Longitudinal Single-Cell Immune Profiling Reveals Distinct Dynamics of Response to Teclistamab in Patients with High-Risk Smoldering Myeloma Compared to Relapsed-Refractory Multiple Myeloma

Background

Teclistamab (TEC) is a bispecific T-cell engaging antibody that binds CD3 and B-cell maturation antigen (BCMA) and prior studies have highlighted the importance of T cell fitness in determining response to therapy. Given that immune dysregulation may worsen with disease progression, we hypothesized that response to immunotherapy may be better if administered early. We therefore examined immune single cell sequencing and T cell receptor (TCR) sequencing on peripheral blood mononuclear cells (PBMCs) samples obtained from patients treated with TEC in two different disease states of multiple myeloma (MM), an early disease state of high-risk SMM (HR-SMM) vs. a late state of relapsed-refractory MM (RRMM).

Methods

We performed single-cell RNA and TCR sequencing on 46 peripheral blood samples collected from patients with HR-SMM and RRMM treated with TEC. Specifically, we included 27 pre-treatment samples (n=19 HR-SMM and n= 8 RRMM) and 19 paired samples following TEC treatment (n=11 HR-SMM and n=8 RRMM). The post-treatment samples were collected at the time of MRD negativity for patients with HR-SMM or at the time of best overall response (BOR) for patients with RRMM. The time interval between samples was approximately 4-5 months for both groups. Single-cell Gene expression and TCR libraries were prepared with the Chromium Next GEM Single Cell 5’ kit. Overall, we profiled 123,321 immune cells (~2,078 cells/sample), comprising T cells, NK cells, B cells, Monocytes and Dendritic cells. For each cell population, feature selection, dimensionality reduction, and clustering were performed separately, to identify cell subpopulations. In measuring TCR repertoire diversity we accounted for variability in T cell numbers obtained by downsampling 100 cells per sample with 10,000 iterations and computing the Chao index by taking the mean diversity estimate across iterations.

Results

To investigate the degree to which clonal T cell expansion occurs post-treatment with TEC we systematically compared the TCR repertoire’s diversity between the baseline and post-treatment timepoints for patients with HR-SMM vs. RRMM using paired Wilcoxon’s rank-sum tests. Notably, we observed significant clonal T cell expansion post-TEC in patients with HR-SMM (fold change of the median 1/Chao Index Mean=3.3, p=0.02) which was present at the time of MRD negative state, but not in patients with RRMM who presented a more mixed pattern of changes in diversity post-TEC (p=0.63) and mixed disease response. This result suggests that T cells from earlier disease states such as HR-SMM may indeed have a higher propensity to expand post-T-cell engager therapy. To assess for differences in T cell functionality and how those may impact response to TEC, we isolated clonally expanded CD8+ effector memory T cells (TEMs) from post-therapy samples and performed differential expression analysis between cells from patients with HR-SMM and cells from patients with RRMM. We observed a significant upregulation of T cell activation markers post-therapy in clonally expanded CD8+ TEMs from patients with HR-SMM, such as the canonical activation marker CD69, and immediate early genes such as JUN and EGR1, as well as stemness and memory markers, such as TCF7 and SELL. In addition, compared to patients with RRMM with BOR < VGPR (n=4), the post-treatment CD8+ TEM clones of patients with HR-SMM showed decreased expression of TOX, a well-known exhaustion marker. These results indicate increased T cell activation post-therapy in patients with HR-SMM compared to patients with RRMM and further support the hypothesis that T cells from patients with HR-SMM may be more fit to respond to immunotherapy with T cell engagers.

Conclusion

Single-cell RNA and TCR sequencing of peripheral blood immune cells before TEC and at the time of MRD negativity/BOR revealed significantly higher clonal T cell expansion and increased expression of activation and stemness genes in patients with HR-SMM compared to RRMM. These findings suggest that patients with HR-SMM may exhibit more robust T cell responses to Teclistamab therapy which may lead to deeper responses in earlier disease settings such as SMM.