T(11;14) and High BCL2 Expression Are Predictive Biomarkers of Response to Venetoclax in Combination with Bortezomib and Dexamethasone in Patients with Relapsed/Refractory Multiple Myeloma: Biomarker Analyses from the Phase 3 Bellini Study
Harrison, S; Cavo, M; De La Rubia, J; Popat, R; Gasparetto, C; Hungria, VTM; Salwender, H; Suzuki, K; Kim, I; Moreau, P; Spencer, A; Garg, M ...
Published in: Blood
Background: Overexpression of the anti-apoptotic BCL-2 protein promotes multiple myeloma (MM) cell survival. Venetoclax (Ven) is a highly selective, potent, oral BCL-2 inhibitor that induces apoptosis and has shown synergistic activity with bortezomib (B) and dexamethasone (d). Phase 1 studies in relapsed/refractory (RR) MM demonstrated encouraging clinical efficacy of Ven + d in t(11;14) MM and in a broader patient (pt) population in combination with Bd. Recent results from the Phase 3 BELLINI study of Ven vs placebo (Pbo) + Bd in pts with RRMM demonstrated that pts treated with Ven + Bd had improved clinical response rates and progression-free survival (PFS) vs Pbo, although the overall survival (OS) result was in favor of Pbo. Subgroup analyses showed different efficacy and survival outcomes based on tumor cytogenetics and BCL-2 expression. Results of pre-specified subgroup analyses and additional retrospective correlative biomarker analyses in the Phase 3 BELLINI study are described herein.Methods: BELLINI (NCT02755597) was a randomized, double-blind, multicenter Phase 3 study of Ven or Pbo + Bd in pts with RRMM who received 1-3 prior therapies and were either sensitive or naïve to PIs. Pts were randomized 2:1 to receive Ven 800 mg/day or Pbo + Bd. The following biomarker analyses were performed by central laboratory assessments of pre-treatment tumor samples: BCL-2 protein expression by immunohistochemistry (IHC) analysis of bone marrow (BM) core biopsies; BCL2 gene expression by quantitative PCR (qPCR) and cytogenetic abnormalities by interphase fluorescence in situ hybridization (FISH) analysis of CD138-enriched BM mononuclear cells. Correlation between BCL-2 (protein and gene) expression, cytogenetics, and outcomes were examined by Kruskal-Wallis tests and by hazard ratio (HR) using the Cox proportional hazard model.Results: As of the data cut-off of 18 Mar 2019, 291 pts were randomized, 194 to the Ven arm and 97 to the Pbo arm. Out of the 291 pts randomized, 177 pts (61%) were evaluable by IHC, 257 pts (88%) by qPCR, and 262 pts (90%) by FISH. A broad range of BCL2 gene expression was observed (median 2-DCt: 0.212 [range: 0-5.21]), which strongly correlated with protein expression (median 2-DCt 0.115 in BCL-2 IHC Low vs 0.277 in BCL-2 IHC High, p=0.0021). t(11;14) MM had the highest levels of BCL-2 expression (23/23 BCL-2 High by IHC; median 2-DCt 0.406 vs 0.212 in t(11;14) negative, p=0.0132), however high BCL-2 expression was not limited to the t(11;14) subgroup. Univariate analyses showed higher BCL2 expression in pt tumor samples with del(13q) (median 2-DCt 0.333 vs 0.159 in pts without del(13q), p=0.0008) and gain(1q) (median 2-DCt 0. 295 vs 0.180 in pts without gain(1q), p=0.0059). Bootstrapping and aggregating thresholds from trees (BATTing) was used retrospectively to identify an estimated threshold value for BCL2 expression (2-DCt ≥0.323) that could provide optimum selection of pts with maximum improvement in PFS when treated with Ven+Bd.Biomarker subgroups with the greatest PFS improvement were t(11;14) (HR=0.10; 95% CI: 0.02-0.46, p=0.003) and High BCL2 by qPCR (HR=0.26; 95% CI: 0.13-0.51, p<0.001; Table 1). Since the t(11;14) and High BCL2 patient populations do not completely overlap (20% of High BCL2 pts were t(11;14) and 54% of t(11;14) were High BCL2), a combined subgroup analysis was performed. For pts with t(11;14) or High BCL2, the median PFS was not reached in the Ven arm vs 9.9 mo in Pbo (HR=0.26, 95% CI: 0.14-0.48, p<0.001; Table 1). Higher overall response (ORR, 88% vs 70%), very good partial response or better (≥VGPR, 73% vs 33%), and complete response or better (≥CR, 42% vs 3%) rates were observed in the Ven vs Pbo arm (Table 2). Minimal residual disease negativity (MRD-, 10-5) rate was also higher for t(11;14) or High BCL2 pts in the Ven vs Pbo arm (19% vs 0%). Median overall survival (OS) was not reached in either arm but was similar between treatment arms for the combined group with t(11;14) or High BCL2 pts (HR=0.92, 95% CI=0.39-2.16, p=0.85). In contrast, in the t(11;14) negative and Low BCL2 pts, OS favored Pbo (HR=3.13, 95% CI=1.2-8.13, p=0.019; Table 1).Conclusions: Adding Ven to Bd demonstrates significant efficacy in pts with RRMM harboring either t(11;14) or tumor cells expressing high levels of BCL2. The benefit-risk profile appears to be favorable in these subsets of pts and additional studies to gain further safety and efficacy information are warranted.