Expanding the universe of T-PLL targets Free

In this issue of Blood, Pohly and colleagues present data from systematic, high-throughput drug screens for T-cell prolymphocytic leukemia (T-PLL), demonstrating differential sensitivity of T-PLL cells to therapeutic agents targeting nuclear export, autophagy, and apoptosis.¹ 

T-PLL is a rare malignancy of mature T-cell origin, with poor clinical outcome. Alemtuzumab-based chemoimmunotherapeutic regimens, consolidated whenever appropriate with autologous or allogeneic hematopoietic cell transplantation, remain a cornerstone of contemporary therapeutic management of T-PLL. Notwithstanding an initial response rate of ∼80% with alemtuzumab, the frequency of relapse and rapidity of progression renders current treatment strategies for T-PLL inadequate.² Novel therapeutic approaches could emerge from the targeting of recurrent genetic lesions that contribute to T-PLL leukemogenesis. In this respect, a recent study demonstrated p53 activators, such as MDM2 inhibitors and histone deacetylase inhibitors, to be among the most pharmacologically active agents against T-PLL cells.³ Although promising, these agents are associated with several potential limitations that could hamper their clinical translation, including treatment-related toxicity arising from p53 activation in healthy cells, therapeutic resistance in tumors harboring inactivating TP53 mutations, and selection for TP53-mutant clones in the presence of continuous p53 activation.⁴ In the light of these limitations, the continuous search for additional therapeutic targets in T-PLL remains highly relevant.

Advances in genome-scale technology, laboratory automation, and machine-learning approaches have facilitated the advent of unbiased, high-throughput drug screens that are gaining traction as essential tools for cancer drug discovery.⁵ With the potential to uncover hitherto unrecognized disease-specific therapeutic vulnerabilities and novel oncogenic mechanisms, these drug screens take various forms spanning CRISPR/Cas9-based genome-scale editing, to the systematic testing of hundreds of tumor biospecimens against thousands of therapeutic compounds using well-defined readouts as surrogate markers for tumor cytotoxicity. Given that high-throughput genome editing of primary tumor cells remains technically challenging, the latter approach predominates in ex vivo screens. Indeed, a previous small-scale ex vivo drug screen involving biospecimens from 39 T-PLL patients revealed cyclin-dependent kinase inhibitors, BCL2 inhibitors, histone deacetylase inhibitors, and p53 activators to be among the most efficacious drug classes against T-PLL samples.⁶ 

In their investigation, Pohly and colleagues extend these studies through an integrative analysis of 5 ex vivo discovery drug screens featuring >2500 unique compounds and tumor samples from 764 patients, including 61 T-PLL patients. Such an extensive repertoire of therapeutic agents and primary biospecimens represents a considerable scaling up of previous efforts, allowing the identification of additional therapeutic vulnerabilities that may not have been detected with earlier smaller studies. In particular, the testing of multiple compounds targeting individual oncogenic pathways provide added confidence of their functional relevance. Moreover, the inclusion of biospecimens from other T-cell malignancies allows identification of pathway dependencies that are shared across T-PLL and other T-cell lymphomas, as well as those that may be distinct to T-PLL. Finally, drugs initially found to display T-PLL–specific activity were further validated in a secondary screen attesting to the robustness of the experimental approach. All in all, in addition to confirming previously described T-PLL sensitivities toward MDM2 inhibitors, BCL2 inhibitors, and epigenetic modifiers, the authors have uncovered profound T-PLL–specific vulnerability to compounds targeting autophagy (eg, bafilomycin A1), nuclear export (eg, selinexor), and inhibitor-of-apoptosis proteins (eg, birinapant) that represents novel pathway dependencies (see figure). These therapeutic vulnerabilities were recapitulated across a range of T-cell malignancies but not in B-cell tumors such as chronic lymphocytic leukemia (CLL) or mantle cell lymphoma (MCL).

To discern the mechanistic basis of these novel T-PLL targets and identify biomarkers of sensitivity, the investigators employed an elaborate, high-throughput, systems-based approach incorporating bulk- and single-cell RNA sequencing and spectral flow cytometry to interrogate the effect of drug perturbation(s). Focusing on birinapant as the compound with a unique mode of action,^7,8 they observed a trend toward its cytotoxic effect being dependent on the level of expression of cIAP2 (BIRC3) and cFLIPL (CFLAR) in T-PLL cells, although this finding necessitates further validation. Furthermore, transcriptomic analysis identified a birinapant-induced upregulation of the tumor necrosis factor-α/NF-κB signaling pathway that likely underpins its leukemia/lymphoma-specific activity. Accordingly, the mechanism of tumor cell death was tumor necrosis factor-α–dependent and involved inflammation-promoting necroptosis. Importantly, after exposure of T-PLL cells to birinapant, the normal T-cell fraction was preserved, with a shift toward naive and central memory T-cell states. Although these initial forays into addressing questions of drug mechanisms are informative, lingering questions remain. For instance, it is not entirely clear why birinapant induces necroptosis in T-PLL but caspase-dependent apoptosis in CLL and MCL. Moreover, drug screens that evaluate tumor responses outside their native microenvironmental niches may not sufficiently reflect the drug effect on tumor-immune interactions that could contribute to its in vivo efficacy or resistance. Thus, the reported effect of birinapant on healthy T- and NK-cell subsets and their cytokine production, while fascinating, needs further exploration in relation to its impact on T-PLL cells. Nevertheless, Pohly et al provide compelling rationale for additional mechanistic study and the eventual clinical evaluation of this potentially important target.

What is next for drug discovery studies in T-PLL? With the ever-increasing scale and sophistication of drug discovery screens, more promising therapeutic targets are likely to emerge. We must not, however, become lost in the “surfeit of riches.” The holy grail of personalized cancer medicine requires treatments to be appropriately matched to the biology of individual tumors. As evident from the drug screens reported by Pohly et al, there is marked variation of response to individual drugs reflecting the biological heterogeneity of T-PLL. In CLL and lymphoma, our acknowledgment of tumor heterogeneity is resulting in an avalanche of effort to stratify tumors according to their genomic, transcriptomic, epigenetic, and clinical properties,⁹ and drug screens are being directed toward understanding the unique drug sensitivities of each biological subtype.¹⁰ Replicating such an effort in a rare disease like T-PLL will be a formidable task, but perhaps one that is necessary to achieve tangible therapeutic advance. International, multicenter, collaborative studies that provide the essential “critical mass” of clinical samples and resources will be indispensable to achieving this goal.

Conflict-of-interest disclosure: The authors declare no competing financial interests.