A bendamustine resistance gene signature in diffuse large B-cell lymphoma and multiple myeloma

Abstract

Aim: Bendamustine is primarily used for treatment of indolent lymphomas but has shown efficacy in some patients with diffuse large B-cell lymphoma (DLBCL) and multiple myeloma (MM). Molecular-based patient stratification for identification of resistant patients, who will benefit from alternative treatments, is important. The aim of this study was to develop a resistance gene signature (REGS) from bendamustine dose-response assays in cultures of DLBCL and MM cell lines, enabling prediction of bendamustine response in DLBCL and MM patients. Methods: Bendamustine response was determined in 14 DLBCL and 11 MM cell lines. Using baseline gene expression profiles and degree of growth inhibition after bendamustine exposure, a bendamustine REGS was developed and examined for the risk stratification potential in DLBCL (n = 971) and MM (n = 1,126) patients divided into prognostic subtypes. Results: Bendamustine resistance significantly correlated with resistance to cyclophosphamide in DLBCL and melphalan in MM cell lines. The bendamustine REGS showed significantly lower bendamustine resistance probabilities in DLBCL patients with GCB subtype tumors and in tumors of the differentiation dependent centrocyte and plasmablast subtypes. In MM patients, pre-BII classified tumors displayed high bendamustine resistance probabilities and the plasma cell subtype had lower bendamustine resistance probability than memory cells. Furthermore, tumors belonging to the 4p14, MAF, and D2 TC subclasses consistently displayed high bendamustine resistance probabilities. Conclusion: Significant differences in predicted response to bendamustine were found in molecular subtypes of DLBCL and MM, encouraging validation in prospective bendamustine-treated cohorts with available gene expression profiles and follow-up data.

Keywords: Bendamustine; diffuse large B-cell lymphoma; multiple myeloma; resistance gene signature.

Figure 1

Dose-response curves for the bendamustine screens in DLBCL and MM cell lines. A: dose-response curves for the 14 DLBCL cell lines (DB, FARAGE, HBL-1, MC-116, NU-DHL-1, NU-DUL-1, OCI-Ly3, OCI-Ly7, OCI-Ly8, RIVA, SU-DHL-5, SU-DHL-8, SU-DHL-10, and U-2932) and 11 MM cell lines (AMO-1, KMS-11, KMS-12-BM, KMS-12-PE, LP-1, MOLP-2, MOLP-8, NCI-H929, OPM-2, RPMI-8226, and U266); B: ranked 50% growth inhibition (GI₅₀) corrected for individual growth rates of the cell lines with DLBCL cell lines to the left and MM cell lines to the right; C: ranked area under the positive part of the curve (AUC₀) values corrected for individual growth rates of the cell lines with DLBCL cell lines to the left and MM cell lines to the right. DLBCL: diffuse large B-cell lymphoma; MM: multiple myeloma

Figure 2

Correlations between drug response in HBCCL. A generalized pairs plot, showing correlations in response between two drugs in DLBCL and MM cell lines. Distributions and plots are based on values for area under the positive part of the curve (AUC₀) which are displayed on the axes, and the shaded area shows the 95% confidence interval. Data for melphalan in DLBCL was not included since there were too few melphalan treated cell lines available. Corr: correlation. _•P < 0.1, *P < 0.05, **P < 0.01, ***P < 0.001. DLBCL: diffuse large B-cell lymphoma; MM: multiple myeloma; HBCCL: human B-cell cancer cell lines

Figure 3

Bendamustine resistance probability in COO subtypes of patient cohorts. A: bendamustine REGS in ABC/GCB subtypes of DLBCL; B: bendamustine resistance probabilities using the bendamustine REGS in BAGS-DLBCL subtypes; C: bendamustine resistance probability using the REGS in BAGS-MM subtypes separately for each dataset; D: bendamustine resistance probability using the REGS in TC subclasses separately for each dataset. Significant differences between subtypes were tested using Kruskal-Wallis tests. DLBCL: diffuse large B-cell lymphoma; MM: multiple myeloma; BAGS: B-cell associated gene signature; REGS: resistance gene signature; COO: cell-of-origin; ABC: activated B-cell-like; GCB: germinal center B-cell-like