Alzheimer's disease is one of the notorious rocks of drug development. Around it are piles of debris, shipwrecks of clinical research programs large and small. But late last week, as yet another company sailed in close, something new happened.
Unfortunately, that may not mean "new" an in "good". A small company called AFFiRiS (that's really how they spell it) was testing a vaccine against beta-amyloid (an idea that has been tried numerous times before, although with the immune system you never know what'll happen next). Here's what happened next:
On June 4, AFFiRiS AG offered a smattering of results from its Phase 2 clinical trial of AD02, an active vaccine for Alzheimer’s disease derived from the company’s proprietary method of making synthetic antigens based on the Aβ peptide. At a press conference in Vienna, company scientists reported that among older people with early Alzheimer’s, a placebo group fared better than any other. Patients in this group reportedly had less cognitive decline over the course of 18 months, correlating with less hippocampal shrinkage. This group had been injected not with any Aβ-based antigen, but only with what the company calls an immunomodulator that was part of the AD02 formulation. Company scientists then named this placebo formulation AD04, and said they planned to explore options for clinical development.
Now, not every story about this actually picks up on this switcheroo. Try this one: you'd think that the company marched into the clinic with several vaccine ideas and got one of them to work. But that's not what happened.
I'm having a lot of trouble with this idea. Serendipitous discoveries there are, and this may be one of them. But I very much doubt it. The company provided no real data in their announcement - no error bars, no actual numbers. The "immunomodulator" was present in the actual vaccine formulation, but those patients (apparently) showed no effect. That Alzforum story linked above also notes that the company took two clinical rating scales, usually used separately and combined them into their own composite score. It is generally a safe bet that no one does that unless that was the only way that their results look promising.
So no, I have to disagree with the company that these results represent some sort of breakthrough. Breakthroughs in the clinic have clearly stated sample sizes, and error bars, and don't require any mixing and stirring of the numbers. Odds are excellent that this is noise. AFFiRiS (can't say I enjoy typing that) can move ahead with its mysterious immunomodulator if they like, and for the sake of Alzheimer's patients everywhere we can hope that it does something. But when a small company tries to say "Oh look, turns out our control group is actually a new Alzheimer's therapy! Isn't that great!", well, I think some skepticism is appropriate.