Neuro-Oncology · Nov 2026
CNSC-50 Cortical brain aging reveals treatment- and sex-dependent effects in melanoma patients
By Safari N, Gule-Monroe MK, Wintermark M et al.
Abstract Accelerated brain aging may occur in melanoma patients, potentially related to treatment effects, including surgery and systemic therapies, along with cortical alterations. Given known treatment- and sex-related differences in brain aging trajectories, cortical thickness (CT)-derived brain aging may show treatment- and sex-dependent effects and associations with disease duration. Structural MRI data from 633 healthy adults (mean age = 56.17 years; 279 males) were used to train Random Forest (RF) and Elastic Net (ENET) brain-age models. An independent cohort of treated melanoma patients (n = 64; mean age = 61.29 years; 33 males) underwent structural MRI following either chemotherapy (n = 51) or surgery (n = 13). CT across 68 Desikan–Killiany regions was extracted using FreeSurfer. Models were trained with nested cross-validation and applied to the melanoma cohort. Brain Age Gap (BAG) was calculated as predicted minus chronological age, with analyses adjusted for sex, imaging site, and time since first diagnosis (TFD). RF showed better generalization to the melanoma cohort than ENET, with lower prediction error (MAE = 10.42 vs. 17.81 years). Melanoma patients showed elevated BAG relative to controls (4.67 vs. -0.57 years), with increased BAG in both chemotherapy-treated (4.07 years) and surgery-treated (7.67 years) patients, without significant group differences. Key model contributors included bilateral superior temporal and posterior cingulate cortices, and the right pars opercularis. A significant interaction between TFD and sex was observed in surgery-treated patients (p = .035), but not in chemotherapy-treated patients (p = .41), with a significant positive association between TFD and BAG observed only in surgery-treated males (p = .027), not females. Melanoma patients show accelerated brain aging, potentially more pronounced in surgery-treated patients. Sex-stratified analysis further highlights the relevance of treatment modality and sex differences in brain aging. Given the small surgery-treated sample, larger-sampled studies are needed to confirm these findings.