Atypical Alzheimer's disease fails the eligibility test on the wrong grounds

Design

Retrospective eligibility analysis applying trial and appropriate-use criteria at initial clinical evaluation

Population

184 patients with biomarker-confirmed atypical Alzheimer's disease at a single US centre

Primary outcome

Proportion theoretically eligible for anti-amyloid therapy, and reasons for exclusion

Effect

70-85% ineligible depending on framework; bedside cognitive thresholds drove 50-67% of exclusions despite 82% having global Clinical Dementia Rating 0.5-1; imaging exclusions 22-27%, severity thresholds 19-23%

Anti-amyloid trials recruited patients with mild, amnestic-predominant Alzheimer's disease. The syndromes that present visually, linguistically or dysexecutively were not the trial population, and it has never been clear how the eligibility criteria land on them. This retrospective analysis applied the CLARITY-AD and TRAILBLAZER-ALZ2 criteria, and the appropriate use criteria for lecanemab and donanemab, to 184 patients with biomarker-confirmed atypical Alzheimer's disease at one centre: posterior cortical atrophy, logopenic variant primary progressive aphasia, dysexecutive Alzheimer's disease and corticobasal syndrome.

Between 70% and 85% would not qualify, depending on which framework was used. The reason matters more than the number. Bedside cognitive thresholds — chiefly the Mini-Mental State Examination — accounted for 50 to 67% of exclusions, while 82% of the cohort had a global Clinical Dementia Rating of 0.5 to 1, which is early symptomatic disease. Imaging exclusions accounted for 22 to 27% and formal severity thresholds for 19 to 23%.

The mismatch is mechanical. The MMSE is heavily weighted towards the visuospatial and language functions that posterior cortical atrophy and logopenic aphasia destroy first, so these patients score as moderately demented while remaining functionally early. Where anti-amyloid therapy is available, an MMSE of 21 in posterior cortical atrophy should prompt functional staging, not a closed door — and the case for that has to be made at the point of referral, because the threshold is applied before anyone looks further.

• Stage atypical Alzheimer's disease functionally — Clinical Dementia Rating, not MMSE alone
• Record why the bedside score is low: the deficit domain, not global severity
• Expect posterior cortical atrophy and logopenic aphasia to score disproportionately low
• Document biomarker confirmation clearly, since these patients are often doubted twice
• Raise the staging discrepancy at referral rather than after a refusal

Why it matters

It shows that the gatekeeping test for a disease-modifying therapy penalises exactly the phenotypes it was not designed to measure.

Don't overread it

Retrospective and single-centre, applying criteria on paper — it does not describe what happened to these patients in practice.

The statistics, in plain English

The 70 to 85% range is not uncertainty in the measurement; it is the spread produced by applying four different eligibility frameworks to the same patients, which tells you how much the criteria themselves vary. This is theoretical eligibility assessed retrospectively at first visit, so it estimates who would be screened out on paper, not who was actually refused treatment. With only seven patients in the corticobasal group, the between-phenotype comparisons are weak.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.