In this issue of Blood, Zelenetz et al1 report the results of the phase 1b portion of the CAVALLI study, one of the first trials to explore a novel strategy of chemosensitization, by adding the B-cell leukemia/lymphoma-2 (BCL2) inhibitor venetoclax to chemoimmunotherapy for patients with non-Hodgkin lymphoma (NHL).
Although chemoimmunotherapy with rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone (R-CHOP) can be highly effective for patients with NHL, it is generally not curative in follicular lymphoma (FL),2 and even where there is curative potential, such as in diffuse large B-cell lymphoma (DLBCL), cure will be achieved in only a little over half of patients.3 Recently, several targeted small molecule agents have been approved in B-cell malignancies, and although each has antilymphoma activity, none of them are curative as single agents. Therefore, a concerted effort is underway to explore combinations of these novel agents, including with R-CHOP, in an attempt to enhance efficacy without significantly increasing toxicity. To date, most such efforts have been disappointing, with promising early results with drugs such as ibrutinib,4 lenalidomide,5 or bortezomib6 not translating into clear clinical benefits when explored in larger studies. Another strategy of substituting the newer CD20 monoclonal antibody obinutuzumab (formerly GA-101 [G]) for rituximab also failed to demonstrate significant benefit in DLBCL.7
Recently, a phase 1 study of the BCL2 inhibitor venetoclax demonstrated single-agent activity in both FL and DLBCL, where the overall response rates were 38% and 18%, respectively, with some patients with each histology achieving complete remission.8 These response rates are similar to, or even somewhat lower than, the response rates of other targeted agents that have later gone on to show no significant benefit in combination studies with chemotherapy when compared with chemotherapy alone. This raises the question: why would the fate of venetoclax be any different than that of these other active agents?
The reason may be related to the underlying molecular biology of the BCL-2 family of proteins, which controls the intrinsic, mitochondrial apoptotic pathway that is fundamental to the pathophysiology of both FL and at least a subset of DLBCL.9 Most lymphoma cells are highly primed to undergo apoptosis, meaning that these vulnerable tumor cells lie close to the apoptotic threshold. Therefore, when NHL patients are treated with R-CHOP, nearly all will have at least some initial response; however, patients may also harbor heartier tumor cells that are unprimed for apoptosis. These resistant cells can survive the initial onslaught by R-CHOP and later lead to relapse and progressive disease.
As a BCL2 selective inhibitor, venetoclax has the potential to increase the apoptotic priming of these previously unprimed cells. In other words, it could push these tumor cells closer to the apoptotic threshold, thereby making the proapoptotic signals induced by chemotherapy more effective, and eradicating even previously unprimed lymphoma cells. Unlike with other novel agent plus chemotherapy approaches, this novel chemosensitization approach of using venetoclax to boost the effects of chemotherapy may directly propel even the most resistant lymphoma cells toward apoptotic demise (see figure). Although this strategy has theoretical appeal, its significance in NHL patients had not previously been explored.
In the CAVALLI study, Zelenetz and colleagues report the results of a phase 1b trial combining R- or G-CHOP with venetoclax. Fifty-six patients were enrolled, most with FL (43%) or DLBCL (32%). Dose-limiting toxicities were reported in 3 of 14 patients at the first venetoclax dose (200 mg/d), after which dosing was changed from daily to 10 days per cycle and escalated to 800 mg. Cytopenias were the most common grade 3/4 adverse events reported and occurred at a higher rate than expected, particularly in the G-CHOP arm; however, the safety profile was generally manageable. Overall response rates in both arms were high at 87.5%. Notably, 7 of 8 DLBCL patients with the poor prognostic double-expressor immunophenotype (BCL2+ and MYC+) achieved complete remission. Although the maximum tolerated dose was not reached, the recommended phase 2 dose for venetoclax was established at 800 mg days 4 to 10 of cycle 1 and days 1 to 10 of cycles 2 to 8. Rituximab was chosen for further study mainly because the GOYA study did not demonstrate superiority of G-CHOP over R-CHOP in DLBCL.7
The current report focuses only on the phase 1b portion of this phase 1b/2 trial, and additional phase 2 data were presented at 2018 American Society of Hematology meeting.10 The presentation included an interesting comparison of the results for BCL2 immunohistochemistry-positive DLBCL patients compared with matched controls from the GOYA study of R-/G-CHOP alone. End of treatment complete response rate was improved in BCL2-positive subgroups, specifically in BCL2 fluorescence in situ hybridization–positive (70.0% vs 47.5%) and double-hit (71.4% vs 25.0%) patients, suggesting that this chemosensitization strategy with venetoclax may be of particular benefit in these poor prognostic DLBCL subgroups.
With a median follow-up of only 22 months in the responders in this report of the phase 1b portion of the CAVALLI study, it is uncertain whether the promisingly high overall response rates observed will translate into meaningful survival improvement compared with chemoimmunotherapy alone. A larger, comparative study will be needed to answer this question. Despite these caveats, Zelenetz et al provide compelling new data for the field of NHL, demonstrating the feasibility of combining venetoclax with chemoimmunotherapy and identifying a modified venetoclax dosing schedule to move forward. A similar approach is also being examined in studies combining venetoclax with R-EPOCH (rituximab, etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin) in Richter syndrome (#NCT03054896) and double-hit lymphoma (#NCT03036904). Chemosensitization with venetoclax should be further explored not only in patients with NHL but also for patients with other hematologic malignancies, and possibly also with solid tumors, because the same principles may also apply in some of those diseases.
Conflict-of-interest disclosure: C.H. received consulting fees from Gilead, Janssen, AbbVie, Roche, Genentech, and BMS. M.S.D. received consulting fees from AbbVie, Acerta Pharma, Adaptive Biotechnologies, Astra-Zeneca, Celgene, Genentech, Gilead Sciences, Janssen, MEI Pharma, Merck, Pharmacyclics, Roche, Syros Pharmaceuticals, TG Therapeutics, and Verastem; received research funding from Acerta Pharma, Bristol-Myers Squibb, Genentech, MEI Pharma, Pharmacyclics, Surface Oncology, TG Therapeutics, and Verastem; and received honoraria from Research to Practice. M.S.D. declares no competing financial interests.