The buzz among investors is that a new Food & Drug Administration draft guidance on diabetes drug development makes an already tough regulatory standard even tougher.
An article by Forbes summarizes the reaction, describing the guidance as proposing “tougher standards for how and when diabetes drugs will be tested for risks to the heart” in response to concerns raised by the Avandia controversy. It comes complete with comments from Cleveland Clinic cardiologist and Avandia meta-analyzer Steve Nissen praising the guidance as a step in the right direction—a notion sure to reinforce the view that the standards must be very tough indeed.
It seems to us that this is making a mountain out of a molehill—or perhaps, confusing a molehill (the FDA draft guidance) with the mountain (FDA’s authority to mandate post-marketing authorities).
Here (we think) is the section of the guidance raising concerns:
“Although a recommendation to demonstrate macrovascular risk reduction premarketing may delay availability of many effective antidiabetic drugs for a progressive disease that often requires multiple drug therapy, sponsors should conduct large outcomes trials before submission of marketing applications for drugs in development that show nonclinical or clinical evidence of increasing macrovascular risk. Therapies that have not demonstrated a deleterious effect on cardiovascular outcome during extensive premarketing evaluation may need further post-approval assessment for their effects on long-term macrovascular disease.”
This is news?
FDA would certainly say it would never have approved a glucose-lowering drug with a clear signal of cardiovascular risk without demanding long-term safety trials. The agency’s critics may dispute that, citing Avandia itself as an example.
We'll leave that debate to others. The point is that whatever FDA might have done in 1999, does anyone really think the agency would approve such a drug today? That question may sound rhetorical, but it doesn’t have to be. Look not at Avandia, but at muraglitazar—the Bristol-Myers Squibb diabetes drug that the agency declined to approve in 2005 after Dr. Nissen helpfully pointed out a cardiovascular safety signal in the Phase III trial database. The agency asked Bristol to do a long term study prior to approval, and Bristol opted to discontinue development.
FDA says Nissen’s outside review had no impact on its decision with muraglitazar, and we’ll let others debate that point as well. What the agency did, though, is not debatable—nor is the fact that in doing so it set a standard for requiring more than glucose reduction as an endpoint when a drug carries a significant cardiovascular risk signal.
In any event, if the bar for type 2 diabetes had been raised, it was raised then. Bristol certainly reached that conclusion, opting to partner its Phase III DPP4 inhibitor saxagliptin with AstraZeneca to help share risk in light of the tough climate. (We’ve argued elsewhere that, while Bristol undeniably did very well financially in this deal, it may have been overestimating the regulatory risk based on its experience with muraglitazar.)
We would argue that, in fact, the diabetes guidance is a sign that the regulatory risk in type 2 diabetes is a bit lower than most investors might think.
First, lay to rest any idea that FDA put this guidance out in response to Avandia. Guidance development does not move that quickly at the agency—to say the least. In this case, FDA began drafting a diabetes drug development guideline in 1996, and presented a draft to an FDA advisory committee in 1998.
After that, the agency’s Endocrine & Metabolism Division veered off into related issues like diabetes claims for weight loss drugs, standards for diabetes prevention claims, and discussion of metabolic syndrome as a potential therapeutic indication. The new draft guidance references a different starting point: a 2004 FDA/National Institutes of Health forum on diabetes. In any event, the guidance is most definitely not, on the whole, a reaction to Avandia or even muraglitazar.
Then there is a misinterpretation of what FDA guidances do: they attempt to lay out the agency’s thinking on drug development—not change that thinking. In other words, the guidance tells you what the agency already thinks. By definition, it doesn’t raise the bar on its own.
That in turn is one reason why issuing the guidance probably lowers regulatory risk: at least sponsors can read for themselves what the agency’s thinking is—and can frame their submissions appropriately to address the issues the agency has identified.
And then there is what the guidance doesn’t do: back away from HbA1c as a fully validated surrogate endpoint. “For purposes of drug approval and labeling, final demonstration of efficacy should be based on reduction in HbA1c (i.e., HbA1c is the primary endpoint of choice, albeit a surrogate), which will support an indication of glycemic control,” the agency says. That is nice and clear.
We suspect that the real issue here has nothing to do with the guidance, and everything to do with the fundamental changes underway in the regulation of drugs prompted by the new safety law enacted in 2007.
One of the key changes in the law: FDA now has the authority to mandate post-marketing studies—and to levy fines for companies that fail to follow through. That means that the wink-wink Phase IV system in place for many years, under which sponsors of products like Avandia agreed to post-marketing commitments as a condition for approval and then failed to complete the trials. Those commitments were typically “negotiated” in the final weeks before an approval deadline, under circumstances that encouraged sponsors to agree to anything FDA requested—and which even FDA recognized seldom led to useful data to resolve legitimate scientific questions.
In other words, if FDA was really worried about a safety question, it would demand more data prior to approval.
Those days are over. Now FDA and sponsors have to agree on realistic, real-world post-marketing studies. For sponsors, that means recognizing that commitments to develop long-term outcomes data in diabetes are indeed commitments.
And for FDA, that means the agency can have the confidence to approve a new anti-diabetic without demanding long-term outcomes data up front.