Just when you think everything that can be said about Avandia has been said, along comes an impromptu, verbal sparring match between high-level officials of the U.S. and UK drug regulatory agencies over the role and value of hemoglobin A1c reduction in diabetes drug approval.
In one corner: Sir Alasdair Breckenridge, chairman of the UK’s Medicines and Healthcare products Regulatory Agency, better known as the MHRA.
And in the other corner: FDA Center for Drug Evaluation and Research Director Janet Woodcock.
The setting: the Third Annual Risk Management and Drug Safety Summit in Washington, D.C. on Oct. 18.
Woodcock, who was the first presenter at the meeting, spoke about CDER’s efforts to improve risk management and drug safety since passage of the FDA Amendments Act. She was followed at the podium by Breckenridge, who presented the European perspective on risk management and a pharmacovigilance “tool kit” for assessing and mitigating drug risks.
Near the end of his presentation, Breckenridge turned to the recent regulatory decisions on Avandia, GlaxoSmithKline’s beleaguered thiazolidinedione. Even though FDA and the European Medicines Agency took different regulatory paths – with FDA restricting distribution under a Risk Evaluation and Mitigation Strategy, and EMA suspending rosiglitazone’s license – Breckenridge stressed the extensive collaboration between the two agencies that culminated in simultaneous announcements on Sept. 23.
“If you think about the difference between what Europe has done and what the U.S. has done, in fact I would suggest there was very little difference indeed, and it was an example of regulatory authorities working together in a global manner,” he said. ("The Pink Sheet" offers an analysis of why they diverged in their final judgment.)
Following these glowing remarks about alignment among regulators on both sides of the Atlantic, Breckenridge took the Avandia post-mortem a step further, and perhaps one too far for Woodcock.
“The question I’ve asked myself is if Avandia came through for licensing today, with the information we had, what should be done, what would we have done? How has regulation advanced? Well firstly, it wouldn’t have been approved for efficacy on a surrogate marker [HbA1c]. That would not be accepted,” he said. Some in industry concur that there is now a higher hurdle, at least commercially, for diabetes products.
Fortunately for the audience, Woodcock hung around after her presentation to hear Breckenridge’s speech, and during a question and answer session, while still sitting in the audience, Woodcock pounced on the British knight’s skepticism toward HbA1c. Here is an abbreviated transcript of the exchange, along with some first-hand, editorial observations noted in brackets:
Woodcock: “If you’re not going to use hemoglobin A1c or serum glucose … what are you going to use for efficacy in diabetes? No drug in type 2 diabetes has ever been shown to improve cardiovascular outcomes.”
Breckenridge: “I believe this illustrates the problem with antidiabetic drugs. I believe it’s going to be increasingly difficult to develop any drug for diabetes which has got a suggestion that Avandia did have, and I think drugs like Avandia are going to die at a much earlier stage and be killed at a much earlier stage than developed.”
Woodcock: “I would say the question with rosiglitazone is a safety issue, it’s not an efficacy issue … I think hemoglobin A1c is more than a surrogate … I date as an internist from the era when people walked around with untreated type 2 diabetes. They hit my emergency room they were in hyperosmolar coma. That’s a life-threatening disease. Or they had severe invasive soft tissue infection with gram negative organisms, or they were dehydrated and had blurry vision and CNS issues.”
Breckenridge: [apparently attempting to explain that not all type 2 diabetics are in such dire straits] “I can remember as well, and I’m not sort of swapping stories with you, but patients with type 2 diabetes are the rather large ladies who you see walking around in the United Kingdom and I’m afraid Washington as well.” [disapproving murmurs from the audience]
Woodcock: “But if you go untreated long enough with type 2 diabetes that’s what you get into. It’s a progressive disease. So the idea that you don’t need treatments for type 2 diabetes I think is an incorrect …. ”
Breckenridge: “I’m not suggesting that ….”
Woodcock: “You will get renal failure, you’ll get amputation … It’s a symptomatic disease. People have studied this and they’ve looked at central nervous system effects of hyperglycemia … There are people walking around with blood sugar 300, 400 and so on. That is not good for you, acutely. And sub-acutely, glycemic control has been shown to be correlated with progression of retinopathy, renal failure and so forth, and neuropathy to some extent. [By now standing, holding the microphone and looking as comfortable as a talk show host on a TV production set] So I would take issue with the fact that hemoglobin A1c is a bad surrogate. I think it’s a very good surrogate for efficacy. I don’t think it tells you anything about safety of a drug just like most surrogates for efficacy.”
Breckendridge: “I’m afraid I disagree with you there, Janet. I think by the definition of a surrogate, hemoglobin A1c fulfills all the criteria … and the point I was trying to make was that if you take, in the development of a drug in the latter phase of the drug, and you had a drug which was effective by affecting the surrogate, but it had some other not just potential but huge changes, big changes which were known at the time in a possible adverse event which diabetics are already prone to, the manufacturers, I would suggest, would have a very, very careful look at that before continuing with its development.”
Woodcock: “I don’t think we’re in disagreement, I’m simply saying I thought the earlier definitions [of a surrogate] were mainly done by statisticians, like Prentiss and others … that it should contain all outcomes. That’s completely naive from a biological perspective, because you may perfectly control the disease and kill people from something else. It’s unrelated to the pathway of the disease. So I think expectation that a surrogate for efficacy would take care of your safety evaluation is unrealistic, and I think we’re saying the same thing, which is for chronic diseases there’s going to have to be a much more thorough safety evaluation, it’s longer term, includes more patients, looks for more outcomes than we have traditionally had.”
Breckenridge: “And I think diabetes is an especially difficult case for the reason I described. If you’ve got a disease whose natural history is to develop vascular disease anyway, then a drug which is going to influence that in any kind of adverse way is not good news.”
Woodcock: “The sulfonylureas have long had a warning in the United States for cardiovascular disease because the only time that was studied long-term there was a signal.”
Breckenridge: “And so do the thiazide diuretics, too.”
Woodcock: [laughing] “So there’s a lot of things we don’t know.”
During a break in the meeting later in the day, Breckenridge was overheard describing Woodcock as “feisty.”
Feisty? Perhaps. But definitely defensive of the view strongly held within CDER’s Office of New Drugs that HbA1C reduction, not cardiovascular benefit, is an appropriate efficacy endpoint for new antidiabetics. Woodcok's eagerness to enter the ring on the issue is especially interesting given that outcomes data is now essentially required to demonstrate the safety of the products.
– Sue SutterPhoto "Natalya" by flickr user Snerkie used under Creative Commons License.