I’m not an academic. Never will be.
Dr. Allan M. Block
I’m also a crappy statistician. Neither my university nor medical school required statistics classes, so I never really learned them. In medicine you pick up an idea of how to interpret them as part of the job, but I’m certainly not a pro with numbers.
Which brings me to the word of the day, Aduhelm, AKA aducanumab.
A lot of drugs have come and gone in the 30 years since my medical school pharmacology class, but very few with this one’s degree of uncertainty.
Clearly its mechanism works: It removes amyloid from the brain. I don’t think anyone will argue that. But the real question is whether this translates into actual clinical benefit.
The water is murky here, and even its most ardent supporters admit the evidence isn’t exactly overwhelming. To some extent the approval basically puts it in a huge open-label clinical trial, with the Food and Drug Administration saying that it will be withdrawn if success isn’t seen in follow-up studies.
I’m not a statistics person, but I understand that, when numbers are marginal, they can be spun to mean whatever someone wants them to mean. And the stakes here, both medically and financially, are pretty high.
Alzheimer’s disease, unquestionably, is a devastating illness. The best treatments we have for it are modest at best. The demand for new treatments is huge.
But “new” doesn’t mean the same as “effective.” This is where the statistics, and their supporters and detractors, come in.
Patients and their families aren’t (usually) doctors. They want a treatment that’s both effective and reasonably safe, especially for a disease where a tragic prognosis is well established. With this drug (and similar ones in development) we face a balance between uncertain benefits and a clear risk of amyloid-related imaging abnormalities. The best we can do is explain these vagaries to people so they understand the uncertainties involved.
Perhaps more troubling is the possibility lurking in the background: The amyloid comes out, but the prognosis doesn’t improve. This brings us to the possibility (already voiced in journals) that the whole amyloid theory is wrong, and we’ve spent all this time and money chasing the wrong villain. As Morpheus, in The Matrix, implies, our whole reality on this may not be real.
Regrettably, in science (and medicine is a science) the only way to find out what works and what doesn’t is through trial and error. Computer modeling can take us only so far.
But if it (and similar agents) fail in the general population, then it may be time to accept that we’re chasing the wrong bad guy.
That’s what data and statistics do.
Dr. Block has a solo neurology practice in Scottsdale, Ariz.