Abstract
Tumor mutational burden (TMB) and driver mutations are potential biomarkers to predict efficacy of immune checkpoint inhibitors (ICPI). Studies demonstrated malignant gliomas with high TMB in children and adults may preferentially benefit from treatment with ICPI. Higher TMB has been found to relate to lower incidence of driver mutations, but this relationship has not been studied in pediatric brain tumors. To characterize this association, comprehensive genomic profiling (CGP) was performed on 723 pediatric (≤21 years) brain tumor samples using DNA extracted from 40 microns of formalin-fixed paraffin-embedded tissue. CGP utilized hybridization-captured, adaptor ligation based libraries sequenced to a mean coverage depth of >500X for up to 315 cancer-related genes. TMB was calculated as mutations per megabase and categorized as low (0–6), intermediate (6–20), or high (20+). Analysis included 79 clinically relevant driver mutations; genomic alterations known to confer a selective growth advantage. Of 723 brain tumors, TMB was low in 91.8%, intermediate in 6.1%, and high in 2.1%. When excluding tumor suppressor genes (TSGs), there was an association between decreased incidence of driver mutations in tumors with high TMB (p<0.001). Additionally, we determined that BRAF alterations were not identified in high TMB tumors, but were enriched in low TMB tumors (p<0.01). When including TSGs, however, 93% of tumors in the high TMB cohort harbored a driver mutation; 63% and 70% in the low and intermediate TMB cohorts, respectively (p<0.05). There is an association between high TMB and incidence of TSG alterations (p<10-10), especially TP53 mutations (p<10–13). Of the 15 tumors with high TMB, 14 were high grade gliomas and had alterations in TP53. Three biallelic mismatch repair mutations identified were MSH2, MSH6, and PMS2; they were associated with a higher TMB (p<0.01). These represent populations in which ICPI may be more or less effective and further studies are needed.