[House Hearing, 110 Congress]
[From the U.S. Government Printing Office]



                               before the

                         COMMITTEE ON OVERSIGHT
                         AND GOVERNMENT REFORM

                        HOUSE OF REPRESENTATIVES

                       ONE HUNDRED TENTH CONGRESS

                             FIRST SESSION


                              JUNE 6, 2007


                           Serial No. 110-76


Printed for the use of the Committee on Oversight and Government Reform

  Available via the World Wide Web: http://www.gpoaccess.gov/congress/

44-429 PDF                  WASHINGTON : 2009
For sale by the Superintendent of Documents, U.S. Government Printing 
Office Internet: bookstore.gpo.gov Phone: toll free(866) 512-1800; DC 
area (202) 512-1800 Fax: (202) 512-2104  Mail: Stop IDCC, 
Washington, DC 20402-0001


                 HENRY A. WAXMAN, California, Chairman
TOM LANTOS, California               TOM DAVIS, Virginia
EDOLPHUS TOWNS, New York             DAN BURTON, Indiana
PAUL E. KANJORSKI, Pennsylvania      CHRISTOPHER SHAYS, Connecticut
CAROLYN B. MALONEY, New York         JOHN M. McHUGH, New York
ELIJAH E. CUMMINGS, Maryland         JOHN L. MICA, Florida
DENNIS J. KUCINICH, Ohio             MARK E. SOUDER, Indiana
DANNY K. DAVIS, Illinois             TODD RUSSELL PLATTS, Pennsylvania
JOHN F. TIERNEY, Massachusetts       CHRIS CANNON, Utah
WM. LACY CLAY, Missouri              JOHN J. DUNCAN, Jr., Tennessee
DIANE E. WATSON, California          MICHAEL R. TURNER, Ohio
STEPHEN F. LYNCH, Massachusetts      DARRELL E. ISSA, California
BRIAN HIGGINS, New York              KENNY MARCHANT, Texas
JOHN A. YARMUTH, Kentucky            LYNN A. WESTMORELAND, Georgia
BRUCE L. BRALEY, Iowa                PATRICK T. McHENRY, North Carolina
    Columbia                         BRIAN P. BILBRAY, California
BETTY McCOLLUM, Minnesota            BILL SALI, Idaho
JIM COOPER, Tennessee                JIM JORDAN, Ohio
PAUL W. HODES, New Hampshire

                     Phil Schiliro, Chief of Staff
                      Phil Barnett, Staff Director
                       Earley Green, Chief Clerk
                  David Marin, Minority Staff Director

                            C O N T E N T S

Hearing held on June 6, 2007.....................................     1
Statement of:
    Nissen, Steven, M.D., F.A.C.C., chairman, Department of 
      Cardiovascular Medicine, Cleveland Clinic; John B. Buse, 
      M.D., Ph.D., professor, University of North Carolina School 
      of Medicine; and Bruce M. Psaty, M.D., Ph.D., co-director, 
      Cardiovascular Health Research Unit, professor of medicine, 
      Epidemiology and Health Services, University of Washington, 
      investigator, Center for Health Studies, Group Health, 
      Seattle, WA................................................    87
        Buse, John B.............................................    92
        Nissen, Steven...........................................    87
        Psaty, Bruce M...........................................    97
    Slaoui, Moncef M., Ph.D., chairman, research and development, 
      GlaxoSmithKline............................................   121
    von Eschenbach, Andrew C., M.D., Commissioner, Food and Drug 
      Administration, accompanied by John K. Jenkins, M.D., 
      director, Office of New Drugs, Food and Drug 
      Administration; and Gerald Dal Pan, M.D., Office of 
      Surveillance and Epidemiology, Food and Drug Administration    35
Letters, statements, etc., submitted for the record by:
    Buse, John B., M.D., Ph.D., professor, University of North 
      Carolina School of Medicine, prepared statement of.........    94
    Davis, Hon. Tom, a Representative in Congress from the State 
      of Virginia:
        Prepared statement of....................................    25
        Prepared statement of Brian Storm........................    16
    Nissen, Steven, M.D., F.A.C.C., chairman, Department of 
      Cardiovascular Medicine, Cleveland Clinic, prepared 
      statement of...............................................    90
    Psaty, Bruce M., M.D., Ph.D., co-director, Cardiovascular 
      Health Research Unit, professor of medicine, Epidemiology 
      and Health Services, University of Washington, 
      investigator, Center for Health Studies, Group Health, 
      Seattle, WA, prepared statement of.........................    99
    Sali, Hon. Bill, a Representative in Congress from the State 
      of Idaho, prepared statement of............................    33
    Slaoui, Moncef M., Ph.D., chairman, research and development, 
      GlaxoSmithKline, prepared statement of.....................   124
    von Eschenbach, Andrew C., M.D., Commissioner, Food and Drug 
      Administration, prepared statement of......................    38
    Waxman, Chairman Henry A., a Representative in Congress from 
      the State of California, prepared statement of.............     5
    Welch, Hon. Peter, a Representative in Congress from the 
      State of Vermont, prepared statement of....................   137
    Yarmuth, Hon. John A., a Representative in Congress from the 
      State of Kentucky, fax from Mr. Buse to Mr. Yamada.........   108



                        WEDNESDAY, JUNE 6, 2007

                          House of Representatives,
              Committee on Oversight and Government Reform,
                                                    Washington, DC.
    The committee met, pursuant to notice, at 10 a.m. in room 
2154, Rayburn House Office Building, Hon. Henry A. Waxman 
(chairman of the committee) presiding.
    Present: Representatives Waxman, Towns, Cummings, Kucinich, 
Davis of Illinois, Tierney, Clay, Watson, Yarmuth, Cooper, 
Hodes, Sarbanes, Davis of Virginia, Shays, Cannon, Issa, 
McHenry, Foxx, and Sali.
    Staff present: Phil Barnett, staff director and chief 
counsel; Kristen Amerling, general counsel; Karen Nelson, 
health policy director; Karen Lightfoot, communications 
director and senior policy advisor; Andy Schneider, chief 
health counsel; Sarah Despres, senior health counsel; Molly 
Gulland, assistant communications director; Steve Cha, 
professional staff member; Earley Green, chief clerk; Teresa 
Coufal, deputy clerk; Caren Auchman, press assistant; Zhongrui 
``JR'' Deng, chief information officer; Leneal Scott, 
information systems manager; Rachel Sher, counsel; William 
Ragland and Kerry Gutknecht, staff assistants; David Marin, 
minority staff director; Larry Halloran, minority deputy staff 
director; Jennifer Safavian, minority chief counsel for 
oversight and investigations; Keith Ausbrook, minority general 
counsel; Ellen Brown, minority legislative director and senior 
policy counsel; Anne Marie Turner, minority counsel; Victoria 
Proctor and Susie Schulte, minority senior professional staff 
members; John Cuaderes, minority senior investigator and policy 
advisor; Patrick Lyden, minority parliamentarian and member 
services coordinator; Brian McNicoll, minority communications 
director; and Benjamin Chance, minority clerk.
    Chairman Waxman. The meeting of the committee will come to 
    Today we are holding a hearing about an important 
medication that is being used by a million Americans to control 
their diabetes. Diabetes is a terrible disease. Diabetics are 
unable to control their blood sugar. High blood sugar affects 
nearly every part of the body and can cause blindness, kidney 
failure, heart attack and stroke. Heart attacks and stroke 
caused by high blood sugar levels end up killing two out of 
every three diabetics.
    Diabetes can't be cured. But with proper medical attention 
and effective drugs, it can be controlled, and the devastating 
consequences of diabetes can be delayed or even prevented. 
Endocrinologists who specialize in the treatment of diabetes 
believe that drugs that lower blood sugar levels are especially 
important to prevent the long-term complications of this 
disease. Avandia was approved in 1999 because of clinical 
evidence that it effectively lowers the blood sugar levels in 
diabetics. Trials conducted since then confirm that Avandia is 
indeed effective in lowering blood sugar levels. That is why it 
has been so widely prescribed by doctors across the Nation.
    Avandia, however, is a sophisticated and complicated drug. 
It works at the gene level and has multiple effects on the 
body. For instance, it may increase weight and cholesterol. 
That is why from the outset, concerns have been raised about 
whether Avandia could increase the risk of heart attacks.
    I have struggled with the right tone for today's hearings. 
Diabetes is a serious illness and Avandia is an effective 
medication for lowering blood sugar. Sounding a false alarm 
about the dangers of the drug has a potential to cause serious 
harm to patients.
    On the other hand, there have been repeated warnings from 
the day of approval forward about the potential cardiac risks 
associated with Avandia. And these should not be ignored.
    It is not Congress' role to adjudicate these medical 
issues. But it is our role to assure that the Federal Food and 
Drug Administration is taking these concerns seriously and 
providing doctors and patients with the guidance they need to 
make informed decisions.
    That is why we are holding this hearing today. Although 
Avandia has been marketed for 8 years and has been used by 
millions of Americans, the post-market studies have not been 
done to say conclusively whether Avandia increases or decreases 
the risk of heart attacks. That is a major failure of our 
system, and that is what is causing so much confusion and worry 
among the patients who are taking Avandia today.
    Avandia was approved on May 25, 1999. The primary medical 
reviewer at FDA recommended approval of the drug because 
clinical trials showed it to be effective at reducing blood 
sugar. That was justified and appropriate. The medical reviewer 
also noticed that the clinical data raised questions about 
Avandia's effect on the heart. I would like to introduce the 
findings of the medical reviewer into the record and read an 
    The excerpt is technical and long, but it reveals how our 
system is supposed to work, and the quote I want to read is: 
``Whether Avandia favorably affects the natural history of type 
2 diabetes is open to question. Long-term improvement in HbA1c, 
a measure of blood sugar, should decrease the risk of 
retinopathy, eye problems, nephropathy, kidney problems, and 
neuropathy, nerve problems. However, the increase in body 
weight and undesirable effects on serum lipids, cholesterol, is 
cause for concern. Heart disease due to atherosclerosis is a 
major cause of morbidity and mortality in patients with type 2 
diabetes, and it cannot be assumed that treatment with Avandia 
would decrease the risk.''
    Well, because of this concern about the potential for 
``deleterious long-term effects on the heart,'' the medical 
reviewer recommended that ``a post-marketing study to address 
these concerns needs to be a condition of approval.'' The 
medical reviewer did everything right. He recognized that 
Avandia held great promise because of its impact on blood 
sugars, and he recognized there were questions about its side 
effects that could be answered conclusively only through a 
properly designed post-market trial. Unfortunately, at that 
point, FDA dropped the ball.
    FDA and the drug manufacturer did agree on a post-market 
study called ADOPT. But it was designed to show whether Avandia 
provided long-term control of blood sugar levels, not to assess 
whether Avandia increases the risk of heart attacks. ADOPT did 
show that Avandia is an excellent drug for keeping blood sugar 
under control, but it did not answer the medical reviewer's 
questions about heart risks.
    FDA did receive several warnings about a potential link 
between Avandia and heart attacks. In March 2000, Dr. John 
Buse, who will testify on the second panel today, wrote FDA to 
request cardiovascular safety trials of high-risk populations. 
In February 2003, the World Health Organization issued a 
warning of the potential cardiac risks associated with drugs 
like Avandia. A year later, a review in the New England Journal 
of Medicine stated that ``Data about the effects of TZDs, drugs 
like Avandia, on cardiovascular disease, are urgently needed.''
    Then in October 2005, the drug manufacturer GlaxoSmithKline 
informed FDA that an internal company analysis showed that 
Avandia may be associated with increased risk of myocardial 
ischemia, a medical term that includes heart attacks. The drug 
manufacturer gave the FDA this analysis 11 months later, along 
with a second study the company sponsored that did not show 
increased risks.
    Yet despite the FDA medical reviewer's recommendation, 
despite additional warnings by outside experts, despite the 
millions of patients who rely on Avandia to control their blood 
sugar, and despite the potential risks involved, FDA never 
required the manufacturer to conduct a thorough post-market 
study of Avandia's heart risks. In fact, it took the 
publication of an article last month in the New England Journal 
of Medicine to spur the agency to public action.
    European regulators were not so negligent. Over 6 years 
ago, they required GlaxoSmithKline to initiate a study called 
RECORD, which is designed to assess cardiovascular risks. The 
company published partial results from this study yesterday. 
Unfortunately, as we will hear from the experts on our second 
panel, the results to date are inconclusive and RECORD does not 
appear to be large enough to answer the key questions about 
Avandia's cardiac risks. It was not designed to be completed 
until 2009.
    While many people watching this hearing today will be 
looking for answers about whether Avandia is safe, and I 
understand and share their desire for answers, but because of 
the lack of data, there may be no definitive conclusions. By 
examining Avandia, however, we can learn a lot about the drug 
approval and post-market surveillance process. Avandia is a 
case study of the need for reform of our drug safety laws.
    As a Member of Congress, I am not qualified to judge 
whether the risks of Avandia outweigh its benefits. But I do 
know that the millions of diabetics who have taken Avandia have 
not been well served by our regulatory system. Doctors and 
their patients should be able to turn to FDA for guidance about 
the safety of the drugs they take. But in the case of Avandia, 
FDA did not insist upon the data it needs to consider their 
questions definitively.
    Legislation has passed the Senate and is pending in the 
House that would give FDA new powers to require post-market 
studies of drugs like Avandia. This hearing will show why these 
reforms are urgently needed. FDA needs the will, the resources 
and the authority to be a more effective watchdog of drug 
    I look forward to the testimony we will receive and I want 
to thank all of the witnesses for being here today.
    I want to now call upon the ranking Republican member of 
the committee, Mr. Davis, for his statement.
    [The prepared statement of Chairman Henry A. Waxman 









    Mr. Davis of Virginia. Thank you, Mr. Chairman, and good 
    Once again, this committee meets to consider serious 
questions about how the Food and Drug Administration and drug 
makers monitor the long-term safety of approved pharmaceutical 
products. In 2004 and 2005, we led an extensive bipartisan 
investigation into the pain reliever Vioxx, confronting many of 
the same questions we face today.
    How effective are programs by the FDA and industry to 
gather timely and useful data on lingering safety concerns 
about approved products? When those safety concerns emerge, how 
should preliminary, often anecdotal information be used by 
regulators, clinicians and patients? And how do we strike the 
correct balance between speedy approval of life-saving or life-
enhancing therapies that patients want and the much slower 
process of amassing statistically valid data sets on long-term 
health outcomes?
    Today's hearing was prompted by recent warnings the 
diabetes medication Avandia, manufactured by GlaxoSmithKline, 
may increase the risk of cardiovascular disease in some 
patients, patients already uniquely vulnerable to heart 
problems. An admittedly limited meta-analysis of disparate 
research findings suggests that increase may be substantial. 
But other studies point to little, if any, measurable increase 
in heart risks.
    So patients and doctors are left with conflicting or 
incomplete information upon which to base delicate judgments 
about the net benefits of various treatment options.
    But this hearing, as the chairman notes, is not about one 
product. At least, it shouldn't be. It is about the 
effectiveness of the overall drug approval and the monitoring 
process. As the chairman's memo to Members cautioned, this 
hearing is not about whether Avandia makes patients healthier 
or harms them. We are not here to substitute our judgment for 
that of scientists and regulators still evaluating clinical 
safety data.
    But we are here to ask whether current post-marketing 
surveillance programs and protocols are both robust and 
sensitive enough to detect emerging evidence of deleterious 
health effects and how that evidence informs regulatory 
research and treatment decisions.
    Taken by almost 1 million Americans today, Avandia was 
approved in 1999 because it lowers harmful blood sugar levels 
in patients suffering type 2 diabetes. Managing type 2 diabetes 
by lowering blood sugar can decrease the patient's chance of 
having diabetes-related problems later in life, such as kidney 
failure, heart disease, stroke and limb amputation.
    But the so-called surrogate endpoint of reduced blood 
glucose is only an indirect measure of the drug's overall 
impact on health. Questions about the extent of any increase 
cardiovascular risk posed by Avandia were raised 8 years ago. 
So the FDA required Glaxo to compare the safety and 
effectiveness of Avandia with other oral anti-diabetes 
medicines. In 2000, the company initiated another large, long-
term clinical trial to look specifically at cardiovascular 
outcomes in people with diabetes using Avandia to manage the 
    So far, results from that study have not shown increased 
health risks at levels suggested by the meta-analysis that 
would require discontinuation of the research for safety 
reasons. Nevertheless, last year, based on data from a study 
involving patients with existing congestive heart failure, the 
FDA required a labeling change for the drug to include a new 
warning about a potential increase in heart attacks and heart-
related chest pain in some individuals.
    The FDA will convene an advisory committee as early as next 
month to review this matter. That committee's findings should 
provide health care providers and patients with a better 
understanding of any cardiovascular risks involved with the use 
of Avandia.
    It is not clear if the advisory committee will also look at 
the entire class of oral anti-diabetes medications that operate 
like Avandia. Perhaps FDA can answer that question today.
    This muddled post-marketing picture is not unique. A recent 
New England Journal of Medicine editorial called the FDA 
approach to post-approval or Phase 4 research ``desultory,'' 
because during the period from 1998 through 2003, only about a 
quarter of the required Phase 4 trials were completed. And as 
of September 30, 2006, a total of 899 Phase 4 studies remain 
pending. As a result, the safety profile of some drugs, 
particularly those approved using surrogate endpoints, can 
remain incomplete for years.
    Most Americans believe once the FDA approves a drug, it 
carries the medical equivalent of the Good Housekeeping seal of 
approval and can be used with little or no risk. But the 
process of developing, marketing, regulating, prescribing and 
using modern pharmaceuticals involves some, at times 
considerable risk, at every stage. Those risks have to be 
acknowledged frankly and managed responsibly.
    Adverse event surveillance and research have to be 
sensitive enough to detect potential safety problems but 
discrete enough to distinguish between well-publicized 
anecdotes and scientific evidence. Otherwise, public confidence 
in both the FDA and the pharmaceutical industry will be 
undermined by conflicting data and allegations no one is 
protecting the long-term welfare of patients.
    I look forward to hearing from our panels of expert 
witnesses today on how we can strengthen FDA approval and post-
marketing surveillance systems. I would ask unanimous consent 
that the statement of Dr. Brian Strom, the chairman of 
biostatistics and epidemiology and director of the Center for 
Clinical Epidemiology and Biostatistics at the University of 
Pennsylvania be included in the official hearing record.
    Chairman Waxman. Without objection, that will be the 
    [The information referred to follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.010
    [GRAPHIC] [TIFF OMITTED] 44429.011
    [GRAPHIC] [TIFF OMITTED] 44429.012
    [GRAPHIC] [TIFF OMITTED] 44429.013
    [GRAPHIC] [TIFF OMITTED] 44429.014
    [GRAPHIC] [TIFF OMITTED] 44429.015
    [GRAPHIC] [TIFF OMITTED] 44429.016
    [GRAPHIC] [TIFF OMITTED] 44429.017
    Mr. Davis of Virginia. Thank you.
    [The prepared statement of Hon. Tom Davis follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.018
    [GRAPHIC] [TIFF OMITTED] 44429.019
    Chairman Waxman. We have a number of witnesses to present 
testimony to us today. So we did not invite Members to give 
opening statements. Of course, all of the Members' opening 
statements that they wish to submit will be made part of the 
    But we do have a request from Congressman Towns and I do 
want to recognize him. In doing so, I will invite any other 
Member who wants to make a very brief statement to do so. But 
do recognize the fact that we will keep it brief, and you may 
submit a fuller statement for the record.
    Mr. Towns.
    Mr. Towns. Thank you very much, Mr. Chairman. I thank you 
for calling this hearing on patient safety.
    As you know, diabetes and heart disease occur in the 
African American population at a rate disproportionate to the 
general population. That is also true of Hispanic Americans. 
Death rates for strokes are about 25 percent higher for African 
American males and about 20 percent higher for African American 
women. African Americans develop high blood pressure at an 
early age, and heart disease death rates are 1.5 times higher 
and 1.8 times greater for fatal strokes.
    Yet, despite the disproportionately higher mortality and 
morbidity of cardiovascular disease, Latinos and African 
Americans are significantly less likely than Whites to undergo 
treatment for their conditions, and less likely to receive the 
most advanced cardiac procedures. Despite having the same 
insurance status and disease severity rates, diabetes rates are 
also significantly higher for African Americans and Hispanic 
Americans. These are also not one at a time conditions. If you 
have one, there is a greater likelihood that you may have them 
    The published higher death rates from the May 16th New 
England Journal of Medicine study is of course what brings us 
here today. However, Mr. Chairman, while I am certainly 
concerned about the possibility or the potential higher level 
of risk for cardiovascular causes that has been associated in 
this single study of Avandia, I am more concerned with the 
likelihood of the low levels of participation of African 
Americans and other people of color in the clinical trials 
associated with Avandia.
    I am certainly aware of the large number of clinical trials 
associated with it. However, I am particularly concerned that 
the findings have not had sufficient data to make a 
determination as to the effects of this drug on African 
Americans and Hispanics, whether they associate Avandia with 
the higher levels of risk for death from cardiovascular causes 
or not.
    While we are not here today, Mr. Chairman, to discuss the 
reauthorization of the Prescription Drug User Fee Act, a number 
of us serve on the Committee on Oversight and the Committee on 
Energy and Commerce, as you and I do. I am here today to make 
sure that both the Food and Drug Administration and the 
pharmaceutical and medical devices industry takes the expansion 
of the numbers of African Americans and Hispanic Americans in 
drug and medical devices studies seriously.
    I am therefore proposing in the PDUFA reauthorization a 
more verifiable alternative for minorities than the pediatric 
exclusion and an office of diverse population within the Office 
of the FDA Commissioner that will have the authority and 
responsibility of increasing the numbers of racially and 
ethnically diverse populations within the FDA.
    Mr. Chairman, I believe that we need to get to the bottom 
of whether or not there is associated risk with Avandia. 
However, that risk should have scientific evidence that applies 
to ethnically and racially diverse communities, as well as the 
general population.
    And on that note, I yield back, Mr. Chairman, and thank you 
for the special consideration.
    Chairman Waxman. Thank you, Mr. Towns.
    Does any other Member wish to make an opening statement? 
Mr. Issa.
    Mr. Issa. Thank you, Mr. Chairman. I will be brief.
    But I think it is important, first of all, I would like to 
thank you for your opening statement. I think it helped balance 
perhaps what started off very much as imbalance in this 
hearing. I am concerned today that we not tread too closely 
toward the hypocrisy that I believe this hearing begins to look 
    Just a few months ago, this committee held a hearing in 
which the Bush administration was accused of politicizing 
science, of censoring and editing research and politicizing 
science is exactly what we could be doing here today. This is 
not global warming, this is in fact, though, an ongoing 
investigation on a current drug early in the questioning 
period. I believe that the anecdotal evidence that came out 
from the New England Journal of Medicine, which we now 
understand included some consulting to the majority members of 
this committee, is in fact a very dangerous pattern.
    A few weeks ago, the New England Journal of Medicine 
questioned something. We now hold a hearing on that drug and 
consistent with that drug. As the chairman said, rightfully, 
and I appreciate his saying it, none of us here is qualified to 
evaluate this drug. As a matter of fact, none of the people 
speaking before us today, without a vast group of people not 
present, is capable of evaluating the safety and side effects 
of this drug. It is in fact the FDA and science's community 
responsibility to get all the research in, and in fact then to 
go through that as a panel, not as one individual speaking 
before this committee.
    I appreciate that this is the Committee on Oversight and 
Government Reform. If we are doing oversight, I believe that it 
is OK to look at something if it is a clear and present danger. 
That is not the case here. This drug is very much still 
effective and on the market for patients today and should not 
be artificially called into question as to its safety or side 
effects as a result of anecdotal information presented here.
    Vioxx, Celebrex and other drugs certainly have gone through 
a much more exhaustive study and could be just as easily used 
to show the need for reform and in fact, as an oversight 
agency, to look at past failures. I believe that we are 
treading very close to exactly the hypocrisy that this 
committee can easily be drawn into, politicizing science while 
saying that we don't want to politicize science. So I 
appreciate the chairman's opening remarks. Hopefully that has 
set a tenor for not only what is being said by the witnesses 
today, but in fact for our questions, that we not allow this to 
be about one drug or one limited study, and that we try to stay 
toward the settled science, toward the settled cases of the FDA 
in our oversight and potential reforms.
    I thank the chairman for his opening statement, because 
hopefully it brought us a little closer--and the ranking 
member--a little closer toward the correct reason for this 
committee to hold these types of hearings. I yield back and 
thank the chairman.
    Chairman Waxman. Mr. Issa, I am pleased you attacked the 
hypocrisy that you admitted did not exist. I don't know if the 
New England Journal of Medicine would resent being categorized 
as a magazine that simply puts together a bunch of anecdotes, 
but I certainly resent the statement that there was any kind of 
consultation between the people that wrote the article in the 
New England Journal of Medicine and the majority of this 
committee. It is just absolutely not true.
    Mr. Issa. Mr. Chairman, the author of the study published 
in the New England Journal of Medicine admitted to the Wall 
Street Journal that he had talked to people on the Hill while 
preparing his analysis. Yet the FDA says that no one has 
consulted them. So in fact, I believe that this is dangerously 
close to that question of politicizing science. And like I say, 
I appreciate the fact that your opening statement was balanced. 
But we have to look at the underlying premise of bringing a 
hearing on a drug 3 weeks after an article comes out and the 
author of that article admits that he's been talking to people 
on the Hill.
    This is one of those times in which I want to make sure 
that this is not an attack on the practice of a particular 
company, or a chilling effect on companies, but rather, 
legitimate oversight and legitimate effort to find reform. I 
appreciate the chairman's effort to try to lead at that 
direction. I wanted to make sure that I supported him in 
pushing this hearing in that correct direction.
    Chairman Waxman. I thank you for your explanation of your 
conclusion. And it will stand for all to review. And I 
appreciate your statements.
    Any other Member wish to make an opening statement? Yes, 
Mr. Davis.
    Mr. Davis of Illinois. Thank you, Mr. Chairman. I do not 
have a written statement, but I do want to, as a member of the 
committee, thank you for calling the hearing. And also as a 
person who has been diagnosed as a type 2 diabetic, I want to 
emphasize the particular personal interest that I have in this 
hearing. I agree with the conclusion in your opening statement 
that I hope that we will move toward, and we do in fact need a 
stronger and more resourceful Food and Drug Administration, so 
that they have not only the authority but also the resources 
that are needed to do extensive research and oversight to try 
and assure that the pharmaceutical drugs that we use for 
medical treatment are as safe as humanly possible.
    So again, I thank you for calling the hearing and look 
forward to hearing the witnesses.
    Chairman Waxman. Thank you very much, Mr. Davis.
    Any other Member wish to make a very brief statement? Ms. 
    Ms. Foxx. Thank you, Mr. Chairman. I appreciate it very 
    My background is as a social scientist. I worked for many 
years in medical research. So in reading the material about 
today's hearing, I tried to bring back some of my experiences 
of some time ago. And I wanted to get a definition of the term 
``meta-analysis.'' I think that it is really important that in 
this hearing we keep in mind what a meta-analysis is.
    The purpose of it is to raise questions but not to draw a 
conclusion. Let me read you a definition from Taber's 
Cyclopedic Medical Dictionary. It says, ``Meta-analysis, a 
statistical procedure for combining data from a number of 
studies and investigations in order to analyze the therapeutic 
effectiveness of specific treatments''--and this is the really 
important part--``and plan future studies.''
    The meta-analysis does not actually do research. It does 
not gather the data that is so important to gather when drug 
companies are searching for the effectiveness of the drugs 
they're working with. So I think it's extremely important that 
we keep in mind what a meta-analysis is.
    Now, Mr. Chairman, on May 21st, Dr. Nissen's study was 
published by the New England Journal of Medicine, along with a 
Journal editorial encouraging physicians to stop prescribing 
the drug and encouraging the FDA to take regulatory action. 
Then there were alarming headlines pronouncing an increased 
risk of death for anyone taking this drug.
    According to a very interesting article entitled Political 
Defibrillator, published in the May 28, 2007 issue of 
Biocentury, a journal providing analysis for the biotechnical 
community, soon after the release of Dr. Nissen's study, some 
of my congressional colleagues in the House and Senate issued 
statements to the press suggesting that they knew ahead of time 
about this study. Included among the press releases, there was 
an apparent attempt to manufacture a scandal, including the 
statement that ``Both the drug company and the FDA have some 
major explaining to do about what they knew about Avandia, when 
they knew it and why they didn't take immediate action to 
protect patients.'' These statements were made with disregard 
for the limits of this study and the impact that these 
statements and actions could have on public safety or the 
reputation of the company involved.
    Let me read the opening paragraph of the Biocentury piece: 
``The circumstances surrounding the publication by the New 
England Journal of Medicine of a meta-analysis of safety data 
from studies of Avandia and an accompanying commentary 
suggesting that FDA critics on Capitol Hill have collaborated 
with whistleblowers in the agency and pharmaceutical industry 
critics and academia to create a controversy over Avandia's 
safety in order to advance a political agenda.'' According to 
this article, even though Members of the Senate and House and 
their staffs were apparently aware of this study and that it 
was going to be published, the author never notified the FDA. 
Yet the FDA is the one agency that holds the key to action if 
this study in fact reveals data about an immediate threat to 
the public.
    The British medical journal, the Lancet, published May 23, 
2007, took issue with how this was handled, stating that ``To 
avoid unnecessary panic among patients, a calmer and more 
considered approach to the safety of rosiglitazone is needed. 
Alarmist headlines and confident declarations help nobody.''
    Mr. Chairman, while there is no need to manufacture a 
scandal here, it appears that there may already be one that 
needs investigating, at least by the press. I would like to see 
the press determine what Members of Congress and their staff 
knew about this study, when they knew it and whether there was 
a coordinated effort among the author, disgruntled FDA staff 
and staff at the New England Journal of Medicine to develop and 
publish this study in a way that would create a sensation in 
the press and maximum embarrassment for the FDA.
    My husband is diabetic. So I am very interested in this 
disease and very interested in our finding treatments for it. 
It is a very pernicious disease and one of the most expensive 
in our country.
    However, we serve no purpose by scaring people about drugs. 
And I have no dog in this fight, as they say. I am not here as 
an apologist for Glaxo, but I think we should be very careful 
when we talk about scientific issues and make sure that we have 
a balanced approach to this. Thank you, Mr. Chairman.
    Chairman Waxman. The gentlelady's time has concluded.
    I would like to get to the witnesses. Does any Member feel 
compelled to say anything further? Yes, the gentleman from 
    Mr. Lynch. Thank you, Mr. Chairman, and I will be brief.
    I just wanted to address a couple of things. First of all, 
there has been the allegation that this study was anecdotal. I 
just want to point to the editorial itself and the reports and 
the concerns that have been cited by the doctors. They were 
based on 40 different studies, and I think they are very 
    Second, I agree with the sentiment, although I am not sure 
it is shared, that this shouldn't be dragged down into some 
type of partisan politics issue. However, I think when you 
begin the hearing by criticizing the New England Journal of 
Medicine because of something that has been published there, 
which is, I think, a very thoughtful view, it is just one view, 
but very thoughtful, but to impugn their character that it is 
somehow in league politically to take down a drug company, I 
think you immediately drag down the debate to that level. I 
would just caution against it.
    The second comment I want to address is the idea that 
somehow folks that come to the Oversight Committee because of 
an issue of genuine concern have done so for political purposes 
and not for legitimate reasons has not been proven here, and 
should not be suggested. This is where people should come. It 
should not be circumstantial evidence to the disingenuousness 
of people who come to this committee that they have come to us 
with an issue. This is the Oversight Committee. This is where 
they should be coming. And we should have the intelligence and 
the balance here to just let the evidence be presented and not 
suggest that it is being done for a disingenuous reason and 
then have it presented in that context.
    This is a tremendously important issue. My family has 
diabetes, I know thousands and thousands of families that are 
dealing with this problem. We should approach this as adults. 
And at the end of the day, it may prove that the concern was 
elevated. It may prove that the concern was understated, but we 
should receive the evidence in an open and honest discussion. 
That is the way we should have it, and I yield back.
    Chairman Waxman. The gentleman's time has expired. We will 
now go to our witnesses.
    Mr. Sali. Mr. Chairman, may I make a brief statement?
    Chairman Waxman. The gentleman is recognized for a brief 
    Mr. Sali. Mr. Chairman, it appears to me, in hearing the 
opening statements and kind of thinking through this, that the 
real concern is that there may be a side effect from this drug. 
And we don't know if that side effect is present based on this 
meta-study, that it may be a side effect.
    I also understand that, according to the FDA, no approved 
diabetes drug has ever shown any kind of reduction in 
macrovascular risk, the kinds of risk that may exist here 
today. So I guess in the testimony, I am hoping that it becomes 
clear, No. 1, whether we can really say that the side effect 
does exist from this drug, and if it doesn't, then I think our 
job of oversight may be done at that point.
    Second, even if it does exist, does it exist in such a 
significant number of cases that we know about that we can say 
the FDA is off track and this committee, with its oversight 
capability, should intervene?
    Finally, Mr. Chairman, I think the question is, knowing 
that there is a side effect, is it appropriate for doctors to 
prescribe it anyway? There are plenty of drugs that have known 
side effects. If patients are better off if this drug is 
prescribed, perhaps it will change prescribing patterns for 
physicians that are involved. But if there is a known side 
effect, if everybody takes that into account in making the 
decision whether to take the drug, prescribe the drug, are the 
people better off who can take this drug by prescription? And 
if they are, again, this committee has no business in providing 
    [The prepared statement of Hon. Bill Sali follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.020
    [GRAPHIC] [TIFF OMITTED] 44429.021
    Chairman Waxman. Well, perhaps we can get some answers to 
those questions from the scientists.
    I would like to welcome our first witnesses. Dr. von 
Eschenbach is the current Commissioner of the Food and Drug 
Administration. He is the former head of the National Cancer 
Institute and is a renowned cancer specialist. We are delighted 
to have you here to testify.
    Accompanying Dr. von Eschenbach is Dr. Dal Pan, who is the 
head of the Office and Surveillance and Epidemiology at the 
Food and Drug Administration. And Dr. Jenkins is the head of 
the Office of New Drugs at FDA. We want to welcome each of you 
to our hearing today. We are looking forward to your views on 
some of these scientific and regulatory questions that Members 
have on their minds.
    It is the practice of this committee to ask all witnesses 
to take an oath. I would like to ask you to rise.
    [Witnesses sworn.]
    Chairman Waxman. Thank you very much. The record will 
indicate that each of the witnesses answered in the 
affirmative. Dr. von Eschenbach, why don't we start with you?
    We ordinarily ask witnesses to be limited to 5 minutes in 
their oral presentation. Your full statement will be part of 
the record. We will run the clock, if you need a little bit 
more time, we will certainly provide it to you.


