Incidence and risk factors for central nervous system relapse in patients with diffuse large B-cell lymphoma: The impact of the addition of rituximab to CHOP chemotherapy

8037 Background: The addition of rituximab to CHOP (CHOP-R) chemotherapy improves outcome in patients with diffuse large B- cell lymphoma (DLBCL). However, limited data suggests that it may not impact the risk of central nervous system (CNS) relapse. We evaluated the risk of CNS relapse and associated predictive factors in a large population-based cohort of patients with DLBCL. Methods: All patients with DLBCL diagnosed from September 1, 1999 to January 14, 2005 at the British Columbia Cancer Agency were identified, including those from a recently published report (Sehn, J Clin Onc. 2005). Patients were included if they were ≥16 years old with advanced stage (stage III /IV, or stage I /II with B symptoms or bulky disease (≥10 cm)), or limited stage with testicular involvement and excluded if HIV-positive or known to have CNS involvement at diagnosis. From Sept 1999 to Feb 2001 patients were treated with CHOP- type chemotherapy. After March 1, 2001 patients received CHOP-R. Results: A total of 435 patients with DLBCL were identified; 125 (29%) were treated with CHOP and 310 (71%) with CHOP-R. There were 28 CNS relapses with no significant difference in frequency between CHOP-treated (11 (9%)) or CHOP-R-treated (17 (5%)) patients (p=0.202). The median time to CNS relapse (8.1 mos vs 6.7 mos) and median survival after CNS relapse (2.7 mos vs 3.8 mos, p=0.665) were similar in the CHOP and CHOP-R treated patients, respectively. In univariate analysis, testicular involvement, renal involvement, advanced stage, and LDH ≥2x ULN were associated with increased risk of CNS relapse. In a multivariate analysis, testicular (OR 7.44, p=0.013) or renal (OR 6.23, p=0.012) involvement, and stage IV disease (OR 4.43, p=0.034) were independent predictors of CNS relapse. Conclusions: Consistent with prior reports, the addition of rituximab does not appear to significantly influence the risk of CNS relapse in patients with DLBCL, possibly due to poor CNS penetration. Risk factors for CNS relapse in the rituximab era are similar to those noted in prior reports based on anthracycline-treated patients. The overall survival of patients with CNS relapse is dismal and new strategies to reduce its incidence warrant further evaluation. No significant financial relationships to disclose.