Primary mediastinal B-cell lymphoma (PMBCL) is an uncommon lymphoma for which existing data is limited. We utilized the National Cancer Database (NCDB) to evaluate PMBCL and the impact of radiotherapy (RT) on outcomes in the years following FDA approval of rituximab. We queried the NCDB for patients with PMBCL diagnosed from 2006 to 2011 and treated with multiagent chemotherapy. Kaplan-Meier overall survival (OS) estimates, univariate (UVA), and multivariate (MVA) Cox proportional hazards regression analyses were performed. Propensity score matched analysis (PSMA) was performed to account for indication bias and mitigate heterogeneity between treatment groups. 465 patients were identified with a median follow-up of 36 months. Median age was 36 years; 43% received RT. 5-year OS for the entire cohort was 87%, and for the no-RT and RT groups, 83% versus 93%, respectively. On UVA, OS was improved with RT (HR 0.34, P = 0.002). On MVA, RT remained significantly associated with improved OS (HR 0.44, P = 0.028) while Medicaid insurance status and increasing stage remained significantly associated with OS decrement. PSMA confirmed the OS benefit associated with RT. This analysis is the largest PMBCL dataset to date and demonstrates a significant survival benefit associated with RT in patients receiving multiagent chemotherapy in the rituximab era. More than half of patients treated in the United States during this time period did not receive RT. In the absence of phase III data to support omission, combined modality therapy with its associated survival benefit should be the benchmark against which other therapies are compared.
© 2016 Wiley Periodicals, Inc.