Low Amyloid Burden Bodes Well for Future Dementia Risk

AMSTERDAM -- An offshoot of the closely watched A4 trial, which mainly tested whether the anti-amyloid drug solanezumab could delay onset of Alzheimer's disease in high-risk individuals, showed that older people with low initial levels of amyloid brain plaques maintained normal cognition over almost 5 years.

The primary analysis, reported here at the Alzheimer's Association International Conference (AAIC) by Reisa Sperling, MD, of Brigham and Women's Hospital in Boston, indicated that solanezumab provided no benefit relative to placebo as tracked for 240 weeks.

While the investigators had hoped for a different result, the fact that solanezumab had previously failed to show a significant effect in a large phase III trial involving patients with mild cognitive impairment meant that A4 faced a heavy lift. At the time A4 was conceived in 2014, however, solanezumab was the most effective anti-amyloid drug then available, having shown "some evidence of benefit" in trials, explained co-investigator Paul Aisen, MD, of the University of Southern California Alzheimer's Therapeutic Research Institute in San Diego. "We hoped for a larger effect in a preclinical population," he told AAIC attendees.

That did not happen. Patients assigned to solanezumab showed only slightly slower cognitive decline that did not achieve statistical significance relative to a placebo group, as measured by so-called Preclinical Alzheimer Cognitive Composite (PACC) scores or other measures such as the Clinical Dementia Rating (CDR). Findings from A4 were also published Monday in the New England Journal of Medicine.

However, the offshoot dubbed LEARN provides new validation that, if nothing else, measurements of amyloid plaque burden could be a valuable clinical tool in assessing future risk.

A4 enrolled 1,169 older patients (mean age 72) with normal cognition and "elevated" levels of beta-amyloid protein in their brains as measured with PET scans. (Amyloid burdens in the cortex greater by 10%-15% than in the cerebellum, where plaques generally don't appear, were considered elevated.)

But, as Aisen explained, the investigators had to perform PET scans on some 4,500 people to find those meeting the amyloid elevation criteria. So, they decided to enroll another cohort from those without elevated amyloid to track alongside the A4 participants with the same clinical evaluations, for a companion study they called LEARN, for Longitudinal Evaluation of Amyloid Risk and Neurodegeneration. A total of 541 people were eventually enrolled.

Participants in LEARN (who were told they had no amyloid burden at enrollment) actually showed slight increases in PACC scores over the first couple years, indicating improved cognitive performance, after which it did track downward. But not nearly as fast as did A4 participants -- mean scores in the LEARN group were virtually the same at week 240 as at baseline, whereas in A4, participants' scores had fallen by a full point, a substantial decline. The pattern was the same for more standard dementia assessments including the CDR Sum of Boxes and the Activities of Daily Living scale.

Both A4 and LEARN suffered from several methodological hiccups, to the extent that one might consider A4, at least, a failed study. In 2017, after A4 was well underway, the leaders decided to up the solanezumab dosage substantially, from 400 mg to 1,600 mg per dose, and also to extend the follow-up period from the originally planned 3.2 years to 4.5 years, so that effects of the dosage increase could be captured adequately.

Meanwhile, LEARN commenced at about the same time. That meant that less than 3 years later, both studies were hit by the COVID-19 pandemic and its associated disruptions. Many study sites shut down, and the investigators had to resort to home infusions in some places to keep the study going. Aisen and Sperling acknowledged that these issues complicated data collection, analysis, and interpretation. But there was no suggestion that these problems might have obscured a genuine beneficial effect from solanezumab.

Taken together, said Sperling, the results from A4 and LEARN indicate that "amyloid reduction may be necessary to slow progression even at the stage of preclinical [Alzheimer's disease]."

If this had been the last analysis of anti-amyloid therapy to prevent onset of Alzheimer's disease, others might have concluded that such efforts are futile if any amyloid burden is present. However, since that time, two other amyloid-targeting drugs have come along with clear, if modest, benefits in terms of reducing progression in cognitive impairment. They are aducanumab (Aduhelm) and lecanemab (Leqembi) -- the latter having been awarded full, nonconditional FDA approval just 10 days ago.

Comment