Incorporation of Mini-Mental State Exam Z-Scores May Increase Eligibility for Anti-Amyloid Treatments, Study Suggests

In a recently published, Korean-based study of amyloid-positive patients, findings showed that incorporation of demographically adjusted Mini Mental State Exam (MMSE) z-score thresholds helped increase eligibility for anti-amyloid treatments such as lecanemab (Leqembi; Eisai) and donanemab (Kisulna; Eli Lilly). Despite this, the study still showed that a substantial proportion of otherwise eligible patients were excluded due to MRI findings associated with amyloid-related imaging abnormalities (ARIA), underscoring the clinical relevance of cerebrovascular comorbidities.¹

The study included 2726 individuals evaluated at a Korean memory clinic, of whom 1005 were amyloid positive with available MRI and a Clinical Dementia Rating (CDR) of 0.5 (n = 738; 73.4%) or 1 (n = 267; 26.6%). Eligibility for lecanemab and donanemab was determined using Appropriate Use Recommendations (AUR)-based criteria, including age, MMSE performance, brain MRI findings, apolipoprotein E genotype for donanemab, and anticoagulant use, assessed with both raw score and z-score MMSE thresholds.

To evaluate whether fixed MMSE raw-score cutoffs (≥22 for lecanemab, ≥20 for donanemab) might misclassify patients across demographic groups, the study conducted a sensitivity analysis using MMSE z-scores derived from large-scale Korean normative datasets stratified by age, sex, and education. Z-scores were calculated using the standard formula (z = [X − μ]/σ), and participants were deemed cognitively eligible if they met either the original raw-score thresholds or achieved a z-score ≥ –2.0, a conservative boundary representing performance 2 SD below the mean. All remaining eligibility criteria mirrored the main analysis.

Led by Dong Young Lee, Department of Psychiatry, Seoul National University College of Medicine, around a fourth (24.6% for lecanemab; 28.0% for donanemab) of amyloid-positive patients were eligible for antiamyloid treatments before applying MMSE z-scores. After adding this sensitivity analysis, the proportion of eligible patients jumped to 38.4% and 33.7%, respectively.

"Applying AUR-based criteria and incorporating z-score-adjusted MMSE thresholds can expand eligibility especially for older adults with lower educational attainment," Lee et al wrote. "This approach enhances equity in treatment access and reflects a more inclusive strategy for identifying candidates for anti-amyloid treatment. However, expanding the eligibility pool also raises important safety considerations."

The study authors noted that older adults with lower education levels often carry higher vascular and comorbidity risks, which may increase their susceptibility to treatment-related adverse events such as ARIA. In addition, there were several limitations to the study, including the fact that non-clinical factors such as cost, insurance coverage, access to infusion and imaging centers, and patient or caregiver hesitancy toward PET scans or infusion-based therapies were not evaluated.

READ MORE: Expanding Neurology Care in South Miami: Dr. Abdelmalik’s Vision

Using the revised cognitive criteria, participants were categorized into 3 groups based on whether they qualified by both MMSE raw scores and z-scores, by z-scores alone, or by raw scores alone. Among those assessed for lecanemab eligibility (n = 653), most fell into Group 1, while about one third qualified only through z-scores and a smaller subset qualified only through raw scores. Clear demographic differences emerged: individuals in the z-score–only group were the oldest and had the fewest years of education, while those eligible by raw score alone were younger and more highly educated. Of note, participants meeting both criteria showed intermediate profiles.

A similar pattern appeared among donanemab-eligible participants (n = 620). The majority met both MMSE thresholds, with smaller proportions qualifying through z-scores or raw scores alone. As with the lecanemab cohort, the z-score–only group tended to be older with lower educational attainment, and the raw-score–only group was younger and more educated. All told, these consistent patterns highlighted how demographic factors influence cognitive classification across both therapies.

The team also examined how shifting the MMSE raw-score cutoff alone affected eligibility by lowering the threshold from 22 to 18 in one-point steps. Eligibility increased steadily with each reduction, rising from about one quarter of participants at a cutoff of 22 to nearly one third at 20, with further gains at 19 and 18. These results show that even small, single-point changes in MMSE criteria can meaningfully influence who qualifies for treatment.

In total, 36.6% and 37.3% of participants eligible for lecanemab and donanemab, respectively, were excluded because of brain MRI-based findings. The most common exclusions were severe WMHs, lacunar infarcts, and number of microbleeds.

Lee et al wrote, “Given that the SWI sequence is more sensitive to microbleeds than the gradient-recalled echo (GRE) sequence, the use of SWI in our study may have contributed to a relatively higher proportion of patients excluded based on MRI findings. This contrasts with previous clinical trials such as CLARITY-AD and TRAILBLAZER-ALZ 2, which employed GRE sequences for microbleeds detection."

REFERENCE
1. Jeon SY, Byun MS, Choi HJ, et al. Eligibility for lecanemab and donanemab in Korea under Appropriate Use Recommendations. Alzheim & Dement. 2025;21(11):e70875. doi:10.1002/alz.70875