    Dr. von Eschenbach. Thank you very much, Mr. Chairman and 
Ranking Member Davis and members of the committee. I really 
want to express our appreciation for allowing us to appear 
before you today.
    My written testimony provides important details about the 
scientific facts and many post-marketing trials that are 
involved in FDA's ongoing multi-faceted regulation of the 
diabetes drug rosiglitazone, perhaps better known as Avandia. 
Rather than recount those details, I would like to focus my 
oral statement on the process used at FDA to do the right thing 
for patients by making decisions using a comprehensive, 
multidisciplinary approach that incorporates all the data 
available and addresses the best interest of all patients 
affected by that decision.
    With me are two senior and expert FDA colleagues: Dr. John 
Jenkins, the Director of the Office of New Drugs; and Dr. 
Gerald Dal Pan, the Director of the Office of Surveillance and 
Epidemiology, formerly the Office of Drug Safety. Both of these 
offices are part of FDA's Center for Drug Evaluation and 
Research. Their presence this morning is important regarding 
the FDA's decisionmaking process, because they represent the 
close interaction between the FDA office that reviews marketing 
applications for new drugs and the office that monitors their 
safety profile.
    We are here as partners, reflecting the management and the 
professionals at the FDA who are dedicated to collaborating 
even more closely, not simply to approve products, disapprove 
them or defer decisions, but rather, to do the right thing, so 
that our actions will both promote and protect the health of 
    Mr. Chairman, I know that you called this hearing because 
of your deep concern for the welfare of Americans, a motivation 
that transcends politics and that is shared by every member of 
this committee. I know you and Members of Congress want and 
even demand that the FDA do its utmost to protect and promote 
the health of all Americans, including those millions of 
Americans affected by diabetes, and the hundreds of thousands 
that are perhaps using the drug Avandia.
    Let me be clear at the outset. Our focus in the decisions 
FDA has made and will make on Avandia is to serve an 
approximate 18 to 20 million Americans who are at risk of 
blindness, kidney failure, limb amputation and death from 
diabetes. We will carry out that mission by thoughtfully 
weighing the potential effect of FDA's actions on the entire 
patient and on all patients. It is our goal to not just make 
the right decision about a drug like Avandia; but more 
importantly, to always do the right thing for patients.
    How do we do the right thing? First, by doing it as a team 
that embraces the diversity of all points of view and weighs 
all points of view to arrive at an FDA decision. Second, by 
using decision standards that are science-based, drawing upon 
all the scientific data that bears on an issue and by demanding 
of ourselves and others rigor, precision and accuracy in the 
analysis of that data. Because our decision that weighs both 
the benefits and the risks of a drug will affect not one or a 
few, but often millions of lives.
    Third, by committing to a standard of excellence that 
requires us to constantly improve the processes by which we 
make decisions. Since I arrived at FDA, we have specifically 
addressed process improvement as it relates to decisions 
regarding drug safety. We have completed or are rapidly putting 
in place more than 40 drug safety initiatives that are in 
keeping with the recommendations of the Institute of Medicine 
report that we commissioned.
    A few recent examples of process improvement are the fact 
that we have issued a guidance on communicating drug safety 
information, announced the creation of a risk communications 
advisory committee, proposed tougher procedures for membership 
on FDA advisory committees, and our critical path initiative 
promises to provide the modern tools needed to improve the 
predictability of the processes by which products are 
discovered, developed and monitored after delivery to patients.
    We have acknowledged that increasing demands and the 
complexity of the products we regulate requires increasing 
resources. We are grateful for the administration's proposals 
and the congressional consideration given to the additional 
resources in fiscal year 2007 and those being considered for 
    Among the many needs, we must especially use these 
resources to build a more robust FDA infrastructure of 
information technology to obtain and analyze all the data 
required for timely and accurate decisions. We need to focus on 
product safety throughout the entire life cycle of the product, 
including stronger post-market surveillance and pharmaco-
vigilance. In fact, a robust pharmaco-vigilance system 
supported through a public-private arrangement such as an 
institute or a foundation could provide considerable benefit 
and would be most welcome as part of the congressional 
consideration of pending FDA legislation.
    In closing, Mr. Chairman, let me emphasize that as we deal 
with drug safety, we encourage those with an interest to bring 
to us comments, ideas and data from all sources. FDA is 
committed to appropriate scientific dialog and discussion about 
the making of decisions. And in the end, we must always be true 
to our mission to both protect and promote the health of all 
    Mr. Chairman and members of the committee, thank you for 
your time, your interest and your commitment to this mission. 
My colleagues and I would be pleased now to answer any 
    [The prepared statement of Dr. von Eschenbach follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.022
    [GRAPHIC] [TIFF OMITTED] 44429.023
    [GRAPHIC] [TIFF OMITTED] 44429.024
    [GRAPHIC] [TIFF OMITTED] 44429.025
    [GRAPHIC] [TIFF OMITTED] 44429.026
    [GRAPHIC] [TIFF OMITTED] 44429.027
    [GRAPHIC] [TIFF OMITTED] 44429.028
    [GRAPHIC] [TIFF OMITTED] 44429.029
    [GRAPHIC] [TIFF OMITTED] 44429.030
    [GRAPHIC] [TIFF OMITTED] 44429.031
    [GRAPHIC] [TIFF OMITTED] 44429.032
    [GRAPHIC] [TIFF OMITTED] 44429.033
    [GRAPHIC] [TIFF OMITTED] 44429.034
    [GRAPHIC] [TIFF OMITTED] 44429.035
    [GRAPHIC] [TIFF OMITTED] 44429.036
    [GRAPHIC] [TIFF OMITTED] 44429.037
    [GRAPHIC] [TIFF OMITTED] 44429.038
    [GRAPHIC] [TIFF OMITTED] 44429.039
    Chairman Waxman. Thank you very much, Dr. von Eschenbach.
    We are going to start with 10 minutes on each side. I want 
to thank you very much for your testimony. You are very 
distinguished scientists and I know that you have a job at FDA 
that you are trying to see through and relying on good science 
and recognizing the public interest. Of course, I have been a 
strong supporter of the FDA, because I think the American 
public expects the FDA to make sure that drugs that are 
available to them are safe and effective, not just at the time 
they are approved, but throughout the time the drug is 
available and going to be used. And that information is to be 
based on science, not rumors, not anecdotes, not demagoguery, 
but science.
    The issue with Avandia, like so many other drugs, it was 
approved without the full knowledge of all the impacts it might 
have. This is not unusual, because many drugs need to be 
watched carefully after they are approved. But there has been a 
pressure at FDA to get drugs approved as quickly as possible. 
In fact, we even have user fees that can help FDA have more 
resources to get those drugs approved.
    The question that I am looking at is the post-marketing 
surveillance of this drug as it reflects post-marketing 
surveillance of other drugs. This particular drug was approved 
in 1999. And your reviewer at the FDA did, as I mentioned in my 
opening statement, exactly what he should do. He looked at the 
effectiveness, whether it lowers blood sugar, and he found that 
there was enough clinical evidence to show that it did.
    But he was concerned about the possibility of increased 
heart attacks, strokes, because of some evidence that he saw in 
the data, and suggested that there be a post-marketing 
surveillance of that issue. So in 1999, we had this opportunity 
for FDA to make sure that the post-marketing study was being 
    But it wasn't done. And then later, in 2000 and 2003, you 
mention in your statement, you welcome the input from those who 
have concerns, well, FDA got input from people who had 
concerns. Dr. Buse wrote to FDA to express his concern about 
Avandia's potential cardiovascular risks. And he urged the FDA 
to conduct a cardiovascular safety trial on high risk 
populations. It is still not being done.
    In February 2003, the World Health Organization issued a 
warning of potential cardiac risks associated with Avandia's 
class of drugs. And this was another opportunity for FDA to 
insist that a post-market study be done by the manufacturer on 
this potential danger, and nothing was done. Not until we got 
this report in the New England Journal of Medicine has there 
been this great concern expressed in the public, which I must 
state to you, I had nothing to do with, nor did any member of 
my staff have anything to do with, nor would the distinguished 
journal welcome us to get involved in their scientific 
    So there are a number of missed opportunities. What 
happened? Wy didn't FDA insist on the post-marketing 
surveillance to look at the risk for heart attacks and strokes?
    Dr. von Eschenbach. Thank you, Mr. Chairman. First of all, 
I would like to echo your important emphasis on the fact that 
we are in fact looking at these issues from the point of view 
of the total life cycle of product. We are building in much 
more opportunity to assure the safety of these drugs, even 
before they are allowed to be applied to patients in the 
general population.
    We are doing that in the most efficient and effective way 
we can, so that it is more rapid, so that we can get these 
life-saving and life-enhancing drugs to people. But that 
rapidity does not mean it is reckless. We are applying the 
rigor and precision and discipline in the internal processes, 
and also recognizing, as you pointed out, that once that drug 
goes out into a much larger population, no clinical study or 
trial could ever give us all the information we need. So we are 
engaged in rigorous post-market surveillance.
    With regard to this drug, there were post-marketing studies 
being conducted. FDA continued to be engaged in acquiring, 
analyzing and assessing data coming in with regard to the 
experience that was being developed with Avandia and these 
large populations, both here and in Europe, and did in fact 
take regulatory action. I would like to ask Dr. Jenkins----
    Chairman Waxman. Before you talk about the regulatory 
action, did you ask for and did you get a study on the 
potential side effects dealing with the heart, as was 
recommended by so many others that I mentioned. Did you 
actually tell the manufacturer to do the study so you could 
have a definitive study?
    Dr. von Eschenbach. I am going to let Dr. Jenkins talk 
about the approval and what was involved, and Dr. Dal Pan 
describe the post-market assessment.
    Chairman Waxman. I am more interested in the post-market. 
Because the approvals seem to be reasonable. You have enough 
evidence. The reviewer saw the studies, said, this drug merits 
approval from what we have seen so far. But raised a concern 
about the possible heart attack problem. And he recommended 
that there be a followup post-market review.
    Dr. Dal Pan, why wasn't one done? Which of you----
    Dr. von Eschenbach. This is on the approval, Dr. Jenkins.
    Dr. Jenkins. Thank you, Mr. Chairman. Let me try and 
address that point.
    I was the senior member of the review team that reviewed 
Avandia back in 1999. I actually signed the approval letter for 
Avandia in 1999. And the approval did have a phase 4 commitment 
for a long-term, 4-year safety and efficacy study titled ADOPT, 
which was designed to look at the long-term efficacy of the 
drug, but also long-term safety and specifically reading from 
our post-marketing commitment Web site, we talked about long-
term safety, including hepatic effects, cardiovascular and 
hematologic effects, changes in body weight and serum lipids.
    So the medical officer that you are describing who, in his 
review called for the study, this is the same study that he was 
calling for that we actually got as a post-marketing 
    Chairman Waxman. Did the study, the ADOPT study look at the 
specific concerns about potential heart attack? I know you 
requested it. But in my understanding, the ADOPT study only 
confirmed that the drug was effective in lowering blood sugar.
    Dr. Jenkins. At the time we approved Avandia, there were 
quite a number of different questions we had that we were 
looking for answers for. One of them was about its long-term 
efficacy in comparison to other drugs. There were concerns 
about its hepatic safety, because the previous member of this 
class had proven to have a liver toxicity signal. There were 
also concerns about congestive heart failure and edema.
    Chairman Waxman. Did the study give you the answers you 
needed on the question of the safety matters that involved the 
larger population using the drug? Did we have the answers from 
that study that we can now cite as showing us, on this specific 
issue of cardiac problems, that we now know the risks?
    Dr. Jenkins. The study was not specifically designed to be 
a study to evaluate myocardial infarction or heart attack in 
and of itself. It was designed to look at cardiovascular 
outcomes. We now have the data from that study. It was 
published last fall, it is currently under review by FDA----
    Chairman Waxman. You are talking about ADOPT?
    Dr. Jenkins. ADOPT. It provided a lot of very valuable 
information about the cardiovascular safety of Avandia, as well 
as its liver safety, its effectiveness in long-term use. So I 
think it was a very useful study.
    Chairman Waxman. And when was that study concluded?
    Dr. Jenkins. I can't give you the exact date when it was 
concluded. It was published last fall and it was submitted to 
the FDA as a final study report earlier this year. It is 
currently under a complete review by the FDA.
    Chairman Waxman. Did it show that there were more heart 
    Dr. Jenkins. The overall data did not seem to suggest that 
there was a difference between Avandia and Metformin, another 
commonly used drug, or a sulfonylurea, I think it was 
glyburide, in that study.
    Chairman Waxman. So you didn't have any reason as a result 
of that study to think anything more needed to be done?
    Dr. Jenkins. We only got the final study report of ADOPT 
earlier this year. It is still under review. We have not 
completed our review of that study. The results I am describing 
are what are in the published article from last fall.
    Chairman Waxman. The company says that they have the study 
RECORD. They weren't told to do that study by the FDA, but by 
the Europeans.
    Dr. Jenkins. Right.
    Chairman Waxman. And they cited some preliminary data from 
that study which was specifically on the cardiac problems. And 
they said, well, this shows that it is not a problem. But some 
of the critics say, well, it wasn't a big enough population 
covered in that study.
    Why did they do a second study if ADOPT resolved this 
    Dr. Jenkins. The RECORD study was requested as a post-
marketing commitment by the European regulatory agency when 
they approved the drug shortly after we did. So it was designed 
to address different questions. As I said, at the time of 
approval, there were multiple questions that could be answered 
by different studies. They chose to try to address a 
cardiovascular outcome study. Those data just recently became 
available and are under review at FDA. As you know, they were 
just published online in the New England Journal of Medicine 
    Chairman Waxman. My time is up, but I would submit to you, 
Dr. Jenkins and Dr. von Eschenbach, that the study, ADOPT, did 
not have a sample big enough, from what I understand, of the 
cardiac issues. It was not conclusive on that question. Even 
accepting what you had to say, it took 8 years before you got 
that study. And there had been enough warning signs that this 
is a problem, even before the New England Journal of Medicine 
article finally came out with their report.
    You had a number of instances where FDA's intention should 
have been to ask for a genuine study looking at this specific 
issue. Because after all, heart attacks and strokes are one of 
the leading causes of death for people with diabetes. We want 
to know if this drug is reducing the risk or increasing the 
risk. That is an issue that I don't think we fully resolved, or 
do you believe we resolved?
    Dr. Jenkins. If I could respond to that, we did ask for a 
study to look at the long-term safety of Avandia. And we have 
the results of that study under review. The Europeans asked for 
a different study. We now have an interim analysis from that 
    There were several different issues related to the cardiac 
effects of Avandia that were of interest in 1999 and 2000 when 
those studies were designed, including congestive heart 
failure. So you are probably correct that the RECORD study 
doesn't look like it is going to be adequately powered for the 
endpoint of myocardial infarction or heart attack alone. That 
was not the primary concern in 2000 when the study was 
    Chairman Waxman. But there are others who have raised that 
    Dr. Jenkins. We do have very valuable information coming to 
bear on this question.
    Chairman Waxman. Dr. Dal Pan, you reviewed this ADOPT 
study, and other studies post-market?
    Dr. Dal Pan. Right.
    Chairman Waxman. Do you think we have concluded this issue 
as a result of this ADOPT study?
    Dr. Dal Pan. I don't think we have come to a conclusion as 
a result of this or any study. I think we are still looking at 
all the data. We are looking at exactly how the study was 
designed, conducted, taking apart the data, if you would. We 
are also doing that for RECORD. We are taking a careful look at 
how the study was designed, what it can and can't answer. We 
only have data that is essentially what we have in the online 
publication from the New England Journal about RECORD. We don't 
have the data sets or anything like that to look at it more 
thoroughly. But we are looking at the design and the end term 
analysis results.
    Chairman Waxman. Thank you.
    Mr. Davis.
    Mr. Davis of Virginia. Thank you. I want to thank you all 
for your time.
    There is controversy in the medical community about the use 
of surrogate endpoints because drugs approved on this basis are 
not required to demonstrate actual clinical benefit. Is that 
    Dr. von Eschenbach. The expectation is that we look at a 
clinical endpoint that will reflect the favorable outcome of 
survival, the improvement.
    Mr. Davis of Virginia. But we don't test for survival, we 
just look at the endpoint and assume the rest, basically.
    Dr. von Eschenbach. Correct.
    Mr. Davis of Virginia. Some argue that the Avandia was 
approved on a surrogate endpoint, and while the drug is clearly 
efficacious, the health benefits haven't been demonstrated for 
exactly that reason. If you were to sit through the whole 
process it could take years to get any kind of approval.
    Dr. von Eschenbach. That is correct. It was also approved 
in the context of the overall experience with diabetes, both 
type 1 and type 2, where it is recognized that control of blood 
sugar is an extremely important part of care, resulting then in 
the ability to reduce the complications and problems that then 
would reduce the risk of death and----
    Mr. Davis of Virginia. I guess my question is, what effect 
would abandoning glycemic control as an endpoint have on the 
approval process for a diabetes drug?
    Dr. von Eschenbach. If we were to eliminate that and go to 
a model that said we could not make a decision about a drug 
until we had absolute outcomes with regard to death, you would 
be looking at studies that would have to go on for decades, 25, 
30 years perhaps, before you would get an answer.
    Mr. Davis of Virginia. So if you went to that to get a 
diabetes drug approved, if the outcome trials were needed pre-
approval, you are talking decades?
    Dr. von Eschenbach. There would literally be millions of 
people or hundreds of thousands of people dying in the interim 
until we got that answer.
    Mr. Davis of Virginia. Some in the medical community have 
been critical in recent weeks that Dr. Nissen's study was 
rushed to publication, and created unnecessary confusion and 
concern among diabetics. How has the meta-analysis published in 
May in the New England Journal of Medicine contributed to our 
understanding of the balance of risks and benefits of Avandia?
    Dr. von Eschenbach. We view the publication of this meta-
analysis, along with all of the other pieces and data of 
information that we had, both from other meta-analyses as well 
as data and information from controlled clinical trials. So we 
welcome the additional contribution, recognizing that like 
other meta-analyses, there are limitations of these kinds of 
studies. That is factored in, obviously, to the weight we apply 
to a meta-analysis.
    But the important point is, it was one piece of information 
in a large portfolio of data and information that we, the FDA, 
have available to us upon which to make ultimate decisions 
about the right thing.
    Mr. Davis of Virginia. In fact, the editorial itself notes 
the study has a number of weaknesses, only summary trial level 
data rather than patient level data were available. So it was 
not possible to conduct time to event analyses or to evaluate 
the time course of risks. And they note in this setting the 
possibility that the findings were due to chance cannot be 
excluded. So the meta-analysis could be basically irrelevant.
    Dr. von Eschenbach. As you are very well pointing out, 
there are limitations to any study. There are particular 
limitations to a meta-analysis. We took the opportunity to 
recognize this, along with other information, were clues in any 
kind of detective game. But we had to look at all the clues, 
all the information, all the data from all sources.
    Mr. Davis of Virginia. Now, you had done your own meta-
analysis, am I right on that?
    Dr. von Eschenbach. That is correct.
    Mr. Davis of Virginia. Prior to this article?
    Dr. Dal Pan. Dr. Dal Pan can speak specifically to our 
analysis on that, Mr. Davis.
    Mr. Davis of Virginia. That is what I am interested in.
    Dr. Dal Pan. So in August 2006, the company submitted what 
was called a pool of clinical trial analysis, essentially a 
meta-analysis. That was one of two studies they submitted. They 
also submitted a large observational epidemiologic study. The 
pooled clinical trial analysis, the meta-analysis, suggested a 
risk of heart attacks, let's call it, while the observational 
study did not suggest that risk. So one of our challenges was 
to try to reconcile this apparent difference.
    As part of that, we looked into both of these studies and 
we realized that there were some methods that the company used 
that we didn't think were the best methods, given the data they 
had. We had the data and our statisticians have recently 
completed their own meta-analysis of the data.
    Mr. Davis of Virginia. And what have your statisticians 
    Dr. Dal Pan. The statisticians came up with a numerical 
finding that is similar to the company's and similar to Dr. 
Nissen's, approximately a relative risk of 1.4. Now, the job of 
the FDA at this point is to look at those data in, how can I 
put it, in a more granular level, to look to see if there are 
sub-groups of patients who may be at particular risk, to 
analyze the data more to see what's contributing to that, and 
also to put it in the context of all the other data we have. So 
that is an ongoing process.
    Mr. Davis of Virginia. So you haven't reached any 
conclusions yet, is that fair to say?
    Dr. Dal Pan. No, the agency hasn't reached a conclusion on 
    Mr. Davis of Virginia. Would you say even with your 
setting, looking at both of them, that the findings could be 
due to chance?
    Dr. Dal Pan. I think that is a question more for a 
statistician. I think that from someone who is interested in 
drug safety, I always have to consider that possibility, but I 
have to actually look at what the data are telling me as well 
about the numerical evidence of risk.
    Mr. Davis of Virginia. Your testimony also mentioned that 
FDA is going to convene an advisory committee in the near 
future. When do you plan to convene the panel?
    Dr. von Eschenbach. The advisory committee meeting is now 
scheduled for July 30th. It is the end of July, it has been 
published in the Federal Register.
    Mr. Davis of Virginia. Are they going to look strictly at 
Avandia, or is it going to examine other drugs in its class?
    Dr. Dal Pan. The focus will be on Avandia. But because of 
the nature of the studies, we are going to be looking at other 
oral agents to treat diabetes. They are all involved in the 
same studies.
    Mr. Davis of Virginia. People get very confused when this 
stuff gets out in the media and it gets very unfiltered. Some 
others in the medical community have argued that too many 
warnings on a drug label can lead to as much harm as too few 
warnings, because it leads to the under-use or the under-
prescribing of effective drugs to treat certain conditions. How 
does FDA reach an appropriate balance between caution about 
safety and unnecessary concern?
    Dr. von Eschenbach. Mr. Davis, I think you are making an 
extremely important point that I tried to emphasize in my oral 
statement. Our challenge, first of all, is to take the data 
associated with this particular drug, which is in fact very 
voluminous, very complex and very complicated, come to an 
analysis and an understanding of what has it told us about this 
specific drug as it relates to its complications. Also, what 
has it told us about drugs that may be very similar to it.
    Second, then take that information and put it in the 
context of what should be our appropriate action, what is the 
right thing to do for patients. If we have to in that regard 
weigh the benefit of what would occur if we continued to use 
this drug under certain circumstances and provide information 
to patients and doctors, or if we were to withdraw this drug 
and everything else like it, what would that mean to patients 
who were now deprived of an important therapy to control their 
diabetes, and what would the alternatives be and what were the 
complications of those alternatives, for example, if they had 
to go on insulin.
    So we, the FDA, are not looking at one slice or one piece 
in isolation.
    Mr. Davis of Virginia. You are looking at the big picture.
    Dr. von Eschenbach. We are looking at every piece and 
putting it together into a comprehensive decision of what the 
right thing to do is for patients.
    Mr. Davis of Virginia. Have similar drugs also been subject 
to meta-analysis by either you or anyone else? And if so, what 
have they found?
    Dr. Jenkins. We have requested that the manufacturer of the 
other drug in this class, pioglitazone, which is marketed as 
Actos, perform a similar meta-analysis of their short-term 
studies. Other than that, I am not aware if there have been 
other published meta-analyses for the other drugs. Gerald may 
    Dr. Dal Pan. I am not aware of published meta-analyses for 
diabetes drugs.
    Mr. Davis of Virginia. Could you give me a scientific 
reason why you might have that cause and effect that the Nissen 
report, their meta-analysis brought up? Why the cause and 
effect would be a higher risk of heart attacks?
    Dr. Dal Pan. I am sorry, I don't really understand the 
    Mr. Davis of Virginia. We understand what the meta-analysis 
and the article in the New England Journal of Medicine said. 
Can you give me a scientific reason why you would get that 
conclusion with higher incidence of heart attack, given your 
understanding of the drug?
    Dr. Dal Pan. I think that is what the meta-analysis does, 
it is a technique to bring together smaller trials, which each 
    Mr. Davis of Virginia. Well, it shows the results, but I am 
asking, not the results, I am asking then what is the reason? 
Why does this happen?
    Dr. von Eschenbach. One of the things I think your question 
is pointing out, Mr. Davis, is the need for us to understand 
more about the mechanisms of these drugs.
    Mr. Davis of Virginia. That is what I am trying to get at. 
I am a lawyer.
    Dr. von Eschenbach. And as we know more about the 
mechanisms, as well as observe the effects that they are having 
on patients, then we will be in a much better position to make 
decisions about safety.
    Mr. Davis of Virginia. So you don't know at this point, in 
other words?
    Dr. von Eschenbach. No, in fact, one might suggest it is a 
little paradoxical. You might conclude that the effect on 
microvasculature would be to have improved it, rather than to 
predispose to infarction.
    Mr. Davis of Virginia. I have one last question. In your 
testimony, you say that the FDA approves a drug only after a 
sponsor demonstrates that drug's benefits outweigh its risks 
for a specific population and a specific indication and it 
shows that the drug meets the statutory standard for safety and 
effectiveness. Does the FDA still believe that Avandia 
continues to meet those statutory standards?
    Dr. von Eschenbach. We are in the midst of an analysis as 
we speak, and we have not arrived at a conclusion regarding 
that final decision. Up to this point in time, we clearly have 
believed that it was an important part of the armamentarium. We 
have issued changes in the label to provide appropriate 
warnings, as we had the data to support it. And we will 
continue to do that. And if the data changes or alters after 
our decision after this current analysis that we are in the 
midst of, we will take appropriate action.
    Mr. Davis of Virginia. I guess my question is, it meets the 
standards until you conclude otherwise, basically?
    Dr. von Eschenbach. Correct.
    Chairman Waxman. Thank you, Mr. Davis.
    Mr. Davis.
    Mr. Davis of Illinois. Thank you very much, Mr. Chairman. 
Dr. von Eschenbach, it is good to see you again. I want to 
thank you for being here and thank you for your testimony.
    On May 21st, the Food and Drug Administration issued a 
safety alert on Avandia. Could you tell us, as close to 
possible, exactly what that means?
    Dr. von Eschenbach. I am going to let Dr. Jenkins and Dr. 
Dal Pan speak specifically to that.
    Dr. Jenkins. Mr. Davis, the intent of the announcement from 
the FDA was to communicate to physicians and patients and other 
health care providers about the status of the information, so 
they could be aware of the findings from the meta-analysis, 
aware of other data that FDA was reviewing from other trials 
that we have talked about a bit already this morning, as well 
as to give advice to physicians and patients about how we felt 
they should respond to this new information.
    We particularly wanted to make sure that patients got the 
message that they should not stop taking the drug 
precipitously. If they had concerns, they should speak with 
their doctor. Because going off of a drug for diabetes without 
careful attention can lead to your diabetes being out of 
control, which has its own health risks.
    Mr. Davis of Illinois. The Food and Drug Administration, of 
course, knew prior to this article and prior to the issuance of 
this information that there were potential side effects for the 
use of the drug, is that correct?
    Dr. Jenkins. Yes.
    Mr. Davis of Illinois. What has the Food and Drug 
Administration done, if anything, to help make the general 
public more aware of these side effects?
    Dr. Jenkins. The primary vehicle by which we communicate 
about the risks and benefits of drugs is through the approved 
labeling for the product. And we have made numerous changes to 
the Avandia labeling over the years since it has been approved 
to reflect emerging information and new information about the 
risks. When we make those changes to the labeling, we share 
those through a system we have with many stakeholder groups and 
public patient groups, professional societies, so that they are 
aware of the changes. They are often communicated to the 
physicians through letters from the company and through the 
promotional materials.
    So those are the primary vehicles that we have utilized for 
    Dr. von Eschenbach. Mr. Davis, also, if you will allow me, 
this is an extremely important issue for the FDA in the future, 
in terms of our continuous improvement of how we communicate 
both to professionals and most importantly, to patients and to 
patients of a diverse population. We are approaching that, 
first of all, to learn more about how to do that even better. 
And we have issued guidances with regard to communicating drug 
safety information.
    We now have put in place a risk communications advisory 
committee to help us learn how to do that. We are paying 
particular attention to the vehicles we use, including our Web 
site, and we are engaged in a major overhaul of the FDA Web 
site and the initial project. And that overhaul is to address 
the part of our Web site that is prepared for consumers, for 
patients, so that they can come to the FDA and get information 
in a form that is understandable and useful to them as they 
need to make informed decisions about their health care, but to 
do that in the context of a relationship with their physician.
    Mr. Davis of Illinois. Are we of the opinion that this 
causes physicians now to know anything that they did not 
already know? If I am a physician and I have studied and I have 
paid close attention to what I prescribe and what I do, would I 
learn anything from this that I didn't already know?
    Dr. von Eschenbach. What we hopefully have done, and even 
going back to April 2006, when we added a warning in the 
labeling of Avandia, is that as doctors are caring for patients 
and they are looking at those patients with diabetes who they 
believe are at greater risk of cardiovascular problems or 
already have an underlying cardiovascular history, that they 
will be able to make much better informed decisions about 
whether this drug or some alternative drug is the most 
appropriate treatment for that specific patient.
    So it arms them with more information and more awareness to 
make patient by patient decisions.
    Mr. Davis of Illinois. I know that my time is about to 
expire, Mr. Chairman. Let me just ask this one question, 
