MDB-67. MEDULLOBLASTOMA MOLECULAR SUBGROUPING AND OUTCOMES DATA OF A SINGLE CENTER FROM A LOW-MIDDLE-INCOME COUNTRY

Abstract

Listen

Translate article

Abstract BACKGROUND Medulloblastoma is the most common malignant childhood brain tumor. The molecular subgrouping of medulloblastoma is a major determinant of management in high-income countries. Subgrouping is still very limited in low- and middle-income countries, and its relevance to management has yet to be evaluated in this setting. We describe molecular findings and their implications in the outcome of a single tertiary care center. METHODS Children between 3 and 18 years diagnosed with medulloblastoma between April 2014 and December 2020 at Aga Khan University Hospital were included. Subgrouping was performed by nanostring through a collaboration with the Hospital for Sick Children, Toronto. RESULTS Thirty-five children were identified (85.17% male); the median age was 8. Nineteen patients were high-risk, including 9 with metastatic disease. In 28 children, there was a clear molecular subgroup: 4 WNT (11.4%), 7 Sonic Hedgehog (20%), 3 Group 3 (8.6%), and 14 Group 4 (40%) medulloblastomas. Molecular subgrouping was inconclusive for two patients (5.7%) and not done in 5 patients (14.3%). All patients underwent surgery; 91.4% received radiation therapy and chemotherapy. Of 35 children, overall Survival at 5 years was 100%, 57.1%, 66.7%, and 85.7% for WNT, Sonic Hedgehog, Group 3, and Group 4, respectively. WNT and Group 3 had 100% Progression-free survival, whereas Sonic Hedgehog and Group 4 had decreased progression-free survival of 42.9% and 85.7%, respectively. Average-risk patients had an overall survival of 93.8%, whereas high-risk patients exhibited a statistically significant lower overall survival of 63.2% (P=0.014). CONCLUSION Our cohort demonstrates excellent outcomes for WNT and Group 4, despite one WNT being metastatic and four Group 4 patients being high-risk. Modern risk stratification is an excellent predictor of survival, suggesting that treatment can be tailored according to subgrouping in the LMIC setting.