Background: Brain tumors are the leading cause of death from disease in children. Racial/ethnic minority children have poorer outcomes than White children; however, it is not clear whether this association is mediated by treatment received.
Methods: Children (aged 0-19 years) diagnosed with brain tumors in the National Cancer Database (2004-2016) were identified. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) between race/ethnicity (Black, Hispanic, Asian/Pacific Islander, American Indian/Alaska Native, or White [reference]) and death. An inverse odds weighted mediation analysis was performed with treatment received as the mediator.
Results: Among 22,469 cases, White children (69% of the sample) had significantly better overall 12.5-year survival (P < .01). Black children (13% of the sample) and Hispanic children (14% of the sample) had an increased risk of death overall and for glioblastoma and oligodendroglioma. Compared with Whites, Asian/Pacific Islander children had a higher risk of death from choroid plexus tumors and a lower risk of death from medulloblastoma. There were no statistically significant meditating effects by treatment received, although the estimate was borderline in Hispanic children (indirect HR, 1.08; 95% CI, 0.99-1.18). A treatment-independent association between race/ethnicity and death remained for Hispanic children (direct HR, 1.18; 95% CI, 1.04-1.33) and Black children (direct HR, 1.28; 95% CI, 1.13-1.45). If deaths in minorities had equaled those in White children, 5% fewer total deaths and 15% fewer minority deaths would have occurred.
Conclusions: Survival disparities exist in pediatric brain tumors and are largely independent of treatment received, but other mechanisms linked to race/ethnicity remain important.
Keywords: mediation analysis; pediatric brain tumors; racial disparities; survival; treatment differences.
© 2022 American Cancer Society.