be given
only to patients demonstrated to be responsive
to it; see DOSAGE AND ADMINISTRATION for dis'
cussion of secondary failure. Short term
PAGENO="0463"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 13713
administration of (drug) may be sufficient
during periods of transient loss of control.
Concomitant Thera~p~ withaBiguanide:
(Drug) may be ~ised in conjunction with phen'
formin to control symptoms due to hyperglycemia in
patients with maturity~onset nonketotic dia-'
betes mellitus whose symptoms cannot be controlled
by diet and maximum recommended doses of either
drug alone and in whom insulin cannot be used
for any of the reasons cited above.
In considering the use of concomitant
therapy, it should be noted that both a
sulfonylurea drug (tolbutamide) and a biguanide;
drug (phenforinin) have been reported to be
associated with increased cardiovascular mortality;
see WARNINGS. In addition, phenformin can
PAGENO="0464"
13714 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
produce lethal lactic acidosis in son~e patients.
Thus the use of (drug) in association with
phenfortnin carries a greater risk than the use
of (drug) alone.
If a judgment is made that (drug) and phenformin
are to be used together in a particular patient,
it should be established that the patient is
* responsive to both drugs. This may be accomplished
either by a trial of each drug separately or
by adding the second drug and then tapering the
dosage of the first, observing for diminished
* control of blood glucose. Once the need for both
drugs is established, the desired contr9l of
blood sugar may be obtained by adjusting the
dose of either drug. The possibility of hypoglycemia
should be anticipated and appropriate precautions
taken. See package insert for phenfortnin hy4rochloride
for CONTRAINDICATIONS, WARNINGS, PRECAUTIONS,
ADVERSE REACTIONS, and DOSAGE AND ADMINISTRATION.
CONTRAINDICATIONS
(Drug) is contraindicated in patients with:
1. Known hypersensitivity or allergy to the
drug.
2. Diabetic ketoacidosis, with or without
coma. Such patients should be treated with insulin.
PAGENO="0465"
COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY 1~71 5
WARNINGS
SPECIAL WARNINQS ON CMDIOVASCULAR MORTALITY
(This subsection of labeling to be boxed, set in
boldface type, and placed at the beginning of
WARNINGS section of labeling:.)
The administration of oral hypoglycemic drugs may
be associated with J~ncreased cardiovascular mortality
as compared to treatment with diet alone or diet plui.
insulin.
This warning is based on the study conducted by
the University Group Diabetes~Program (UGD?), a long
term prospective clinical trialdesigrted to evaluate
the effectiveness of glucose-4owering drugs in prevent-
lug or delaying vascular complications in patients with
maturity-onset nonketotic diabetes. The study involved
1,027 patients who were randomly assigned to one of
five treatment groups ~ betes, 19 (supp. 2):
747~-83O, 1970; Diabetes, 24 (supp. l):65-184,
1975),
PAGENO="0466"
13716 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The IJGDP reported that patients treated for 5
to 8 years with diet plus a fixed.dose of tolbut-
amide (1.5 grams per day) or diet plus a fixed dose of
phenforinin (100 milligrams per day) had a rate of cardio-~
vascular mortality approximately twice that of patients
treated with diet alone or diet plus insulin.
Total mortality was increased in both the to1but~
amide- and phenformin-treated groups, but this
increase was statistically significant only for
phenfortain. Despite controversy regarding the
interpretation of these results, the findings
of the UCDP study provide adequate scientific
basis for this warning.
Although only one drug in the sulfonylurea
category (tolbutamide) and. one in th~ biguanide
category (phenfortnin) were included in this
study, it is prudent from a safety standpoint
to consider that this result may also apply to other
oral hypoglycemic drugs in these categories, in
view of the close similarities in mode of action
and chemical structure among the drugs in each
category.
PAGENO="0467"
COMPETITIVE PROBLEMSIN THE DRUG INDUSTRY 13717
(Drug) should be used in preference to insulin
only In patients with maturity-onset diabetes
* whose symptoms or blood glucose level cannot be con-
* trolled by diet alone and only when the advantages
in the individual patient justify the potential
risk; see INDICATIONS. The patient should be
informed of the advantages and potential risks
of (drug) and of alternative modes of therapy
and should participate in the decision to use
this drug.
(Drug) is not effective in patients with juvenile
diabetes or insulin-dependent diabetes at any age. Sucl~
patients should be treated with insulin. The concomitant
long term use of insulin and (drug) in an individual
patient is, In vi of the potential risk of increased c~rdio-
vascular mortality with (drug), less safe on a benef it-
risk basis than the use of Insulin alone.
PAGENO="0468"
13718 COMPETITIVE PROBLEMS IN THE DRUG INDUSTRY
The effectiveness of any ~rai hypoglycemi'~ drug,
including (drug), In lowering blood ~1ucose to c desired
level decreases in a large numb2r of patiento as tie
drug is administered over a period of months or years,
in part because the patient's blood glucose tends to
rise over time and in part because of diminished
responsiveness to the drug. This phenomenon is known
as secondary failure to distinguish it from primary
failure in which the drug is ineffective in an
individual patient at the time of its initial adminis-
tration. See DOSAGE AND AD~flNISTRATION.
Itenal or hepatic insufficiency may cause
elevated blood level~ of (drug) and increase the risk
of serious hypoglycemic reactions.
P~~nan~çy: (Data and interpretation related to
reproduction and teratology studies to be supplied by
manufacturer).
Prolonged severe hypoglycemia (4 to 10 days) has
been reported in neonates born to mothers who were
receiving a sulfonylurea drug at the time of delivery.
Neonatal hypoglycemia has been reported more frequently
following use of the 1/9n~'er-acting agents. If (drug)
is t~sed during pregnancy, it should be discon~in"ed
(time period to be supplied b7 rnanufactu'er, he~ore
* the expected delivery date.
0