(OA-06) Inactivation of TRAF3 and CYLD Drive Resistance to T-Cell Based Therapies in Multiple Myeloma by Promoting Inflammation and an Immunosuppressive Microenvironment that Impair T Cell Function

Adaptive T cell therapies using chimeric antigen receptor (CAR) T cells and T cell engagers (TCE) targeting BCMA on plasma cells have recently demonstrated encouraging responses in multiple myeloma (MM). However, responses are not universal and acquired resistance does invariably occur. We have recently reported (Lee H et al, Nat Med 2023) an enrichment of NF-κB activating mutations, including the deletion of its negative regulators, TRAF3 and CYLD, in patients with acquired BCMA antigen escape. Therefore, we herein investigated the impact of NF-κB alterations on the development of resistance to these therapies in MM.

Methods:

Through CRISPR Cas9 gene editing, we generated stable knockout (KO) clones of TRAF3 and CYLD in OPM2 and ARP1 MM cells, along with non-targeting Cas9 transduced cells as a negative control. TRAF3 and CYLD KO protein expression were confirmed by western blot analysis and the activation of the NF-κB canonical (p65) or non-canonical (p52) pathways validated by ELISA.

Results:

In flow cytometry-based cytotoxicity assays, we have first demonstrated that the inactivation of TRAF3 and CYLD drives relative resistance to T-cell-based therapies, including anti-BCMA CAR-T cells (ide-cel) or anti-BCMA TCE (elranatamab), as well as anti-GPRC5D (talquetamab). Transcriptome profiling (scRNA-seq) of TRAF3 and CYLD KO cells and their isogenic controls (Cas9 only) was then performed to define the factors mediating this resistance. Several differentially activating pathways, including the interferon type I and II and TNF-α signaling, were enriched in the KO cells when compared to their respective controls, suggesting a possible contribution of these pathways in the development of this acquired intrinsic resistance. Furthermore, in a multiplex cytokine assay on the supernatant of the KO cells, we observed elevated levels of multiple inhibitory factors and cytokines, such as sFas, sFasL, TGF-β3, TNF-β, IL-10, and IP-10, compared to the control cells. IP-10 (encoded by CXCL10), in particular, is known to suppress T cell function. Therefore, we next co-cultured OPM2 cells with ide-cel in the presence of IP-10 (200ng/mL) and observed a significant decrease in T-cell-mediated killing following the addition of this chemokine. Lastly, soluble Fas (sFas) is also recognized to interfere with T cell function by acting as a decoy receptor by blocking the interaction between Fas ligand on T cells and Fas on tumor cells. We have therefore evaluated and confirmed by flow cytometry the upregulation of Fas expression in TRAF3 or CYLD KO as the source for the increased levels of sFas.

Conclusions:

In summary, we have identified a novel mechanism of resistance to T-cell-adoptive therapies mediated by a tumor intrinsic inactivation of TRAF3 and CYLD that results in the upregulation of T cells' suppressive chemokine CXCL10 and the suppression of T cells' death receptor-mediated cell death through the shedding of sFas.