Introduction B-cell maturation antigen (TNFRSF17; BCMA) targeted immunotherapy has revolutionized treatment options for multiple myeloma (MM) patients. BCMA/CD3-targeting T cell engagers (TCE) are a recently developed immunotherapy that shows potent tumor killing activity in MM. While still under intense clinical development, TCE alone are not curative, and therapy relapse remains a clinical challenge. Our well-established, immunocompetent Vk*MYC mouse model provides an ideal controllable platform to model the full heterogeneity of the human bone marrow niche, identify mechanisms of relapse after TCE treatment, and test efficacy and toxicity of novel immunotherapy combinations. We recently published preclinical investigations using a murine BCMA/CD3-targeting TCE in the Vk*MYC model. Here, the TCE was safe and efficacious in most mice but failed in those with high tumor burden, consistent with clinical reports of TCE in leukemia. Our attempts to use pomalidomide with TCE, ultimately led to T cell exhaustion that compromised disease control. Alternatively, we reasoned that combination of TCE with a more effective cytotoxic therapy could debulk the tumor and help maintain immune function long-term. We found that a combination with cyclophosphamide was safe and allowed for a tempered proinflammatory response associated with long-lasting complete remission. Now we provide single cell RNA-seq data describing in granular detail the cellular flux underlying the preclinical success of TCE and cyclophosphamide combination.

Methods We treated tumor bearing mice with BCMA/CD3 TCE or a negative control antibody alone, or with cyclophosphamide. We performed single cell 5’ RNA-sequencing (scRNAseq) of splenic tissue containing tumor at the peak of immunotherapy response. scRNA-seq is a powerful tool for examining complete gene expression profiles at an unprecedented level of granularity. We examined CellRanger-generated count matrices with the Seurat R package to generate subsets of homogeneous cells according to nearest neighbor clustering of principal components, then visualized with uniform manifold approximation and projection (UMAP). We derived cell clusters based on expression data as well as the correlation structure between a large set of custom reference samples derived from the ARCHS4 repository of over 500,000 mouse RNA samples. This less-biased method informs us about definitive, and sometimes lesser-known, immune typologies in our high dimensional analysis of microenvironmental remodeling. We tested for differential gene expression between cell clusters using Wilcoxon Rank Sum tests. We pair these scRNA-seq data with flow cytometry and immunohistochemistry.

Results To characterize the features of the T cell subsets that contribute to the noted complete remission with TCE cyclophosphamide combination, we focused on the nine clusters of T cell subsets. Two clusters of T cells were enriched after cyclophosphamide lymphodepletion regardless of whether they were treated with TCE or TCE control: (i) a naïve CD8+ T cell cluster of stem-like cytotoxic cells expressing genes of recent thymic emigrants, and (ii) CD4+ T follicular helper cells. These cells help with germinal centers and memory, which may promote the protection and curative response we observed. Importantly regulatory T cells, as a ratio with total T cells, were reduced in mice treated with cyclophosphamide, TCE, and the combination compared to controls. Finally, we defined signatures of the T cell types activated by and proliferating in response to TCE and cyclophosphamide combination: (i) highly proliferative CD8+ T cells, and (ii) tumor infiltrating CD8+ T cells with high expression of cytotoxicity genes, cell migration genes, and genes associated with recent thymic emigrants. These data suggest that the combination therapy leads to an influx of supportive and highly functional T cell subsets that are optimal for redirection with TCE. Additional comparisons will be presented at the meeting.

Conclusions We discovered a curative combination of BCMA/CD3 TCE with cyclophosphamide in the Vk*MYC mouse model. These data suggest that the unique ability of cyclophosphamide in combination with TCE to reshape the tumor microenvironment by depleting suppressive and exhausted T cells while replenishing with supportive and cytotoxic stem-like T cell subsets to provide a tempered and sustainable anti-MM therapy.

Wheeler:Bristol Myers Squibb: Current Employment, Current equity holder in private company. Bergsagel:Janssen: Consultancy; Oncopeptides: Consultancy; GSK: Consultancy; Novartis: Consultancy; Pfizer: Consultancy. Chesi:Abcuro, Palleon Pharmaceuticals, Pi Therapeutics.: Patents & Royalties: Genetically engineered mouse model of myeloma.; Pfizer, Novartis.: Consultancy.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution