Lower symptom burden strengthens the association between DTI-ALPS index and brain aging in head and neck cancer survivors.
Authors
Affiliations (7)
Affiliations (7)
- Department of Cancer Systems Imaging, Division of Diagnostic Imaging, The University of Texas MD Anderson Cancer Center, Houston, Texas, United States of America.
- Cancer Neuroimaging Research Program, The University of Texas MD Anderson Cancer Center, Houston, Texas, United States of America.
- Cancer Neuroscience Program, The University of Texas MD Anderson Cancer Center, Houston, Texas, United States of America.
- Cancer Survivorship Program, The University of Texas MD Anderson Cancer Center, Houston, Texas, United States of America.
- Department of Radiology, The University of Texas Medical Branch (UTMB), Galveston, Texas, United States of America.
- Department of Neuro-Oncology, Division of Cancer Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas, United States of America.
- Department of Pediatrics, Division of Pediatrics, The University of Texas MD Anderson Cancer Center, Houston, Texas, United States of America.
Abstract
Head and neck (HN) cancer survivorship is associated with persistent symptom burden and long-term brain-behavioral health alterations, yet mechanisms underlying brain aging remain poorly understood. We investigated whether HN cancer survivors exhibit accelerated brain aging, whether limbic white matter (WM) pathways contribute to these changes, and whether measures associated with glymphatic function are associated with reduced brain aging. Diffusion MRI and symptom burden data were analyzed in 135 HN cancer survivors and compared with a normative cohort from the Human Connectome Project (n = 1,228). Two machine learning models (elastic net [ENET] and random forest [RF]) using fractional anisotropy (FA) and quantitative anisotropy (QA) from 45 WM tracts were trained to estimate brain age and compute brain age gap (BAG). Glymphatic-related processes were indirectly assessed using diffusion tensor imaging analysis along the perivascular spaces (DTI-ALPS). Moderation analyses examined whether symptom burden influenced associations among DTI-ALPS, BAG, and WM integrity. Survivors showed a negative mean BAG of -1.64 years in the ENET model, indicating a nonsignificant tendency toward delayed brain aging, whereas the RF model showed a positive mean BAG of 3.59 years, indicating accelerated brain aging. The RF model demonstrated stronger external generalizability and greater biological coherence, including associations with ALPS and symptom-related measures. The bilateral fornix and right U-fibers were among the strongest contributors to RF-based brain age prediction. Higher DTI-ALPS index values were associated with lower BAG and greater fornix integrity, particularly in survivors with lower symptom burden and reduced activity-related interference. HN cancer survivorship was associated with accelerated brain aging in the model demonstrating greater external generalizability and biological relevance, alongside alterations in limbic WM pathways. Higher DTI-ALPS was associated with reduced brain aging, with symptom burden modifying this relationship. These findings highlight brain health as an important component of cancer survivorship.