Abstract

Purpose Diffuse large B-cell lymphoma (DLBCL) includes at least two prognostically important subtypes (ie, germinal center B-cell–like [GCB] and activated B-cell–like [ABC] DLBCL), which initially were characterized by gene expression profiling and subsequently were confirmed by immunostaining. However, with the addition of rituximab to standard chemotherapy, the prognostic significance of this subclassification of DLBCL is unclear. Patients and Methods We studied 243 patient cases of de novo DLBCL, which included 131 patient cases treated with rituximab plus standard chemotherapy (rituximab group) and 112 patient cases treated with only standard chemotherapy (control group). The cases were assigned to GCB or non-GCB subgroups (the latter of which included both ABC DLBCL and unclassifiable DLBCL) on the basis of immunophenotype by using the Hans method. Clinical characteristics and survival outcomes of the two patient groups were compared. Results The clinical characteristics of the patients in the rituximab and the control groups were similar. Compared with the control group, addition of rituximab improved the 3-year overall survival (OS; 42% v 77%; P < .001) of patients with DLBCL. Rituximab-treated patients in either the GCB or the non-GCB subgroups also had a significantly improved 3-year OS compared with their respective subgroups in the control group (P < .001). In the rituximab group, the GCB subgroup had a significantly better 3-year OS than the non-GCB subgroup (85% v 69%; P = .032). Multivariate analyses confirmed that rituximab treatment was predictive for survival in both the GCB and the non-GCB subgroups. Conclusion In this retrospective study, we have shown that the subclassification of DLBCL on the basis of the cell of origin continues to have prognostic importance in the rituximab era.