Differential efficacy of bortezomib plus chemotherapy within molecular subtypes of diffuse large B-cell lymphoma

Abstract

Gene expression profiling of diffuse large B-cell lymphoma (DLBCL) has revealed distinct molecular subtypes that include germinal center B cell-like (GCB) and activated B cell-like (ABC) DLBCL. ABC DLBCL has a worse survival after upfront chemotherapy and is characterized by constitutive activation of the antiapoptotic nuclear factor-kappa B (NF-kappaB) pathway, which can inhibit chemotherapy. We hypothesized that inhibition of NF-kappaB might sensitize ABC but not GCB DLBCL to chemotherapy and improve outcome. As the proteasome inhibitor bortezomib can inhibit NF-kappaB through blocking IkappaBalpha degradation, we investigated bortezomib alone followed by bortezomib and doxorubicin-based chemotherapy in recurrent DLBCL. Tumor tissue was analyzed by gene expression profiling and/or immunohistochemistry to identify molecular DLBCL subtypes. As a control, we showed that relapsed/refractory ABC and GCB DLBCL have equally poor survivals after upfront chemotherapy. Bortezomib alone had no activity in DLBCL, but when combined with chemotherapy, it demonstrated a significantly higher response (83% vs 13%; P < .001) and median overall survival (10.8 vs 3.4 months; P = .003) in ABC compared with GCB DLBCL, respectively. These results suggest bortezomib enhances the activity of chemotherapy in ABC but not GCB DLBCL, and provide a rational therapeutic approach based on genetically distinct DLBCL subtypes. This trial is registered with http://www.ClinicalTrials.gov under identifier NCT00057902.

Figures

Figure 1

Study schema. (A) Clinical treatment paradigm. Patients initially received bortezomib alone at 1.3 mg/m2 on days 1, 4, 8, and 11 every 21 days (Part A) unless they had disease that the investigators judged to require immediate chemotherapy, as in cases of impending or ongoing organ compromise; these patients received only Part B. Patients with progressive disease in Part A later received bortezomib with DA-EPOCH (Part B). Molecular classification. Of 31 DLBCL cases analyzed by gene expression profiling, 16 were excluded due to ineligible subtype by classification or did not receive Part A, leaving 5 ABC and 10 GCB cases eligible for analysis of outcome. Of 24 paraffin- embedded tumor biopsies analyzed by immunohistochemistry, 12 of each were categorized as GCB and ABC (non-GCB) type. By combining both methods, cases were identified as GCB in 15 and ABC in 12 and included in the analysis of outcome with Part B. (B) Gene expression profiling for 15 biopsy samples that were classified as GCB (10) or ABC (5) DLBCL. Relative mRNA expression levels for 100 genes that distinguish ABC and GCB DLBCL are depicted according to the color scale shown. The probability that a sample is ABC or GCB DLBCL based on the Bayesian gene expression-based classifier is shown at the top.

Figure 2

Overall survival in patients with DLBCL. (A) Overall survival of 31 patients with de novo DLBCL who received DA-EPOCH-B. With a median potential follow-up of 49 months, the median survival was 8 months. B. (B) Overall survival of 27 patients with ABC or GCB DLBCL who received DA-EPOCH-B showed a median survival of 10.8 and 3.4 months, respectively (P = .0026). (C) Overall survival of 54 patients with de novo DLBCL whose disease progressed after R-CHOP. With median potential follow-up of 22.5 months for survivors, the median survival was 5.6 months. (D) Overall survival of 50 patients with ABC or GCB DLBCL whose disease progressed after R-CHOP showed a median survival of 5.8 and 9.5 months, respectively (P = .93).