The Food and Drug Administration announced Wednesday it will add its toughest type of warning to prescribing instructions for the widely sold diabetes drug Avandia instead of taking it off the market. Although a review of 42 studies of the drug earlier this year linked it to a 43 percent greater risk of heart attacks, other longer-term data did not confirm a heightened risk of heart attack, the agency said when it announced the new black box warnings.

Agency officials also said in a midday telephone press briefing that there isn’t enough evidence to indicate that Avandia carries a greater risk of heart attack or death than some other types of drugs that, like Avandia, are taken orally to help prevent complications from Type 2 diabetes. “It’s important to understand that people with diabetes need medications to prevent those complications,” said Dr. John K. Jenkins, director of the FDA Office of New Drugs, explaining the decision to keep the drug on the market despite its potential risks.

Concern about Avandia’s potential link to heart attacks spiked earlier this year when the analysis of 42 studies by Dr. Steven Nissen of the Cleveland Clinic was published in the New England Journal of Medicine (See related story, CQ HealthBeat, May 23, 2007). Nissen’s analytical method was a ‘meta-analysis,’ in which data from a number of studies are analyzed together to shed light on a particular research question, a procedure which may not be as accurate as a dedicated study designed to answer that specific question.

FDA’s black box warning calls doctors’ attention to the meta-analysis of 42 studies, but also to three other longer term studies comparing Avandia to other diabetes drugs or to a placebo. The three studies neither confirmed nor excluded the heightened risk found in the meta-analysis, the new black box warning language says. The 42 studies on average lasted 6 months and included 14,237 patients, while the 3 studies had a total of 14,067 patients and lasted an average of 41 months.

“In their entirety, the available data on the risk of myocardial ischemia are inconclusive,” the black box warning states. Myocardial ischemia refers to medical conditions including heart attacks and chest pain.

FDA said that GlaxoSmithKline has agreed to conduct a study of up to five years comparing Avandia to other diabetes drugs, including Takeda Pharmaceutical’s Actos, which hasn’t been linked to the heart attack risk and whose sales have prospered in the wake of the Avandia meta-research data. GlaxoSmithKline must submit the design for the study by the end of July 2008, and the study itself will begin by the end of November 2008, FDA officials said.

Final results aren’t due until 2014, but interim results before then could lead to regulatory actions involving Avandia and other drugs in the study. “It isn’t that we’re going to be clueless until 2014,” said Janet Woodcock, FDA’s deputy commissioner for scientific and medical programs.

FDA’s handling of Avandia has drawn strong criticism from lawmakers including Republican Senator Charles E. Grassley of Iowa and Henry A. Waxman, a Democratic Representative from California. Grassley said in May that FDA managers overruled a recommendation by agency safety officers that a black box warning be added to Avandia for a different type of cardiovascular problem: congestive heart failure. FDA did finally add the black box warning for congestive heart failure in August, however. The warning also was added to other drugs in the same class as Avandia, including Actos.

Waxman held a hearing in June highlighting the alarming results on heart attack risk from the meta-analysis. Among the questions on Capitol Hill about FDA’s handling of the drug were why the agency took so long to issue a safety alert on the findings of the meta-analysis showing a heightened heart attack risk. Data on the 42 studies had been submitted to FDA in August 2006, but the safety alert wasn’t issued until May, when the Nissen data was published in the New England Journal of Medicine (the safety alert is a less serious warning than a black box warning). But some Republicans challenged Waxman’s attack on FDA, questioning the validity of the meta-analysis technique and accusing Waxman of collaborating with Nissen to manufacture a scandal.

Asked Wednesday whether FDA’s latest action on Avandia will give fodder to critics who say the agency too often overrides the recommendations of its drug safety staff in deciding to keep risky drugs on the market, Woodcock noted that a panel of outside advisors voted 22 to 1 in July to keep the drug on the market pending a review of additional data. FDA officials and scientists then met internally on the matter, Woodcock said. She characterized the agency’s drug safety oversight board as “split” on whether Avandia should stay on the market, but then added that a majority of the board did favor continued sales. The final decision to keep Avandia on the market was the result of broad consultation, she said.

But Grassley issued a statement Wednesday expressing continued dissatisfaction with the agency. “Overall, this case is a clear example of the problem within the FDA, where the views of the office that reviews drugs after they’re on the market must play second fiddle to the position of the FDA office that approved a drug ,” he said. And Senate Finance Committee Chairman Max Baucus, D-Mont., said “its about time” FDA weighed in with a firm warning. He said the committee will closely monitor FDA’s handling of the comparison study.