In this issue of Blood, 1 report the results of the open-label phase 2 CONTRALTO study, which evaluated the additional benefit of adding the BCL-2 inhibitor, venetoclax (VEN), to common treatment backbones, such as rituximab and bendamustine plus rituximab, in patients with relapsed/refractory follicular lymphoma (FL).
The BCL-2 family is a key player in the balance between cell survival and cell death. Family members such as BCL-2, BCLXL, and MCL-1 are well known to be dysregulated in cancer, leading to a prosurvival advantage. Although attempts to inhibit specific BCL-2 member proteins have been partially successful, evolution of protein targets through upregulation/mutation, and signaling from the TME may ultimately lead to cell survival/progression.
The BCL-2 family is a key player in the balance between cell survival and cell death. Family members such as BCL-2, BCLXL, and MCL-1 are well known to be dysregulated in cancer, leading to a prosurvival advantage. Although attempts to inhibit specific BCL-2 member proteins have been partially successful, evolution of protein targets through upregulation/mutation, and signaling from the TME may ultimately lead to cell survival/progression.
Historically, various cytotoxic regimens coupled with a monoclonal antibody have been the mainstay of both frontline and relapsed therapy in FL. Recently, as understanding of pathogenic drivers of the malignant B cell evolved, alternative therapeutic strategies have been tested. Clinical trials utilizing agents targeting the tumor microenvironment (TME; lenalidomide) as well as select cell survival pathways (PI3K, BTK, SYK, EZH2) have demonstrated promising activity in relapsed disease.2 The fine balance between cancer cell death and survival is regulated in large through mitochondrial apoptosis and is tightly controlled through dynamic interactions between BCL-2 protein family members3 (see figure). BCL-2 proteins include antiapoptotic members such as BCL-2, MCL-1, and BCL-XL, as well as the proapoptotic BIM, BID, BIK, NOXA, BAX, and BAK. Not surprisingly, cancer cells have developed multiple prosurvival mechanisms resulting in dysregulation of multiple BCL-2 family proteins. In fact, member aberrations are found at various levels in nearly all B-cell malignancies, and the hallmark BCL-2 translocation occurs in >90% of FL.4 Early efforts to target BCL-2 were met with variable success, often due to ineffective target affinity and specificity as well as intolerable on-/off-target side effects.
VEN is a selective oral inhibitor of BCL-2 (BH-3 mimetic) that is currently approved for use in chronic lymphocytic leukemia (CLL) and acute myeloid leukemia. In 2017, Davids et al reported a phase 1 study of VEN in a range of non-Hodgkin lymphomas, including FL, mantle cell lymphoma (MCL), and diffuse large cell lymphoma (DLBCL). Overall responses were noted in 75% of MCL, 38% of FL, and 18% of DLBCL. Interestingly, the level of BCL-2 expression was not predictive of tumor lysis or overall response, suggesting a potential role of alternative BCL-2 family members in treatment resistance and/or survival.5 For this reason, and to target BCL-2’s role in chemotherapy resistance, subsequent studies explored the additional benefit of BCL-2 inhibition when given as part of a combination regimen. de Vos and colleagues6 treated 60 patients (32 with follicular) with VEN plus bendamustine and rituximab (BR) in a phase 1b study, reporting 75% overall response rate (ORR) with 38% CR in FL. The recommended phase 2 dose of VEN with BR was 800 mg continuously, which was associated with grade 3 neutropenia in 59% and grade 3 thrombocytopenia in 36% of patients.
In the CONTRALTO study, Zinzani and colleagues report the results of a multicohort phase 2 study of VEN combinations in relapsed/refractory FL. The trial enrolled patients to receive either ventoclax (800 mg daily) plus rituximab (n = 52) or to be randomized into a substudy (111 patients) comparing 6 cycles of BR plus VEN (800 mg daily) vs 6 cycles of BR alone (following a safety run-in). In the VEN plus rituximab arm, the ORR and CR rates were 35% and 17%, respectively. Following 6 cycles of therapy, both BR and BR plus VEN cohorts showed an 84% ORR, with 49% of BR plus VEN patients maintaining response at 1 year compared with 57% in the BR-alone arm. The addition of VEN to BR did not appear to significantly improve either complete remission rates (75% vs 69%) or progression-free survival. Furthermore, the addition of VEN to BR resulted in higher rates of neutropenia and thrombocytopenia compared with BR alone (34% vs 61% and 16% vs 57%). These and other adverse events led to higher rates of treatment discontinuation in the VEN plus BR arm (40% vs 4%). In addition, dose reductions and interruptions resulted in dramatically disproportionate drug delivery between arms. In the VEN plus BR arm, only 25%, 85%, and 67% of patients received 90% of planned VEN, rituximab, and bendamustine, compared with 98% and 97% in the BR arm.
The inability of the study to demonstrate a clear benefit of VEN addition to BR may in part be secondary to poor drug exposure, especially in the chemoimmunotherapy backbone, and future trials could benefit from exploring alternate dosing and schedules. However, the addition of VEN to rituximab monotherapy also demonstrated only marginal improvement over previously published studies with rituximab alone in similar populations. These results lie in stark contrast to recent studies in CLL demonstrating profound activity of VEN both as a single agent and in combination with anti-CD20 therapy.7
The reasons behind FL’s puzzling lack of response to BCL-2 inhibition remains an area of ongoing study. FL is characterized by intratumoral heterogeneity of BCL-2 expression and an increasingly recognized dependence on BCL-Xl. It is also possible that although t(14;18) is required for lymphogenesis, clonal evolution and subsequent genetic aberrations lead to redundancy, reducing the lymphoma’s dependence on BCL-2 for survival and maintenance.8 The TME also plays a key role in FL survival and evolution.9 Competing prosurvival signals arising from the TME could potentially further negate select inhibition of BCL-2 family members and contribute to VEN resistance.
Unfortunately, CONTRALTO adds to an ever-growing list of randomized trials in FL that have failed to show a significant benefit of novel agents to various clinical outcomes when substituted or added to traditional chemotherapy backbones (especially overall survival!). These experiences highlight not only the complexity of clinical trial design but also the biologic heterogeneity of FL. As such, there is a growing need for predictive biomarkers to help select appropriate populations for subsequent prospective studies, ultimately resulting in a higher chance of trial success, and more importantly, to better patient outcomes.
Conflict-of-interest disclosure: The author serves on the advisory board for Roche and receives research funding from Roche.
