» Drug, Devices, and Supplements

» Physician Accountability

» Consumer Product Safety

» Worker Safety

» Health Care Delivery

» Auto and Truck Safety

» Global Access to Medicines

» Infant Formula Marketing


More Information on rosiglitazone (Avandia)

Testimony on Rosiglitazone (Avandia)

July 30, 2007

FDA Endocrine Metabolic Drugs Advisory Committee Meeting on Rosiglitazone
Testimony of Sidney Wolfe M.D., Elizabeth Barbehenn Ph.D. and Ben Wolpaw, Health Research Group of Public Citizen

I will focus on the fourth question to the committee: Does the overall risk-benefit profile of rosiglitazone (Avandia) support its continued marketing in the US?

The evidence presented today concerning the increased risk of ischemic heart disease including myocardial infarctions appears to justify the removal of this drug from the market, but due to the ubiquitous nature of PPAR gamma receptor sites in so many parts of the body, it is hardly surprising that there are many other significant kinds of damage this drug is causing to patients.

Preclinical (Pre-Approval) and Post-Approval Evidence of Cardiac Toxicity

In a 1999 FDA pharmacology review of animal toxicity, "rosiglitazone produced various toxicities such as left atrial thrombosis, hydrothorax, cardio hypertrophy and elevations of hepatic enzymes in the high dose group. In this reviewer's opinion, it is not possible to anticipate potential human toxicities" and stated that "these findings appear as long term clinical concern." The final recommendation states "Pharmacology recommends not to approve rosiglitazone for the proposed indication for long-term human use."

In dogs, at doses only 1.2 times higher than the human dose, there was evidence of significant cardiac hypertrophy after 26 weeks.

FDA Pharmacology Review: April 26, 1999

In her review of animal and human evidence of cardiac toxicity of PPAR gamma agonists such as rosiglitazone last year, former FDA Pharmacologist Dr. Jeri El-Hage stated:

  • There was fluid accumulation in all species (mouse, rat, dog, rabbit, monkey, human).
  • Fluid accumulation leads to weight gain, edema, cardiac hypertrophy with resultant heart failure in all species.
  • Drug-induced heart failure and death were observed with chronic treatment (>6 months in animals and man).
  • In people, the longer a patient was on a PPAR gamma, the lower the dose needed to produce edema or CHF.

In a recently published meta-analysis of randomized trials with rosiglitazone (2 trials) and pioglitazone (1 trial) co-authored by Dr. Curt Furberg, the increased risk of heart failure compared to placebo was 2.1 (95% CI of 1.08-4.08)

Finally, in FDA adverse reaction reports filed since marketing began through the end of last year, there were 698 cases of heart failure reported with Avandia compared with 39 for the older diabetes drug, Glucotrol. Adjusting for differences in the number of prescriptions for the two drugs, the rate of heart failure reports for Avandia compared with Glucotrol was 15.2 times higher.

FDA AERS Reports: 5/25/99 - 12/31/06 

Preclinical (Pre-Approval) and Post-Approval Evidence of Liver Toxicity

In the pre-clinical 1999 FDA pharmacology review, evidence of liver toxicity was seen in dogs at a dose only 1.2 times higher than the human dose that included ALT, AST and LDH in males that increased to being statistically significant in males and females (ALT only) by 25 weeks in the next higher dose.

In eight published case reports of rosiglitazone-caused liver toxicity, with human exposures ranging from two weeks to 15 months, for the seven patients for whom follow-up data was available, five recovered within two weeks to four months after the drug was stopped. Two of the patients died. Bilirubin levels in these patients, consistent with the similar hepatocellular toxicity seen with troglitazone (Rezulin), ranged from 2.9 to 22.3 with an average of 10.5.

Published Case Reports of Hepatic Toxicity from Rosiglitazone

Finally, in FDA adverse reaction reports filed since marketing began through the end of last year, there were 594 cases of hepatic toxicity (including 122 cases of liver failure) reported with Avandia compared with 53 cases of toxicity (and only seven cases of liver failure) for Glucotrol. Adjusting for differences in the number of prescriptions for the two drugs, the increased adjusted rate of reports for Avandia compared with Glucotrol was 9.5 times higher than Glucotrol for all reports of hepatic toxicity and 14.8 times higher for hepatic failure.