following up on the opening statement of Representative Towns. 
Is there anything that the Food and Drug Administration can do 
to help assure that there is greater diversity in the clinical 
trials that are often used to determine the viability of 
pharmaceutical drugs? We all know that when it comes to African 
Americans and some other population groups, there is a paucity, 
it is very difficult to have data that actually reflects the 
impact on this particular population group.
    Dr. von Eschenbach. Absolutely, Mr. Davis. And we are 
approaching that from a number of perspectives. One, as you are 
well aware from our previous conversations, even our 
relationship with NIH and continuing to find ways to encourage 
participation of minority and under-served populations in 
clinical trials, so that we can learn about that in specific.
    Also, we have been reaching out at the FDA as a part of our 
overarching diversity initiative. I have had meetings with the 
National Medical Association leadership specifically to address 
the issue of how can we get representation, especially from the 
African American community in this situation, in the FDA as 
part of our advisory process, as part of our committee 
structure, so that there is the richness of their 
representation as we go about the process of our regulatory 
    So we are coming at it from both ends of that spectrum, the 
leadership that is required, the involvement at the FDA level, 
and then promoting opportunities at the clinical trials level, 
so that we learn, understand and can serve those populations 
more appropriately.
    Mr. Davis of Illinois. Thank you very much, and thank you, 
Mr. Chairman.
    Chairman Waxman. Thank you, Mr. Davis.
    Mr. Issa.
    Mr. Issa. Thank you, Mr. Chairman.
    Dr. von Eschenbach, I am going to try and summarize what I 
think I heard. You don't know whether or not there are any, in 
this class of drugs or in this particular one drug, if there 
are any side effects that essentially say, we will help you 
with your blood sugar, but we may hurt your heart? That is what 
I heard, particularly from Dr. Dal Pan.
    Dr. von Eschenbach. What we have tried to communicate, Mr. 
Issa, is the fact that we have had signals and indications 
about this drug. As those signals and indications have had the 
adequate scientific data in support of a conclusion, we have 
made that conclusion and taken steps to inform the public and 
physicians about what we have known.
    For example, the warning----
    Mr. Issa. My time is limited. My summary is the one that I 
wanted the question answered on. Basically, you are saying here 
today that, and I used the word anecdotal, and maybe that is 
not perfect, but Dr. Nissen in his upcoming testimony is going 
to say that there were several small and medium size clinical 
trials that are insufficient to answer a scientific question. 
He is going to observe that this group already has a high risk 
of heart disease, and that in fact, his own study, which he 
published, which caused this hearing to be rushed here today 3 
weeks later, is not in fact based on sufficient study to 
reach--it looks like my time is coming and going, Mr. Chairman.
    Dr. von Eschenbach. I apologize. I misunderstood your 
question. You are correct in the sense that we are in the midst 
of making that decision right now. Up to this point in time, we 
have not had sufficient data of a nature that we could rely 
upon to draw that conclusion. But we are assessing that as we 
speak, and we are taking that to an advisory committee at the 
end of July.
    Mr. Issa. Then let me change my line of questioning. If it 
is insufficient and premature for us to be having this hearing 
on this drug and this line of drugs, which I think it is, I 
think this is not settled science, you are certainly not here 
to tell us it is, then let's go through--I don't have a family 
history of diabetes, but I do have a family history of heart 
disease. So I just want to go through real quickly my 
understanding of a little bit of the history of heart disease, 
so that something that is much more settled you can comment on.
    When you were in medical school, or maybe before, they used 
to open somebody's chest and sprinkle talc in there in hopes 
that it would promote growth of arteries and veins and so on. 
And that was the best medical science they had at the time. 
This is not a pharmaceutical, per se, there was no prescription 
there. But that is what they did, because that was the best 
they could do. And looking back, it undoubtedly killed more 
than it saved, because of the risk of opening somebody's chest. 
Is that right? Is that fair to say?
    Dr. von Eschenbach. That is a fair assessment.
    Mr. Issa. OK. And then we went through a long period of 
time of yanking out one vein and putting it into another part 
in hopes that patching in a new one was going to take care of 
it. And we thought we were doing better, but now the studies 
show that in at least some categories of patients, they are 
more likely to die on the table or as a result of it later than 
they are to be saved or get a longer quality of life. And 
having had my father go through that and then die, I am acutely 
aware of it.
    Now, in my own district, it is no longer Guidant 
Pharmaceutical, but Guidant was a major manufacturer of stents. 
So I have had the coated/uncoated stent question going on and 
on and on. And it appears as though you approved, in good 
faith, both coated and uncoated stents and in both cases felt 
they were going to do certain things. And now that the studies 
are in, at least on certain ones, historically, some of them 
simply are not going to do a very good job for a long period of 
time, and you would be better off not having them than having 
them. Isn't that correct?
    Dr. von Eschenbach. Right.
    Mr. Issa. So isn't the pattern and the likely future, based 
on that past, I am just using that anecdotally myself, based on 
that past, you are going to always be in a position in which 
you have to face allowing a drug which shows promise, and then 
in fact recognizing that in the long run, maybe 15, 20 years 
later, the alternative to paralysis by analysis is that you go 
forward with drugs that have promise, as this one does, that 
show in clinical trials it does one thing good.
    And then unfortunately, over a long period of time, you may 
find out, as a matter of fact, about the time it is an obsolete 
drug and there is another one, you may find out that on 
balance, you wouldn't have done it if you knew everything that 
you can only know 10 years later. Isn't that right?
    Dr. von Eschenbach. That is absolutely correct.
    Mr. Issa. OK. So when I am looking at this hearing today, 
because I am a dedicated member of this Committee on Oversight 
and Government Reform, I am seeing two things. One is, from an 
oversight standpoint, we shouldn't be second guessing your 
science, even though I just went through that sort of in the 
case of heart disease, that we have to accept that as long as 
your function--just a moment, Mr. Chairman--as long as your 
functional system is as good as science and minds can be, that 
we have to accept that those risks are going to be part of the 
process, and that 10 years from now, a number of drugs or a 
number of procedures that are common today will no longer be 
common because of what we learned over time.
    Thank you, Mr. Chairman. I yield back.
    Chairman Waxman. Thank you, Mr. Issa. I am sorry the system 
is not working, but we gave you the time.
    Before I recognize the next Member, just to clarify 
something that Members ought to be aware of, Dr. von 
Eschenbach, before a drug is approved, you can demand any test 
from the manufacturer that you think is pertinent to safety and 
effectiveness, isn't that true?
    Dr. von Eschenbach. Correct. Dr. Jenkins may want to 
comment on that.
    Chairman Waxman. Well, it is just yes or no. Do you have 
the power to say, we need more information on this or we need 
more information on that?
    Dr. von Eschenbach. That is true.
    Chairman Waxman. Give us a test on it.
    Dr. Jenkins. The statute says all tests reasonably 
    Mr. Issa. Mr. Chairman, point of privilege. Whose time are 
you speaking on?
    Chairman Waxman. If the gentleman would permit, I just 
think we ought to have this clarification.
    Now, after the drug is approved, can FDA demand that a test 
be done on anything related to efficacy or safety, or do they 
have to negotiate it with the company to get the company to do 
    Dr. Jenkins. Mr. Chairman, there are certain places where 
we do have the authority to require studies after approval. In 
other places the studies are negotiated agreements between us 
and the manufacturer.
    Chairman Waxman. And this particular drug, and I am sure it 
is true of a lot of others, for the approval, there was a 
strong recommendation that the test be done on heart attack 
risks. Could you have demanded such a test be done?
    Dr. Jenkins. At the time of approval, we did in fact have a 
post-marketing commitment for the long-term safety study to 
address the medical concerns.
    Chairman Waxman. What if those commitments aren't kept? 
Could you demand they be kept?
    Dr. Jenkins. Well, we certainly monitor those comments and 
expect them to be kept. They are written commitments to the 
agency and we expect them to be honored. In this case, the 
company did do the study in a timely manner and reported it to 
us earlier this year.
    Dr. von Eschenbach. I think the point that perhaps we 
should emphasize, Mr. Chairman, is that if we by virtue of the 
absence of that data believe that drug should no longer be 
available to patients in terms of our ability to assure and 
protect them and in promoting the public health, we can require 
that drug to be withdrawn.
    Chairman Waxman. Right. Some people call that a very strong 
nuclear option. But that is your option at that point. I did 
want to clarify that issue of the FDA law.
    Mr. Tierney, you are next.
    Mr. Tierney. Thank you, Mr. Chairman. It is exactly the 
line of questioning I wanted to proceed on, Doctors, if I 
could. Your FDA physician, originally, the one who looked at 
the original application, were concerned about adverse effects 
on the heart. As I understand it, he was concerned about bad 
cholesterol increases and increases in weight, and concluded 
that a post-approval study of cardiac effects should be a 
condition of approval. Am I right so far?
    Dr. Jenkins. That is what the medical officer recommended, 
and that is what we implemented with the ADOPT post-marketing 
    Mr. Tierney. Your approval letter stated that?
    Dr. Jenkins. Yes.
    Mr. Tierney. That it wanted a study after approval looking 
at cardiovascular risks?
    Dr. Jenkins. Well, the approval letter said what I said 
earlier. It asked for a 4-year long-term safety and efficacy 
study including looking at cardiovascular and hematologic 
events, the liver events.
    Mr. Tierney. Right. So including the safety and the 
cardiovascular events on that.
    Dr. Jenkins. Yes.
    Mr. Tierney. Now, GlaxoSmithKline in their ADOPT study 
didn't really do that. What they did on the ADOPT study was 
they looked at the control, whether or not it controlled 
elevated blood sugar.
    Dr. Jenkins. The primary endpoint for the ADOPT study was 
an efficacy endpoint comparing how well rosiglitazone compared 
to two other commonly used medications. But they also did 
specifically collect information and submit and analyze 
information about safety of the liver, the heart and other 
aspects, yes.
    Mr. Tierney. People tell us, and I think you will agree, 
that the study was too small, really, to get at heart risk, and 
it also had no independent panel to even look at the heart-
related matters, right?
    Dr. Jenkins. The study was never designed to be a specific 
study for heart attack at the time it was designed in 1999.
    Mr. Tierney. All right. So let me bring you back to your 
FDA physician who had the original application. He was 
concerned about heart attack.
    Dr. Jenkins. He was concerned about various heart effects.
    Mr. Tierney. Including heart attack, right?
    Dr. Jenkins. Including heart attack, but also including 
congestive heart failure.
    Mr. Tierney. So we didn't have in the ADOPT study enough 
information to really give us an answer on heart attacks on 
that. And I guess my question is, with the stakes being so 
high, and if in fact Dr. Nissen is correct in his analysis of 
30 to 40 percent increase in heart attack possible from this, 
we could have a serious health problem here.
    So why didn't we have a clinical test or the data designed 
on a post-marketing study? The FDA as I understand it did not 
insist on the particularity of that, on whether we got the 
heart attacks, but afterwards, you don't have the power to do a 
post-study except in very isolated incidents, if I am correct. 
So Dr. von Eschenbach, do you believe the FDA ought to have the 
authority to require more specific and better post-approval 
    Dr. von Eschenbach. I think the point that Dr. Jenkins was 
making was that the concern at the time was with regard to 
toxicity across a number of organs. With the issue of the 
heart, concerns because of the nature of the drug would be more 
around the idea of heart failure. Those things were included in 
the study.
    Mr. Tierney. I am sorry, you are telling me now that you 
think your FDA, the original doctor was concerned with heart 
failure but not heart attack?
    Dr. von Eschenbach. I think he was concerned about cardiac 
events. But what we know about these drugs would make you think 
that would be more likely heart failure, fluid accumulation and 
edema that could put stress on the heart.
    Mr. Tierney. I guess I am having trouble with that. Because 
the impression that we had clearly from the physician was that 
he was concerned about heart attack, long range, as a result of 
bad cholesterol increase, and the increase in weight. You are 
saying that is not the case, he was just worried about a little 
bit of heart trouble?
    Dr. von Eschenbach. I can't speak specifically to that 
particular individual's concerns. I am raising a general 
concern that in retrospect, now that we have the data that we 
are discussing today, this issue of heart attacks, as in 
different or separate from heart failure, is an important area 
that needs to be explored, and a concern. That is apparent to 
us now. I don't know that it was as obvious to everyone back in 
    Mr. Tierney. Doctor, do you support legislation that would 
give you and your agency the authority to require post-market 
    Dr. von Eschenbach. As I have indicated, Congressman, I 
believe very strongly that we have to be engaged in post-market 
surveillance and pharmaco-vigilance. There is legislation that 
is underway that is addressing those specific issues. I am 
looking forward to working with you on that.
    Mr. Tierney. So it would be, I am trying not to be 
impolite, but it is a very straightforward question. Do you 
support legislation that would give your agency the authority 
to require post-market studies?
    Dr. von Eschenbach. I would look forward to discussing that 
legislation in an effort to get us to a point where we will be 
able to get opportunities to collect appropriate data in the 
appropriate way. And the complexity of that----
    Mr. Tierney. Well, wouldn't the post-market studies, 
wouldn't that do it?
    Dr. von Eschenbach. A post-market study is an extremely 
important tool. The information technologies are extremely 
important tools.
    Mr. Tierney. So if it is an extremely important tool, would 
you not support legislation that would give you that extremely 
important tool?
    Dr. von Eschenbach. I am in support of legislation that 
would give us the resources to be able to have those tools and 
be able to implement them. [Laughter.]
    Mr. Tierney. You know, I am going to take that as a yes, 
because what the hell, why not. [Laughter.]
    I would understand the drug companies running us around the 
rosie like that, but I am not sure I understand your reluctance 
to be direct on that. It is your job to protect public health.
    Dr. von Eschenbach. It is legislation that is currently in 
    Mr. Tierney. I know, I filed it.
    Dr. von Eschenbach. I know, and I am engaged--we are 
engaged in providing technical assistance in that legislation. 
I look forward to continuing to participate in that process.
    Mr. Tierney. So I can look forward to your assistance in 
writing legislation that will give your agency the authority to 
require post-market studies? [Laughter.]
    And I would be happy to sit down and talk about that with 
    Dr. von Eschenbach. I will look forward to that, sir.
    Mr. Tierney. Good. Thank you. Thank you, Mr. Chairman.
    Chairman Waxman. Thank you. Your time is up, even though 
the light is still green.
    Ms. Foxx. Thank you, Mr. Chairman.
    I have a fairly brief comment and my colleague may want to 
use the remainder of my time.
    Commissioner, your written testimony states that while 
meta-analyses are often informative, they have important 
limitations. And FDA has been historically cautious in the use 
of meta-analyses in support of regulatory decisions. To your 
knowledge, has the FDA ever acted solely on the basis of a 
    Dr. von Eschenbach. Congresswoman, I am going to ask the 
two experts on either side. In terms of ever having acted on 
it, I quite frankly cannot answer that factually right now.
    Dr. Jenkins. Yes, I can provide some insight to that. We 
are very cautious about the use of meta-analysis to demonstrate 
the efficacy of a drug. So I am not aware that we have ever 
used a meta-analysis to form the basis of showing a drug is 
    We do consider pooled analyses of studies or meta-analyses, 
as they are sometimes called, when we are looking at safety 
data. In fact, that is one of the primary ways we look at 
safety data in an application, is we pool it all together. 
Because any one study is usually not adequate to provide us 
with the information.
    We did recently make a regulatory decision about a drug 
called Zelnorm that was primarily based on a safety signal that 
was derived from a pooled analysis of their clinical trials, 
where the evidence of the risk of a heart effect was very 
large, and we thought it was so convincing that it was 
actionable to recommend that drug come off the market.
    Ms. Foxx. Thank you, Mr. Chairman. I yield back the 
remainder of my time to my colleague, Mr. McHenry, if I may, 
Mr. Chairman.
    Mr. McHenry. Thank you, Mr. Chairman, and I thank my 
colleague from North Carolina.
    There was a stakeholder meeting on May 29th, regarding the 
safety alert on Avandia. Who participated in that meeting and 
what was the outcome?
    Dr. Jenkins. Dr. von Eschenbach participated in that 
meeting, I participated in that meeting, several others from 
the center, including the center director. We invited, I think 
over 40 stakeholder organizations, professional societies, 
patient groups, etc. I think approximately somewhere in the 
teens were the number of groups that were actually represented. 
Some were in the room with us, some were on the phone.
    Mr. McHenry. What was the outcome?
    Dr. Jenkins. We had a discussion to help them understand 
where we were in our analysis of the data, the scope of the 
large number of trials that we were evaluating to try to come 
to our decision about Avandia. They expressed their interest in 
assisting us in better communicating this information to 
patients in particular parts of society that may not get access 
to the information through the usual pathway.
    So it was a discussion and an information sharing meeting, 
not an action meeting per se.
    Dr. von Eschenbach. And if I may, Congressman, just from 
the perspective of the Commissioner, I believe very strongly in 
the need for FDA to be open, transparent and proactive in our 
communications. One of the things we wanted to accomplish in 
this meeting was to address with stakeholders, especially 
patient groups, the FDA's ongoing investment commitment and 
involvement in coming to a scientific conclusion and answer, 
and then whatever action that deemed appropriate.
    In the meantime, to also have them understand that 
communicating, to prematurely and abruptly stop this 
medication, where patients might choose to do that on their 
own, could lead to other serious problems if their diabetes was 
uncontrolled, and to always re-emphasize the need for these 
decisions to be made in a doctor-patient relationship. It was 
an important part of our communication strategy.
    Mr. McHenry. And a final question to you, Dr. von 
Eschenbach. What do you think the implications are of elevating 
a safety review office within FDA? What do you think those 
implications are? And could that possibly offset the balance of 
benefits to patients and life-saving medications?
    Dr. von Eschenbach. I think we need, as you see from the 
two gentlemen on either side of me, the diversity of focus 
within the FDA that looks at these issues from different 
perspectives, but does it in an integrated and coordinated way. 
And more and more, science is moving us in the direction that 
information data, scientific data is telling us both about the 
effectiveness of a drug and the safety or adverse events 
associated with that drug simultaneously.
    Mr. McHenry. So rather than stovepiping it, it would be 
    Dr. von Eschenbach. It would be, in my opinion, moving into 
the modern era, that would be more destructive than 
constructive to what we want as an ultimate outcome. I look for 
greater integration rather than separation.
    Chairman Waxman. The gentlelady's time has expired. Mr. 
Tierney, you are recognized next. Not Mr. Tierney, Mr. Lynch.
    Mr. Lynch. Thank you, Mr. Chairman.
    I want to thank the witnesses for coming before this 
committee and helping us with our work. I would like to ask 
about the warning labels connected with Avandia. Dr. von 
Eschenbach, in your written testimony you said that in April 
2006, the labeling for Avandia was updated to include new data 
in the warnings section about potential increases in heart 
attacks and heart-related chest pain in some patients. You also 
told USA Today with regard to the risk for heart attacks that 
``About a year ago, we began warning the public about possible 
risks in Avandia's labeling.''
    Again, Dr. von Eschenbach, perhaps you can assist the 
committee right now. There is a Physicians' Desk Reference 
being provided to you, which as you know contains all the 
updated labels for prescription drugs. A new version of the 
3,500 page book is printed each year. We have actually flagged 
the section for Avandia for your convenience.
    Now, can you tell me and can you tell the committee where 
the risk for heart attack warning is in the text of the label? 
Because I read it, and I actually had a couple of physicians 
read it and they couldn't tell me either. I remember the 
earlier statement you had about the warnings of heart attacks 
and chest pain. If you could just tell me in the text there, I 
couldn't find it.
    Dr. von Eschenbach. Congressman, we are looking at that as 
you are questioning us. But I would in the meantime emphasize 
the point you are making. As a physician, I recognize the 
inadequacy of the portrayal of this kind of information. And in 
fact, earlier this year, the Food and Drug Administration 
initiated a revision of the label in terms of our ability to 
provide the meaningful, important information that a physician 
and patient needs to get to immediately at the front end of 
this process, so that it would be easily available to any 
physician who had to find it.
    At the same time, we are moving toward an electronic label 
that would not depend upon the publication of desk references, 
but would be immediately available in real time electronically, 
so that when we make a change, it isn't a delay in another 
publication of a hard copy, but something that would be 
available in real time.
    Mr. Lynch. Have you found it, Doctor? Because even after I 
read through it and read the applicable text, I couldn't define 
    Dr. von Eschenbach. I draw your attention to page 1,387 and 
1,388. There is a section, warnings, cardiac failure and other 
cardiac events.
    Mr. Lynch. OK, can you just read the language that is 
supposed to warn me about a heart attack? That is what I am 
interested in.
    Dr. von Eschenbach. Placebo v. Avandia ischemic adverse 
effects, myocardial infarction, 2 percent with regard to 
placebo, 5 percent with regard to Avandia.
    Mr. Lynch. Is that in the table or is that--where is that?
    Dr. von Eschenbach. It is in the table in this drug label.
    Mr. Lynch. That is it?
    Dr. von Eschenbach. There is a whole section on cardiac 
failure and cardiac events.
    Mr. Lynch. That study of that table is for a couple of 
hundred people, 2 non-Avandia and 5 in Avandia. I mean, you are 
not seriously telling me that is it?
    Dr. von Eschenbach. Actually, the power--well, the point 
    Mr. Lynch. Doctor----
    Dr. von Eschenbach [continuing]. At page 1387 there is a 
long section on contraindications and warnings, cardiac failure 
and cardiac events. I drew your attention specifically to the 
    Mr. Lynch. Cardiac events is not heart attack, though. 
Congestive heart failure is something gradual, over time. I am 
asking you where the--I understand infarction, that comes in 
under, it is in four point type, it is one line in a table. You 
are not seriously suggesting that is the warning?
    Dr. von Eschenbach. I am going to ask Dr. Jenkins to 
describe, perhaps better than I am able to do right now to you, 
Congressman, about this information.
    Dr. Jenkins. This language was added in April 2006. It 
specifically refers to a study that was done in patients with 
pre-existing congestive heart failure to look primarily at the 
function of the heart, how well did the heart function----
    Mr. Lynch. Was it----
    Dr. Jenkins. Let me please finish. As an outcome of that 
study, when we reviewed it, we noticed that there was an 
imbalance in the events for heart attack and heart-related 
chest pain, but they were not conclusive, because as you 
pointed out, the study was small. So we put the study in the 
labeling as a warning. And it says, ``Although in treatment a 
difference in change from baseline of ejection fractions was 
observed, more cardiovascular adverse events were observed with 
Avandia treatment compared to placebo during the 52 week study. 
See Table 7.'' Table 7 is the table that Dr. von Eschenbach 
just pointed to where it shows ischemic adverse events, 
myocardial infarction----
    Mr. Lynch. My time is limited. You are repeating what the 
doctor already said.
    Look, all I am saying is that, you cannot be serious about 
locating the warning in a label referred to, four point type, 
it is this small, in an adjacent table to the warning. And the 
warning, the study that you selected, you have thousands and 
thousands and thousands of people who have gone through these 
various studies. You select a very small portion of them and 
you are warning people who have been in on insulin or who have 
had heart failure.
    What about the millions of other people who are diabetic 
and have not been on insulin and who have not experienced heart 
failure, congestive heart failure? What about all those folks?
    I read the label, the warning, and it talks about just 
those two groups. Then it refers to another, very obscure 
reference in a table. I mean, this is really absurd. This is 
ridiculous, what you are saying is a warning. If I wanted to 
hide something, I would do this.
    Chairman Waxman. Mr. Lynch, your time has expired.
    Mr. Lynch. Thank you, Mr. Chairman.
    Dr. von Eschenbach. Mr. Chairman?
    Chairman Waxman. Yes, Dr. von Eschenbach.
    Dr. von Eschenbach. I fully appreciate the concerns and the 
criticisms of what we have used for decades in the practice of 
medicine, the Physicians' Desk Reference. But the type size 
with regard to this warning is absolutely no different than the 
type size in any of the other drugs on the other 3,500 pages in 
this book. It is not an intent to sequester or hide. It is just 
the vehicle that we have to work with.
    Chairman Waxman. Thank you. Mr. Cannon.
    Mr. Cannon. Thank you, Mr. Chairman. We had a lot of 
pictures clicking there, but I am not sure the record is going 
to reflect the size of the book that you were just holding up, 
Dr. von Eschenbach. That is the kind of thing you could have 
stood on the parapet of a castle and thrown on the attacking 
enemy and crushed their heads, it is so big. [Laughter.]
    This questioning, I think, really reflects the underlying 
problem of the complexity of how we deal with drugs that deal 
with the human body in complex ways and how we identify what 
the issues are and therefore, deal with them through the FDA. I 
appreciate the chairman's holding this hearing.
    We had earlier some discussions among Members about the 
role of the New England Journal of Medicine. I think one of the 
points that was missed there is that the New England Journal of 
Medicine, this enormously important journal, has an editorial 
position that they would like to see the FDA change the nature 
of the way we do business in America. That is acceptable. That 
is a great debate.
    My concern is the sensationalization of the process that 
scares people when we have a problem with drugs. Virtually all 
drugs are going to be helpful, but they will also have sidebar 
problems. Now, Dr. von Eschenbach, you and I have spoken 
personally on these issues. You know that I am committed to 
change and improvement in the FDA. We have also spoken in 
public hearings and said pretty much the same thing. And we 
recognized opportunities, but I am concerned about how do we go 
from here to there. In other words, I think doing basion 
studies instead of double blind studies is an important step 
that we need to take. But we have to do it in the context of 
procedures that work.
    Here, what we have is some alarmism that is extraordinarily 
important to many people who are suffering from a disease that 
is difficult and for whom this drug is helpful.
    Taken together, these results, although based on very small 
numbers of events, certainly raise a signal of concern. Now, 
signal is, I think, a term of art in the system here, which 
means, we ought to look at it. There is something that we ought 
to be looking at. So it raises a signal of concern and 
indicates the need for more reliable information about--I can't 
say this name, I will call it the drug at hand, rosiglitazone. 
Pardon me. [Laughter.]
    It is not the one we use when we are asking the pharmacist 
about it.
    But the FDA physicians and patients can reasonably weight 
the results of record, a phase 3 trial designed specifically to 
study cardiovascular outcomes. Until the results of record are 
in, it would be premature to over-interpret a meta-analysis 
that the authors and the New England Journal of Medicine 
editorialists all acknowledge contains important weaknesses. To 
avoid unnecessary panic among patients, a calmer and more 
considered approach to the safety of Avandia is--that is not 
what they say here, but I will call it Avandia--is needed. 
Alarmist headlines and confident declarations help nobody.
    This is not a matter of confidence. This is a matter of 
what happens to people when they take this drug. Now, the 
problem here is what I think are called surrogate endpoints, 
like controlling blood sugar levels with Avandia and other 
drugs. It takes 10 to 15 years to discover and develop a new 
medicine. Without such endpoints for evaluating a diabetes 
medicine, for example, what would the development and approval 
process, wouldn't it take much longer? And how much longer 
would it take, if it does? Do you agree with the value of using 
surrogate endpoints?
    Dr. von Eschenbach. Yes, sir, I do. And I also echo your 
important point about the need for continuous improvement. We 
are seeing revolutions in science and technology around us that 
are going to enable FDA to continuously improve, including how 
we use clinical trials, new clinical trial type designs that 
will be much more informative. We will also be using many more 
tools of science and biomarkers and genomics etc. that is going 
to help us with regard to the ability to use these biomarkers 
and these intermediate endpoints.
    Mr. Cannon. I see my time is about to expire. But let me 
just ask about this study in particular. The meta-analysis by 
Dr. Nissen excluded studies in which there were no adverse 
events. From a layman's point of view, of not including studies 
where there were no heart attacks or other heart problems, that 
would seem to skew the results a little. But more specifically 
with respect to heart attacks, I understand that six studies 
were not used, because none of the patients had a heart attack. 
Even more studies, approximately half of the overall available 
were not used, because there were no deaths. Yet headlines 
screamed about a 43 percent increased chance of death.
    Is that a responsible way to communicate to the public?
    Dr. von Eschenbach. We value all data and all input with 
regard to these issues. This study, like other meta-analyses, 
has both strengths and weaknesses that have been discussed and 
pointed out by others. And we use it as an additional piece of 
information, but not necessarily one upon which decisions in 
and by themselves would be made.
    I will let Dr. Dal Pan speak specifically to how we use 
data and meta-analyses.
    Mr. Cannon. Mr. Chairman, I see my time has expired. But 
the question I asked is, is it responsible to use this meta-
data to create what is essentially a public panic?
    Dr. von Eschenbach. I believe that the data was being 
presented in the Journal as in a contribution and an additional 
piece of information. We have all done that in our careers in 
terms of publishing information and data that we believe was a 
valuable contribution. We leave it then to the entire 
scientific domain to weigh that, add that, evaluate that in the 
larger context. I believe that is what was hopefully going to 
occur here.
    Other people reacted, perhaps responded to that information 
and perhaps created some of the concerns that you are alluding 
    Mr. Cannon. If the Chair would indulge just one followup, 
there is something different from publishing and awaiting a 
reaction and publishing and promoting. Would that be different 
in your mind?
    Dr. von Eschenbach. I can't speak to the author's intent. I 
have not had any conversations with Dr. Nissen.
    Mr. Cannon. Mr. Chairman, I see my time has expired and I 
yield back.
    Chairman Waxman. The gentleman's time has expired.
