The Alz Assn statement dated Nov 29 includes the idea that “this treatment can meaningfully change the course of the disease for people in the earliest stages of Alzheimer’s disease (AD).”
Most news service headlines are declaring it a win, but the magnitude of improvement is reportedly minor according to a number of commenters. NPR news does report “People who got infusions of lecanemab scored about half a point better on a zero-to-18-point scale of mental functioning, a slight but statistically significant difference.”
The concurrently published NEJM paper clearly shows that all patients had cognitive decline, even those who received lecanemab, almost with the same pace or trajectory. In those who received the drug, the decline was continuous, but delayed ~ 3-6 months to get to the same level of deterioration (my interpretation). On the other hand, imaging by PET scan did show less amyloid.
The study reported on a randomized trial of ~ 1800 “early” Alzheimer subjects (~60% were actually MCI, i.e less symptomatic, rather than mild AD, using NIA-AA criteria). About half of the lecanemab group were on AD symptom meds, presumably donepezil.
About 26% had infusion related problems, ~13% had brain edema or micro-hemorrhages, although none were considered severe. There were no deaths in the NEJM report, but perhaps two have been reported subsequently, one with dramatic details reported in Science.
How does lecanemab compare to donepezil? Lecanemab graph curves (top) may be compared to those below it, published in 2016 from a Swedish group. The latter show scores for Aricept / donepezil and others in that class (AChEI), in Mild vs. Moderate Alzheimer patients (before the use of PET scans to define / enroll subjects). The right hand graph “b” below also uses the ADAS-Cog performance measuring instrument (modified as ADAS-Cog14 as used for lecanemab).
There is no placebo control curve in the Swedish study, but the shapes of the curves show similar decline, with maybe a 2 point better performance at 18M in moderate disease for the lecanemab group compared to those with AChEI, but for those with mild disease, lecanemab seemed about 2 points worse in performance at 18M. The lecanemab data is relatively short term, of course.
The primary endpoint of the lecanemab study was a different assessment tool known as CDR-SB, and it was placebo-controlled, whereas the Swedish study was a longitudinal observation study. Those on meds like donepezil who got lecanemab may have done better (NEJM S2. forest plot in the Supplementary Appendix). That’s like a built-in combination study, but also makes it harder to tease out the contribution of each class of meds.
According to the a lecanemab press release, the company quoted a simulation study that pegged a “value based price” anywhere from ~$9K to $36K per year. Aduhelm ™, another amyloid antibody, was originally priced at $56K/year at launch but was cut to $28K a year ago.
Dr. Thambisetty, a senior clinical MD PhD investigator at the Nat’l Inst Aging (NIA), was worried about previously reported brain shrinkage or atrophy with lecanemab, and wrote that as an opinion piece in STATnews on Nov 28 (Monday). I could not find the brain volume data in the NEJM report, even in the Supplementary Appendix online.
Reuters Nov 29 quoted an author of the lecanemab study and noted researcher, Dr. Paul Aisen of USC: “I believe it’s an important benefit that will justify full approval. But of course, we want a bigger benefit.”
Dr. Thambisetty is quoted in the NPR story cited, calling the results (in his personal opinion, i.e not an NIA opinion) “a very small effect…It’s very unlikely that these differences are going to be noticeable by individual patients in their everyday lives.”
[thank you Eric and Dan for suggesting the donepezil comparison]