Well, Biogen has released more data on its Alzheimer’s antibody, aducanumab. The people (like me) who were doubtful (or worse) that they had enough to make a case for FDA approval remain doubtful. And the people (there are some) who think that it’s approvable haven’t changed their minds, either, from what I can see. Frankly, if you wanted it approved after the very limited information we had before, you’re probably not a mind-changing sort of person to start with, but that might be another topic.

As you may recall, Biogen had two Phase III trials going, Engage and Emerge, and both were halted back in March. They now say that one of these generated positive data on a re-analysis, and yesterday the company made their case that the difference was due to more patients getting the higher dose of the drug in the Emerge trial. But I honestly have a hard time seeing how this works. There was a protocol change during the studies that resulted in more patients getting the high dose, true – but after that change, sure, the high-dose patients managed to scrape over the line into statistical significance, but even the low-dose patients also showed better responses. Which makes no sense. That looks like an artifact to me; I don’t see any reason to assume that it isn’t, and I think that the burden is on Biogen to explain why it isn’t. The only way I can see making that case with a straight face is to run another trial with that hypothesis in mind from the start.

If Biogen is right, that new trial design would generate solid data that no one could brush off – a major triumph for the company and a major advance in Alzheimer’s therapy. If, though, these results are just accidental byproducts of squeezing the data too hard in a couple of halted trials, then no such signal will appear. This is why we run trials. When Biogen people stand up and make the case for aducanumab, they are (whether they are trying to or not) making the case for running a confirmatory trial. Not for getting the FDA to approve it as is.

Why not? Why not give those Alzheimer’s patients and their families hope? If it doesn’t work for Alzheimer’s, though, then aducanumab will likely do real harm to some people (all drugs have side effects) and do good to no one. And I think that the odds are better – much better – that it indeed doesn’t work for Alzheimer’s. Giving “hope” under these circumstances is, in my opinion, more like cruelty dressed up. On top of that, there is harm in making everyone pay for an expensive placebo as well. I would expect the drug to rapidly become one of the largest prescription expenditures in the country, landing especially hard on the elderly population (Medicare, the VA system, etc.) and it would be helpful, to say the least, if we did not do that on the basis of maybe-sorta-squint-hard data. Which is what we have now. It’s in the Scottish-court category of “Not Proven”.

No, the burden of proof is still on Biogen, and the way to provide that proof is to immediately start enrolling patients in a confirmatory trial. That’s what others have done under these circumstances (think Eli Lilly and solanezumab, their own anti-amyloid antibody), and by golly the confirmatory trials confirmed that these drugs didn’t work. If Biogen thinks that they can beat that track record, they should step up and do it. The company said last month that it would start re-dosing patients: have they? How many? Why shouldn’t we wait to see what happens to them? And if it’s worth going back to the people in the trial like this, isn’t it worth starting with a fresh cohort dosed in what the company thinks is the right way from the start, the way that patients would be in the real world after an FDA approval? Why not, exactly?

Here’s where I blow my chances of ever being employed by Biogen. The kinds of statements that some people at the company have been making – that aducanumab is ready for approval now, that the FDA would be remiss in not granting it, all that stuff – are to my eyes just irresponsible. Saying that you have an Alzheimer’s drug when you don’t quite know if you have one or not is wrong. As for the FDA, approving a drug when you don’t quite know if it will help anyone is wrong. And counterproductive, and a bad precedent, and expensive and all that other stuff, but first off, it is wrong.

That’s all I have to say.