    Now I would recognize Mr. Yarmuth.
    Mr. Yarmuth. Thank you, Mr. Chairman, and I thank Dr. von 
    I have a question that relates to the scope of the risk 
that we are talking about. I think any of us who have watched 
television commercials and have taken medications and see these 
percentages have a hard time getting our arms around it. Your 
staff, when they briefed the committee on this particular 
situation, indicated that if these numbers are real, this is a 
big deal. I think that was one of the direct quotes. And you 
said, these data, if confirmed, would be of significant concern 
because patients with diabetes are already at an increased risk 
of heart disease.
    I want to understand this study. The GSK data that was 
presented in August 2006 basically said, and I think you 
confirmed this, that those numbers indicate that the risk went 
from approximately 1.5 percent to approximately 2 percent, 
which was approximately a third increase in the risk.
    But that body of data, 13,000 or so cases, included a lot 
of different combinations of regimens that were being used. As 
I understand it, some were taking Avandia by itself, some with 
insulin, some with nothing else. So in fact, am I not correct 
in saying that for some patients, presumably the conclusion 
would be that the risk is much higher than the 2 percent, but 
we don't know, because we didn't have a breakout of those 
    Dr. von Eschenbach. There are confidence, what we call 
confidence intervals around that number, which means there 
could be a range of lower and slightly higher risk. I will let 
Dr. Dal Pan speak specifically to those statistical 
considerations as we are trying to make these decisions.
    Dr. Dal Pan. I think what you are asking, Congressman, is, 
are there patients or combinations of medications that can 
confer higher risk and could there be some situations where the 
risk is lower. That is the kind of thing our statistical 
analysis is focusing on. We are trying to answer those 
questions and put the answers to those questions into the 
larger context to make our decision.
    Mr. Yarmuth. So you don't know that yet, and you are trying 
to break it down?
    Dr. Dal Pan. Right. Our statistician has finished her 
review, I haven't finished looking at it extensively. But this 
is the kind of thing that we are actively engaged in now, yes.
    Mr. Yarmuth. But presumably in this case, say a patient who 
was taking Avandia and insulin, might have a risk of 5 percent 
of a heart attack as opposed to 2 percent or 1 percent?
    Dr. Dal Pan. Right. There are risks that could be higher 
than the overall summary risks for certain patients.
    Mr. Yarmuth. And of course, what we are dealing with is a 
situation in which if a million people are taking a particular 
medication, a 0.5 percent increase in risk amounts to 5,000 
people who are adversely affected who otherwise wouldn't be. So 
it does become a significant risk.
    Now, at what point would you consider that risk to be of 
significant peril that some dramatic action needed to be taken, 
whether it was the nuclear option or advising doctors to 
immediately take patients off the medication?
    Dr. von Eschenbach. Well, you are pointing out, 
Congressman, an extremely important part of what FDA's role is 
in this whole process. First of all, it is to absolutely, 
critically, vigorously assess the scientific data. Do patient 
individual analyses, for example, the kinds of things you were 
alluding to. But then put that into a larger context. That 
brings into play what is the implication of that risk as it 
relates to the total population of patients with diabetes who 
might be affected.
    Are there other alternatives that would be available to 
them that would get a benefit and perhaps at less risk? Or if 
there is no other option available, what risk do we deem is 
appropriate and under what circumstances? Can we advise doctors 
and patients to be more selective about who should, who should 
not get that particular treatment. That becomes an important 
part of our overall decisionmaking process to that end of both 
protect and promote the public health.
    Mr. Yarmuth. And I am concerned because as we watch 
television commercials and we talk about warnings, at a certain 
point the public becomes numb to these things, because they 
really don't mean anything. But if you told me that if I went 
to the grocery in my car and I had a 2 percent risk of being in 
an accident, I might still take the chance. If I had a 10 
percent risk of it, I might not drive my car to the grocery.
    I am concerned that what information that FDA provides to 
the public and what we do here as well gives the public 
adequate explanation of the risks they are taking. Because for 
those 5,000 people presumably it was a 100 percent risk.
    Dr. von Eschenbach. Right. And to your point, we are 
attempting to do that even better than we have done it, as I 
indicated to you, the initiatives that we have with regard to 
risk communication, the vehicles that we use. But your point is 
extremely well taken. There are issues in which our decision 
will always be based on the standards of rigorous, scientific 
analysis, whether it is a drug for hay fever or whether it is a 
drug for diabetes or for cancer.
    However, from the patient's perspective, the risk-benefit 
ratio is dramatically different, whether you are thinking about 
taking a drug for sniffles or whether you are taking a drug for 
terminal cancer for which there is no other option available to 
you. And that is an important part of this equation that we 
can't lose sight of.
    Chairman Waxman. Thank you, Mr. Yarmuth.
    Mr. Yarmuth. Thank you, Mr. Chairman.
    Chairman Waxman. Mr. Hodes.
    Mr. McHenry. Excuse me, Mr. Chairman? I have not been 
    Chairman Waxman. I didn't see you. You are recognized for 
your time.
    Mr. McHenry. I appreciate it. At this time I would like to 
yield my time to my colleague from California, Mr. Issa.
    Mr. Issa. I thank you, Mr. McHenry. I just want to followup 
on two more things. I know you are going to be leaving shortly. 
Mr. Cannon's question, it sort of prompted my wanting to delve 
a little further.
    If you have the study, the study at hand, the study that 
led to today's hearing, if you have a study taking out, and 
maybe this is a statistical question, but it doesn't seem like 
a complex one, taking out those in which nobody died of heart 
attack, in which nobody got a heart attack, if you take those 
out, by definition, you put them back in and the 43 percent 
becomes lower. We may not know how much lower, but 
significantly lower, isn't that correct, inevitably?
    Dr. Dal Pan. Let me say, none of the three of us here is an 
expert on the statistics methods of----
    Mr. Issa. No, no, no, wait a second.
    Dr. Dal Pan. But there are statistical issues----
    Mr. Issa. But let's--I only took 2 years in statistics in 
college. It doesn't make me a statistician, but I know that if 
you leave the zeroes out of a zero through 10 and you are 
averaging, you are going to get a lower amount if you put the 
zeroes in, isn't that right?
    Dr. Dal Pan. One of the things our statistician is doing is 
to see if there are techniques that she could use to actually 
address that issue. I can say conclusively that it would make 
that risk go away, though.
    Mr. Issa. OK. Do you know of any reason, though, for 
leaving out those who did not suffer? I mean, other than 
promoting panic, other than getting people to think that this 
drug had a higher incidence of heart attack, is there any 
reason to leave out other groups who took the drug and didn't 
have heart attacks? Is there any valid reason that you can 
think of, without knowing anything more than what we have heard 
    Dr. Dal Pan. I think it is the statistical issue. But then 
the issue then becomes looking at all the available data to put 
it together. But I think all these techniques have their 
statistical basis. And those statistical bases have to be 
respected to do the study.
    Mr. Issa. Well, maybe I will go back to what we did a 
couple of weeks ago. We did global warming. I happen to believe 
in global warming, I have been a promotor of reducing 
CO2 emissions. But I am trying to understand, if I 
only took the days of the year that were cooler and I left out 
the days that were hotter, I could prove the earth is cooling, 
not heating. So I am a little shocked that you are not more 
concerned that a study published not for peer review but in 
fact published for the public and widely reported on and linked 
to this hearing today deliberately ignored those other patients 
who could have brought the number more to zero.
    Dr. von Eschenbach. Mr. Issa, I cannot comment on why and 
how this particular study was done and designed and developed. 
That is something for the author to comment on. But your point 
is extremely well taken, that with regard to a meta-analysis, 
it is well recognized that they are fraught with problems, 
statistical problems, in terms of how you do them. And in this 
case, whether you did fixed events or random events, in terms 
of how you analyze the information and data.
    And that points out, whether it is this meta-analysis or 
any other meta-analysis, the problem and concern about making 
definitive, explicit decisions with regard to just a meta-
analysis. You have to be mindful of the dangers that could 
involve. And that is why the FDA chose to go much further since 
we had individual patient data, which the author was not 
available to him. And we have expanded and used our expertise 
of our biostatisticians to take this to an appropriate level, 
which we are in the midst of doing right now.
    Mr. Issa. OK. I am going to yield back to the gentleman. I 
just want to make sure something gets in the record, though.
    The American Enterprise Institute published something that 
I think says a lot about the author that we are going to hear 
from in a few minutes. The study's primary author, Cleveland 
Clinic cardiologist, Steven Nissen, admitted to the Wall Street 
Journal that he was in touch with Congress while preparing his 
analysis. Three days after the study was submitted to the New 
England Journal of Medicine and before it was published, the 
FDA Commissioner received a letter about Avandia from members 
of the House Energy and Commerce Committee that seemed to 
reference the New England Journal of Medicine study. I just 
want to make sure that is in the record, and I will yield back 
to the gentleman.
    Mr. McHenry. I thank my friend from California.
    Let me just ask a broader question, I would like you to 
touch on this. I know your struggles at the FDA to make sure 
that we have safe drugs on the market, there is a proper 
balance between patient safety and life-saving medicine. It is 
an ongoing struggle.
    Do you think our regulatory hurdles are too high or just 
about right, or too low? There is a lot of debate going on 
right now and I know the chairman is very interested in this 
issue and actually wants to increase the regulatory hurdles to 
get drugs on the market. I would like you all, all three of 
you, to comment upon this, on whether or not that is 
appropriate or our regulatory level to get a drug on the 
market, is about right or too high?
    Dr. von Eschenbach. Congressman, I believe that the 
regulatory levels are appropriate for the individual 
circumstances in which the regulatory barrier has to be 
extraordinarily high with regard to this risk and benefit 
ratio. I have alluded to that, the reasons why that might be 
the case whether you are dealing with hay fever or whether you 
are dealing with cancer.
    So I think they have to be applicable to the individual 
situation and circumstance. I think it is important to point 
out, as I did in my oral testimony, that the world around us is 
radically changing, rapidly changing. Science and technology, 
the complexity of the products, the circumstances. We need, at 
the FDA, to continue to adapt and respond to those changes. The 
resources that we are looking forward to are designed to 
specifically enable us to do that and continuously improve.
    So I think it is an issue of using the regulatory framework 
but continuously improving it and improving our ability to 
apply it. I think the standards are appropriate.
    Chairman Waxman. The gentleman's time has expired.
    Would Dr. Jenkins or Dr. Dal Pan like to respond to the 
question, or do you agree with Dr. von Eschenbach?
    Dr. Jenkins. Congressman, I head the Office of New Drugs 
that makes these decisions every day. So my staff and I make 
these decisions every day. It is always a weighing, of 
balancing the certainty you know about the drug versus the 
uncertainty of things you don't know about the drug. I think we 
strike that balance very well and within the framework of the 
regulations and the statute that have been given to us by 
Congress to operate in. So I do think we have struck the right 
    This is clearly a societal, public policy question as far 
as how much certainty do you need to know about a drug before 
you approve it, how much uncertainty are you willing to accept 
at the time of approval. You can never know everything about a 
drug at time of approval. I think it is a public policy debate 
about where that standard should be set. I think we adhere to 
the standard that has been set for us by Congress in the 
    Chairman Waxman. Dr. Dal Pan.
    Dr. Dal Pan. Let me just add on to what Dr. Jenkins has 
stated. There always is this residual uncertainty at a time 
when a drug is approved. I think for that reason, as Dr. von 
Eschenbach said, it is important to have a strong post-
marketing system as well, to be able to monitor that 
uncertainty and come up with better understanding of the drug's 
risks as time goes on.
    Chairman Waxman. Thank you.
    Mr. Hodes.
    Mr. Hodes. Thank you, Mr. Chairman.
    Gentlemen, thank you for your testimony. Much of the focus 
of this hearing has been on post-market surveillance, what does 
the FDA do after a drug is approved. I would like to direct 
your attention to a slightly different question. I am 
specifically concerned with what the FDA does to ensure the 
accuracy of the pharmaceutical direct to consumer drug ads 
after the company's drug has gone to market.
    I note in Dr. von Eschenbach's written testimony the 
statement ``In April 2006, the labeling for Avandia was updated 
to include new data in the warning section about a potential 
increase in heart attacks.'' That was the language you used, 
Dr. von Eschenbach.
    There was questioning by my colleague Mr. Lynch about 
warnings. Now, yesterday, in both the New York Times and the 
Washington Post, GSK, the maker of the drug, took out full-page 
advertisements about Avandia. In fact, a page and a half in the 
New York Times, I have it here. I think you have it in front of 
you. There is a full page which has something on top, and then 
they have important safety information on the bottom. And then 
in another half page, there is the patient information.
    Now, I am concerned about the gap we seem to have between 
concern about heart attacks and warnings about heart failure. 
Because if you are a consumer, plain ordinary guy like me, a 
heart attack means something very different than heart failure, 
which happens to be, could be the inability of the heart to 
pump blood, could be a long-term thing. Heart attack is a 
rather sudden and specific event.
    Now, despite that you say there were label warnings for 
heart attacks, if I read the language in both the New York 
Times and the Washington Post, what I see is a warning that 
says if you have heart problems or heart failure, tell your 
doctor. Avandia can cause your body to keep extra fluid, which 
leads to swelling and weight gain. Well, that is a problem. 
Extra body fluid can make some heart problems worse or lead to 
heart failure. The word heart attack, which is what consumers 
understand, does not appear.
    Now, GSK has spent $42 million on advertisements to 
consumers for Avandia. Its revenue has increased 25 percent in 
recent years. If I am right, and if this doesn't contain the 
concerns about heart attacks, do you believe that consumers 
understand this warning by GSK to be a warning that there is an 
increased risk of heart attacks from Avandia?
    Dr. von Eschenbach. No, sir, I do not believe that looking 
at an ad like this in a newspaper really helps to provide the 
kind of depth and understanding that you just described. I 
think that this does not occur by looking at these kinds of 
    Mr. Hodes. So this ad doesn't use the word heart attacks, 
does it?
    Dr. von Eschenbach. I haven't read the complete ad, sir, 
but I will take your word that it does not.
    Mr. Hodes. Because I am happy to represent to you with 
absolute assurance that it doesn't use the word heart attacks.
    Dr. von Eschenbach. I will accept that.
    Mr. Hodes. Now, in that light, if there is concern as we 
now know about the increased risk of heart attacks, and that is 
what you talked about in your testimony, that is what has now 
come out. And yesterday, this company is still not warning 
consumers about the increased risk of heart attacks.
    My question to you, as the regulatory agency, is do you 
have enough power now to do something about the manufacturers 
and what they are doing with post-consumer advertising? Do you 
need more power? Do you need different power? What needs to be 
done for you to adequately regulate how the manufacturers are 
communicating in simple, plain terms that consumers will 
    Dr. von Eschenbach. As part of the negotiations and 
discussions with regard to PDUFA IV reauthorization, which is 
currently in place, we have sought the resources to be able to 
expand our ability to review, survey and therefore take action 
against direct to consumer advertising.
    Mr. Hodes. Sir, with great respect, this reminds me of your 
answer to my colleague Mr. Tierney's question, when he asked 
you a direct question, you said, we are looking for more 
resources. Now, to me, resources means maybe people, maybe it 
means money. By resources, do you mean some more regulatory 
power that you currently do not have to interface with the drug 
manufacturers to make sure that they are doing what they need 
to do to tell consumers about the risks you are flagging?
    Dr. von Eschenbach. I believe right now the most serious 
concern for me is having adequate numbers of people to be able 
to monitor and take action against direct to consumer 
advertising when it is inappropriate. That for me is a major 
area that needs to be addressed.
    The ability to then affect that, if that becomes a problem 
that requires legislation, is something that, as I indicated, I 
think we need to address. But I am not prepared at the present 
time to say that is absolutely the answer that I need in order 
to fix the concern or problem that is being raised.
    Mr. Hodes. I am not sure I understand you. If I may just 
followup briefly with one question. Are you telling me you 
don't have enough people to read this ad and see whether or not 
the ad adequately, in your expert opinion, warns the consumer 
of the increased risk of heart attack? Are you telling me you 
don't have enough people to do that?
    Dr. von Eschenbach. Yes, sir. I am telling you that I need 
more resources to be able to direct to the issue of the FDA's 
oversight of direct to consumer advertising.
    Chairman Waxman. The gentleman's time has expired.
    Mr. Hodes. May I just have one last question, Mr. Chairman? 
Thank you.
    You need more people to read the ad. Fine. Do you have the 
power that you need to say to the drug manufacturer, fix the 
    Dr. von Eschenbach. I believe at the present time I do have 
the ability to get that accomplished and get that done. I would 
certainly, if that is not adequate, after we have done our 
appropriate intervention, I would then welcome any legislative 
action that would require that to be a fix. But at this point 
in time, I don't believe that is at the core of the problem for 
    Mr. Hodes. Thank you very much. Thank you, Mr. Chairman.
    Chairman Waxman. Thank you, Mr. Hodes.
    Ms. Watson.
    Ms. Watson. Thank you so much, and I thank the panelists 
for indulging us.
    I too have the same concern. I myself have diabetes 2. I 
had a complete health examination before I took my post as 
Ambassador, no problems. Now I develop diabetes 2 after 2 
years. All of a sudden, I had a heart murmur, a heart problem. 
I went to my cardiologist and he examined me, he said, what are 
you taking. Avandia. He said, get off of it. I myself, no 
history in the family. I have a history of diabetes, yes. He 
said, get off of Avandia. There are other options out there.
    Now, here is my concern, listening to the testimony. Why 
has it taken FDA so long to come and say, we need more 
resources? Why did so much time pass after your approval? And 
the post-marketing studies seem to me to be a way to reduce the 
risks that millions of people are under in this country. I 
heard your response to Representative Hodes, I heard your 
response to Mr. Tierney.
    But I didn't hear a plea to give us that authority. You 
ought to have heart attack on the label, because that would 
have been understood. It looked like I was heading toward just 
that when I went to my physician.
    Dr. von Eschenbach. I believe at the core and the heart of 
the question that you have just placed before me, 
Congresswoman, is the issue of the fact that we have attempted 
to provide information that when a doctor is caring for a 
patient such as yourself, and there seems to be a problem or 
concern, that is addressed. And it may require a change in your 
    Ms. Watson. Doctor, let me take back my time because I will 
be out of it in just a second. Would you have anything against 
putting on the label, there is a high risk of heart attack?
    Dr. von Eschenbach. That is precisely what we are engaged 
in determining as we speak. The comprehensive analysis of all 
of the data related to heart attack, both from meta-analyses as 
well as other studies. And the deliberation that will occur at 
the advisory committee at the end of July will lead us to the 
answer to that specific question.
    Ms. Watson. All right. Thank you. The stakes are very high.
    Dr. von Eschenbach. I agree.
    Ms. Watson. And you represent us who give permission for 
these drugs to go on the market, and too many people are at 
    Now, let me shift my questioning. I am an African American. 
And diabetes is spreading higher among African Americans and 
now Hispanic Americans than any other group. But I find there 
are too few of us in the test. So what can you do to be sure 
that Americans of all ethnicity become part of your test?
    Dr. von Eschenbach. I fully support and concur. We are 
approaching this from one, the perspective of working with, for 
example, our sister agency, the National Institutes of Health, 
to be able to promote the participation of more minorities and 
under-served in the clinical trials themselves. Two, we are 
approaching this from the perspective of I am engaging, with 
the National Medical Association and have met with them to lay 
out specific plans to address that issue, to bring 
representation from the African American community specifically 
into the FDA's processes. Participation in committees and the 
ability for us to address in the appropriate way the way in 
which the community believes is most appropriate and effective. 
But to get to the endpoint, we absolutely need to serve 
patients better by having them participate in these clinical 
    Ms. Watson. Thank you for that response. I just want to end 
up by saying, the American Diabetes Association had to be 
forced by a group of us, I represent Los Angeles, to do 
outreach into these communities. So we had to hold our own 
outreach informational sessions, ourselves. So we need a whole 
reform in how we meet and reach Americans of various 
    Thank you, Mr. Chairman.
    Chairman Waxman. Thank you very much, Ms. Watson.
    Dr. Jenkins, Dr. Dal Pan, Dr. von Eschenbach, thank you 
very much for your appearance today and your willingness to 
answer the questions that we had to ask you. We are of course 
interested in the process used to inform the American public 
about the efficacy and safety of these drugs. I think your 
contribution today is helpful to us. We want to of course 
review this situation in the context of legislation that is 
pending in both the House and the Senate.
    Dr. von Eschenbach. Thank you, Mr. Chairman. On behalf of 
my colleagues and the entire FDA, let me thank you and the rest 
of the members of the committee for your consideration and your 
openness to our perspective. Thank you.
    Chairman Waxman. Well, I was a little premature in thanking 
you and expecting that we would move on, because we have 
another distinguished member of our committee who is eager to 
ask questions. So I do want to recognize him. Mr. Cummings.
    Mr. Cummings. Thank you very much, Mr. Chairman.
    Dr. von Eschenbach, I want to ask you about the actions of 
your press office over the past 2 weeks. On May 21st, the New 
England Journal of Medicine published an analysis of clinical 
trial data about Avandia that started a vigorous scientific and 
medical debate that continues today. The analysis provided a 
signal that Avandia may be associated with increased risk of 
heart attack. As you acknowledge in your written testimony, if 
confirmed, this signal ``would be of significant concern, 
because patients with diabetes are already at an increased risk 
of heart disease.''
    You told us in your written testimony how the FDA is 
committed to ``early communication of emerging information 
about the safety of drugs,'' stressing that ``any communication 
must be responsible and measured, taking into account the 
impact that the message will have on patients and practitioners 
alike to encourage good health care choices and help avoid bad 
ones.'' This seems like an appropriate communication strategy.
    What I want to know is why it was not followed in the case 
of Dr. Nissen, the author of the study in the New England 
Journal article.
    Dr. von Eschenbach. I am sorry, Mr. Cummings, could you be 
more specific about----
    Mr. Cummings. On May 24th, just 3 days after the 
publication of Dr. Nissen's analysis, at least two individuals 
in the FDA press office forwarded to reporters in the national 
media and trade press an article from the Web site, heart.org, 
that contains derogatory comments about Dr. Nissen. 
Specifically, the article contained accusations from an 
anonymous commenter to a blog posting in the Wall Street 
Journal that questioned Dr. Nissen's motives in undertaking and 
publishing his analysis, implying that he was only interested 
in hurting companies that did not work with him and the 
Cleveland Clinic.
    The accusations were so baseless that the Web site itself 
later retracted the comments. It said that the accusations ``do 
not meet the highest standards of journalistic or scientific 
integrity or credibility.'' Even worse, one of your press 
consultants, Douglas Aberfell [phonetically], sent out these 
articles with bizarre titles. One e-mail title was ``What are 
St. Steven's feet made of? Clay, perhaps?''
    Another one read, ``Did you ask Nissen if the Pope called 
yet?'' Are you familiar with this? Are you following me so far?
    Dr. von Eschenbach. Yes, sir, I understand the point that 
you are----
    Mr. Cummings. I would like to request that a copy of Mr. 
Aberfell's [phonetically] e-mail be included in the record, Mr. 
    Mr. McHenry. Reserving the right to object.
    Chairman Waxman. The gentleman reserves the right to 
    Mr. McHenry. I have not seen the e-mail. I would love to 
see a copy of the e-mail before I agree that this should be 
entered into the record.
    Chairman Waxman. The gentleman will withhold his unanimous 
consent request and----
    Mr. Cummings. Very well.
    Well, since I will have to work with what I have, do you 
believe that these actions represent responsible and measured 
communication to which your agency is committed?
    Dr. von Eschenbach. No, sir.
    Mr. Cummings. Let me finish. I am almost finished. Is it 
really an appropriate use of Federal, Federal taxpayer dollars 
to use the FDA press office as a vehicle for attacking 
scientists who raise important signals about potential public 
health dangers in prestigious scientific journals?
    Dr. von Eschenbach. Mr. Cummings, this was not an action on 
the part of the FDA or the FDA's press office. This was an 
action of an individual within the FDA. I completely concur 
with you that it is inappropriate and unacceptable. That 
individual's supervisor has taken appropriate action with that 
individual. I would not condone or accept that kind of 
    Mr. Cummings. Is that individual still working with the 
    Dr. von Eschenbach. That individual is still employed by 
the Government. His action was addressed.
    Mr. Cummings. What action was taken?
    Dr. von Eschenbach. This action has been addressed by the 
individual's superior, a letter of reprimand is in his file.
    Mr. Cummings. But we are still paying him?
    Dr. von Eschenbach. It was an inappropriate and unfortunate 
action on the part of an individual, and I believe that is 
being appropriately addressed from a disciplinary point of 
    Mr. Cummings. The medical experts who are appearing before 
this committee this morning have distinguished professional 
careers. They and their institutions should be proud of the 
work they have done. And we as a country should not tolerate 
efforts by either private or public entities that engage in 
intimidation and smear campaigns against experts who act in the 
service of the public.
    Thank you very much.
    Dr. von Eschenbach. Thank you, Mr. Cummings. Let me 
reassure you and other members of the committee, there is 
absolutely no intention nor has there been any action on the 
part of the FDA to take and behave or participate in any kind 
of campaign with regard to Nissen. We have welcomed his 
information and his data as a part of our ongoing assessment 
and analysis. Although I have never had the opportunity to 
discuss things with him personally or directly, I would look 
forward to doing so at any time.
    Chairman Waxman. Thank you, Mr. Cummings. Another Member 
seeks recognition, Mr. Shays.
    Mr. Shays. Thank you, Mr. Chairman. I don't usually seek 
recognition when I have come so late in the panel and I don't 
have a question to ask, but I know that Mr. McHenry would like 
to ask a brief question, so I would yield to him.
    Mr. McHenry. I thank my colleague.
    I would like to followup with you and give you an 
opportunity to respond to this. With complex scientific 
research, it is important that a balanced perspective is given 
on a study that has been released? Is that an important 
    Dr. von Eschenbach. Yes. Yes, it is.
    Mr. McHenry. Now, an additional followup to this. Is it 
necessary for the FDA to perhaps, in order to quell 
overreaction about a release of a study, to provide a balanced 
perspective on that study?
    Dr. von Eschenbach. I believe the FDA must accept 
information and data from a variety of sources, analyze it 
appropriately and then take what we believe to be the 
appropriate action.
    Mr. McHenry. An additional comment here. After the release 
of the study, there have been a number of articles written 
about the failure in the study. Is that something important for 
consumers to be aware of?
    Dr. von Eschenbach. I think it is important for everyone to 
be aware of balance and where there is legitimate scientific 
debate, that should be something that people are aware of. 
There were issues here where, for example, two journals that 
are each highly reputable had differing perspectives and points 
of view with regard to this particular study. I think that is 
an important part of an open and healthy dialog and discussion.
    Mr. McHenry. Thank you. I yield back.
    Mr. Shays. I yield back.
    Chairman Waxman. Thank you very much again. Thank you, 
gentlemen, for your testimony. We appreciate your being here.
    Dr. von Eschenbach. Thank you, sir.
    Chairman Waxman. We are now pleased to call forward for our 
second panel Dr. Steven Nissen, who is the chairman of the 
Department of Cardiovascular Medicine at the Cleveland Clinic, 
one of the Nation's most respected academic medical centers. He 
is the immediate past president of the American College of 
Cardiology. And from 2000 to 2005, Dr. Nissen served as a 
member of the FDA's cardio-renal advisory panel and chaired the 
committee during his final year.
    Dr. Nissen was the lead author of the May 21, 2007, New 
England Journal of Medicine article that drew a connection 
between Avandia and increased cardiac risks.
    We have also Dr. Bruce M. Psaty, who is professor of 
medicine, epidemiology and health services and co-director of 
the cardiovascular health research unit at the University of 
Washington. From 2000 to 2006, he was a member of the Institute 
of Medicine's Committee on the Assessment of the U.S. Drug 
Safety System. Dr. Psaty was the lead author for the May 21st 
editorial in the New England Journal of Medicine, commenting on 
Dr. Nissen's study, and is a lead author of one of the June 5th 
editorials in the same journal commenting on the newly released 
RECORD study.
    And Dr. John Buse is a professor of medicine at the 
University of North Carolina School of Medicine in Chapel Hill, 
NC, where he serves as the chief of the Division of 
Endocrinology. One of our Nation's most highly respected 
experts on diabetes care, Dr. Buse is president-elect of the 
American Diabetes Association. He has received numerous awards 
and honors, including citation in Best Doctors of America every 
year since 2001.
    Dr. Buse was the first physician in the country to raise 
concerns about the cardiovascular safety of Avandia in a letter 
he wrote to the FDA in 2000.
    We welcome the three of you. It is the practice of our 
committee to ask all witnesses to take an oath. I would like 
you to rise.
    [Witnesses sworn.]
    Chairman Waxman. The record will indicate that each of the 
witnesses answered in the affirmative.
    Dr. Nissen, why don't we start with you. We have your full 
statements in the record. We would like to ask you to summarize 
your testimony in around 5 minutes. We have a clock that I hope 
will work appropriately to let you know. Yellow light means 
that 1 minute is left, red light means the time is up. We would 
like to ask you to, when you see the red light, to conclude.
    There is a button on the base of the mic. Be sure it is 
pressed in. We want to hear from you.
    Dr. Nissen.

