Abstract A030: The Montreal Cognitive Assessment (MoCA) administered virtually or via Duo APP in primary brain tumor patients: a preliminary analysis

Abstract Background: Primary brain tumor (PBT) patients experience cognitive dysfunction (CD) because of both their tumor and its treatment. Evaluating CD in PBT populations may inform patient outcomes and clinical management and allow exploration of biologic underpinnings associated with its occurrence. The Montreal Cognitive Assessment (MoCA) is a brief assessment of objective measures of cognition that evaluates 8 specific cognitive domains. Our group has previously reported the feasibility of MoCA, and this study includes using the newly developed MoCA Duo APP to evaluate and report the cognitive function of patients with PBT. Associations between total MoCA scores and clinical and demographic characteristics were explored. Methods: This cohort consisted of adult PBT patients (n=172) enrolled in the NCI-NOB Natural History Study (NCT02851706: PI T. Armstrong). Assessments were performed between February 2020 and July 2023 by MoCA certified assessors and administered in the clinic or via telehealth. MoCA scores range from 0 to 30 (normal cognition ≥ 26). Healthcare providers collected demographic and clinical characteristics, with Karnofsky performance status (KPS) scores categorized as poor (≤ 80) or good (≥ 90). Descriptive statistics, independent t-tests, one-way ANOVAs, and Pearson’s correlation were conducted using IBM SPSS Statistics software. Results: The majority of patients were male (58%), white (81%), had ≥12 years of education (87%), had high-grade (3/4) tumors (65%), and a good KPS score (68%). The mean MoCA score was 25 (median: 26; range: 6-30) with a mean completion time of 10.7 minutes (median: 10; range: 6.1-41.1). Forty-two percent (n=73) were classified as abnormal. The median scores were lowest in the visuospatial/executive (4/5; range: 0-5), abstraction (2/3; range: 0-3) and delayed recall (4/5; range: 0-5) domains. A univariate analysis identified lower MoCA scores in patients with high-grade tumors (p<0.001), poor KPS scores (p<0.001), who underwent two or more surgeries (p=0.004) or treatments (p=0.038), had recurrence (p=0.002), had progression on current imaging (p=0.017), and had current anticonvulsant (p=0.014) and corticosteroid use (p=0.036). Age at diagnosis (r=-0.157, p=0.021) and at visit (r=-0.193, p=0.005) had a weak negative correlation with MoCA scores. Conclusions: Our results indicate that patient age, tumor progression, treatment, and concomitant medications are associated with MoCA scores. Further analysis will evaluate tumor methylation status and blood-based biomarkers associated with occurrence to develop prediction models that can be evaluated in larger cohorts. Future investigations should compare MoCA scores to other forms of cognitive assessment, including patient-reported outcomes, incorporating multiple timepoints to monitor CD in patients throughout their treatment and survivorship care. Citation Format: McKenzie C. Kauss, Elizabeth Vera, Kimberly Reinhart, Hope Miller, Jaime Garcia, Morgan Johnson, Madhura V. Managoli, Maeve Pascoe, Kaitlynn Slattery, Alex R. Wollet, Mark R. Gilbert, Alvina Acquaye-Mallory, Terri S. Armstrong, Vivian A. Guedes. The Montreal Cognitive Assessment (MoCA) administered virtually or via Duo APP in primary brain tumor patients: a preliminary analysis [abstract]. In: Proceedings of the AACR Special Conference on Brain Cancer; 2023 Oct 19-22; Minneapolis, Minnesota. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_1):Abstract nr A030.