Aduhelm and Leqimbi were hot news a few years ago. They are both antibodies that work as anti-Alzheimer’s disease drugs by binding to and hopefully destroying amyloid beta. The hypothesis that amyloid beta is the causative agent of Alzheimer’s, and that reducing amyloid beta will lessen the disease, is known as the Amyloid Hypothesis. And while the Amyloid Hypothesis is still the most widely supported, I wonder if the failures of Aduhelm and Leqimbi to make much of a dent to Alzheimer’s disease has damaged the hypothesis somewhat.

Because think about it, the whole job of an antibody is to help your body clear a foreign object. When antibodies bind to something, they trigger your immune system to destroy it. And this is why you get inflammation whenever you get a cut or scrape, antibodies will bind to whatever microscopic dirt and bacteria that enter your body, and your immune system flooding that area to destroy them is felt by you as inflammation.

And we know that Aduhelm and Leqimbi are working as antibodies against amyloid beta. They bind strongly to amyloid beta, they induce inflammation when given to Alzheimer’s patients (although inflammation in the brain can cause multiple side effects), and tests show that they seem to be reducing the amount of amyloid beta in the patients who take them.

Yet the prognosis for Alzheimer’s is not much better with these drugs than without them. Maybe they just aren’t destroying *enough* amyloid beta, but they are barely reducing the rate at which Alzheimer’s patients decline in mental faculty, and are not at all causing patients to improve and regain their mental state. Maybe the brain just *can’t* be fixed once it’s been damaged by amyloid beta, but you’d hope that there would at least be some improvement for patients if the Amyloid hypothesis is correct.

This has caused the field to seemingly split, with many still supporting the Amyloid hypothesis but saying these drugs don’t target amyloid beta correctly, with others now fractured in trying to study the many, many other possible causes of Alzheimer’s diesease. Tau, ApoE, neurotransmitters, there’s lots of other stuff that might cause this disease, but I want to focus on the final hail mary of the Amyloid hypothesis: that the drugs aren’t targeting amyloid beta correctly.

Because it’s honestly not the stupidest idea. One thing I learned when I researched this topic was the variety of forms and flavors that *any* protein can come in, and amyloid beta is no different.

When it’s normally synthesized, amyloid beta is an unfolded protein, called “intrinsically disordered” because it doesn’t take a defined shape. Through some unknown mechanism, multiple proteins can then cluster together to form aggregates, again of no defined shape. But these aggregates can fold into a very stable structure called a protofilament, and protofilaments can further stabilize into large, long filaments.

Each of these different structures of amyloid beta, from the monomers to the aggregates to the filaments, will have a slightly different overall shape and will bind slightly differently to antibodies. One reason given for why Aduheim causes more brain bleeds than Leqimbi is because Aduheim binds to the large filaments of amyloid beta, which are often found in the blood vessels of the brain. By siccing the body’s immune system on these large filaments, the blood vessels get caught in the crossfire, and bleeding often results.

Meanwhile other antibodies are more prone to target other forms of amyloid beta, such as the protofilaments or the amorphous aggregates.

But what amyloid beta does or what it looks like in its intrinsically disordered state is still unknown, and still very hard to study. All our techniques for studying small proteins like this require them to have a defined shape. Our instruments are like a camera, and amyloid beta is like a hummingbird flapping its wings too fast. We can’t see what those wings look like because they just look like a blur to our cameras.

So maybe we’ve been looking at the wrong forms of amyloid beta, rather than the filaments and protofilaments which are easy to extract, see, and study, maybe we should have been looking at the intrinsically disordered monomers all along, and we only studied the filaments and protofilaments because we were *able* to study them, not because they were actually important.

There’s a parable I heard in philosophy class about a drunk man looking for his keys. He keeps searching under the bright streetlight but can never seem to find them. But he’s only searching under the streetlight because *that’s where he can see*, he isn’t searching because *that’s where his keys are*.

Endlessly searching the only places you *can* search won’t necessarily bring results, you may instead need to alter your methods to search where you currently can’t. And if the Amyloid hypothesis is to be proven true, that will probably be necessary. Because right now I’ve heard nothing to write home about Aduheim and Leqimbi, many doctors won’t even proscribe them because the risk of brain bleeds is greater than the reward of very marginally slowing a patient’s mental decline, not even reversing the decline.

I no longer directly research Alzheimer’s disease, but the field is in a sad place when just 4 years ago it seemed like it was on the cusp of a breakthrough.