                      HEALTH, SEATTLE, WA


    Dr. Nissen. Thank you very much, Mr. Waxman.
    My name is Steven E. Nissen, M.D. I am chairman of the 
Department of Cardiovascular Medicine at the Cleveland Clinic, 
and the immediate past president of the American College of 
Cardiology. My testimony does not reflect the views of either 
the Cleveland Clinic or the ACC.
    Before I begin, I want to thank the committee, I want to 
thank the bipartisan efforts of this committee to look into 
issues of drug safety and the FDA. This is an extremely 
important issue. It affects all 300 million Americans, and I 
applaud you for looking into this. I think it is clearly the 
right thing to do.
    I have been asked to summarize for the committee the 
sequence of events and the scientific basis for our manuscript 
describing the potential cardiovascular risks of Avandia. In 
September 2006, a clinical trial called DREAM was published in 
the British medical journal, the Lancet. In the study, patients 
at high risk for developing diabetes were assigned to receive 
either Avandia or an inactive placebo. Avandia did indeed 
reduce the incidence of new onset diabetes.
    However, the DREAM study also showed a numerical excess of 
heart-related adverse events, including 15 heart attacks in the 
Avandia group compared with 9 in the placebo group. The number 
of heart attacks was too few to reach statistical significance, 
but they were trending in the wrong direction. This was 
potentially an important observation, because the reason for 
giving a drug to prevent diabetes is to reduce the 
complications of diabetes, the most serious of which is heart 
    Then in December 2006, a clinical trial known as ADOPT was 
published in the New England Journal of Medicine. This study 
was designed to show whether Avandia had a more durable effect 
at reducing blood sugar than two generic diabetes medications. 
The study indeed showed a more long-lasting reduction in blood 
sugar with Avandia, but heart-related complications were also 
trending in the wrong direction. The heart attack rate was 33 
percent greater in Avandia-treated patients, but again, there 
were too few events to reach statistical significance.
    After reviewing DREAM and ADOPT, I was concerned, because 
these were the only long-term large-scale clinical trials 
comparing Avandia with other therapies. And both studies showed 
an excess of heart attacks. When you have several small or 
medium-size clinical trials that are insufficient to answer a 
scientific question, the logical next approach is to combine 
these trials to try to address the issue. This process is known 
as a meta-analysis.
    Using this method, I asked one of my colleagues, a 
statistician, to combine DREAM and ADOPT. We noted a 40 percent 
excess of heart attacks, which was not statistically 
significant, but showed a strong trend in the wrong direction. 
And it was approaching statistical significance.
    This observation was particularly concerning, because heart 
disease is highly prevalent in diabetics, comprising between 65 
and 80 percent of all diabetic deaths. A diabetes drug that may 
increase the risk of heart disease would represent a 
potentially important public health concern.
    We sought more data to objectively address this scientific 
question. Eventually we located on the FDA Web site the 
original group of clinical trials submitted to the agency to 
support approval of the drug in 1999. There were five clinical 
trials comparing Avandia to other diabetes drugs or placebo. We 
again noted that there were more heart-related complications in 
the Avandia treatment group in these initial clinical trials. 
But we still did not have enough clinical trial data to form 
any reasonable scientific conclusions.
    Eventually, in April 2007, we discovered a GlaxoSmithKline 
Web site that disclosed basic information and summary results 
for clinical trials conducted by the company. Now we had access 
to the heart attack and death rates for all relevant 42 Avandia 
clinical trials completed before or after drug approval. We 
competed the meta-analysis, which showed a 43 percent excess 
incidence of heart attack in Avandia-treated patients, which 
was statistically significant with a p value of 0.03. A p value 
of 03 means that there is a 97 percent probability that the 
results of the study are not due to chance alone. We submitted 
a manuscript reporting our findings to the New England Journal 
of Medicine, where the manuscript was peer-reviewed and 
published online on May 21, 2007.
    In our manuscript, we were careful to point out the 
strengths and limitations of our analysis. Because our access 
to data was limited to publicly available clinical trial data, 
we could not analyze original patient-level information. In 
addition, as we pointed out, a meta-analysis is always less 
convincing than a large, prospective trial designed to answer a 
specific scientific question. Nonetheless, we thought the 
findings were sufficiently important to warrant prompt 
publication and concluded ``Until more precise estimates of the 
cardiovascular risk of this treatment can be delineated in 
patients with diabetes, patients and providers should carefully 
consider the potential risks of rosiglitazone in the treatment 
of type 2 diabetes.''
    The same 42 trials that we included in our analysis are 
available to the company and to the FDA. Because both of these 
organizations have access to raw patient data, they can perform 
more statistically powerful analyses which can help clarify the 
extent of risk. GSK has reported the basic results of their own 
patient-level meta-analysis on their clinical trials Web site, 
which confirms a statistically significant increase in heart-
related complications in patients who received Avandia.
    The FDA also recently announced that their own internal 
analysis of patient-level data confirms an approximately 40 
percent excess of heart-related complications. However, neither 
the GSK nor FDA analyses have been published and it is 
therefore not possible to directly compare the results for all 
three of these analyses.
    I look forward to discussing these findings and the policy 
implications with the committee during the course of today's 
    [The prepared statement of Dr. Nissen follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.040
    [GRAPHIC] [TIFF OMITTED] 44429.041
    Chairman Waxman. Thank you, Dr. Nissen.
    Dr. Buse.

                     STATEMENT OF JOHN BUSE

    Dr. Buse. Chairman Waxman, members of the committee, it is 
really an honor to be called to testify before this committee. 
Before I tell you what I am really here for, I do want to make 
two introductory points as a matter of disclosure.
    First, this statement and my testimony do not reflect the 
opinions of my employer, the University of North Carolina 
School of Medicine, nor the American Diabetes Association, a 
voluntary health agency for which I serve as an officer.
    Second, I have been working in the glitazone class since 
approximately 1992. I have a number of conflicts of interest in 
that regard, and I have tried to expand those a bit in my 
written statement, but I don't want to go through that in 
detail, because of my time limitations.
    So I do want to give some background as to how I got 
involved in this process. In June 1999, I was invited to give 
about six presentations at the American Diabetes Association 
meetings and the Endocrine Society's meetings, and dug around 
through the same databases with the same materials that Dr. 
Nissen spoke of earlier.
    I was concerned about the potential of cardiovascular 
safety because of what I perceived to be an increase in 
cholesterol that was relatively specific to Avandia among the 
three agents that have been marketed in the United States, 
Avandia, Actos and Rezulin. Because of that, I looked for 
signals of cardiovascular safety and found a signal with regard 
to a comparison between Avandia and so-called active 
comparators in the initial Avandia data set.
    I realized that was a potentially explosive issue, reviewed 
these data with colleagues and with scientists from SmithKline 
Beecham, the manufacturer of Avandia. Those discussions were 
very helpful. Couched with many caveats, in June 1999, on two 
occasions, I presented this information, including, among many, 
many things, this potential signal of increased risk of 
cardiovascular disease.
    Subsequent to that, I received a phone call from an 
employee of SmithKline Beecham, suggesting that people in the 
company were very upset. I explained to him that I had 
discussed it with people in the company before. He mentioned 
that there was a notion that market capitalization of the 
company had decreased by approximately $4 billion, and that the 
company, there were people in the company that felt that I 
might be liable for that.
    Similar discussions were held with the chairman of my 
department. And over the next few days, I made an agreement to 
sign a statement to be used with the investment community to 
clarify some of my statements and offered to help with further 
analysis with regard to this problem.
    In March 2000, I was aware of ongoing discussions with the 
Food and Drug Administration regarding the safety of Rezulin. 
Because I was concerned about the safety of each of the agents 
for different reasons, I wanted to make sure that the Food and 
Drug Administration was careful in considering withdrawing one 
agent when we didn't have robust safety data with the other 
agents. So I made the FDA Commissioner aware of the concerns 
that I have just mentioned to you, and called for greater 
enforcement of marketing regulations, as well as additional 
    By their very nature, the observations I made in 1999 and 
the more sophisticated analyses by Dr. Nissen are only useful 
to generate questions, not to provide answers. And the most 
important question is today, what should patients and doctors 
do with regard to Avandia. I think the data are sufficient that 
there is a reason for concern. But I think if a patient is very 
well controlled on Avandia with good cholesterol control, good 
blood pressure control, good diabetes control, that with the 
available data, there might be greater risk to switching than 
to staying. Unfortunately, most patients with diabetes are not 
well controlled across the board.
    To be fair, there is no currently available drug for 
diabetes that is known to reduce cardiovascular risks. That 
said, there is certainly no diabetes drug that is marketed 
where we are aware of a signal to increase cardiovascular 
events, except for possibly Avandia. If there is a lesson from 
the events of the last weeks and years, perhaps it is that upon 
filing a new drug application, pharmaceutical manufacturers 
should make every effort to make adequately powered, 
independently executed studies that examine clinically 
meaningful endpoints, such as heart attack or loss of vision. 
In parallel with regulatory approval, such a study should be 
reviewed with attention to design, oversight, funding plan and 
timeline, recognizing that such studies are very expensive and 
will take many years to complete. Direct to consumer 
advertising and medical marketing should be constrained until 
such studies are completed.
    Thank you.
    [The prepared statement of Dr. Buse follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.042
    [GRAPHIC] [TIFF OMITTED] 44429.043
    [GRAPHIC] [TIFF OMITTED] 44429.044
    Chairman Waxman. Thank you very much, Dr. Buse.
    Dr. Psaty.

