What evidence is required by the FDA to approve a drug for Alzheimer’s or cognitive disorders?

1. What kinds of trials the FDA expects: randomized, double‑blind, placebo‑controlled evidence of clinical benefit

Approval decisions for recent Alzheimer’s drugs have relied on large, double‑blind, placebo‑controlled trials that measure change on accepted cognitive and functional scales over substantial follow‑up periods — for example, donanemab’s efficacy was demonstrated by statistically significant differences on the Integrated Alzheimer’s Disease Rating Scale (iADRS) and component scales at Week 76 versus placebo in a pivotal trial ^[1], and the FDA convened an advisory committee to scrutinize whether those trials showed true clinical benefit ^[2].

2. When surrogate endpoints are acceptable: the accelerated‑approval pathway

When there is an unmet medical need, the FDA can approve drugs under Accelerated Approval based on surrogate endpoints “reasonably likely” to predict clinical benefit; aducanumab (Aduhelm) received accelerated approval after showing reduction in brain amyloid plaques despite weak evidence of cognitive benefit at the time, illustrating how biomarker changes have been used as the primary evidence basis ^{[3] [5]}. Those approvals require rigorous, well‑designed confirmatory trials to verify actual clinical benefit within an agreed timeframe ^[2].

3. Confirmatory trials and conversion to traditional approval

If a drug was granted accelerated approval, the sponsor must complete a confirmatory trial that demonstrates real, patient‑meaningful clinical benefit — upon which the FDA can convert an accelerated approval to a traditional approval, as happened when the agency determined a confirmatory trial for lecanemab verified clinical benefit and converted its status to traditional approval ^[2].

4. Biomarkers, eligibility, and labeling limits

For the new class of anti‑amyloid antibodies the FDA has required confirmation of amyloid pathology (via PET scan, cerebrospinal fluid, or validated tests) in subjects studied and limited labels to the disease stages actually tested — typically mild cognitive impairment or mild dementia due to Alzheimer’s — with explicit labeling language that there are no safety or effectiveness data outside those studied stages ^{[6] [2] [1]}.

5. Safety evidence and monitoring requirements

Because anti‑amyloid antibodies can cause amyloid‑related imaging abnormalities (ARIA) including brain swelling and microhemorrhages, FDA decisions have hinged not only on efficacy but on safety profiles and provisions for genetic or imaging risk stratification (for example, APOE4 testing is recommended on labels) and careful MRI monitoring strategies outlined in prescribing information ^{[6] [7] [3]}.

6. Statistical significance, clinical meaningfulness, and the controversy over “modest” benefit

The agency evaluates statistical significance on prespecified endpoints (e.g., iADRS, ADAS‑Cog13, ADCS‑iADL) and must judge whether the magnitude of change is clinically meaningful; recent approvals (lecanemab, donanemab) produced modest but statistically significant slowing of decline, prompting debate among clinicians, payers and patient groups about real‑world benefit versus risks and cost ^{[1] [8] [5]}.

7. Broader regulatory and payer context affects approval impact

FDA approval is distinct from payer coverage decisions; Medicare and other payers may limit access based on their assessments, which in turn influence the practical availability and further development incentives for Alzheimer’s drugs — an unresolved but material part of the evidence ecosystem around these approvals ^[5].