When last heard from, Biogen and Eisai’s aducanumab (another amyloid-targeting antibody for Alzheimer’s) had failed in Phase III and the whole effort was being terminated. Then came Tuesday. Biogen then startled everyone by announcing that a review of the clinical data had convinced them that the drug had actually worked – or at least worked enough to be worth an FDA submission, which they say is exactly what they’re going to do next. Here are their slides (start at about #11), where they make their case.
There were two Phase III trials, ENGAGE and EMERGE, which started at nearly the same time in the fall of 2015. By June of 2017, patients had been enrolled for the futility analysis (which would read out earlier), and by June of 2018, both studies reached full enrollment. In December 2018 the last data for the futility check were collected and began to be worked up, and in March of this year the results were completed. That, by golly, is when things came to a shuddering halt, because that analysis suggested that further dosing in the treatment groups had no chance of reaching a useful endpoint.
Comes now Biogen, noting that at the time of that futility analysis, EMERGE was “trending positive” while “ENGAGE” was not. The company does not have a solid explanation for the divergence. There were, though, amendments to the trial protocols along the way (some of which allowed for higher dosing later in the trials), and they hypothesize that these could have made a difference due to the enrollment schedules. The company says that now they they have access to more data than they had at the discontinuation, they believe that the new larger dataset’s results disagree with the results of the futility analysis. The company believes that the cohort of people in the ENGAGE study who had sufficiently large exposure were trending along with the (better-looking) EMERGE data, and it’s my impression that they want to use these together to make their case for efficacy. Differences in exposure to the high-dose aducanumab, they say, “largely explain” the differences between the futility analysis and this new interpretation, and that “following consultation with the FDA” they believe that they have sufficient data for a regulatory filing. (No doubt they do, but do they have enough data for a successful approval?)
This news spilled a lot of coffee in the pre-market. BIIB shares had closed Monday at about $223 (pretty much where they were after aducanumab hit the skids in March) and opened Tuesday at $310 (pretty close to the $323 it had been before the March crash), and trading down to $282 at the close. Clearly a lot of investors suddenly believe in the drug again and are willing to put money on that proposition. Are they right?
I’m going to ignore all the data on those slides about amyloid level in the brain via PET scans and suchlike. I honestly do not care; what’s important are the measures of dementia. And while the new larger EMERGE data do look like there’s a dose-responsive effect in the primary measurement (the CDR-SB scale) and three secondary dementia scores, the ENGAGE data actually show worse than placebo for CDR-SB at the higher dose (see slide 19 in Biogen’s presentation). They’re also worse than placebo in the MMSE scoring, but both the ADAS-Cog13 and the ADAS-ADL-MCI scales show apparent dose-responsive improvement. I have no idea why four different dementia rating systems should give such different results, and to a first approximation, no one else seems to have a good idea, either. So we’ve got the EMERGE/ENGAGE difference to explain (you’ve seen Biogen’s explanation, higher dosing later in the trial), and the variable results within the ENGAGE trial itself. And who are these high-dose patients in the ENGAGE study that Biogen says line up with the EMERGE trial? Ed Silverman at Stat quotes an analyst saying that it looks like fewer than 100 patients out of the whole set. That reminds me a bit of Biogen and Eisai’s tease of “We’re seeing positive data!” during the clinical trials, which also turned out to be based on a smaller number of patients than one would want when making a call on Alzheimer’s efficacy.
Admittedly, the EMERGE data look better, but Biogen’s explanation for the differences between the two is necessarily post hoc. It makes a good story, but the FDA would be well within their rights to ask the company to prove it with another Phase III, this time run under the conditions that are hypothesized to work the best. That, to me, seems like the most likely outcome of any regulatory submission, and this could happen at several points. It could come out of an advisory committee meeting that decided that it couldn’t make a clear call based on the package as it stands, or the agency could make that call on its own regardless of the Ad-Comm. You also have the nastier possibility of a Complete Response Letter even before a thorough review of the submission, but Biogen clearly thinks that they’ll get further than that, and they’re probably right (those “discussions with the FDA” they made sure to mention).
Outright approval at the end of the process is possible, too: none of those people who piled into BIIB stock yesterday were hoping for just another Phase III trial. I think that’s only an outside chance, though. Edit: there’s another possibility: conditional approval by the FDA, with a revisit-in-X-years provision, similar to what was done with Avastin in breast cancer. I can’t say I like that prospect much. Getting that pulled back was enough of a mess, in spite of clear evidence that the drug did no good, and this would be even worse.
There are a *lot* of factors at work here, and for better or worse not all of them are scientific. I wrote about some of them while Eli Lilly was thrashing around with their own anti-amyloid, solanuzemab; here’s the short version: there is nothing that really helps for Alzheimer’s. The first actual disease-modifying therapy will very quickly show revenues of over ten billion dollars a year and rising, probably to levels we have never seen before. There are huge numbers of desperate patients and their families who will immediately demand any such drug, and who could blame them? Meanwhile, there is what I can only describe as a moral hazard for drug developers in this area: if you could only get something approved, it would become the biggest drug in the world even if it doesn’t really help people all that much. I mean. . .
My point is, if you’re going to unleash something like this, it had better be good. An Alzheimer’s drug like this has the potential to alter the entire drug-expenditure landscape in the US – every private insurance company, employer contributions, state-sponsored programs, Medicare, the VA, ACA participants, the lot. You will want clear evidence of benefit before you do that. The FDA does not consider drug pricing in its decisions, but it does have an explicit mandate to consider efficacy (and efficacy versus safety). They could very well decide that they want a clearer read on that efficacy before we start dosing millions of Alzheimer’s patients with a drug that had already been killed off by its own developers. And before we start paying for such dosing.
There’s a moral hazard for the FDA, too. The pressure will be on to approve something, anything for Alzheimer’s. It looks like aducanumab will be the first time a big (potentially) disease-modifying therapy will get as far as an actual regulatory submission – everything else has blown up before that point. Now comes the hard part: dealing with the actual data. It would be so much easier with a clear clinical victory, but that’s not what we have here. The Biogen/Eisai data are internally contradictory and very much open to argument. Argue we shall.