Impaired Death Receptor Signaling Mediates Cross-Resistance to Immunotherapy in MM

Abstract

The treatment landscape in various cancers is currently shifting from genotoxic drugs to immunotherapy, such as monoclonal antibodies, immunoconjugates, T-cell engaging antibodies (TCE) or CAR T-cells (CARTs). This is in particular true for multiple myeloma (MM), in which unprecedented treatment efficiencies are induced by these novel therapies; however, patients continue to relapse and a subset even faces primary resistance, with the underlying mechanisms poorly understood. The role of the death receptor signaling proteins FADD and BID in MM resistance towards immunotherapy has not yet been explored. Therefore, we first generated FADD and BID CRISPR-Cas9 knock out (KO) models in two independent MM cell lines, AMO1 and L363. We confirmed the functional impact of the KO genes on the apoptotic activity using apoptosis inducing, CD95 activating antibody (anti-FAS Ab). As hypothesized, no apoptosis was observed in our FADD KO clones following 24 hours treatment with 15µg/ml Fas antibody, compared to a 50% increase of apoptosis in the WT cells via Annexin-V PI staining. Activity of Caspase 3/7 and 8 increased 3-4 folds in the WT cells and no increase was seen in our KO models by using commercially available fluorometric assays. Detection of cleaved PARP in WT and its absence in KO clones via Western blotting further validated the inactivation of the apoptotic pathways in our KO models. Next, we treated our KO cells with 0,1µg/ml Daratumumab (Dara) along with 10:1 ratio of PBMCs from healthy donors and 5:1 and 10:1 effector to target ratio of BCMA-CARTs: Of note, killing efficacy by Daratumumab was significantly lower in our KO models compared to the WT cells (22% vs 47% cell death) after 24 hours of treatment. Similarly, BCMA-CARTs efficiency was dramatically reduced with 58% survival in our KO models compared to around 10% in WT cells. Strikingly, we did not observe such a difference in IC50 between the KO and the WT cells when we treated for 3-5 days with Bortezomib, Lenalidomide, Doxorubicin, Melphalan, or the approved ADC Belantamab mafodotin. To determine the functional profile of the CART cells we performed a commercially available cytokine quantification assay (Human IL-2, Human IFN- γ) in response to 20 hours of antigen stimulation with WT and KO cells and no difference in IL-2 and IFN-γ production was observed. In addition, we excluded CD38 or BCMA antigen loss in our KO model by Direct Stochastic Optical Reconstruction Microscopy (dSTORM) and confirmed no significant differences in receptor density /µm2 or antigen surface distribution between the WT and the KO cells. To determine the clinical relevance of our findings we screened for decreased FADD and BID expression levels in a pilot cohort of 24 newly diagnosed (NDMM) and 54 relapsed refractory MM (RRMM) patients and normalized expression levels with 26 CD138+ bone marrow plasma cell samples from healthy donors. We did not find significant expression differences between the cohorts (that did not include patients post TCE or CARTs). However, when we ranked the patients according to their gene expression of FADD and BID the ten patients with the lowest values were predominantly Daratumumab resistant (7/10 and 9/10, respectively), whereas in the patients with highest FADD BID expression Daratumumab resistance was less pronounced (3/10 and 4/10 respectively). In conclusion, our study highlights induction of apoptosis as the main mode of action of T-cell based immunotherapies in MM. Furthermore, we provide first evidence, that gene expression of FADD and BID is essential for the action of Daratumumab and BCMA CARTs but may be dispensable for the anti-tumor effects in MM of ADC, IMiD, PI or chemotherapeutic agents. In our in vivo screen, we identified patients with low expression of FADD and BID being resistant to Daratumumab, suggesting a potential resistance mechanism.