Abstract
Primarily due to concerns about severe neurocognitive sequelae, immediate post-surgical craniospinal radiotherapy (CSRT) is often avoided in young children (<5 years of age) with MB; the price paid for such an approach in relative loss of disease control or the cognitive benefits are poorly characterized. In ACNS0331, 226 participants, aged 3-7, were randomized to either 18 or 23.4 Gy CSRT, along with local-boost RT and chemotherapy. We report outcomes for the subset between 3-5 years at diagnosis (n=83). 5-year EFS was 72 ± 5% and was similar in the 32 less than 4 years (67 ± 9%) and the 51 between 4 and 5 years of age (76 ± 7%). Although not adequately powered, there was no statistical difference in EFS based on CSRT dose (p=0.88). 81 neurocognitive assessments were obtained on 56 children. Initial median IQ was 97 (range 50-127) and dropped a median of 3 points at the second timepoint (27-48 months from diagnosis). Those <4 had a 9-point median drop compared to 2.5 in older children; there was extensive individual variability in the degree of IQ change (range = -23 to 13). There was no statistical difference in IQ as a function of CSRT, although sample size was small (p=0.52). Omissions in CSRT in this age cohort should be done cautiously, especially in those between 4 and 5, given the 5 year EFS and the concerning, but not devastating, neurocognitive declines. Pending molecular subgrouping may help in deciding which patients have the most to lose or gain by CSRT avoiding regimens.