In this issue of Blood, Boidol et al report a highly T-cell prolymphocytic leukemia (T-PLL)-specific response to the B-cell lymphoma 2 (BCL-2) inhibitor venetoclax in an ex vivo high-throughput drug screen, as well as partial remissions (PRs) in 2 patients with relapsed/refractory (R/R) T-PLL and high disease burden, both of whom were previously treated with multiagent chemotherapy and alemtuzumab.1
T-PLL is a rare, clinically heterogenous neoplasm, initial management of which ranges from observation for indolent, asymptomatic disease to alemtuzumab-based induction in patients with symptoms, followed by consolidative stem cell transplantation.2 There are no randomized controlled trials that inform the management of T-PLL; pentostatin is active, and is often added in patients who do not attain complete remission with alemtuzumab alone, but the treatment of R/R T-PLL is an area of clear unmet need.2 There is interest in therapeutically targeting the JAK-STAT pathway, mutations in which are frequent in T-PLL,3 but clinical data are lacking.
In the report by Boidol et al, single-cell suspensions from fresh bone marrow, peripheral blood, or lymph node samples from 86 patients were exposed to 106 different compounds and functional drug screening was performed, including assessment of cell viability after 72 hours and tissue microarray generation. Venetoclax displayed the strongest differential response for T-PLL; expectedly, chronic lymphocytic leukemia (CLL) samples were exquisitely sensitive to venetoclax, whereas aggressive lymphoma and acute myeloid leukemia (AML) samples were variably sensitive. T-PLL samples were more sensitive than AML samples but less so than CLL samples. Ex vivo sensitivity to venetoclax correlated with levels of expression of BCL-2 protein, but not with levels of BCL extra long (BCL-XL) or myeloid cell leukemia 1 (MCL-1), the 2 other major antiapoptotic proteins in the BCL-2 family. Levels of MCL-1, a major mediator of resistance to venetoclax,4 inversely correlated with those of BCL-2. Based on these preclinical efficacy data, the investigators administered venetoclax on a compassionate-use basis to 2 patients with R/R T-PLL, both of whose cells exhibited ex vivo sensitivity to venetoclax. One had failed 3 lines of therapy with a very short (<3 months) response to the last treatment (bendamustine). This patient had a fairly dramatic clinical response to the first 20-mg dose of venetoclax, given as part of the standard dose ramp-up, accompanied by laboratory evidence of tumor lysis syndrome (TLS). Venetoclax had to be stopped prematurely because of sepsis, with a best clinical response of PR. The other patient, who had received 2 prior lines of therapy, did not experience TLS and only achieved clinical response at a venetoclax dose of 800 mg daily, which had to be escalated further, up to 1200 mg daily. This patient, also with a best clinical response of PR, experienced disease progression in his fifth month of venetoclax single-agent therapy. Venetoclax induced BCL-2 and BCL-XL expression in both patients, but had no impact on MCL-1 protein levels.
Venetoclax was developed by reverse engineering of its predecessor molecule, navitoclax, an antagonist of both BCL-2 and BCL-XL, to selectively inhibit the prosurvival function of BCL-2 and induce apoptosis in tumor cells while sparing megakaryocytes, which depend on BCL-XL for their survival.5 It is currently approved as a single agent for the treatment of patients with relapsed 17p-deleted CLL, but is being studied in combination with CD20 monoclonal antibodies and/or ibrutinib in both the frontline and salvage settings in CLL, while also being developed for use in patients with B-cell non-Hodgkin lymphomas and multiple myeloma.6 As monotherapy, venetoclax yielded a modest response rate of 19% in patients with R/R AML,7 although highly promising results have been reported in treatment-naive elderly AML patients in combination with hypomethylating agents or low-dose cytarabine. Indeed, given the central role of BCL-2 in the regulation of apoptosis, it is not surprising that venetoclax acts as a universal apoptosis sensitizer, lending itself to a very large number of rational combinations across tumor types.4 As far as single-agent activity of venetoclax is concerned, an important determinant of the same is the extent of dependence of the specific tumor type on BCL-2 for survival, and the relative importance of other antiapoptotic BCL-2 family proteins, specifically BCL-XL and MCL-1.4 For example, BH3 profiling has revealed a high degree of dependence of early T-cell precursor acute lymphoblastic leukemia (ALL) on BCL-2, correlating with in vitro and in vivo sensitivity to venetoclax.8 Similarly, even though the majority of pediatric ALLs in xenograft studies seem to require concurrent inhibition of both BCL-2 and BCL-XL (eg, with navitoclax) for optimal cell killing, mixed lineage leukemia (MLL)-rearranged ALL appears to be an exception, being highly susceptible to BCL-2 inhibition alone with venetoclax.9 Based on the importance and wide applicability of BCL-2 as a therapeutic target, the clinical indications for venetoclax will undoubtedly expand.
Other investigators have used high-throughput ex vivo drug-testing approaches in T-PLL.10 Interestingly, in these studies, the cyclin-dependent kinase (CDK) inhibitor SNS-032 emerged as the agent that all T-PLL samples were sensitive to, although many other drugs, including venetoclax, were effective. Surprisingly, despite a high prevalence of mutations affecting the JAK-STAT pathway, ex vivo responses to JAK-STAT inhibitors did not correlate with the same. One of the mechanisms through which CDK9 inhibitors like SNS-032 induce cell death is by transcriptionally downregulating MCL-1, and the combination of CDK9 inhibitors with venetoclax or navitoclax is a well-recognized synergistic strategy.4 The strength of the study by Boidol and colleagues lies in the clinical responses to venetoclax they describe. However, the underlying biologic heterogeneity of T-PLL is apparent even in these 2 cases, which differed markedly in terms of time to response and dose dependence of the clinical response. That relapse eventually occurred in the patient who lived long enough after initiation of venetoclax suggests activation of compensatory cellular survival pathways, arguing for the need for rational venetoclax-based combinations in this disease.
Conflict-of-interest disclosure: M.Y.K. reports receiving research support from AbbVie and Genentech, and consulting fees from Roche, AbbVie, and Genentech. P.B. declares no competing financial interests.