How do FDA‑approved Alzheimer’s drugs perform in randomized trials on cognition and disease progression?
Executive summary
Randomized, double‑blind trials of the newest FDA‑approved Alzheimer’s drugs—principally the anti‑amyloid monoclonal antibodies lecanemab (Leqembi) and donanemab (Kisunla)—have demonstrated statistically significant slowing of cognitive and functional decline in people with early Alzheimer’s disease (mild cognitive impairment or mild dementia) who have biomarker‑confirmed amyloid pathology, but the clinical magnitudes are modest and accompanied by nontrivial safety risks that limit generalizability [1] [2] [3] [4]. Older symptomatic agents (cholinesterase inhibitors) improve cognition transiently without evidence of altering underlying disease progression [5] [6].
1. Trial design and the patient population that showed benefit
The pivotal randomized trials were large, multicenter, double‑blind, placebo‑controlled studies enrolling people with early symptomatic Alzheimer’s disease confirmed by amyloid PET or CSF testing; lecanemab’s CLARITY‑AD (Study 301) randomized about 1,795 participants over 18 months and donanemab’s TRAILBLAZER‑ALZ2 program enrolled roughly 1,736 participants up to 72 weeks in its phase‑3 trial framework, demonstrating that observed effects are restricted to early‑stage, biomarker‑positive populations—the exact groups studied and approved by regulators [1] [3] [7] [2].
2. What the trials measured and the size of the benefit
Sponsors and regulators prioritized composite and standard clinical scales of cognition and function (e.g., iADRS, ADAS‑Cog13, CDR‑SB) and reported statistically significant differences favoring anti‑amyloid antibodies: donanemab (Kisunla) produced a 2.92‑point difference on the integrated Alzheimer’s Disease Rating Scale (iADRS) at week 76 versus placebo with highly significant p‑values, and lecanemab showed slowed clinical decline on pre‑specified cognitive and functional outcomes in its phase‑3 confirmatory trial [2] [3] [1]. These numeric gains are real in trial populations but are described by experts as modest in magnitude relative to the long natural history of the illness [8] [9].
3. Disease modification claims and the role of biomarkers
Regulators converted lecanemab to traditional approval after a confirmatory randomized trial verified that amyloid‑lowering by the drug translated into slower clinical decline—marking, according to the FDA, the first verification that an Alzheimer’s drug targeting an underlying process showed clinical benefit [3]. Trials consistently showed drug‑induced amyloid plaque reduction coincident with clinical slowing over 18 months, supporting the amyloid‑targeting strategy for early disease stages, but whether these interventions truly alter long‑term disease trajectory beyond measured trial windows remains an open question in the literature [8] [1].
4. Safety, controversies and limits to applicability
Anti‑amyloid antibodies carry a recognized risk of amyloid‑related imaging abnormalities (ARIA), including brain swelling and hemorrhage; phase‑3 data reported ARIA rates substantially higher in treated groups (e.g., lecanemab arm versus placebo) and such adverse events shaped debate about risk‑benefit and patient selection [4]. Aducanumab’s earlier accelerated approval—based on amyloid reduction as a surrogate—sparked controversy because clinical benefit was less clear, illustrating how endpoints and regulatory pathways affect public and scientific trust [9] [4]. Crucially, the randomized trials enrolled early‑stage, biomarker‑positive individuals, so findings cannot be extrapolated to moderate or late‑stage Alzheimer’s, and older symptomatic drugs remain useful only for short‑term symptomatic relief without evidence of disease modification [5] [6] [3].
5. Bottom line for clinicians and patients from randomized trials
Randomized evidence establishes that certain monoclonal anti‑amyloid therapies can slow cognitive and functional decline modestly in carefully selected early‑stage, amyloid‑positive patients—an advance over prior symptomatic treatments—but benefits are limited in size, accompanied by real safety concerns (ARIA), and anchored to specific trial populations and 18‑ to 72‑week follow‑up windows; long‑term disease‑altering effects beyond these windows remain to be fully demonstrated in broader populations [3] [1] [2] [4]. Where trials are silent—such as reversal of established dementia in humans—claims remain unsupported by randomized clinical data [10].