Pre-Clinical Activity of Navitoclax in TCR-Driven and Non-ALCL Mature T-Cell Lymphomas

Background: In contrast to many B-cell lymphomas that highly express BCL-2, BCL-2 expression is observed in <50% of most mature T-cell lymphoma (MTCL) subtypes, although a clinical trial investigating venetoclax in BCL-2 positive relapsed/refractory PTCL is ongoing (NCT03552692). In contrast, BCL-xL and/or MCL-1 are expressed in the majority of MTCL. Consistent with these findings, MCL-1 expressing anaplastic large cell lymphomas (ALCL) and typical (non-ETP) T-ALL, which highly express BCL-xL, have been shown by BH3 profiling to be dependent upon MCL-1 or BCL-xL, respectively. Among the more common MTCL subtypes, including cutaneous T-cell lymphomas (CTCL), the extent to which BCL-2 family members are a therapeutic vulnerability is uncertain. However, we have previously demonstrated that antigen- and costimulatory receptor signaling promote the growth and survival of these lymphomas and confers resistance to chemotherapy, and in conventional (non-malignant) T cells, these same signals upregulate BCL-xL, which is required for their clonal expansion and escape from cell death following T-cell activation. Therefore, given the increasing availability of selective BH3 mimetics and the need for improved therapeutic strategies, we were motivated to examine BCL-2 family member dependencies in these lymphomas.

Methods: Peptides generated from BH3-only proteins and selective BH3 mimetics were utilized to perform “BH3 profilng” in MTCL. Cell viability upon exposure to BH3 mimetics was determined using standard approaches.

Results: To pre-clinically assess BCL-2 family member dependencies and the mitochondrial apoptotic pathway in T-cell lymphoproliferative neoplasms, we performed BH3 profiling in 17 cell lines. Peptides derived from BIM and PUMA, which are promiscuous, were utilized to assess the extent to which cells are “primed” for apoptosis. Differences in apoptotic priming were not observed across subtypes, and most cell lines examined were highly primed. We next examined the dependence on specific antiapoptotic proteins by BH3 profiling using BAD (antagonizes BCL-2, BCL-xL, and BCL-w), HRK (antagonizes BCL-xL), MS1 (antagonizes MCL-1), and FS1 (antagonizes A1/BFL-1) peptides. BCL-2 dependence was not observed in the T-cell lymphoma cell lines we examined. However, and in stark contrast, most non-ALCL cell lines were BCL-xL dependent, whereas ALCL cell lines were largely MCL-1 dependent. In an effort to further confirm these results, and examine the extent to which peptide-based BH3 profiling predicts sensitivity to selective BH3 mimetics in MTCL, cell lines were treated with venetoclax, A1155463 (a BCL-xL selective antagonist), navitoclax (a BCL-2/BCL-xL antagonist), or S63845 (MCL-1 antagonist), and cell viability determined. Consistent with peptide-based BH3 profiling, all cell lines examined were resistant to venetoclax. ALCL cell lines were largely MCL-1 dependent, whereas the non-ALCL cell lines examined were largely BCL-xL dependent. Therefore, we examined BCL-xL and MCL-1 expression in these cells and observed a strong association between BCL-xL expression and both cytochrome c release in response to HRK peptide and cell viability in the presence of A1155463. A similar association was observed between MCL-1 expression and sensitivity to MS1 and S63845. Transcriptionally profiled MTCL specimens were stratified by MTCL classification (ALCL vs. non-ALCL) and in keeping with the data obtained using cell lines increased BCL-xL expression was observed in non-ALCL MTCL subtypes. Given the observed association between BCL-xL expression and vulnerability to BCL-xL antagonists, we sought to examined mechanisms that promote its expression. Both TCR signaling and CD28 costimulation are exploited by non-ALCL subtypes. We examined the extent to which TCR/CD28 stimulation regulates BCL-xL transcription. In MTCL specimens, a significant increase in BCL-xL transcription and sensitization to Bcl-xL antagonism were observed upon TCR/CD28 stimulation. A TCR-dependent GEM model was utilized, demonstrating that these MTCL are Bcl-xL dependent and sensitive to navitoclax.

Conclusions: These findings suggest that BCL-xL is a therapeutic vulnerability for MTCL subsets, particularly non-ALCL subtypes that are TCR dependent, and provide a robust pre-clinical rationale for future studies investigating navitoclax in these lymphomas.