Angst

The aducanumab debacle (and it is a debacle) continues to inspire soul searching in the research community.  The focus is on whether the ability of a treatment to clean beta amyloid from the brain is valid evidence that the treatment is effective.  In other words, does simply getting rid of amyloid treat the disease?  Moreover, is simply showing that a treatment gets rid of amyloid strong enough evidence to approve the treatment for public use? 

I just read this fascinating article.  It's long and gets kind of deep into the issue, but here's what I got from it.

We have a long history of choosing the wrong therapeutic target when testing a new treatment candidate.  An example offered was treatments to raise HDL ("good cholesterol") as a treatment for heart disease.  We know that people with higher levels of HDL are less prone to heart disease.  So why not use drugs to raise its level in the blood stream?  A couple of candidate drugs went to trial.  How did that work out?  Well, they raised HDL without reducing or preventing anyone's heart disease.  In one case more people on the drug died of heart disease than those on the placebo.  Not a good showing.

The takeaway is that you need to understand the disease mechanisms before you know how to treat it.  And we do not understand Alzheimer's disease well at all.

In Alzheimer's disease we see anther correlation with an uncertain causal relationship.  There's a strong correlation between the presence of beta amyloid plaques and degeneration of the brain.  In Beating the Dementia Monster, we discussed possible pitfalls in assuming a direct causal relationship between the plaques and the disease.  I recalled what one researcher told me.  He said it's as though we correlate the presence of fast food wrappers in the trash with someone's heart disease.  Maybe if we just remove the fast food wrappers from the trash she will get better.  His point was that this may be what we have been doing with aducanumab.  

The authors of the article discussed an idea in this vein that was interesting to me.  In Beating the Dementia Monster, we discussed the origin of beta amyloid from the amyloid precursor protein (APP).  No one knows the purpose of the protein, but it's cut multiple times by some enzymes, notably one called secretase.  Different variants of secretase cut the APP molecule differently.  Ideally, if the APP molecule is cut appropriately (by the right variant), the fragments (peptides) are water soluble.  If they are not, insoluble fragments accumulate on brain cells as those famous plaques. 

I said in the book that the soluble fragments disappear innocuously from the scene, but maybe that's not right.  An idea being proposed is that it's not the presence of the insoluble fragments that are a feature of the disease process, but rather the reduced production of the soluble fragments.  When too many insoluble fragments are being produced not enough soluble ones are produced.  Just getting rid of the insoluble ones won't help. 

But who knows?  These proposals underline how poorly we understand the disease.  More work is needed to increase our understanding.