FDA “Mea Culpa” Part of a Cautionary Tale | Page 2
Published Online: September 25, 2013
Tracey L. Regan
Three years later after Nissen’s report, the FDA announced it would restrict use of Avandia to patients unable to control their diabetes on other medications, ultimately requiring them to enroll in the agency’s Risk Evaluation and Mitigation Strategy program in order to receive it.2,3 “Steve Nissen’s paper came as a complete surprise to many of us who thought the opposite. He said the drug caused heart attacks. And while the medical community was given the impression with his meta-analysis that the drug was dangerous, to be fair, the meta-analysis was incomplete and heavily criticized by statisticians, but that didn’t matter in the ensuing public controversy. Things became politicized, and the debate was no longer based on science,” he recalled.
“The FDA responded by making the ‘safe’ move in sharply restricting Avandia and that killed the drug. The perception was so poor that patients were asking to be taken off it. As a doctor, if you tried to reassure them, you put yourself in a bad position. Lawyers were out there advertising, saying, ‘If you took Avandia, call us.’ I stopped prescribing it.” Fleming described Avandia’s situation as “almost unique along the spectrum of what happens in drug development and regulation.”
Nissen took a look at the publically available Avandia trial data and added up the cardiovascular events associated with Avandia and the control treatment. This approach is called a meta-analysis and is generally regarded as “hypothesisgenerating, never definitive because of some significant limitations,” Fleming said. “There was nothing wrong with Nissen performing the meta-analysis or even the Journal publishing it, though it rarely publishes meta-analyses. The problem was with Nissen’s over-reaching conclusions and the editors allowing the title of the article itself, which implied both definitiveness and that higher death rates were caused by Avandia,” he said.
“Nissen went on to publish conclusions that the closely related and competing drug, pioglitazone (Actos), provided cardiovascular benefits, making Avandia even more untenable.” Multiple FDA advisory panels were called—including one the resulted in the FDA “guidance,” published in record time, that cardiovascular safety trials would have to be completed before any therapy for T2DM could be approved. This requirement suddenly added years and perhaps as much as $500 million to the cost of developing a drug for T2DM, Fleming said.. The controversy around Avandia itself escalated. In the wake of the Nissen paper, at FDA there was a storm of additional reviews and different opinions reached about Avandia.
“In the FDA drug review process, data, analyses, and interpretations are verified by a host of well-qualified professionals. The FDA reviews of Avandia went beyond this standard approval process. It added not 1 but 2 rounds of adjudicating which patients in the Avandia RECORD trial had serious cardiovascular events. Event adjudication is generally the responsibility of an expert panel independent of the FDA and the company (though the company selects the panel). Adjudication during the initial review process may be done when there is a particular reason. What is more unusual is to have the adjudication process come after the drug approval. Even rarer, if not unique, is to have a second adjudication process as occurred with FDA’s commissioning of the Duke Clinical Research Institute. The Duke report absolved Avandia from showing increased cardiovascular risk in the RECORD trial. This undoubtedly led to FDA’s soul-searching.” he said.
Since then, however, new health concerns about the entire class of TZDs have emerged, making the debate over cardiovascular risk essentially a moot point. “There is an even larger story about this class: it’s going away and not because of the CV issue. Over the past several years, data have emerged that point to increased risks of bladder cancer and osteoporosis associated with pioglitazone and ongoing concern about fluid retention caused by TZDs, which could worsen congestive heart failure, TZDs have already been removed from the market in Germany and France,” Fleming said.
“Over the past decade there has been a lot of progress on the drug review and the clinical development process. One of the key lessons learned is that long-term clinical outcomes do need to be verified and not just assumed. Rather than just relying on treating a number such as blood sugar, we now more often look at the longer term measure of a drug in preventing complications of the disease,” he said.
“We won’t see TZDs any more, although perhaps for the wrong reasons,” Ovalle commented. “They are gone and we have moved on.” He said that TZDs have been replaced by new drugs that are working well, including GLP-1 (glucagon-like peptide-1) receptor agonists, DPP-IV (dipeptidyl peptidase IV) inhibitors, and most recently, SGLT-2 (sodium-glucose transport 2) inhibitors, “with no weight gain or edema, and no bad press.” The new drugs accomplish some of what the insulin sensitizers did but through different mechanisms, he noted, including through weight loss. “GLP-1 agonists are drugs that are focused on anti-hyperglycemia, and a nice side effect is that they help people lose weight. Another class of drugs, SGLT-2 inhibitors, also look like good drugs but these are new and we always have to be careful with new drugs as there may be things we don’t know about them yet,” he said.
“In general, when I prescribe a newer drug like any of these, I now say to patients that I think it’s a safe drug, but I now have to include a disclaimer on the possibility of yet unknown side effects. That’s something I never did before.”
Ovalle said that Avandia’s dramatic regulatory reversal, among other controversies in recent years, has left him with a somewhat jaundiced view of the agency’s review process. “I think the FDA is heavily influenced by the media. Regulators probably feel under pressure to protect their jobs,” he said. “On the other hand, people who sit on the FDA’s advisory panels often have close ties to the pharmaceutical industry and to particular drugs, and disclosure of these conflicts of interest is probably not sufficient.”
PDF is available on the last page.