Anti-amyloid translation.

Short answer.

In early symptomatic Alzheimer's disease with amyloid confirmation, lecanemab and donanemab have crossed the positive phase 3 threshold, but the absolute clinical effects are modest and must be interpreted against CDR-SB or iADRS meaningfulness thresholds.

Caveat.

This is not a class-wide endorsement. Aducanumab remains a contested precedent, and the positive antibody results do not rescue failed soluble-amyloid or BACE programs.

What would change this answer.

Prospective evidence showing larger patient-level functional benefit, clearer subgroup benefit-risk separation, or real-world registry outcomes that diverge materially from trial settings.

Short answer.

Amyloid PET and downstream biomarkers move more strongly than cognition. The current state supports target engagement and a bounded clinical signal, not a simple rule that more amyloid lowering equals meaningful cognitive benefit.

Caveat.

Trial enrichment by tau burden and disease stage makes cross-program comparisons fragile. Biomarker movement is necessary for this mechanism but not sufficient as a decision endpoint.

What would change this answer.

An integrated dataset linking individual amyloid change, tau progression, ARIA, and cognition across antibody programs with comparable endpoints and enough follow-up.

Short answer.

BACE inhibition is a negative translation lane for symptomatic and prodromal Alzheimer's disease. Verubecestat, lanabecestat, and atabecestat failed despite target engagement, with cognitive or safety signals that weaken broad amyloid-production suppression as a usable strategy.

Caveat.

These failures do not falsify amyloid clearance by antibodies. They narrow mechanism, timing, and safety constraints for production-side intervention.

What would change this answer.

A safer amyloid-production intervention with prespecified cognitive benefit, no worsening signal, and mechanistic separation from prior BACE substrate liabilities.

Short answer.

ARIA risk is central, not incidental. APOE4 homozygosity, pretreatment hemorrhage markers, anticoagulant exposure, and MRI monitoring capacity materially shape the eligible population and the net-benefit case.

Caveat.

Labels and regional regulators differ in how sharply they restrict APOE4 homozygotes. The frontier stores this as eligibility and monitoring state, not as patient advice.

What would change this answer.

Prospective genotype-stratified risk models and registry data showing which subgroups retain enough expected benefit after ARIA and monitoring burden are included.

Short answer.

BBB delivery is a plausible next constraint, not a solved clinical lever. Transferrin-receptor shuttle designs and focused ultrasound can increase brain exposure or regional amyloid removal in early evidence, but the frontier does not yet show a clinical translation-ready path.

Caveat.

Most delivery findings remain preclinical or very small human studies. They explain why antibody exposure and safety are constrained, but they should not be promoted as current anti-amyloid treatment solutions.

What would change this answer.

Clinical evidence that delivery-enhanced anti-amyloid therapy improves cognition or risk-adjusted benefit beyond current antibodies, not only brain exposure or regional plaque change.

Short answer.

The most valuable next evidence links patient-level treatment exposure, amyloid clearance, tau progression, ARIA, genotype, anticoagulant status, and cognition in the same longitudinal substrate.

Caveat.

More papers alone will not resolve the decision. The key gap is interoperable, patient-level trajectory evidence that can update subgroup benefit-risk rather than add another average trial result.

What would change this answer.

A public proof packet or registry export that lets Vela replay outcome, safety, biomarker, and eligibility updates by subgroup rather than treating each program as an isolated artifact.