Abstract
INTRODUCTION: Brain tumors are the most common solid malignancy and leading cause of cancer-related deaths in infants. Current epidemiological data is limited by low numbers of reported cases. This study used a population-based approach with analysis of contemporary and historical survival curves to provide up-to-date prognostication. METHODS: Observational cohort analysis was performed using the Surveillance, Epidemiology and End Results (SEER) database. Infants with brain tumors diagnosed from 1973 to 2013 were categorized by the most common tumor types (diffuse astrocytic and oligodendroglioma, choroid plexus, embryonal, ependymal, medulloblastoma and pilocytic astrocytoma). The 1, 5 and 10 year survival was stratified by decade, with trends in management and outcomes analyzed. RESULTS: We identified 2996 affected infants satisfying inclusion criteria. All tumor types, except embryonal and choroid plexus, demonstrated improving survival with time. Infants with embryonal tumors showed a decline in survival from the 1970s to 1990s (p = 0.009), whereas infants with choroid plexus tumors had no change in survival. Infants with ependymal tumors experienced the greatest improvement in survival from 1980s to 1990s and 1990s to 2000s (p = 0.0001, p = 0.01), with 5-year survival probability improving from 28% (95% CI 15-42%) in the 1980s to 77% (95% CI 69-83%) the 2000s. The use of radiation declined from 1970 to 2000 for all tumors; however, radiation treatment for embryonal and ependymal subtypes increased after 2000. CONCLUSIONS: While overall survival for infants with brain tumors has improved from the 1970s onwards, not every tumor type has seen a statistically significant change. Given changes in management and survival, prognostication of infants with brain tumor should be updated.