Recently there have been a spate of reports carried in trade media that Commissioner Califf would prefer to limit the voting mechanism customarily included as a key part of FDA Advisory Committee meetings. Traditionally, committee votes can be taken on a range of topics during the course of the meeting. These votes are usually interspersed with questions that the agency poses to the committee for the purpose of eliciting discussion about specific aspects in relation to the approval of a new drug application. Customarily this culminates in the ultimate question whether the committee believes that the benefit/risk ratio is such that approval is warranted. Now the Commissioner appears to be suggesting that the focus be put on discussion with less, if any, inclusion of what he refers to in one report as a “gladiator” vote of approval.
The notion to limit voting seems a bit contrary to the character of the current environment – one which would seem to beg for more transparency rather than less. In November 2020 FDA held a meeting of the Peripheral and Central Nervous System Drugs Advisory Committee to consider a new drug application for an investigational treatment – aducanumab – for Alzheimer’s. The standing voting members of the committee (7) were joined by 4 additional temporary voting members. FDA posed a total of 8 questions for the committee’s input into the agency’s decision, half of which were discussion questions and half of which were voting questions. It went something like this:
- The first question posed to the committee in that meeting centered on one of the clinical studies that had been submitted in the application and asked whether it provided strong evidence of efficacy and the committee overwhelming voted that it did not (1 Yes, 8 No, 2 Uncertain).
- The second voting question focused on a different study, asking the same point regarding efficacy and again the vote was overwhelmingly negative (0 Yes, 7 No, 4 Uncertain).
- Asked whether there was strong evidence of a pharmacodynamic effect related to Alzheimer’s pathophysiology, the outcome was a little more balanced, but still not an endorsement (5 Yes, 0 No, 6 Uncertain).
- Returning to the study voted on in the first voting question, the agency then asked whether it was reasonable to consider the study as evidence of efficacy in the treatment of Alzheimer’s and the vote was once again overwhelming (0 Yes, 10 No, 1 Uncertain).
- Several months later, FDA approved the NDA under accelerated approval provisions, a possibility that was not something discussed at the meeting.
The results were disastrous from nearly any way you look at it. FDA pointedly ignored the overwhelming sentiment of the panel, resulting in the unprecedented resignation of 3 of the advisory committee members. CMS issued a national coverage decision to limit coverage for the drug by implementing a policy that it would only cover monoclonal antibodies approved under accelerated approval for Alzheimer’s for those patients who had been in the clinical trials, pretty much guaranteeing that uptake would be limited. Confidence in the approval process was shaken and FDA’s reputation took a blow. The outcome fostered confusion among patients and physicians. And among the many negative side effects of the decision, there was a Congressional investigation highly critical of FDA’s lack of transparency in the process, among other things.
The Aduhelm case study is an example of why voting is in fact essential.
In referring to the voting tradition in the meetings as reported in trade media, the Commissioner stated words to the effect that the advisory committees exist to give the FDA advice and that approval decisions should be left up to full time civil servants. In the same article he’s quoted as saying that these meetings are like a democracy – where there is not always agreement – and therefore sometimes messy. But while the committees are there to provide FDA advice – and ultimately the decision does rest with FDA, AdComms also serve another very important purpose.
With votes, we get not only the qualitative insight yielded to us by the nature of the discussion, but a quantitative understanding as well. It makes the outcome concrete. We are left with a good perspective on whether the committee is wholeheartedly enthusiastic, fence-sitting, or in the case of aducanumab, very much against approval. We don’t have to read the tea leaves. Votes are a fundamental component of transparency to those of us on the outside and the considerations that go into an approval decision are known and understood by everyone. Part of understanding is provided by the discussion, but a part is also the mechanism of a vote. And to the Commissioner’s point regarding “messy democracy” – in point of fact, voting is a fundamental component of the democracy to which he refers.
Commissioner Califf’s thinking on this has been through various reported conversations and has not been embodied in a speech. The topic merits wider consideration though and a more thoughtful and contextual discussion. Hopefully that will happen.
Can we at least vote on it?