There were 46 reports of deaths from hepatic failure with Avandia, and six with Glucotrol, for an adjusted rate of 6.5 times higher with Avandia.

FDA AERS Liver Toxicity: 5/25/99 - 12/31/06

FDA AERS Death Reports: 5/25/99 - 12/31/06 

Post Approval Evidence of Increased Bone Fractures in Women

The recently completed ADOPT study found statistically significant increases in total and lower limb fractures in women using Avandia compared to those using either glyburide or metformin.

Fractures in Women Data from ADOPT Trial

Fractures in Women Data from ADOPT Trial (cont'd) 

Pre-Approval and Post-Approval Clinical Evidence of Anemia

  • In pre-approval clinical trials, anemia was reported in 1.9% of patients receiving Avandia as monotherapy compared to 0.7% on placebo.
  • Severe anemia (male: Hct less than 31; female: Hct less than 28) was seen in clinical trials in 9/2,121 patients on rosiglitazone, 0/485 patients given placebo. A lesser degree of anemia was seen in many more patients.
  • Anemia was seen most commonly in combination therapy with rosiglitazone plus metformin (7.1%) compared to those receiving placebo plus metformin (2.2%).

Post-approval adverse reaction reports included 407 cases of anemia in patients using Avandia and 26 in patients using Glucotrol for an adjusted reporting rate of 13.3 times higher for Avandia.

In patients already damaged by heart failure or other cardiac risks of this drug, the addition of anemia can significantly worsen their clinical condition.

Macular Edema

Although one of the major worries as far as microvascular complications of diabetes is visual impairment, not only does Avandia lack evidence of improving this situation but, through a mechanism related to heart failure and fluid accumulation, many patients have developed macular edema—swelling in the critical macular portion of the retina—with Avandia with concomitant, usually reversible, damage to their vision.

Post-approval adverse reaction reports included 83 cases of macular edema in patients using Avandia and two in patients using Glucotrol, for an adjusted reporting rate of 35.3 times higher for Avandia.


In a recent editorial in the New England Journal of Medicine, Dr. David Nathan, a diabetes expert from the Mass General Hospital wrote:

In theory, newer classes of antidiabetes medications might be welcome additions to the existing armamentarium; however, those that have been developed recently are generally no more potent, and often less effective in lowering glycemia, than the three oldest classes (insulin, the sulfonylureas, and the biguanides), all of which are more than 50 years old.

Moreover, the newer classes are uniformly more expensive and are associated with adverse effects — some that are shared by the older drugs, but others that are new.

He added that “The failure of clinicians and their patients with diabetes to implement currently available interventions aggressively and effectively is, I suspect, the major barrier to good care. This problem will not be fixed by making more medications available.”

Our answer to the question, "Does the overall risk-benefit profile of Avandia support its continued marketing in the US?" is clearly no.

There is no evidence of any uniquely beneficial clinical outcome for Avandia and growing evidence in multiple organ systems (cardiac, liver, bone, bone marrow) of unique risks. If this drug were up for approval today, based on what is now known, it would be summarily rejected. There should not be a double standard for removing it from the market.

Public Citizen is currently preparing a petition to the FDA to ban Avandia. 

Copyright © 2017 Public Citizen. Some rights reserved. Non-commercial use of text and images in which Public Citizen holds the copyright is permitted, with attribution, under the terms and conditions of a Creative Commons License. This Web site is shared by Public Citizen Inc. and Public Citizen Foundation. Learn More about the distinction between these two components of Public Citizen.

Public Citizen, Inc. and Public Citizen Foundation


You can support the fight for greater government and corporate accountability through a donation to either Public Citizen, Inc., or Public Citizen Foundation, Inc.

Public Citizen lobbies Congress and federal agencies to advance Public Citizen’s mission of advancing government and corporate accountability. When you make a contribution to Public Citizen, you become a member of Public Citizen, showing your support and entitling you to benefits such as Public Citizen News. Contributions to Public Citizen are not tax-deductible.

Public Citizen Foundation focuses on research, public education, and litigation in support of our mission. By law, the Foundation can engage in only very limited lobbying. Contributions to Public Citizen Foundation are tax-deductible.