How do lecanemab, aducanumab, and donanemab compare in clinical trial results?

1. How researchers reached different summaries — head‑to‑head ambiguity and meta‑analytic disagreements

Systematic reviews and network meta‑analyses disagree on whether these antibodies differ meaningfully in cognitive benefit. One network meta‑analysis with 6,547 participants concluded no clear difference in cognitive effectiveness among lecanemab, aducanumab and donanemab, and even suggested lithium might outperform aducanumab—while flagging small sample sizes and lack of direct comparisons as major limitations ^{[4] [5]}. By contrast, a larger frequentist/Bayesian network review covering phase‑III trials (~16,971 participants) reported distinct profiles, ranking donanemab highest for cognition, lecanemab highest for amyloid clearance, and aducanumab intermediate but controversial—underscoring that conclusions depend on included trials, endpoints and analytic choices ^[1]. These differences illustrate that meta‑analytic methodology and trial selection materially change comparative claims, and that direct head‑to‑head randomized trials remain absent.

2. Cognitive benefits: modest gains and the question of clinical meaning

Phase‑III data present statistically significant but modest cognitive slowing for lecanemab and donanemab. Lecanemab’s pivotal Clarity AD trial reported a roughly 27% slower decline on CDR‑SB with an absolute between‑group difference small enough that some analyses argue it is below typical patient‑perceived thresholds (0.45 CDR‑SB points cited) ^{[6] [2]}. Donanemab’s TRAILBLAZER data indicated a larger relative slowing—reported as ~35% on integrated rating scales and stronger ranking for ADAS‑Cog and CDR‑SB in certain meta‑analyses—particularly in participants with lower tau burden, suggesting biomarker‑defined subgroups may drive benefit ^{[1] [2]}. Aducanumab’s EMERGE/ENGAGE program produced inconsistent cognitive signals, with pooled analyses and re‑analyses leaving clinical significance debated ^{[2] [1]}. The upshot: statistical separation exists, but absolute effect sizes are small and their day‑to‑day impact is contested.

3. Biomarkers tell a different story: amyloid clearance versus clinical effect

Biomarker outcomes diverge from cognitive rankings. Lecanemab produced the largest reductions in amyloid PET centiloid/SUVr, positioning it as the most potent at clearing soluble and aggregated Aβ in imaging endpoints ^{[1] [2]}. Donanemab strongly reduced plaque burden too but its cognitive advantage seems concentrated in participants with lower tau pathology, indicating that plaque removal alone does not uniformly translate to clinical benefit ^{[1] [2]}. Aducanumab reduces plaque burden but to a lesser and more variable extent in trials; this contributed to regulatory controversy and disagreement about whether biomarker change equates to meaningful clinical improvement ^{[2] [1]}. Binding specificity data show lecanemab prefers protofibrils whereas aducanumab favors fibrils; such differences plausibly explain divergent biomarker and safety profiles ^[7].

4. Safety trade‑offs — ARIA dominates risk discussions and affects acceptability

All three antibodies increase risk of ARIA‑E (edema) and ARIA‑H (hemorrhagic events), with reported incidence ranges from roughly 12% to 35% depending on drug and trial, and APOE ε4 carriers experiencing higher rates ^{[2] [3] [1]}. Donanemab appears to have the highest ARIA and discontinuation rates, with ARIA‑E reported near ~35% in some datasets; lecanemab’s ARIA rates are lower but still notable (~12–17%), and aducanumab’s ARIA incidence has been reported variably but includes substantial rates in some trials ^{[2] [1]}. Observational reports and commentary emphasize serious, sometimes irreversible harms in a small proportion of patients and call for greater manufacturer transparency and careful monitoring in clinical use ^[3]. These safety differences materially affect clinical acceptability and numbers‑needed‑to‑treat versus numbers‑needed‑to‑harm calculations.

5. Who benefits most — biomarkers, tau status and APOE genotype shape responders

Trial subgroup analyses and meta‑analyses point to heterogeneous treatment effects: donanemab’s cognitive advantage concentrates in participants with low‑to‑moderate tau, suggesting earlier or biomarker‑selected patients derive greater clinical benefit ^{[2] [1]}. APOE ε4 carriers feature a consistent pattern of higher ARIA risk across agents, raising genetic stratification as a practical safety tool ^[2]. Lecanemab’s protofibril binding may favor certain pathological states, while aducanumab’s fibril preference might change both efficacy and harm profiles ^[7]. As a result, precision selection by amyloid/tau imaging and APOE genotyping could shift the perceived balance of benefit and harm in real‑world deployment.

6. Bottom line and what’s missing — transparency, head‑to‑head trials and long‑term outcomes

Current evidence shows distinct trade‑offs: donanemab may offer the largest cognitive slowing at the cost of more ARIA and discontinuations; lecanemab clears amyloid most effectively with a somewhat better tolerability profile; aducanumab’s efficacy remains contested with significant safety concerns ^{[1] [2]}. However, systematic reviews that pool limited and heterogeneous trials sometimes reach different conclusions, underscoring methodological sensitivity and the absence of randomized head‑to‑head trials ^{[4] [5]}. Long‑term durability of benefit, real‑world safety, transparent adverse event reporting and trials that directly compare these agents remain the crucial gaps to resolve before definitive comparative judgments are possible ^{[3] [1]}.