T-cell dynamics, CD8/NK ratio and tumor survival hallmarks determine response to T-cell bispecific forimtamig in myeloma.

T-cell bispecific (TCB) therapies have demonstrated promising therapeutic efficacies for patients with relapsed/refractory multiple myeloma. However, responses are transient and essentially all patients eventually relapse. Cytokine release syndrome and other on-target off-tumor toxicities are common. Together, these challenges ask for new approaches to optimize treatment and to eventually understand resistance mechanisms. This phase 1 study (NCT04557150) investigated mode of action and resistance mechanisms of forimtamig, a novel GPRC5DxCD3 TCB. Biomarker analyses included deep phenotyping of patients' tumor and immune cells, functional T cell characterization, cytokine release, soluble B-cell maturation antigen kinetics and minimal residual disease evaluation. Biomarker data analysis assessed the impact of different TCB treatment approaches on T cell-mediated tumor killing dynamics: subcutaneous vs intravenous administration, weekly vs every 4 days step-up dosing, and low vs high target doses. A machine learning model revealed T-to-natural killer (NK)cell ratio in blood and bone marrow as a strong predictor of response. Single cell transcriptomics showed increased DNA repair and proliferation as major tumor hallmarks associated with resistance, while T cell dysfunction signatures were not predictive of response. T cells from non-responding patients showed an increased cytokine release when exposed to higher forimtamig exposure in an ex vivo setting. T cells from responding and non-responding patients were equally fit in killing tumor cells ex vivo. Our results highlight that forimtamig treatment optimization may be possible by boosting T cell numbers over NK cells, targeting tumor intrinsic resistance mechanisms or individualized dosing approach.