23 Dec 2024

Dear Scott,

Thanks for your kind comments and for posing an important question for debate (see 20 Dec Open Letter). Your choice of “Open Letter” to one individual is a bit whimsical, a first for me. I’m not sure why you singled me out, but I accept the debate in the positive spirit you conveyed. I view your topic as distinct from the Perspective that 12 of us wrote for Alzheimer’s & Dementia on the case for regulatory approval of new AD immunotherapies based on robust biomarker outcomes (lowered amyloid and tau proteins) plus safety. Accordingly, I’ll only respond here to your “Open Letter.” I may respond separately later to some of the additional comments on the A&D Perspective.

As you say, Scott, we agree on several points favoring a key role for the accumulation of Aβ-protein in the development of AD. But you question whether extracellular Aβ accumulation can be considered causative (as I believe) or instead represents “one defining feature of AD that has proven cognitively detrimental,” while the actual causative process is neurodegeneration per se. I disagree. The genetics of AD, to which your and your colleagues’ work on SORL1 has contributed importantly, suggests that dyshomeostasis in the production versus clearance of Aβ is the earliest demonstrated molecular change in these familial forms (FAD). In humans born with missense mutations in APP or presenilin or with trisomy 21, the increased relative production of Aβ42 begins in utero and leads to the quintessential neuropathology of AD, including neurodegeneration, years to decades later. Here, clear evidence of the neurodegenerative process begins after the first occurrence of Aβ dyshomeostasis at birth (e.g., babies born with Down’s syndrome have elevated plasma Aβ42). I have no doubt that the subsequent neurodegenerative process, including early endosomal alterations, tau deposition, and astrocytic and microglial inflammation, occur long before onset of symptoms and help drive them forward, but these features follow Aβ dyshomeostasis.

I do not believe we can prove that early endolysosomal defects, which are well-documented in AD brain tissue, begin before or without any extra- or intracellular accumulation of Aβ oligomers (which cannot be easily imaged by amyloid PET or even by amyloid histochemistry in the brain). We can’t be certain that no Aβ oligomers are present when endosomal changes begin in AD, particularly since Aβ oligomers and fibrils steadily accrue in all people with AD as we define the disease. One could opine that “sporadic” AD is mechanistically unrelated to FAD, but the facts that “sporadic” AD closely phenocopies the mutant APP and PS forms and that some cases once classified as “sporadic AD” may include humans with dysfunction of wild-type presenilin and effects of wild-type apolipoprotein E4 suggest that the “sporadic” AD biological process is mechanistically similar to that of FAD. Neurodegeneration is critical to the production of neurological deficits, but I and many others hypothesize that the initiating factor preceding it is altered Aβ economy.

Another way to convey my view of causality is to consider the etiology of other chronic diseases where scientific agreement about causal events has emerged. Inheritance of specific mutations in hemoglobin is universally considered to be the molecular cause of sickle cell disease, not the hematologic and microvascular events and organ damage that produce its intermittent symptoms. All cases of Huntington’s disease can be identified in utero or at birth as caused by polyglutamine expansion mutations of the huntingtin protein, with neurodegeneration slowly ensuing from this molecular etiology. Having the SARS-CoV-2 virus enter and take hold in one’s body is the beginning of all the complex downstream cytopathology of COVID-19, including long COVID with its “brain fog,” so we prevent the disease by neutralizing the causative virus. Heart attacks commonly involve myocardial injury and dysfunction including low left ventricular ejection fractions, but these are not considered the “cause” of the disorder; rather, they are the tissue consequences (effects) of the causative cholesterol-rich plaques and thrombi that block coronary arteries. In short, we can describe the characteristic tissue pathology of chronic diseases and the organ failure that ensues, but we don’t claim that these events are the proximate cause the disease. And so it is for neurodegeneration in AD.

In my view, “neurodegeneration” comprises an ill-defined, highly complex, slowly evolving spectrum of cellular alterations that is proximal to symptoms but long preceded by biochemical events that initiate it. In AD, multiple lines of evidence from labs and clinics worldwide indicate that elevation, oligomerization, and fibrillization of Aβ42 leads gradually to inflammation, neurodegeneration, and tissue atrophy. We can debate whether the cytotoxicity of Aβ represents principally an extracellular accumulation, as classic neuropathology, brain imaging, and fluid biomarkers all suggest, or also involves Aβ accumulation inside neurons, for which there has long been published evidence. There’s no reason why both cannot contribute.

Cell biological studies have shown that Aβ is largely generated in mildly acidic endosomal vesicles in which the two required aspartyl proteases (b-secretase and presenilin) cleave the APP substrate efficiently, and Aβ peptides are then quantitatively secreted by recycling endosomes, a normal intracellular process. As you rightly say of the FAD gene defects, “they all reliably coax neurons to produce and secrete more Aβ peptides from neuronal endosomes, and … once secreted, they end up forming extracellular aggregates.” It is this abnormal Aβ42 accumulation over time, both extra- and intracellularly, that initiates the neurodegenerative process in the disease we call Alzheimer’s. As you emphasize, Scott, “targeting a causal mechanism promises to have the greatest therapeutic benefits.” While therapeutics that ameliorate the downstream neuronal and glial alterations of AD are highly desirable, lowering the production of Aβ42 (e.g., with GSMs) and/or clearing amyloid deposits (e.g., with antibodies and vaccines) both target the molecular cause.

So, as you say, “the pathologies that occur outside or inside of neurons trigger AD’s defining neurodegenerative process,” but I do not view the latter as causative, in contrast to the telltale accumulation of Aβ dimers/oligomers well before neurodegeneration.

Your reference to an “apparent uncoupling of amyloid and neurodegeneration based on the glaring anatomical mismatch between where extracellular amyloid aggregates are first and foremost found, versus where neurodegeneration occurs first and foremost” is, in my view, an older and inaccurate simplification, in that a) rising diffusible Aβ oligomers likely occur widely across brain regions, including in the entorhinal cortex and hippocampus; and b) both proximal and more anatomically distal effects of extracellular Aβ aggregates could subtly compromise neuronal function (including glutamate receptor recycling) and structure.

I disagree that we can be certain the Aβ42-lowering trials have failed on readouts of neurodegeneration; rather, falling phosphotau, some decreases in tau-PET, and blunting of astrocytosis (falling plasma GFAP)—along with slower clinical progression—are some of the findings suggesting a slowing of neurodegeneration. And not all Aβ42-lowering clinical effects are “mild” as you say, since the analysis of some participants in CLARITY AD who at baseline had no tau deposition beyond the medial temporal lobe and/or had amyloid-PET levels <60 CL appear to have experienced virtual arrest of clinical progression over 36 months in the open label extension (van Dyck et al, AAIC, August 28, 2024), far beyond the “handwringing over 30 percent slowing” you cite. Let’s see what deltas emerge from our practices.

Finally, I agree that “Future trials testing drugs that either clear out amyloid or clear up traffic jams, early in the disease process, are required.” But the results of the former approach are in: It works. And many of us believe it will work even better earlier—in secondary prevention mode.

Thank you, Scott, for this opportunity to exchange opinions, whether or not we are “changing our minds”. This letter comprises my full response to your piece.

All the best,

… And a Response

Quick Links

Tools

Comments

Make a Comment

References

News Citations

Further Reading

Annotate