Abstract
Medulloblastoma recurrence has historically been diagnosed based on identification of new contrast-enhancing lesions on surveillance imaging. We propose reduced diffusion is a more sensitive imaging marker than contrast-enhancement in identifying recurrent metastatic medulloblastoma. We identified 46 pediatric patients with medulloblastoma at our institution. Seventeen patients went on to develop definitive disease recurrence within the brain. In these patients, MR imaging at the time of diagnosis of recurrence and for 6 months prior to diagnosis was evaluated for presence of contrast enhanced lesions, size of contrast enhanced lesions, presence of lesions with reduced diffusion, and quantitative measurement of ADC within the recurrent lesions. All patients with recurrent medulloblastoma demonstrated reduced diffusion within recurrent lesions. ADC measurements were 0.658 ± 0.072 for recurrent lesions as compared to 0.923 ± 0.146 for contralateral normal white matter region of interest (ROI) measurements (p = 0.00001). Eleven patients (65%) with disease recurrence demonstrated contrast enhancement within the recurrent lesions. All six patients with non-enhancing recurrence demonstrated reduced diffusion with mean ADC of 0.680 ± 0.097, which was statistically significantly lower compared to contralateral normal brain with ADC values 0.894 ± 0.089 (p = 0.003). Non-enhancing recurrent disease was found equally within the tumor resection bed and mixed distal and tumor resection bed. In our cohort of patients with recurrent medulloblastoma, recurrent lesions did not uniformly demonstrate contrast enhancement on MR imaging but all recurrent lesions demonstrated reduced diffusion. Our findings support diffusion weighted imaging (DWI) is more sensitive than contrast enhancement for detection of medulloblastoma recurrence. As such, DWI should be considered part of standard of care imaging for medulloblastoma surveillance.