UCSF

Abstract Pediatric brain tumors affect more than 50,000 children in the US with nearly 5,000 new diagnoses annually. Although neurocognitive outcomes for this population show lower scores when compared to population means, there is great variability among individuals. In the current study, we aim to investigate the impact of cognition-related genes, KLOTHO, ApoE4, BDNF, COMT, and KIBRA, on neurocognitive outcomes in pediatric brain tumor patients. Utilizing an existing cohort of pediatric brain tumor patients from a diverse brain tumor population, we performed single nucleotide polymorphism (SNP) genotyping for KLOTHO, ApoE4, BDNF, COMT, and KIBRA. Neurocognitive evaluations were completed for these same patients at baseline and follow-up intervals, using validated, computer-based CogState batteries. Mixed modeling was done to correlate individual SNPs with neurocognitive outcomes over time, adjusting a priori for time from radiation, age at diagnosis, hydrocephalus, seizures, chemotherapy/radiation exposure, sex, tumor type, and tumor location. Lowess smoothing regression was done to assess the impact of individual genes SNPs on neurocognition. Ninety-seven patients were included in the analyses (54 males; median age at diagnosis 7 years). Medulloblastoma was the most frequent diagnosis (29%). Time from radiation, hydrocephalus at diagnosis, and seizures at diagnosis most consistently impacted neurocognition across neurocognitive test batteries (as based on statistical significance at p-value<0.05 across gene SNPs and testing batteries). Using smoothing regression, ApoE4 carrier status consistently negatively impacted neurocognitive function over time, as seen by decreasing neurocognitive scores over time across all batteries. Time from radiation and hydrocephalus and/or seizures at diagnosis illustrated statistically significant impacts on neurocognitive outcomes of pediatric brain tumor patients. KLOTHO, BDNF, COMT, and KIBRA did not appear to impact neurocognition; however, ApoE4 warrants further investigation as a possible predictor of neurocognitive outcomes. Our study highlights factors that may be of importance in anticipatory guidance and potentially, treatment-planning for children with brain tumors.