Abstract 2683▪▪This icon denotes a clinically relevant abstract BackgroundSecondary central nervous system (CNS) involvement in diffuse large B-cell lymphoma (DLBCL) includes CNS relapse or CNS involvement with systemic disease progression. Although many publications have provided information regarding the incidence and risk factors for CNS involvement in DLBCL, its incidence reported across those studies varies widely. It might be related with that the majority of data were from retrospective analyses. Furthermore, the role of CNS prophylaxis for DLBCL has been challenged, especially in the era of R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone). As a result, this rare but fatal clinical problem still remains a therapeutic dilemma in the management of DLBCL. In this study, we prospectively explored the risk factors of CNS involvement and the clinical impact of screening evaluation for CNS involvement. MethodsWe analyzed the incidence of secondary CNS involvement in pathologically confirmed DLBCL patients enrolled in the Prospective Cohort Study with Risk-adapted Central Nervous System Evaluation in Diffuse Large B-cell Lymphoma (PROCESS study, NCT01202448). Patients should be treated with at least one cycle of R-CHOP, and provide written informed consents. We assessed the risk of CNS involvement based on previously reported risk factors: serum LDH elevation, the number of extranodal involvements, serum albumin, bone marrow invasion, HIV positivity, the involvement of testis, breast, paranasal sinus, bone, retroperitoneal lymph nodes, orbit, and epidural space. If patients had any of these risk factors, they underwent CSF study to screen the CNS involvement at diagnosis. If the results were abnormal, additional studies including brain MRI could be done depending on physicians' decision. CNS prophylaxis was done with intrathecal chemotherapy with methotrexate for patients who had positive findings of screening evaluation or were determined to have a risk of CNS involvement based on physicians' decision. Results564 patients were enrolled between 2010 and 2012 from 26 institutions belonged to the Consortium for Improving Survival of Lymphoma (CISL). They were prospectively monitored with the median follow-up duration of 10.5 months. The median age was 59.5 years old (range 20–89 years), and approximately a half of patients had Ann Arbor stage III/IV (n = 276, 48.9%) and 193 patients involved two or more than two extranodal sites (34.2%). Based on the International Prognostic Index (IPI) risk, 192 patients belonged to high or high-intermediate risk (34%). Among patients (n = 368) who had at least one of risk factors for CNS involvement, 243 patients underwent CNS evaluation, and the evidence of CNS involvement was found in16 patients including positive cytology (n = 11), and brain parenchyma lesion (n = 5). The other 78 patients showed equivocal results of CSF analysis including the presence of atypical cells (n = 17). Intrathecal prophylaxis was done for 51 patients whereas high dose methotrexate chemotherapy was combined with R-CHOP for patients with brain lesion. During follow-up, 14 cases of additional CNS involvement including brain parenchyma (n = 8), leptomeningeal (n = 5), and ocular invasion (n = 1) were observed. The median time to CNS event in these 14 patients was 7.5 months (range 1.2 – 15.9 months). Thus, 30 cases of secondary CNS involvement were documented in our study population at the time of analysis (5.3%) including 16 cases at diagnosis and 14 cases during follow-up. The univariate analysis for evaluation of risk factors demonstrated serum LDH, the number of extranodal involvements, bone marrow invasion, and the involvement of retroperitoneal lymph nodes, breast, paranasal sinus and orbit were significantly associated with CNS involvement. The high/high-intermediate risk of IPI was also predictive of CNS involvement (P < 0.05). However, in the multivariate analysis, bone marrow invasion and the involvement of breast, paranasal sinus and orbit were independently predictive for CNS involvement. ConclusionsThe incidence of secondary CNS involvement in DLBCL patients treated with R-CHOP was around 5%, and a half of cases had the evidence of CNS involvement at diagnosis. Considering a particular risk of CNS involvement of disease-related factors, risk-adapted active screening against CNS involvement may help to improve treatment outcome of patients with DLBCL. Disclosures:No relevant conflicts of interest to declare.