OT-05 * H3.3-K27M IS A NEGATIVE PROGNOSTIC MARKER IN THALAMIC PEDIATRIC GLIOMA

Abstract

Pediatric gliomas are the most commonly diagnosed brain cancer in children, accounting for approximately 50% of all cases. Recently, we conducted genome-analysis of diffuse intrinsic pontine glioma (DIPG), which specifically arises in the brainstem. Through this, we identified a histone H3.3 mutation (a lysine-methionine substitution at position 27 [K27M]) in approximately 70% of cases. We then showed that H3.3K27M is a negative prognostic marker for DIPG patients independent of histology. We hypothesized a similar prognostic role for H3.3K27M in thalamic gliomas. Utilizing a newly optimized, ultra-sensitive digital droplet PCR assay, we analyzed 38 radiologically confirmed thalamic gliomas (13 low grade and 25 high grade gliomas). Median survival for patients with high grade thalamic gliomas was 1.35 years vs 8.71 years for low grade thalamic glioma patients. Three (23%) low grade and 11 (44%) high grade gliomas tested positive for the H3.3K27M mutation. 31 patients, clinically annotated for treatment, extent of resection, age and outcome were used to determine the prognostic implications of H3.3-K27M. Kaplan-Meier survival analysis revealed significantly worse overall survival of thalamic glioma patients harbouring the H3.3-K27M mutation versus wild type samples (log rank p < 0.001) with a median survival of 1.04 vs 5.81 years for H3.3K27M vs WT thalamic gliomas, respectively. Multivariate cox analysis demonstrated H3.3K27M mutation status to be independent of WHO grade as a predictor of overall survival with a hazard ratio of 6.5 (95% confidence intervals 1.78-24.4, p = 0.005). These findings provide the first evidence that H3.3K27M status is a negative prognostic indicator for thalamic pediatric glioma. Furthermore, we have optimized a digital droplet PCR assay to assess mutation status in small samples, including formalin-fixed paraffin embedded, with high sensitivity and specificity in a clinical setting.