The impact of Epstein-Barr virus status on primary CNS lymphoma survival.
e13528 Background: Primary CNS lymphoma (PCNSL) incidence is increasing among the elderly and immunocompromised. Especially for these vulnerable populations, a deeper understanding of the prognostic value of tumor biomarkers is necessary to recommend more targeted therapies. Our primary objective is to evaluate the predictive strength of immunohistochemical markers on PCNSL overall survival. Secondary objectives include estimating the impact of these biomarkers on event-free survival. Methods: We retrospectively analyzed PCNSL patients treated at Yale between 2000 and 2018. The primary endpoint is overall survival. Event-free survival, measured by time to relapse or death, is the secondary endpoint. Kaplan-Meier survival curves with log-log confidence intervals were used to estimate survival outcomes. Cox proportional hazards regression models were used to evaluate the statistical significance (at alpha = 0.05) of immunohistochemical markers. Results: One hundred ten subjects were analyzed. Not all biomarkers were available for every subject. The median age of the cohort is 64 years. Sixty-three patients died, and seventy-six experienced an event (progression or death). Surviving patients had a median follow-up of 4.36 years. The median overall survival is 2.63 years (95% confidence interval (CI): 1.13 to 5.98 years). Epstein-Barr virus (EBV) status by immunohistochemistry significantly impacted overall survival. Adjusting for age and immunocompromised status, EBV-positive patients had a higher risk of death than EBV-negative patients (n = 13, hazard ratio (HR) = 33.75, 95% CI: 1.61-708.68). Other significant biomarkers include the GFAP and S100 proteins. The median event-free survival is 1.13 years (95% CI: 0.66 to 2.12 years), and EBV-positive patients had an increased risk of event (HR = 38.23, 95% CI: 2.32-630.88). CD10, BCL2, BCL6, and MUM1 were not significant to predict either survival outcome. Conclusions: We describe a large cohort of PCNSL patients treated at a single institution with comparable survival outcomes to similar research yet conflicting evidence of biomarker significance. Multi-institutional studies may further clarify the prognostic role of immunohistochemistry for improved PCNSL patient survival.