This is Joseph.
Thomas Lumley has a nice piece on the newly approved biologic for treating Alzheimer's Disease:
The drug (or, technically, the ‘biologic’ since it’s an antibody) has been through a lot of ups and downs in its clinical trial history. There were two main trials that were supposed to show it was effective. They failed. A re-analysis of one of them suggested that it might actually work, at least for some patients. Normally, this would be the cue to do a confirmatory trial to see if it does actually help an identifiable group of people. And the FDA did mandate this trial — but they will let the manufacturer sell and promote the medication for nine years while the trial goes on. Given that the the market for aducanumab is conservatively estimated at tens of millions of dollars per day, and there’s only a possible downside to getting trial results, the trial is unlikely to end a day sooner than it has to; it’s not unheard of for these post-approval trials to just never recruit enough participants and drag on longer than ‘allowed’.
While this is a very cynical take, it is clear that the treatment is high burden (given intravenously), expensive ($56,000 USD/year), and with a lot of side effects. There is obviously a huge incentive for the trial results to be generated as quickly as possible, as if the drug is not effective then there are huge costs in money and side effects to be considered.
What is so puzzling about this decision is that it follows the recent responsible approach to covid-19 vaccines. One can argue whether the FDA was too conservative with vaccines in the face of a pandemic, but it did a great job in approving effective and safe vaccine products. But this is a treatment where the sponsor ended the trials early for lack of efficacy, which is not generally seen as a clear vote of confidence in a product. Even if the treatment post-hoc analysis was right by good fortune, the effect size is quite small, which is a big deal for extremely expensive treatments and requires some careful technology assessment work to decide how to handle this treatment.
So while Thomas Lumley is cynical here, I think it is worth considering these points very carefully as a part of any medical system, US or other.
Ok, third try at getting this comment to post.
ReplyDeleteI don't think I'm cynical about either the drug company or the FDA here. I think there is terrible systems design in setting up the incentives.
Usually the FDA is good at incentives: it's more profitable to make a drug that works than one that doesn't, and you'll get easier approval with good evidence than shaky evidence.
Here, the FDA is providing an incentive for the vendor to run a trial slowly. They lose truckloads of money if the trial finishes earlier than necessary and is negative. Everyone else benefits if the trial finishes earlier than necessary and is negative. You don't need to be cynical to think that's a bad setup -- "lead us not into temptation"