                  STATEMENT OF BRUCE M. PSATY

    Dr. Psaty. Mr. Chairman and members of the committee, my 
name is Bruce Psaty. I am a professor of medicine and 
epidemiology at the University of Washington. I wrote the New 
England Journal editorials that accompanied Dr. Nissen's meta-
analysis and the GSK RECORD study. I also served on the IOM 
drug safety committee. This testimony reflects my professional 
views as a public health scientist.
    The crisis in confidence about the safety of medicines in 
America, which started with the withdrawal of rofecoxib in 
September 2004, sadly still awaits resolution. The loss of 
confidence has created an explosive atmosphere around drug 
safety issues. The problems raised by Avandia, the subject of 
the hearing today, point to the importance of several 
recommendations made by the IOM committee. The FDA needs 
leadership and authority to require sponsors to conduct high 
quality post-market trials in a timely fashion. Public posting 
of clinical trial data was crucial to the identification of 
heart attack risk associated with Avandia. Direct to consumer 
advertising increases demand for drugs, some of which, like 
Avandia, may have been incompletely evaluated.
    The FDA needs additional resources, preferably from general 
revenues rather than PDUFA funds. Joint authority for 
regulatory actions in the post-market setting is also essential 
for the Office of Surveillance and Epidemiology. Decisions 
about safety matters need to be turned over in part or in whole 
to a new group with a more robust public health focus.
    Dr. Nissen conducted a meta-analysis, which is a method of 
summarizing previously conducted trials. In that analysis, 
Avandia was associated with a significant increase in the risk 
of heart attacks. In other words, Avandia increases the risk by 
about as much as the statin-lipid lowering drugs reduce the 
risk of heart attacks.
    The main limitations of Dr. Nissen's meta-analysis were the 
quantity and quality of the available data. The responsibility 
for the limited availability of high quality data resides with 
GSK, which did not conduct studies to definitively address 
heart attack risk in a timely fashion. The regulatory history 
of Avandia includes several key missed opportunities. It was 
approved on the basis of the ability to lower blood glucose, 
because high levels of blood glucose increase the risks of 
vascular disease, a glucose-lowering drug is presumed to reduce 
the risk of a heart attack. Paradoxically, Avandia appears to 
increase rather than decrease this risk.
    GSK did not make a serious effort to verify the presumed 
health benefits of Avandia in a timely fashion. The ADOPT and 
the DREAM trials focused largely on marketing questions and 
failed to address directly questions of heart attack risk or 
    For drugs that will be used by millions of people for many 
years, it is essential to document the benefits of therapies 
approved on the basis of surrogate endpoints. If sponsors do 
not voluntarily initiate large, long-term trials of public 
health importance, then the FDA needs the authority to insist 
that they do so in a timely fashion.
    In August 2006, GSK provided the FDA and the European 
Medicines Agency, the European equivalent of the FDA, with the 
results of several studies, including a meta-analysis similar 
to Dr. Nissen's. By October 2006, the product labels in Europe 
were revised to include this information. There was no uproar 
in Europe at this time when the labels were revised. The 
product label in the United States still does not identify 
heart attack risk as a potential adverse event in the general 
population of diabetics.
    It is not clear why the FDA failed to make this information 
public before Dr. Nissen's meta-analysis was published. The 
primary measure of regulatory success is the timeliness of 
information, warnings or withdrawals. With Avandia, FDA failed 
to warn or inform in a timely fashion.
    GSK's RECORD study has several major limitations in design 
and conduct, and even if it continues to its planned 
conclusion, information about heart attack risk is likely to be 
incomplete. Last weekend, after incorporating the interim 
results of the RECORD trial into the meta-analysis, Avandia is 
still associated with a 33 percent increased risk of heart 
attack. The possibility of heart attack benefit seems remote, 
and there is statistically significant evidence of harm.
    Late and incomplete evaluation of the health risks and 
benefits of drugs such as Avandia create concern, confusion and 
uncertainty among patients, physicians and policymakers. The 
House of Representatives, which is about to take up drug safety 
legislation, has a unique opportunity to prevent future drug 
safety problems and to reinvigorate an essential regulatory 
agency that has many outstanding scientists.
    Thank you.
    [The prepared statement of Dr. Psaty follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.045
    [GRAPHIC] [TIFF OMITTED] 44429.046
    [GRAPHIC] [TIFF OMITTED] 44429.047
    Chairman Waxman. Thank you very much, Dr. Psaty.
    I will start the questioning of the three of you. I 
appreciate your being here.
    Dr. Buse, I would like to start with you, because as far as 
I can determine you were the first outside person, outside of 
FDA, to suggest that there be a post-marketing trial to 
determine the risk of heart attacks and stroke in patients that 
were taking Avandia. Specifically, you recommended that the FDA 
should ``encourage cardiovascular in high-risk populations, 
particularly with Avandia, where I believe there is ample cause 
for concern.''
    You sent that letter to FDA. What response did you get from 
the FDA?
    Dr. Buse. I actually don't remember getting any specific 
response. I may have gotten a letter saying thank you for the 
letter. But I don't remember, I certainly don't believe, our 
specific discussion in this regard. I do run into people from 
the FDA from time to time, and have had numerous conversations 
with them over the years. But nothing that specifically 
responded to my letter.
    Chairman Waxman. Well, unfortunately, the FDA did not 
require Avandia's manufacturer to conduct the type of post-
marketing trial you recommended. And here we are 8 years later, 
without that trial having been done, so that we know exactly 
what kind of risks people are taking.
    Why are we in this situation? Do you have any idea of what 
went on in FDA? Dr. von Eschenbach said that they asked for a 
study that would have included that. And that was the ADOPT and 
the DREAM studies. Did those studies give us the answers we 
needed for this issue?
    Dr. Buse. No. As Dr. Nissen indicated, if anything, they 
suggested a trend toward risk of cardiovascular disease. In 
fact, the ADOPT study I don't think adjudicated or very 
carefully looked at heart attacks. I think it was more 
carefully looked at in DREAM. But both of those studies were 
fairly low-risk people, not the high-risk cardiovascular 
patients where my concerns were greatest. And even the RECORD 
study that Dr. Psaty mentioned is a fairly low-risk, though 
higher risk than DREAM and ADOPT.
    Chairman Waxman. I believe that part of the problem is that 
the FDA can't insist that a study be conducted. It can only 
request it. They can negotiate before the drug is approved that 
a study be done. But then if the company doesn't do the study, 
and in fact most of them don't do the studies they commit to, 
then the only recourse the FDA has as an option is to take the 
drug off the market, which seems to me is sometimes called a 
nuclear option, because it deprives people of medicines that 
they are using and they are relying on.
    Dr. Nissen, you did this meta-analysis. You or your people 
informed us that you were doing such an analysis, but we didn't 
tell you to do it, and we didn't tell the New England Journal 
of Medicine to publish it, did we?
    Dr. Nissen. No, and you didn't get to see the manuscript 
until everybody else got to see it, when it was published.
    Chairman Waxman. Do you agree with Dr. Buse that it is 
going to be years before we get the result of an appropriately 
powered cardiovascular outcomes study with Avandia that is 
likely to provide an answer to the questions raised in your 
study, the questions that he has raised?
    Dr. Nissen. I did get a look at the RECORD interim results 
that were published yesterday by the New England Journal of 
Medicine. I agree with Dr. Buse that as currently designed, the 
RECORD study is unlikely to give an answer even when it is 
completed in 2009. And since it is the major ongoing 
cardiovascular outcome study, I think the answer is that we 
will be unlikely to have a definitive answer, even when it is 
completed in 2009.
    Chairman Waxman. Dr. Psaty, how can we avoid this kind of 
problem in the future with drugs? It is going to take so long 
before a specific study can be actually done and give us the 
information we need.
    Dr. Psaty. I think they can be started earlier and designed 
well. It is not clear to me whether the FDA didn't ask for the 
right study or whether the company didn't want to do it. So I 
don't know what happened in those sorts of negotiations. But 
clearly there were concerns about cardiovascular events. Then 
they do a trial where they don't adjudicate cardiovascular 
events. And if you want to not find an answer, that is a way to 
do it.
    So we need the FDA, the FDA needs the authority to be able 
to determine the appropriate design and to insist that the 
company's conduct these studies in a timely fashion.
    Chairman Waxman. I went through a number of timeframes when 
the FDA had the signal that they ought to be looking at this 
issue, starting with their own reviewer who approved the drug, 
Dr. Buse's letter, others who were raising concerns. It doesn't 
appear to me that until Dr. Nissen's mega-study was published 
in the New England Journal of Medicine have we seen real action 
by the FDA on this matter. I hope we can avoid this kind of 
problem in the future.
    Dr. Psaty. Part of the problem is that the way things are 
set up now is we have, the FDA does a terrific job evaluating 
drugs in the pre-approval setting. And then they are approved 
and then it is marketing. And it is partly the responsibility 
of Congress, who set up PDUFA and prevented FDA from using any 
of these funds for drug safety for the first 10 years. We need 
additional attention to drug safety. It needs additional 
funding. And there needs to be a lot of work that takes place 
after the approval process.
    Chairman Waxman. Thank you very much.
    Mr. McHenry.
    Mr. McHenry. Thank you, Mr. Chairman.
    Dr. Nissen, you outline in your testimony a timeline of 
when you found, when you started going through the whole 
process. At what point did you begin your conversations with 
Chairman Waxman and his staff?
    Dr. Nissen. In February, I had looked at the DREAM and the 
ADOPT study. But I didn't have enough information to actually 
answer the question scientifically.
    I wasn't aware that there was a Web site in the United 
Kingdom where GSK had disclosed the results of all their 
trials. So I really had an incomplete set of data. At the time, 
I was discussing with various members in various congressional 
committees the pending legislation around the similar version 
of the Kennedy-Enzi bill on the House side. So I mentioned to 
them that I had concerns about the cardiovascular safety of 
Avandia and actually requested their assistance.
    Mr. McHenry. So February?
    Dr. Nissen. In February. Requested their assistance in 
getting access to the data. I had essentially a scientific 
mystery. I didn't have the means to answer the question in a 
robust, scientific way, and I really was looking for help to be 
able to do that. I was looking to see whether they could use 
their influence and authority----
    Mr. McHenry. Did you provide your interim results to them?
    Dr. Nissen. Well, to get access to any source of 
information. I was really inquiring, was there anything that 
the Congress could do----
    Mr. McHenry. I am going to another question. Did you 
provide your interim analysis results to any member of the Hill 
or staff?
    Dr. Nissen. No. There were no interim results. Basically 
what we had done is, we had a very preliminary analysis, 
nothing formal.
    Mr. McHenry. Did you provide your preliminary analysis to 
people on the Hill?
    Dr. Nissen. I did show them a preliminary analysis, yes. 
That's correct. Yes.
    Mr. McHenry. At what point did you have that and did you 
share it with Mr. Waxman's staff?
    Dr. Nissen. Some time in February.
    Mr. McHenry. February.
    Dr. Nissen. Yes.
    Mr. McHenry. So they were aware of what you were going 
through the process of?
    Dr. Nissen. They were aware of what I was working on, yes.
    Mr. McHenry. Why didn't you discuss your preliminary 
analysis with the Food and Drug Administration?
    Dr. Nissen. Well, the Food and Drug Administration had all 
of these studies already. Remember that when you do a study, 
you submit a study report to the FDA.
    Mr. McHenry. But you were actually submitting to a medical 
journal a new study with meta-analysis, which is aggregating 
what was already public. So you proffer your work as original, 
do you not?
    Dr. Nissen. It is original.
    Mr. McHenry. OK, then, why didn't you share that study with 
the Food and Drug Administration? After all, as Members of 
Congress, we have a regulatory structure that we put in place 
for drug safety. Why didn't you go to the FDA with that 
    Dr. Nissen. This is not how it is done. We have to peer 
    Mr. McHenry. So going to Capitol Hill for a political 
purpose to get publicity here in a hearing is actually the way 
it is done? That's really medical research----
    Dr. Nissen. With all due respect, sir, this is about 
patients. It is not about politics.
    Mr. McHenry. If it is about patients, why would you not go 
to the regulator who has the authority and oversight for drug 
    Dr. Nissen. Please let me finish. This is about patients, 
not politics. I had an incomplete result. I was looking for 
assistance to complete the study. When it was completed, I did 
what any scientist would do. I sent that for peer review and 
for publication. Why? Because it is my scientific, it is my 
ethical and it is my moral obligation to put such information 
into the public domain, so that other physicians, other 
scientists providers, and patients can consider our findings 
when making choices about drugs.
    Mr. McHenry. Thank you, Dr. Nissen.
    My additional question would be, what peers do you have on 
the Oversight and Government Reform staff for the Democrat 
staff? Because you shared your findings with them. Is that what 
you consider peer review? Is that what you consider putting 
patients above politics?
    Dr. Nissen. I did not give a copy of my manuscript to this 
committee or anybody else until it was published.
    Mr. McHenry. Did you provide your initial analysis----
    Dr. Nissen. I provided preliminary suggest--I looked at the 
two trial----
    Mr. McHenry. Did you provide a draft of your----
    Dr. Nissen. You are interrupting me, sir. I really would 
love to be able to answer your questions.
    I provided a preliminary analysis.
    Mr. Issa. I would ask unanimous consent for two additional 
minutes so that this can go on appropriately without----
    Chairman Waxman. No, the gentleman has his time and he 
still has time left.
    Mr. Issa. Then your time is limited.
    Mr. McHenry. Well, my time is limited. And did the editors 
at the New England Journal of Medicine know that you shared 
this analysis with members of the Hill before?
    Dr. Nissen. I don't know what they knew or they didn't 
know. I submitted the manuscript to them.
    Mr. McHenry. So, OK, as a final moment here, because I know 
the chairman will rap me down here, it seems very peculiar to 
me that if you are considering the patients first that you 
would not go to the regulator who is overseeing drug safety, 
that you would go to Capitol Hill, which as we know is a 
political body, and we don't have the authority to take a drug 
off the market, the FDA does. So you can respond to that if you 
like, but my time is up and I yield back the balance of my 
    Dr. Nissen. I would like to respond if I could. The 
regulatory agency had all of the data that I had and much, much 
more. So what I had was a much more limited look at the data 
than what the FDA already had. It would make no sense for me to 
take study level data and submit it to the FDA when they 
already had the patient level data. So I would not have given 
them anything they hadn't had for many, many months.
    Chairman Waxman. The gentleman's time is expired. Mr. 
Yarmuth is now recognized. I would request that the gentleman 
yield to me for just 30 seconds to ask the following question. 
You came to a number of committees, Democratic and Republican 
members of those committees, is that true?
    Dr. Nissen. That is correct.
    Chairman Waxman. And you asked for help to get data to 
complete your evaluation. Did you get any help from anybody on 
the Hill?
    Dr. Nissen. No.
    Chairman Waxman. And wasn't that the reason you came to the 
committees of the Congress?
    Dr. Nissen. Absolutely.
    Chairman Waxman. OK, thanks. The gentleman is recognized.
    Mr. Yarmuth. Thank you, Mr. Chairman. I would like to 
address a question to Dr. Buse and I understand that you have a 
very significant family event tonight, a commencement, and you 
have to leave early. So I want to get this question in. I 
congratulate you on that.
    In your written testimony, you state that as far back as 
1999, you had concerns about Avandia based on your analysis of 
the initial approval studies and your knowledge that Avandia 
might increase levels of bad cholesterol. You explained that 
you had discussed your concerns at a professional meeting in 
1999, and that after you did that, you came under a great deal 
of fire and pressure from the manufacturer at the time, 
SmithKline Beecham, which is now GlaxoSmithKline.
    You said that company representatives complained to your 
department chair. Exactly what did they say to him?
    Dr. Buse. There was a high-ranking member of the company 
that had a longstanding professional relationship before he 
joined the company with my chairman. And I don't know the 
details of the conversation. But it was characterized to me as 
being disturbing, and the two phrases that I remember, or three 
phrases, one involved that number, $4 billion. The second was 
that I was characterized as a liar. And the third was that I 
was characterized as being for sale.
    Mr. Yarmuth. Was this something that happened frequently in 
your capacity as a researcher?
    Dr. Buse. No. That was a fairly unique experience.
    Mr. Yarmuth. Was the company in any position to exert any 
specific pressure on you or your chair or the University of 
North Carolina? Were they funding research through UNC?
    Dr. Buse. I don't know the answer to that question at all.
    Mr. Yarmuth. Was there any evidence, you mentioned the $4 
billion figure as to reduction of market capitalization, was 
there any basis for that statement? Had the stock actually 
taken a hit?
    Dr. Buse. I didn't bother to look.
    Mr. Yarmuth. That would be a lot of money on a professor's 
salary, though, wouldn't it?
    Dr. Buse. It would take a while. [Laughter.]
    Mr. Yarmuth. You also testified that following those 
conversations with your department chair that you signed a 
clarifying statement. Was that statement something that you 
wrote or did the company prepare that?
    Dr. Buse. The company prepared it.
    Mr. Yarmuth. During this committee's preparation, we 
requested documents from GSK relating to their meetings and 
dealings with you. In response, they supplied a copy of a three 
and a half page fax you sent to a Dr. Yamada, the company's 
chairman of pharmaceutical research and development at the 
time. Do you recall writing this letter?
    Dr. Buse. I recall agonizing about writing that letter.
    Mr. Yarmuth. I would like to request unanimous consent that 
a copy of the letter be included in the record, Mr. Chairman.
    Chairman Waxman. Without objection, that will be the order.
    [The information referred to follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.048
    [GRAPHIC] [TIFF OMITTED] 44429.049
    [GRAPHIC] [TIFF OMITTED] 44429.050
    [GRAPHIC] [TIFF OMITTED] 44429.051
    Mr. Yarmuth. I would also like to read an excerpt from the 
letter. It says, ``I may disagree with SB's, that is SmithKline 
Beecham's, interpretation of the data. I am not for sale. I am 
anxious to help in any way that I can to establish Avandia as a 
safe and effective anti-diabetic agent with certain 
stipulations. I cannot change my opinions in the absence of new 
data or understanding, in large part because I am not for sale. 
I look forward to working with SB in the future, but will 
understand and not take offense if I do not. Please call off 
the dogs. I cannot remain civilized much longer under this kind 
of heat.''
    Dr. Buse, I regret that you were the subject of this type 
of intimidation. I certainly hope it has not recurred since you 
sent that letter. It goes without saying that this type of 
conduct is completely unacceptable. We can't have a post-market 
regulatory environment in which manufacturers attempt to 
intimidate science. So I thank you for your testimony.
    Dr. Buse. If I could just add to that. I do think that most 
of the really ugly bits of that interaction were out of 
frustration, anger of a limited number of individuals who felt 
that they were trying to be forthright in presenting the data 
with regard to their drug. I have not had issues since then.
    Mr. Yarmuth. That is comforting. I yield back.
    Chairman Waxman. Mr. Cannon.
    Mr. Cannon. I apologize, we have a markup on energy, in the 
Committee on Natural Resources. So I have been back and forth, 
and I apologize for not being here more. I note that you lose 
your entire status if you leave the dais for a few minutes 
    Thanks for coming. I think you were here earlier when I was 
questioning Dr. von Eschenbach. My concern in this process is 
sensationalization. I think, Dr. Nissen, we probably agree that 
the FDA can do things differently and better. But in this 
process, it has become, I think, well, at least sensational.
    Do you buy stocks yourself, Dr. Nissen?
    Dr. Nissen. I do not.
    Mr. Cannon. Do you have friends that do?
    Dr. Nissen. I am sure I do, but I don't know what they own.
    Mr. Cannon. And of course, that is not what we care about 
really. Are you familiar with what has happened to various drug 
stocks when they have been politicized over, say, the last 8 or 
10 years?
    Dr. Nissen. I really don't follow the stock market.
    Mr. Cannon. When the Clintons took over the Presidency, and 
Mrs. Clinton did her exercise in oversight of the health care 
system, she announced at one point that the drug companies were 
the villains and that the administration was going to go after 
them. Do you have any idea what happened to the stock price of 
those companies?
    Dr. Nissen. I don't.
    Mr. Cannon. Oh, you have to.
    Dr. Nissen. Pardon?
    Mr. Cannon. You have to have an idea. It didn't go up, of 
    Dr. Nissen. Well, again, I don't know. I am not an expert 
on stock prices.
    Mr. Cannon. Stock prices fell by about half in that period 
of time. Then about 2 weeks later she came out and announced 
that the drug companies weren't really the problem and stock 
prices went up, back to their normal state. A huge, multi-
billion dollar transition in a market we try to keep stable and 
we try to have it work for other reasons.
    Have you taken a look at or considered what has happened to 
GlaxoSmithKline's stock?
    Dr. Nissen. I have seen news articles to the extent that 
the stock prices dropped.
    Mr. Cannon. Do you know how much?
    Dr. Nissen. I don't have specific figures.
    Mr. Cannon. It dropped about 20 percent. About that, in 
that range, over one study that is at least, I don't think 
either of you would say that the study is definitive. There are 
