The Expression of Bcl-6 In Diffuse Large B Cell Lymphoma (DLBCL) Is Associated with Improved Progression Free Survival and Overall Survival Following Front-Line Rituximab-Chemotherapy

Abstract 5073

BCL6 proto-oncogene is one of the markers for germinal center origin in DLBCL and encodes a transcriptional repressor factor. Mutation or translocation affecting this proto-oncogene has been implicated in the pathogenesis of certain subtypes of DLBCL. Studies using both gene expression profiling and immunohistochemistry (IHC) have demonstrated that BCL6 gene or Bcl-6 protein expression alone or in combination with other markers for germinal center origin predicts a favorable outcome in DLBCL. However, the prognostic significance of Bcl-6 has been challenged by the incorporation of rituximab into the standard chemotherapy regimens of DLBCL. To elucidate the prognostic value of Bcl-6 in patients treated with rituximab -based chemo-immunotherapy (R+CHOP or R+DA-EPOCH), we retrospectively evaluated the clinical outcomes of DLBCL patients treated at Roswell Park Cancer Institute (RPCI) between 1997 and 2008. De Novo DLBCL patients were identified using the RPCI tumor registry and pharmacy database. Demographic (age, race, sex), clinical (stage, IPI score, ECOG performance status, etc.) and pathological characteristics were obtained for each patient. A total of 242 consecutive DLBCL patients (89F/153M) were included in the study. The median age was 60 years (19-91). Using IHC, Bcl-6 expression was reported to be positive in 123/190 patients (64%) and Bcl-2 expression was detected in 86/125 patients (69%). Demographics and clinical characteristics such as Ann Arbor stage, international prognostic index (IPI) score, performance status (ECOG), and number of cycles of chemotherapy (Median number of cycles=6) were equally distributed between Bcl-6(+) and Bcl-6(-) DLBCL patients. Most patients received R+CHOP (86.4%) or R+ DA-EPOCH (13.6%). The complete response (CR) rate of the entire cohort was 77.7% and there was no difference between Bcl-6(+) and Bcl-6(-) DLBCL patients (P=0.22). After a median follow up of 3.9 years, significant differences in the PFS and OS were observed between Bcl-6(+) and Bcl-6(-) DLBCL. The 5yr-PFS of Bcl-6(+) DLBCL patients was superior to Bcl-6(-) DLBCL (66% vs. 42%, P=0.001). Similarly, the 5yr-OS was better in Bcl-6(+) than in Bcl-6(-) DLBCL patients (78% vs. 63%, P=0.023). A sub-group analysis of DLBCL patients with Bcl-6 and Bcl-2 testing by IHC studies (N=119) demonstrated that Bcl-6 expression was a strong positive predictor of PFS and OS in Bcl-2(-) DLBCL (P=0.002 and P=0.037 respectively). In Bcl-2(+) DLBCL, while Bcl-6 expression was associated with a longer PFS (39 vs. 13 months) and OS (68 vs. 42 months) than in Bcl-2(+)/Bcl-6(-) DLBCL, the difference was not statistically significant (P=0.107 and P=0.174 respectively). In conclusion, our data suggests that, the expression of Bcl-6 by IHC is associated with a better PFS and OS in patients with DLBCL treated with rituximab-chemotherapy in the front-line setting. The positive predictive value of Bcl-6 expression appears to be negatively affected by Bcl-2 over expression. In the post-rituximab era, Bcl-6 negative DLBCL represent a subgroup of DLBCL with poor clinical outcomes. Additional work should be done to determine if the clinical significance of bcl-6 expression is still valid in germinal center B cell like (GC) and non-GC subsets of DLBCL. There is a clinical and scientific need to understand, at the molecular level, differences between DLBCL subtypes and to develop novel treatment strategies targeting cellular pathways driving each particular subtype of DLBCL.