Anti-Amyloid Criteria May Exclude Many With Atypical AD

Up to 85% of patients with biomarker-confirmed atypical Alzheimer’s disease (AD) may be ineligible for anti-amyloid therapy under current trial and appropriate-use criteria, despite most having early-stage disease.Results of a retrospective analysis of 184 patients with biomarker-confirmed atypical AD showed that standard cognitive screening thresholds were the primary driver of ineligibility, particularly in patients whose initial deficits involved visuospatial function, language, or executive function rather than memory.Depending on the criteria applied, only 15%-32% of patients were theoretically eligible for treatment. Bedside cognitive thresholds — particularly Mini-Mental State Examination (MMSE) cutoffs — were the leading reason for exclusion, even though most patients were still at an early symptomatic stage.“This suggests that clinicians should not rely on a single global cognitive score alone. Treatment decisions should also consider the patient’s specific clinical syndrome and level of daily functioning,” first author Dror Shir, MD, of the Mayo Clinic in Jacksonville, Florida, told Medscape Medical News.Shir noted that a low MMSE score does not necessarily mean that a patient with atypical AD has advanced dementia.“These patients often have prominent language or visuospatial impairments that can interfere with their ability to complete these types of tasks, resulting in scores that may overestimate their overall disease stage. Clinical judgment, as emphasized in the appropriate-use recommendations, is especially important when assessing disease stage in patients with atypical presentations,” Shir said.The study was published online August 5 in Neurology.Few Atypical Patients Meet Treatment CriteriaAnti-amyloid trials of lecanemab and donanemab predominantly enrolled patients with mild cognitive impairment or mild dementia and typical amnestic-predominant disease. Patients with atypical forms — including posterior cortical atrophy (PCA), logopenic variant primary progressive aphasia (lvPPA), dysexecutive AD, and corticobasal syndrome due to AD — were largely excluded or underrepresented.These atypical AD phenotypes may initially affect visuospatial processing, language, or executive functioning while relatively sparing episodic memory. They also occur more commonly in younger patients and may be misdiagnosed or recognized late because standard assessments do not adequately capture their dominant deficits.To address this gap, investigators assessed how patients with biomarker-confirmed atypical AD would fare under eligibility criteria from pivotal anti-amyloid trials and current appropriate-use recommendations and examined the factors most commonly responsible for exclusion. For the study, the investigators retrospectively assessed theoretical eligibility for anti-amyloid therapy in 184 patients with biomarker-confirmed atypical AD evaluated at Mayo Clinic. The cohort included 98 patients with PCA, 42 with lvPPA, 37 with dysexecutive AD, and seven with corticobasal syndrome caused by AD. Of the study population, 61% were women, and 77% had symptom onset by age 65.The researchers estimated eligibility at the initial clinical evaluation by applying inclusion and exclusion criteria from the phase 3 CLARITY AD trial of lecanemab and the TRAILBLAZER-ALZ 2 trial of donanemab, as well as published appropriate-use recommendations for both drugs.Overall, 15% of patients met CLARITY AD eligibility criteria, and 17% met TRAILBLAZER-ALZ 2 criteria. Eligibility increased to 27% under lecanemab appropriate-use recommendations and 32% under donanemab recommendations.Eligibility varied by phenotype under the trial criteria. Under CLARITY AD, patients with dysexecutive AD were more likely to qualify than those with lvPPA or PCA. Under TRAILBLAZER-ALZ 2, patients with lvPPA were more likely to qualify than those with PCA.However, no significant eligibility differences across phenotypes emerged when the appropriate-use recommendations were applied, suggesting the criteria broadly restrict access across atypical presentations rather than disproportionately affecting one particular syndrome, the authors said.Cognitive Scores Drive ExclusionsCognitive thresholds, imaging findings, and disease severity were the most common reasons for ineligibility.Under CLARITY AD criteria, the MMSE threshold accounted for 60% of exclusions, followed by imaging abnormalities in 22% and memory or global disease-stage criteria in approximately 19% each.Under the TRAILBLAZER-ALZ 2 framework, age restrictions were the most common exclusion, followed by MMSE scores, imaging findings, and disease stage.MMSE thresholds remained the leading reason for exclusion under the appropriate use recommendations, accounting for 67% of lecanemab exclusions and 56% of donanemab exclusions.Yet low cognitive test scores did not necessarily correspond with advanced functional impairment. Of the 94 patients excluded under both the CLARITY AD criteria and lecanemab appropriate-use criteria because of low MMSE scores, 63 (67%) had a Clinical Dementia Rating (CDR) score of 0.5 or 1, indicating early symptomatic disease.Similarly, 61% of patients excluded from TRAILBLAZER-ALZ 2 because of MMSE scores remained at an early symptomatic stage.Overall, 82% of the cohort presented with a CDR score of 0.5 or 1, yet 70%-85% would have been considered ineligible for treatment depending on the eligibility framework applied.Imaging findings also excluded 20% of patients, including those with severe small-vessel or white matter disease, lacunar infarcts, cerebral hemorrhages, superficial siderosis, encephalomalacia, inflammatory findings, or vascular malformations.Beyond Cognitive CutoffsShir cautioned that the study does not show that all excluded patients should receive treatment.“It was a retrospective, single-center analysis of theoretical eligibility, not a treatment-outcomes study. We still need prospective studies specifically evaluating the safety and effectiveness of anti-amyloid therapies in atypical Alzheimer’s disease,” she said.“The practical message is that treatment decisions should integrate functional status, phenotype-specific cognitive assessment, biomarker confirmation, MRI safety findings, and a clear discussion and shared decision-making process,” Shir added.The priority is to develop “more phenotype-sensitive approaches” to staging disease and identifying appropriate treatment candidates, she added.Borna Bonakdarpour, MD, who was not involved in the study, agreed that reliance on MMSE cutoffs may disadvantage patients with atypical AD.Because the MMSE is “heavily verbal,” patients with aphasia or visuospatial deficits may score poorly despite retaining substantial independence, said Bonakdarpour, neurologist at Northwestern Medicine and associate professor of behavioral neurology at Northwestern University Feinberg School of Medicine in Chicago.Clinicians who do not account for these deficits “may just automatically disqualify patients,” he added. Eligibility “shouldn’t be just by a number” but should also reflect patients’ ability to shop, perform household chores, and manage finances.“It needs to be more of a pluralistic decision rather than just going by number,” Bonakdarpour said.The authors of a linked editorial said the findings “clearly support the need for refining and re-evaluating the validity and clinical relevance of cognitive staging criteria and severity assessments” for patients with nonamnestic AD.Developing cognitive and severity thresholds that rely less on memory impairment and age cutoffs and are more specific to atypical AD phenotypes is a “critical step” toward defining phenotype-specific treatment effects and rigorously weighing treatment risks and benefits, wrote Calin Prodan, MD, and Andriy Yabluchanskiy, MD, PhD, of the University of Oklahoma Health Sciences Center in Oklahoma City.They noted that the impact of anti-amyloid therapy in atypical AD remains “incompletely characterized” and called for larger prospective studies in well-characterized, biomarker-confirmed populations.The study was supported by the National Institutes of Health and the Alzheimer’s Association. Disclosures for study authors are available with the original study publication. Bonakdarpour, Prodan, and Yabluchanskiy reported having no relevant disclosures.