certainly a whole bunch of questions that the study raises. Do 
you have a concern about the kind of sensationalism that 
results in a 20 percent stock movement?
    Dr. Nissen. As a physician-scientist, and first of all, I 
respect your perspective, Mr. Cannon, but as a physician-
scientist, I have to ask different sets of questions. I did 
have concerns about publishing the study and I did have 
concerns about how it would be interpreted. So I have three 
questions I have to ask before publishing a study: is it 
scientifically sound, did I use the right methods, did I 
consider alternatives and did I do a good job.
    Mr. Cannon. And everybody agrees that you are very good at 
that, by the way.
    Dr. Nissen. Thank you. But we can make mistakes. So----
    Mr. Cannon. Sure, so that is why we have a peer-review 
    Dr. Nissen. That is exactly right.
    Mr. Cannon. Oh, I didn't think about that, let's go back. 
But in your case, this case, it was probably not a mistake. You 
had studies that GlaxoSmithKline had already done.
    Dr. Nissen. Yes.
    Mr. Cannon. Their data was available online, it was not 
anything that was being hidden, by any means. So it was a study 
of various studies and a lot of assumptions were made in the 
process, and we came up with a signal.
    Dr. Nissen. That is right. So the first question is 
scientific, and the second question is, is it ethical and 
moral, is it appropriate. And I knew that when we published 
this that it would in fact, there would be concerns on the part 
of patients, that people would be potentially frightened. As a 
consequence I tried to be as measured as I could in how I wrote 
the manuscript. I really would encourage everybody to read what 
I said.
    Mr. Cannon. I understand that, and apparently I have missed 
some of the discussion here. But there is some question about 
whether or not you came to the committee, majority staff, and 
talked to them about this issue.
    Dr. Nissen. What I told them earlier is that I did not 
share the manuscript. I did tell them I was working on it, I 
told them I had concerns. But ultimately, what I wanted to have 
happen was, we had to make a scientific judgment. We came to 
the judgment. I had to make an ethical and a moral judgment.
    Let me tell you what the alternative was. And it was an 
alternative I considered. The alternative would be not to 
publish, to come to the conclusions and say, gee, this is so 
explosive that I just won't put it out there. And I did plenty 
of soul-searching. And I realized that I had absolute, absolute 
ethical and moral obligation to----
    Mr. Cannon. My time is almost gone. Can I just ask this, 
didn't the FDA have that obligation as an institution, and 
wouldn't it have been as well to have gone to them and talked 
to them about the issue?
    Dr. Nissen. Well, the FDA, Mr. Cannon, I think has that 
responsibility, and I recognize that. The FDA, however, had the 
same data that I had.
    Mr. Cannon. Right.
    Dr. Nissen. They actually had more data than I had. As I 
was explaining a little bit earlier, they had all the patient-
level data. They had enough data to do a much more powerful 
analysis than I did. The question obviously on the table here 
is, where were they at in the process. Were they----
    Mr. Cannon. I think the question on the table here is, why 
do we have this sensationalist hearing when everybody agrees 
that the data is indeterminate and you have a really important 
drug and in the middle of all that, you are whacking on a 
business that is doing its job to create a better world for 
people who are sick?
    Dr. Nissen. There is a reason, sir. The reason is that I 
wanted my colleagues who practice medicine and I wanted 
patients who take these drugs to be aware of our analysis. I 
thought that it was my obligation to inform them that there was 
a potential risk. I could not allow patients with diabetes----
    Mr. Cannon. Mr. Chairman, I see my time has expired. If I 
can just make a comment.
    Chairman Waxman. The gentleman's time has expired. And we 
haven't really allowed other Members to extend their time.
    Mr. Cannon. I wouldn't dream of doing that. I yield back, 
Mr. Chairman.
    Chairman Waxman. Thank you, Mr. Cannon.
    Mr. Cummings.
    Mr. Cummings. Thank you, Mr. Chairman.
    I have a lot of questioning, but I have to say that after 
being here for 11 years, I hate it when witnesses are attacked, 
it bothers me. Particularly when they are trying to do the best 
they can, in the words of Thurgood Marshall, with what they 
have. I believe that you all are honorable men, simply trying 
to be the best that you can be. So I am going to ask one or two 
questions to clear this up. And I hate that we have to make, 
that these accusations are made that people are putting 
politics over the health of the American people. That bothers 
    So let me ask it this way. Dr. Buse and Dr. Psaty, you have 
heard this line of questioning, you heard what Dr. Nissen has 
said. Do you all have any issue with the professionalism that 
he has, the way he has gone about doing what he has done to get 
this information published? Dr. Buse first.
    Dr. Buse. I have no issue with it at all. I think he did a 
nice job of organizing the data and setting out that it was 
imperfect but important for people to be aware of.
    Mr. Cummings. Dr. Psaty.
    Dr. Psaty. I agree. I think he did a terrific job in a 
difficult situation. There were opportunities to prevent this. 
GSK could have published their meta-analysis. The FDA has had 
this information for months. It was released in Europe in 
October. I don't know why it takes so long for the FDA to 
release information. Detailed analysis is important, but at 
some point, it looks like a lack of transparency and a lack of 
communication. It would have been perfectly reasonable in 
August to say, we have two studies from GSK, they suggest this 
risk, it is not clear, they contradict each other. It is 
important for people to know this information.
    What Steve is dealing with is a safety issue. And it is 
prudent to warn patients about risks. We have to first do no 
    Mr. Cummings. The reason why I did that is because, you 
guys have to go home. You have to go back to where you came 
from. And I don't want, on national television for folks to 
believe that somebody is doing something that is improper if 
they are not doing it.
    Let me ask you this. Let me say this. In my district, in 
Baltimore, we have a high, high degree of diabetes and heart 
disease. I represent Johns Hopkins. But today, I guarantee you, 
people will die, today, from diabetes. And now I have learned 
something interesting, that they will die from diabetes, but 
probably the heart disease will kill them.
    So today, would you recommend, Dr. Nissen, based upon what 
you see right now, would you, if your physicians came to you 
and said, should we be prescribing this drug, what would you 
say? Just what would you say? If they say, look, Doc, we just 
saw you on C-SPAN and we are kind of concerned about this.
    Dr. Nissen. I deliberately did not answer that question, in 
the manuscript or subsequently. Let me tell you why. With 
science, you have to allow individual physicians to make their 
own minds up about how to interpret the data. My job was to get 
the data into the public domain in the best journal possible, 
carefully reviewed and thoughtfully articulated. What I have 
said is, individual physicians should look at the results, 
discuss it with their patients and make their own minds up 
about what the right thing to do is. We knew that it wasn't the 
definitive end, we knew there were more questions to be asked. 
Rather than come to conclusions, we said, here it is, you 
    Mr. Cummings. What kinds of tests would you recommend that 
give us, would bring you to a conclusion where you would say, 
yes or no?
    Dr. Nissen. What would need to be done is an adequately 
sized, long-term trial, probably in fairly high-risk patients, 
comparing Avandia to other therapies. That would, now, 
unfortunately, because such a trial doesn't exist, it would not 
be completed for probably about another 7 years. So it is a 
long, long way off. The problem is, as Dr. Psaty said, the time 
to have launched such a study would have been 1999 or 2000.
    So we are in a very tough quandary here, in that we don't 
have the data to definitively answer the question. We just have 
the meta-analysis, which is all we are ever going to have, 
because it looks like RECORD isn't going to give the answer, 
    Chairman Waxman. Thank you. Thank you, Mr. Cummings.
    Mr. Issa.
    Mr. Issa. Thank you, Mr. Chairman.
    Dr. Nissen, I guess I am going to keep following up a 
little bit. One thing that was said in the previous panel, and 
it is unfortunate that the FDA you think so little of that you 
go to Congress before you go to the scientists and the doctors 
who we entrust to make these decisions, said, and they weren't 
willing to commit to the statistical likelihood, but you are 
somebody who reads some statistical likelihood. You are 
responsible for this compilation of meta-data.
    Why did you choose to ignore or to leave out meta-data in 
which nobody died, in which nobody had a heart attack? And 
before you answer why you chose to leave it out, by definition, 
if you had put it in, wouldn't it have lowered the conclusions 
that you reached? Please, Dr. Nissen.
    Dr. Nissen. You can't calculate, in a meta-analysis, you 
can't use trials in which there are no events. It simply can't 
be done statistically. Let me explain why. I know you want a 
short answer, but----
    Mr. Issa. Well, no, unfortunately I insist on a short 
answer, so I will rephrase it to help make that happen. If you 
put zeroes in, statistically, yes, you would get a lower 
number. So now, the fact that you can't put it in, anyone with 
common sense says, well, these studies where nobody got sick 
were not something, nobody had heart attacks, those were 
studies in which the public and the doctors that you say you 
are providing this information to, even though you are 
providing, I mean, you might as well just have everyone do 
studies and every doctor evaluate it if we are not going to use 
the FDA.
    But in this case, you left that information out of what the 
doctors got to know, didn't you?
    Dr. Nissen. That information cannot be used to calculate--
    Mr. Issa. No, no, my question was rephrased to make it a 
yes or no. You left that information out so the doctors did not 
have the knowledge that hundreds or thousands, whatever number 
of people were in all those studies, did not have heart 
attacks. You left that out, didn't you?
    Dr. Nissen. That information is publicly available on the 
FDA Web site.
    Mr. Issa. No, no. Of your, of your report, they are relying 
on your report as part of the balancing act, you left it out, 
didn't you?
    Dr. Nissen. Mr. Issa, you can't calculate an effect size 
when there are no events.
    Mr. Issa. OK, look, we already did this----
    Dr. Nissen. The manuscript was----
    Mr. Issa. No, no, sir, I have limited time. You are not 
willing to answer the simple question of did you leave it out, 
were the doctors aware of it. And to say that doctors can pore 
into research that you came to the majority staff and asked for 
help getting back in February as you planned to release this 
very, very earth-shattering effect, whether you intended it to 
be or not. And I suspect you intended it to be. You came to 
Congress, you planned with them to essentially bring this out. 
You asked for additional information and then you are going to 
come here, I am a little disappointed, and tell me that doctors 
can find it out themselves, it is public. I am sorry, but 
leaving that out is the reason that you clearly should have 
gone to the FDA.
    I am going to ask you a question related to that. Did you 
have discussion with the FDA back in January, February or 
March, when you were having discussions with the majority staff 
    Dr. Nissen. No.
    Mr. Issa. OK. So you didn't go to the very body that we 
held here accountable, that we are holding oversight hearings 
on, and yet we are going to ask them why they didn't do their 
job, you didn't even give them the benefit of the doubt. Did 
anyone from the majority staff suggest that you at least bounce 
these off of the FDA?
    Dr. Nissen. That was never discussed.
    Mr. Issa. Did anyone here, as you were trying to get a 
political body to get you more information, did anyone suggest 
that you ask the FDA to assist you?
    Dr. Nissen. No.
    Mr. Issa. OK. So it very much looks like this was a 
political entity designed to make a big, public splash. It is 
clear from letters that I have here that in fact, before your 
study was published, we were asked to ask for a hearing. So in 
fact, didn't you reach a conclusion, back in February, that 
this was in your opinion a potentially dangerous drug, and 
decide that you wanted to shed light on it using this body in a 
public hearing in your article? Didn't you decide that all the 
way back at least in February?
    Dr. Nissen. I did not come to that conclusion until I 
finished the meta-analysis.
    Mr. Issa. OK, so what were you doing in February when you 
were saying you were concerned, and asking for this information 
from a political body rather than in fact from the fundamental 
group that we hold accountable at the end of the day?
    Dr. Nissen. I had incomplete information. I didn't have 
access to all 42 clinical trials. I knew that I needed it.
    Mr. Issa. And you hadn't asked the FDA for it.
    Dr. Nissen. The FDA is not allowed to give the data out.
    Chairman Waxman. How about GSK? Did you ask them?
    Dr. Nissen. I did.
    Chairman Waxman. Did they give you the information?
    Dr. Nissen. No. Well, we were unable to reach agreement on 
getting the information.
    Mr. Issa. When committee staff went with you, with the 
primary drug reviews were raised, did they suggest that they 
could in fact get that information and did you ask them to try 
to get it through other channels, and did you wait for that 
before publishing?
    Dr. Nissen. I am sorry, I didn't hear your question. I 
don't understand your question.
    Mr. Issa. When you met with committee staff, or I am sorry, 
when committee staff met with the FDA, reviewers were raising 
the same concern. You said the FDA included studies with their 
meta-data analysis that you did not. Can you understand why 
they included the studies and you didn't?
    Dr. Nissen. My understanding is, they have not in fact 
announced what studies they have included, so I have no way of 
knowing how they did their analysis. Remember, their analysis 
has not been published or presented. So we have no way of 
comparing the two analyses.
    Chairman Waxman. The gentleman's time has expired. Dr. 
Psaty and Dr. Buse have been raising their hands.
    Mr. Issa. Mr. Chairman, they can do what they want on 
somebody else's time. If you are going to interrupt me during 
my time to ask a question and then you are going to bring it to 
a close, please use somebody else's time to do this. I wish we 
had more time, because this very much does, Mr. Chairman, as I 
said in my opening remarks, this does look like in fact this 
was a political concoction to anecdotally go after a company 
rather than to do legitimate oversight on the FDA. I object to 
    Chairman Waxman. The gentleman is being demagogic. This is 
not anything that is political. Dr. Nissen's paper was peer-
reviewed and published in a very respectable journal. It is 
that article that has raised a lot of concern. It is certainly 
appropriate for this committee to raise these issues and bring 
in the various parties to talk about the issue. You are the one 
who wants to politicize this issue.
    Now, you asked a lot of questions and two of the witnesses 
wanted to respond to your questions. Do you object to having 
them respond?
    Mr. Issa. I asked and did not get answers from one 
individual who continually wanted to evade giving me the proper 
yes or no that I deserved when I rephrased the question.
    Chairman Waxman. That is not my fault. You did what you 
could and he answered to the best of his ability.
    Mr. Issa. Mr. Chairman, in regular order, I would 
appreciate that we can have a second round and certainly those 
can be asked and answered on either one of our times. I would 
look forward to a second round if you think it is appropriate, 
Mr. Chairman.
    Chairman Waxman. Do you object to these two gentlemen 
responding to your----
    Mr. Issa. Mr. Chairman, I would ask for regular order.
    Chairman Waxman. Well, let's go on to, I think Mr. Shays' 
time. Maybe he wants to be recognized.
    Mr. Shays. I would be happy to let Mr. Issa pursue his 
    Chairman Waxman. OK, Mr. Issa----
    Mr. Shays. Beforehand, I just want to, having come late to 
this, Dr. Nissen, and I will allow the two other gentlemen to 
respond to the questions that were asked, because I would like 
to know the answers.
    What I am unclear about, in just one area, is did you come 
to this committee because you wanted this committee to use its 
resources to get data for you?
    Dr. Nissen. That is correct.
    Mr. Shays. And did you feel that this committee had 
legislative ability to get this information that someone else 
didn't have the ability?
    Dr. Nissen. I didn't know what authority it had. But I had 
met the staff, because we had discussed some pending 
legislation. So I said, look, I have a concern here.
    Mr. Shays. What pending legislation was that?
    Dr. Nissen. This is the Waxman-Markey bill that is being 
considered, that is the companion to Kennedy-Enzi.
    Mr. Shays. See, my problem is that sometimes I feel 
Congress has been used to go after companies, and that the 
trial lawyers and everybody else uses the mechanism of Congress 
to then build a case and to be able to get information from the 
company that you wouldn't have a right to unless you mis-used 
Congress to do it. That is where I start to become very 
defensive about the process. I believe that once people come 
before a committee, my colleague on the other side of the aisle 
says he objects to how witnesses are treated. I think it is 
just as important, once you walk into this territory, you have 
to be willing to have the scrutiny and to be able to respond to 
questions. But I would like to the two other gentlemen to 
    Chairman Waxman. Would the gentleman yield to me?
    Mr. Shays. Yes, absolutely.
    Chairman Waxman. I don't know if you were here at the time, 
but Dr. Nissen came to Senator Grassley's staff, our staff, Mr. 
Dingle's staff, others that I might not be aware of, asked for 
help getting data. And he did not get the help with getting the 
data. He asked the company to give him the data. He did not 
eventually get that information.
    So that was the extent of our involvement.
    Mr. Shays. All right, thank you.
    Chairman Waxman. I don't know if there is anything improper 
about it.
    Mr. Shays. I would like the two gentlemen to respond to 
that. And I would be happy to yield.
    Dr. Buse. Just very briefly in response to Congressman 
Issa's questions for Dr. Nissen, I have had the opportunity to 
speak with two statisticians in part of various duties I have 
regarding the analysis that Dr. Nissen did. By the technique, 
he had to leave out those studies and he disclosed in the paper 
that, I left out those studies because I have to to be able to 
do this meta-analysis. And GlaxoSmithKline and the FDA have 
done their own analysis, the best that they could do, and 
basically all the analyses come up with the same result.
    So from my perspective, we don't have to have a big 
discussion about what kind of analysis was done and whether it 
was done properly. Everybody gets the same result.
    Mr. Shays. Is your answer the same, sir?
    Dr. Psaty. It is, but I think I can perhaps, I am a 
biostatistically inclined epidemiologist. If you think about 
it, if a study has no heart attacks, it can add no information 
to a meta-analysis about heart attacks. This is not an effort 
to create incidents routes. It is ratios, and they are not 
affected by leaving out trials that----
    Mr. Shays. Well, to my non-scientific mind, if you do a 
study and there is not an outcome that is negative, it strikes 
me from a non-scientific mind that is certainly important data.
    Dr. Psaty. The studies compare heart attack rates in one 
group to another. And if you have two groups and there are no 
heart attacks, you have no information about heart attack risk. 
This is a standard approach.
    Mr. Shays. Other than they are not getting heart attacks. 
    With all due respect, let me----
    Dr. Psaty. But it is not an incidence rate that you are 
looking at.
    Mr. Shays. I understand there is something I don't get 
because I am not a scientist. And I don't mean that in any way, 
you are just not going to be able to connect with me. 
Logically, if people don't have heart attacks, that is data.
    Mr. Issa. Earlier we heard that there was a study left out 
that had one heart attack, but they didn't die. So I guess if 
you don't die, you don't count, either.
    Dr. Psaty. I think that was in the analysis of 
cardiovascular deaths.
    Mr. Issa. OK, well, the FDA in its review with our staff, 
when we were preparing for this, said that by leaving out that 
data, you did bias the risk assessment, that clearly if you 
take 1,000 people who all took the drug and you say 43 percent 
are more likely to have a heart attack, that 43 percent is a 
relative number and it can be expressed in a number of ways.
    So having said that, my concern here today is not whether 
or not this drug is more dangerous, because I think the science 
is still to be worked out on that, and I look forward to it 
being done. My concern here today, and the chairman is calling 
it demagogy, but it is part of the minority's job, is to second 
guess what is being simply handed to us. And what is being 
handed to us is the various Democrat leadership, you prepared 
for paper in harmony with them. And Doctor, you obviously did 
not intend to get peer review quietly. You intended to get it 
loudly and you are getting it here today.
    I yield back.
    Chairman Waxman. You didn't get peer review, Dr. Nissen, 
from Members of Congress, did you?
    Dr. Nissen. No, they didn't see the manuscript.
    Chairman Waxman. OK. Well, that completes the questioning 
from Members. I want to thank the three of you for your 
presentation here. I note, Dr. Buse, you were reluctant to 
participate in the hearing, so I especially appreciate your 
    Ironically enough, if the FDA and the drug manufacturer, 
GlaxoSmithKline, had listened to you 7 years ago, we would have 
had a more definitive answer on the very important question 
that affects millions of Americans. We don't have the answer to 
it, although some Members of Congress have answers as to how 
the scientific evaluation ought to be done statistically. But 
most of us can't reach these conclusions. The conclusion I 
reach is that we have wasted a lot of time and as a result of 
the information, the meta-analysis, we have an ongoing question 
that people have to grapple with, which is unfortunately not 
    I thank you very much and appreciate your being here.
    Our last witness is Dr. Moncef Slaoui. Dr. Slaoui is the 
chairman of research and development of GlaxoSmithKline. Dr. 
Slaoui has a Ph.D. in molecular biology and immunology in 
Belgium, completed post-doctoral studies at Harvard Medical 
School and Tufts University School of Medicine. In his current 
position at GlaxoSmithKline, he has served on the research and 
development executive team and spearheaded recent changes to 
enhance drug discovery and accelerate product development.
    Dr. Slaoui, we are pleased to welcome you to our hearing 
today. As you might have been aware from earlier witnesses, it 
is the practice of this committee to ask you to rise to take an 
oath, if you would.
    [Witness sworn.]
    Chairman Waxman. The record will indicate you answered 
    We are pleased to have you, and I want to recognize you for 
your oral presentation. Your full statement will be in the 
record in full. We would like to ask you, if you would, to 
limit your presentation to 5 minutes.


