Off BACE: Fascinating Failures, One with Cognitive Recovery?

Shocking cognitive worsening and brain shrinkage (not to mention other toxicities) were seen in subjects receiving BACE inhibitors during recent trials, meant to help early dementia. Could the experience teach us anything about normal brain physiology, and the role of amyloid-β in wellness?

Two major approaches targeting amyloid-β, the presumed culprit in Alzheimer’s disease (AD), have made it to human clinical trials: one with various monoclonal antibodies and one with certain metabolic pathways inhibitors (BACE, Beta Amyloid precursor protein Cleaving Enzyme). As of this writing, after years of trials, none have made a significant positive impact for patients, and none are FDA approved.

As mentioned above, some agents made subjects acutely worse (both monoclonals and BACE agents); some trials paused or were stopped early for futility or safety reasons. Whether some subjects had lasting effects is generally unpublished, but some data are becoming available.

The latest report of a BACE clinical trial failure, atabecestat published this year, is fascinating for its reported “recovery” of cognitive impairment after treatment stopped, with details not easily found in other clinical reports. Subjects (in aggregate) who recovered did NOT do better than the placebo group but did come back to “baseline” (the placebo group seemed to do better than baseline in the same time frame, suggesting learning).

Some details: randomized trial, two dose levels of the drug and placebo, over 500 subjects, all “amyloid positive” (by CSF or PET scan), 61% women, mean age 70.4. Their dementia status was “pre-clinical,” meaning they could be subjectively worried about their own cognition and memory but performed well enough to be considered normal (Clinical Dementia Rating scale global score 0).

At 6- and 12-month timepoints, analysis showed dose-dependent cognitive worsening in the two atabecestat dose groups, along with liver enzyme, lymphocyte count, and neuropsychiatric problems, all worse than placebo. Based on the liver toxicity, “…[the Janssen company]… concluded the benefit/risk was no longer favorable for atabecestat…” The trial was truncated, with no further enrollments.

Using MRI neuroimaging, “Consistent with other BACEis [sic], atabecestat was associated with dose-related and duration-related decreases in whole-brain volume compared with placebo treatment” [i.e., brain shrinkage].

What was striking to this writer, as an oncologist familiar with toxicity readouts, is that the cognitive performance impairment and recovery both happened relatively quickly in this study. Many AD studies, albeit in mild-to-moderate cases, show unrelenting continued decrements in subjects’ performance, even with monoclonals that might “slow” progression.

Many proponents of the Amyloid Cascade Hypothesis suggest that cognitive impairments and AD pathophysiology (amyloid plaque accumulation) start years before they become functional problems, and that one must wait to see any effects. Hence their arguments are for very early diagnosis, and very early “preventive” interventions.

The prevailing opinion seems to be that finding amyloid-β is necessary but perhaps not sufficient to cause dementia. Neuroimagers at Mayo found amyloid-β in about 20% of local 70 year olds, but only 1-2% had clinical dementia: “Most patients with biologically defined [their neuroimaging biomarker definition] Alzheimer disease are not symptomatic, which creates potential confusion around the definition of Alzheimer disease”.

What do neuroscientists think about the BACE endeavors? has news reports on clinical trials; it also gives neuroscientists the chance to comment. When the follow-up data on the failed verubecestat BACE trials became available, December 2020, most commenters still supported the notion of the presumed central role of amyloid-β in the pathophysiology of AD and related dementia.

A naive oncologist might take different tack, since the details of pathophysiology in dementia seem yet to be elucidated. One could speculate that atabecestat acutely interrupted some amyloid-beta process vital to cognitive performance, and that even in amyloid positive 70-somethings, their brains had the capacity to reverse the impairment process when the agent was removed. (The treatment group was also more likely to have brain volume loss, making recovery even more remarkable, although the authors write that it didn’t correlate with performance.)

One idea might be that amyloid-β is somehow necessary for maintenance of cognitive “tone,” and that without its normal processing, cognition suffers, as does brain volume. (In the case of verubecestat, hippocampal volume shrank. Recovery of brain volumes has not been reported for either agent.) Is amyloid positivity simply a sign of the brain’s attempt to compensate or to combat something yet undiscovered, like the body leaving adhesion scars after serious infection?

But what was involved in the recovery process? Was it just allowing APP to be processed into amyloid-β again, or was something else involved? Could the recovery process be exploited to help folks who already have symptomatic dementia, even those with brain shrinkage? Could this be a serendipitous finding that could lead to a new area of therapeutic inquiry?

Of course, the recovery might be something “trivial,” since the AlzForum found that some of the pattern of BACE cognitive deficit seems to be similar to that seen produced by scopolamine. Street drug abuse, alcohol, and THC might also cause acute and chronic impairments; cognitive recovery can also occur in abstinence. The BACE trials did present some evidence that CSF amyloid biomarkers were affected, as a demonstration that APP was targeted.

Again, the fact that subjects in the atabecestat study, even at their advanced age, amyloid positive, could recover from an amyloid-beta drug induced insult to cognitive performance, by itself seems remarkable.

Now that the NIA funding has reached $3B for ADRD, announced last year, perhaps some funds can be used to exploit some novel ideas, and perform follow-up on naive notions gleaned from clinical trial failures. Yes, NIA and AD researchers: patients and families are still that desperate.