ID #224 Prophylactic Computerized Cognitive Training to Mitigate Cognitive Risk in Children undergoing Medulloblastoma Treatment: A Prospective, Multi-site, Randomized, Controlled Trial

Background Childhood medulloblastoma survivors are at significant risk for cognitive deficits that negatively impact quality of life. Computerized working memory training (WMT) is feasible and efficacious for childhood cancer survivors exhibiting cognitive concerns. We investigated prophylactic WMT to mitigate or prevent the emergence of cognitive late effects. Methods In a randomized, single-blind, controlled trial (SJMB12; NCT01878617) children were assigned to computerized WMT or waitlist control groups immediately following cancer-directed treatment. Participants completed cognitive measures at baseline, immediately post training/waitlist (10-12 weeks after baseline), and 6 months post training/waitlist. Results 89 of 121 participants who met a priori complier criteria [completed ≥ 20 of 25 training sessions/in control group with baseline and immediate post training cognitive assessments] were included in analyses (age at randomization= 11.4 ± 4.4 years; 60% male). There were no differences in compliers and noncompliers in age, sex, or disease risk. The rate of compliance in the WMT group was 61% (37/61 randomized). A general linear model adjusting for baseline performance and three stratification factors (age group [5-10 years: ≥ 11 years]; disease risk group; exercise intervention group), revealed greater improvement for the WMT group on the primary outcome (Spatial Span Backwards, p< .006) at immediate post training/waitlist. T-tests comparing change between groups from baseline to immediate post training on additional cognitive measures revealed significant improvements for the WMT group on measures of attention (Spatial Span Forward, p=.031) and working memory (Digit Span Backward, p= .001) but not processing speed, academic fluency, or parent reported cognitive concerns. Linear mixed effects models revealed immediate post training cognitive gains were maintained 6 months post training (p> .100 for slope decline). Conclusions Study findings indicate feasibility of prophylactic cognitive training, with significant gains in targeted skills. Evidence for transfer to distal skills was lacking. Future studies should investigate longer-term maintenance of benefits.