ABSTRACT
Second primary diffuse large B cell lymphoma (spDLBCL) is defined as a metachronous tumor occurring after a first primary cancer. To date, while R-CHOP is the standard first-line treatment for de novo DLBCL, no available data show that R-CHOP is the optimal treatment for spDLBCL. This exploratory study aimed to investigate treatment of spDLBCL. From 2008 to 2015, the Poitou-Charentes general cancer registry recorded 68 cases of spDLBCL ≤ 80 years old, having received a first-line treatment with either R-CHOP (78%) or other regimens (22%). Patients without R-CHOP have worse overall survival in univariate (HR 2.89 [1.33-6.24], P = 0.007) and multivariate (HR 2.98 [1.34-6.67], P = 0.008) analyses. Patients without R-CHOP more frequently had PS > 1 (67% vs. 28%, P = 0.007) and prior chemotherapy (60% vs. 26%, P = 0.02), which suggests that both of these factors influence a clinician's decision to not use R-CHOP. Prior chemotherapy had no prognostic impact in univariate and multivariate analyses; this result could call into question the risk-benefit balance of not using R-CHOP to prevent toxicity. In our study, one DLBCL out of ten occurred after a first primary cancer, and as regards de novo DLBCL, R-CHOP appeared to be the best first-line treatment. Larger series are needed to confirm these results.