作者
Charlotte Sleurs,Catharina M.L. Zegers,M. Ribeiro,Wouter van Elmpt,Jeanette Dijkstra,Alida A. Postma,Laurien De Roeck,Karin Gehring,Wouter De Baene,Margriet M. Sitskoorn,Maarten Lambrecht,Daniëlle B. P. Eekers
摘要
Abstract Background Cranial irradiation is a key component of neuro-oncological treatment, but can result in cognitive side effects. Preserving cognition from radiotherapy-(RT)-induced toxicity remains an ongoing debate. To spatially map radiotoxic effects in patients who underwent cranial RT, this study applied a voxel-based approach. Methods Cognitive assessments (Controlled Word Association (COWA), Hopkins Verbal Learning (HVLT-R), and Trail Making Tests (TMT A,B)) were conducted prospectively before, 6months and 1year post-RT in 111 intracranial tumor patients (18-80years). Reliable change indices indicated cognitive changes across timepoints. CT and T1-weighted MRI scans acquired at diagnosis were co-registered, normalized to standard space, and smoothed. Voxel-wise permutation-based regression analyses examined the relationship between RT dose and cognitive decline (α<.05 at cluster level). Results Images of 111 patients (Mdn age = 55.39 years; 47% male; lesions were gliomas (61%), meningiomas (18%), other (21%); in frontal (33%), temporal (25%), other location (42%)) were analyzed. Reliable decline was most pronounced at 6months, particularly on the TMT A (25.77%), TMT B (24.21%), and HVLT immediate recall (21%). At 1year, 20% of patients continued to show decline in TMT B. Higher RT doses to frontal gyri, temporal, occipital, and para-central regions were associated with declines in verbal fluency, memory, processing speed, and flexibility at both peak- and cluster-level. Conclusion Differential voxel-wise RT dose effects at peak versus cluster level suggest local and network-based recruitment of diverse functional regions and vulnerability to cranial RT. These insights may help re-define key regions at risk from a network-based perspective, preserving cognition in future RT planning.