    Mr. Slaoui. Mr. Chairman and members of the committee, 
thank you for having me here today. My name is Moncef Slaoui, 
and I am the chairman of research and development at 
GlaxoSmithKline [GSK]. I am here to share with you GSK's 
extensive and ongoing efforts to research both the safety and 
the benefits of Avandia, the important medicine that helps 
patients fight the devastating effects of type 2 diabetes.
    GSK has initiated the most comprehensive research program 
for any oral anti-diabetic medicines available today, with 
experience in over 52,000 patients studied in clinical trials. 
By doing so, GSK has already undertaken what Congress has 
suggested all pharmaceutical companies should do; that is, 
rigorous scientific studies of a medicine's safety and benefit 
after it is approved by the FDA.
    The data we have collected from those studies not only 
confirm Avandia's efficacy in controlling blood glucose levels 
in diabetes patients, but those data also show that Avandia 
controls blood sugar for longer periods than other currently 
available oral anti-diabetes medicine. Avandia has shown 30 
percent and 60 percent superior efficacy to Metformin and to 
sulfonyureas, the two most commonly used oral anti-diabetes 
    As concerns the very important point of safety, the 
comparable data that we have generated over the last 8 years 
establishes that when compared to other widely used oral anti-
diabetes medicines, Avandia is not associated with an increased 
risk of death, including death from a cardiovascular event. The 
data also show that except for the well described increased 
risk for congestive heart failure associated with this class of 
medicines, the TZDs, not just with Avandia, Avandia has a 
comparable cardiovascular safety profile to that of the most 
widely used oral anti-diabetes medicine.
    Let me take you through this. From day one, GSK and 
regulatory agencies believed it was important to develop the 
highest level of scientific evidence to assess the 
cardiovascular benefits to the risk profile of Avandia. 
Accordingly, in the year 2000 and again in the year 2001, we 
started two very large prospective long-term clinical trials, 
respectively the ADOPT and the RECORD studies. Both trials 
allowed us to compare over a period of 3 to 4 years the safety 
of Avandia to that of the two most widely used oral anti-
diabetes medicine, each in more than 4,000 diabetes patients.
    Specifically, the primary goal of the RECORD study was to 
compare the risk of cardiovascular deaths and cardiovascular 
hospitalization in these patients, including heart attack, 
stroke, congestive heart failure in patients using Avandia or 
patients using other medicines.
    Importantly, given the length of these prospective clinical 
studies, we did not just sit there and rely on ADOPT and RECORD 
studies to come out. We proactively used other available 
scientific methodologies, albeit less robust than the 
prospective clinical trials, we just heard the discussions 
around that analysis, to assess Avandia's cardiovascular safety 
    We ran our own meta-analysis in 2005 already and also in 
2006, which we knew would be useful for generating hypotheses, 
yes, but not for providing definitive answers. We also ran a 
very large real world epidemiological study in over 33,000 
diabetes patients. That study showed that there was no 
increased risk for Avandia.
    While the meta-analysis conducted in 2005 and 2006 did 
suggest a potential increase in cardiovascular patients using 
Avandia, all other more robust scientific evidence that we 
have, and that is coming from four independent, high-level 
scientific experimentation, three large trials, the ADOPT 
trial, the DREAM trial, the RECORD trial and the large 
epidemiological study that I just spoke about, all those 
studies have shown that the hypothesis is not accurate that 
there is an increase of cardiovascular risk associated with the 
use of Avandia, when we compare it to the two most widely used 
oral anti-diabetes medicines.
    Throughout this time, we also communicated diligently with 
the FDA the data that we received from the meta-analysis. We 
transparently published the DREAM study and the ADOPT study in 
reputable journals and we posted all our clinical trial results 
as well as our meta-analysis on GSK's clinical trial registry, 
actually in October 2006, well before the publication in the 
New England Journal of Medicine.
    We also diligently communicated to physicians and patients 
Avandia's scientifically established safety risks. In summary, 
at every step, GSK examined the questions generated by our 
meta-analysis and by that of others. We determined that more 
robust scientific data consistently conflicted with the signals 
raised. The complete body of evidence available to date clearly 
supports our conviction that the cardiovascular safety of 
Avandia is comparable to that of the two most widely used oral 
anti-diabetes medicines.
    As we all work together here today on these issues, I do 
ask that we all remember that we are working on behalf of 
diabetic patients who are at risk of many major complications. 
They were cited: kidney failure, limb amputation, nerve injury, 
blindness, cardiovascular events, deaths. Unfortunately, the 
worldwide epidemic of type 2 diabetes shows no signs of 
    All medicines have risks. But the benefits of oral anti-
diabetic medicines like Avandia help millions of patients 
control their diabetes and live healthier, more productive 
    I will say that we found the RECORD data which we published 
yesterday in the New England Journal of Medicine very 
reassuring, recognizing that it is interim and therefore not 
fully conclusive. We are extremely disappointed by the 
editorials published yesterday in the New England Journal of 
Medicines that cherry-picked data points when the data taken as 
a whole supports the safety profile of Avandia.
    I thank you very much for your attention, and I would be 
happy to take your questions.
    [The prepared statement of Mr. Slaoui follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.052
    [GRAPHIC] [TIFF OMITTED] 44429.053
    [GRAPHIC] [TIFF OMITTED] 44429.054
    [GRAPHIC] [TIFF OMITTED] 44429.055
    [GRAPHIC] [TIFF OMITTED] 44429.056
    [GRAPHIC] [TIFF OMITTED] 44429.057
    [GRAPHIC] [TIFF OMITTED] 44429.058
    Chairman Waxman. Thank you very much, Dr. Slaoui. I want to 
recognize Mr. Issa for questions.
    Mr. Issa. Thank you, Mr. Chairman.
    I want to note that I appreciate your being here today. The 
first panel was mutually agreed to as being the Commissioner, 
that is common for administration officials. Unfortunately, we 
hoped to have you on the second panel, so that we could have 
the kind of interface that I am afraid we are being denied 
right now. But I will work with what we have.
    Dr. Nissen has been quoted as saying that Avandia as a drug 
has no established health benefits. Would you like to comment 
on that?
    Mr. Slaoui. Well, I completely disagree with that. I think 
that the scientific field has established in the 1990's very 
clearly that if you decrease the blood sugar levels over a 
period of time, you significantly decrease the risk to diabetes 
patients for what is called microvascular disease, which is 
blindness, amputation, renal failure, as well as cardiovascular 
disease. Every single oral anti-diabetes medicine that is today 
approved in the United States by the FDA, including two 
medicines approved last year, have been approved on those 
    Mr. Issa. So essentially by definition, for the FDA to 
approve, your efficacy has already been established and that is 
a really unfortunate statement, since it flies in the face of 
the approval process, isn't that true?
    Mr. Slaoui. That is absolutely true. I would like to add, 
Congressman, that not only is Avandia effective, it is actually 
superior to the most widely used medicines. It as, as I said, 
30 percent and 60 percent superior.
    Mr. Issa. I have been commenting on this being a political 
process. And I am not going to back away from that, because I 
think unfortunately we are playing science here when in fact we 
shouldn't be.
    Let me just ask you one question. How do you believe 
doctors and statisticians should handle meta-analysis results 
prior to receiving data from large clinical trials? We don't 
want to alarm the public unnecessarily or needlessly. But we 
also don't want to sit and let patients not have facts as soon 
as we have them. So how should this have, not only how should 
we do it in general, but how should this have been presented, 
if you don't believe it was presented appropriately by meeting 
with the majority folks behind closed doors and then in fact 
publishing without dealing with your company or with the FDA?
    Mr. Slaoui. Congressman, I would like not to comment on 
exactly what Dr. Nissen has done. I will tell you what I would 
have done, what actually GSK has done. In 2004, we knew that it 
was important for us to continuously look at the cardiovascular 
safety of Avandia. Actually as of 1999, we had a very stringent 
pharmaco-vigilance system that looks at cases of cardiovascular 
deaths or cardiovascular heart attacks, etc., to assess whether 
there is an imbalance. We have not seen such an imbalance.
    Yet there was some report in some patient population, in 
combination with the incident that was cited earlier, that 
attracted our attention to myocardial infarcts. We immediately 
ran a meta-analysis ourselves. However, we knew exactly what we 
were dealing with. These are hypothesis generating 
technologies, methodologies. These are not fact-establishing 
    So we did that analysis and we immediately came with 
another scientific strategy, which was a real life 
epidemiological study on 33,000 patients that has shown 
absolutely no increased risk. We communicated both information 
to the agency and I think we did the right thing.
    Mr. Issa. Now, GlaxoSmithKline, I don't want to get into 
the secret work you are doing, but I am assuming with TZD 
having, we believe, a side effect, in other words, that it can 
have secondary effects as a class, not your drug but all the 
drugs, wouldn't it be reasonable, and say yes if you can, that 
you are working on the next generation that is going to reduce 
that either by changing the basic class of drug or by reducing 
the tendency of TZDs to have those potential side effect, isn't 
that true?
    Mr. Slaoui. Congressman, ourselves as well as many other 
companies have and continue to work on second generations of 
    Mr. Issa. OK, now, there has been a lot of talk about 
statistics. But if in fact this study was normalized for the 
fact that TZDs all have a certain higher risk, at least 
anecdotally, it is believed that they tend to, that you get a 
good and maybe a little bad, if it had been reduced for that, 
wouldn't in fact the study have had different outputs? And I am 
only asking for one reason. Isn't it true you could have sliced 
these statistics several different ways to get much less 
alarming and yet equally accurate statistics?
    Mr. Slaoui. Congressman, meta-analyses are as good as the 
studies you put into them. The studies that we, the FDA and Dr. 
Nissen have put into the meta-analyses, the raw materials, if 
you wish, on which the technology acts, were not designed to 
look for cardiovascular events. You have heard experts here 
talking about adjudication of cases. The cases were not 
    So the starting material, the raw material, is not designed 
for the question that is being asked. The right way to ask the 
question, Congressman, are prospective controlled large 
studies. We have three of them. The three studies do not show a 
significant increase in cardiovascular events. We think that is 
very clear evidence and we seriously look forward to the 
discussion of the FDA advisory committee on the 30th to have an 
in-depth scientific debate around this.
    Mr. Issa. I thank you for that conclusive answer.
    Chairman Waxman. The gentleman's time has expired.
    Mr. McHenry, I will recognize you now for 5 minutes.
    Mr. McHenry. I appreciate the chairman recognizing me.
    I have actually one question to begin with. I know GSK was 
one of the first pharmaceutical companies, I believe the first 
pharmaceutical company to put the company's clinical, to 
actually publicly distribute the clinical trial register, is 
that correct?
    Mr. Slaoui. That is correct, yes.
    Mr. McHenry. And there are some other companies that are 
now following suit. But can you describe what this means for 
patient safety and what this really means for public access?
    Mr. Slaoui. Congressman, it is actually very easy to access 
our clinical trial register. You just need to remember the name 
of the company, GSK, and you put dot com next to it. I am 
disappointed that some may have taken a long time to reach that 
    When you get onto our clinical trial register, you can 
click on the name of a medicine and that takes you to every 
single clinical trial that has been completed, whether it was a 
positive outcome or a negative outcome. The trial is summarized 
there and you can have all the information. I think what this 
means is full transparency. We do not withhold any information 
on a completed study.
    Mr. McHenry. I also know that we have disclaimers on all, 
there are disclaimers available for all prescription medicine. 
And it describes specifically what the manufacturer has found 
in the clinical trials and the research. And Avandia, beginning 
1999, Avandia's label stated it was not indicated for patients 
with moderate or severe symptoms of heart failure.
    Now, that was out of what was derived through your clinical 
trials, is that not correct?
    Mr. Slaoui. That is correct, sir.
    Mr. McHenry. And that was available to the FDA before they 
allowed GSK to take it to the market, is that correct?
    Mr. Slaoui. Absolutely. And discussed very clearly and it 
was a known effect of the whole class of medicines called TZDs.
    Mr. McHenry. I think a larger question here today is beyond 
that. There are short-term studies and long-term studies. GSK 
is very involved through third party sources, I believe, being 
a North Carolina company, I try to pay attention to what Glaxo 
has been doing. But the long-term study about the effectiveness 
and what medicines can do to reduce diabetes. Can you talk 
about some of the data and the difference between a long-term 
study and a short-term study?
    Mr. Slaoui. Yes. Short-term studies, usually lasting about 
6 months observation period, usually allow you to have a very 
thorough and clear assessment of what has been called the 
surrogate marker here for the control of the level of blood 
glucose. Long-term studies allow you to look at somewhat more 
of the clinical events.
    Diabetes is a very long-term chronic disease. It takes 10 
years, 15 years, 20 years, as the expert had said earlier, for 
all the clinical outcomes to unfold. Running a study for 20 
years is simply impractical, and those can be large population 
studies, not clinical trials.
    So we elected to run trials over a period of 3 or 4 years 
that, for instance, one trial was, when you take a diabetes 
medicine, in fact you are condemned to fail on your medicine, 
because your diabetes evolves and all of a sudden your medicine 
doesn't work any more. So you run a trial, we ask, does Avandia 
allow diabetes patients to succeed controlling their glucose 
levels for a longer period of time than all other medicines. 
That is where Avandia was shown to be 30 percent or 60 percent 
better than the other medicine. There is another study where 
people that are going to develop diabetes can be identified, 
and within a year or two you will become a diabetic. When 
tested in this setting, Avandia was shown to prevent 60 percent 
the development of diabetes in such-called pre-diabetes 
    So Avandia has significant public health impact and 
clinical advantages, above and beyond the advantages of the 
other available oral anti-diabetes medicines.
    Mr. McHenry. Additionally, talk about clinical trials. 
Because that is something that GSK, you outsource to a third 
party for verification of your research, do you not?
    Mr. Slaoui. Yes. Actually, when we run the large clinical 
study, we have what we call a steering committee of 
investigators, who are totally independent from GSK, could be 
Dr. Nissen or Dr. Buse, who control the clinical study, control 
the communication around the clinical trial. We also have what 
we call an independent drug safety monitoring board. This is a 
group of experts, again, physician scientists, who look at the 
safety of the patients in the clinical study. And if they see 
an imbalance in any event, they actually have the authority to 
stop the study.
    Every one of our studies has a BSNB. None of the BSNBs who 
have all been informed of all the data we are discussing have 
decided or elected to stop or in any way, shape or form impact 
the course of the studies.
    Mr. McHenry. Thank you for your testimony.
    Chairman Waxman. The gentleman's time is expired.
    I want to ask you a few questions, if I might. Dr. Slaoui, 
we are not here to make the scientific determination of whether 
Avandia makes patients healthier or whether it harms them. That 
is the job of the FDA. Hopefully the new data that you have 
generated will go to the FDA's advisory committee that is going 
to be convened to address this issue and help them.
    But what I am interested in is why it took 8 years after 
Avandia was approved for market that doctors and their patients 
still don't have a clear answer. Now, a major reason we don't 
have the data has been that there is no large, adequately 
designed post-marketing study of whether Avandia increases or 
reduces the risk of heart attack in patients with diabetes. 
ADOPT, the study ADOPT was a post-marketing study that your 
company conducted. And it was not designed to answer these 
    Can you help us understand why, despite the recommendations 
of the FDA's medical reviewer, ADOPT was not designed to 
address the reviewer's concerns about deleterious long-term 
effects on the heart?
    Mr. Slaoui. Certainly, Congressman. I think as the experts 
from the FDA have clearly explained to this committee, and I 
will clarify it further, a clinical trial, in the design, 
addresses more than one question. The questions that the ADOPT 
study addressed were several, of which four very specifically 
were safety questions. At the time Avandia was approved, 
hepatic failure was a very important concern.
    Chairman Waxman. So it wasn't a study just on heart 
disease, it involved other issues? That is what Dr. von 
Eschenbach told us. Do you agree with that?
    Mr. Slaoui. Yes.
    Chairman Waxman. And as a result of that study, did you 
have enough information to tell you specifically on the heart 
attack question that there was no additional risk?
    Mr. Slaoui. I will share with you the data, Congressman, 
because everybody needs to hear it. This study had 4,400 and 
some patients included into it. There were 24 cases of heart 
attacks in the Avandia group and 20 cases in the Metformin 
group, the control medication. These are 4 out of 4,400 
patients treated with--this a 4 individual difference. The 
reason we conclude that this is not a demonstration, it is a 
statistical methodology, is because the number of events is so 
small that we cannot conclude.
    Chairman Waxman. Right.
    Mr. Slaoui. Let me share with you other information, if I 
may. You know and you are aware we ran a second study, the 
RECORD study, where the primary input for cardiovascular----
    Chairman Waxman. That wasn't requested by FDA. That was 
requested by the Europeans, isn't that accurate?
    Mr. Slaoui. Yes, England.
    Chairman Waxman. And that hasn't been completed.
    Mr. Slaoui. Yes. But I have great news for diabetes 
    Chairman Waxman. I know you have some preliminary 
information. But let me ask you, because I only have limited 
time and we also have votes on the Floor, you might have heard 
the bells, in 2005 and then later in 2006, you did a meta-
study. And of course, your meta-study could be more complete 
than Dr. Nissen's, because you have information that he didn't 
    As I understand it, as a result of your 2006 meta-study, 
you reported to the FDA, not you personally, but the company, 
that there was a 31 percent increased risk of heart attack and 
that was statistically significant. Is that an accurate 
    Mr. Slaoui. That is accurate. And as you have heard from 
every expert, including Dr. Nissen, meta-analyses generate 
hypotheses. They do not provide answers. We immediately acted 
on that information. We took it extremely seriously. We ran an 
epidemiological study on 33,000 patients. We analyzed the ADOPT 
and the DREAM studies. These are higher quality standards, 
scientific experimentation. When you can take a plane to 
Europe, you don't take a bus or a boat. Meta-analysis is a 
    Chairman Waxman. Dr. Nissen's study was peer-reviewed. You 
didn't have to have yours peer-reviewed. Would you be willing 
to make available to our committee the data and the information 
on the meta-studies that you did in 2006 and 2005?
    Mr. Slaoui. Congressman, I would be of course very happy. 
Actually, for your information, this data has been available in 
full as of October 2006 on our Web site. And Dr. Nissen knows 
    Chairman Waxman. OK, that is very good. He had asked you 
for some information that would have made his analysis more 
complete. Did you ever give him that information?
    Mr. Slaoui. No, sir, but I believe that this committee has 
a full report on our communication with Dr. Nissen.
    Chairman Waxman. The information on your Web site is not 
patient-level data. Will you make that available to us?
    Mr. Slaoui. We will provide that to this committee.
    Chairman Waxman. We appreciate it.
    I thank you very much for being here. I think your 
presentation was important for us to hear. We didn't have 
anybody request you to be on the second panel as opposed to the 
third panel. My staff asked you or your representatives if you 
minded being on the third panel or if you wanted to be on the 
second panel. So I would just point that out, because it is 
hard to keep up with these grievances that suddenly come up. I 
find hard to believe there is a partisan oversight 
    But we are trying to get the truth, as all Members want us 
to get. My time is up and I am going to have to leave. But I do 
want to point out that I think it was pretty shocking the way 
Dr. Buse was treated when he came in with his complaints. Did 
you, did GSK ever apologize to Dr. Buse?
    Mr. Slaoui. Dr. Buse, as he stated, made actually a mistake 
in a very balanced and good presentation that he made in 1999. 
GSK, I think appropriately, requested that the mistake be 
corrected. There was a lot of passion, as Dr. Buse expressed at 
the time, on his side and on the side of the scientists which 
were involved----
    Chairman Waxman. He has described intimidations. He was 
going to have to personally pay the $4 billion in drop in stock 
prices, that his university was going to be complained, the 
department was going to get complaints from the company. It 
sounded like real intimidation. You heard what he had to say, 
didn't you?
    Mr. Slaoui. I know the person that Dr. Buse was referring 
to. That person was my boss for the last 4 years, I succeeded 
him in this role.
    Chairman Waxman. Who was?
    Mr. Slaoui. Dr. Yamada, who is a world-renowned scientist 
and currently dedicating his life to the Bill and Melinda Gates 
Foundation to help children and patients in the developing 
world. He is passionate about his work. He dedicated his life 
to developing drugs. And as scientists, they had quite a hefty 
debate and I probably would not have done it the same way. We 
regret that Dr. Buse felt pressured, absolutely.
    Chairman Waxman. Thank you.
    Well, I appreciate your being here. Your testimony 
concludes our hearing, so we stand adjourned.
    [Whereupon, at 2:10 p.m., the committee was adjourned.]
    [The prepared statement of Hon. Peter Welch follows:]
    [GRAPHIC] [TIFF OMITTED] 44429.059
    [GRAPHIC] [TIFF OMITTED] 44429.060