Alec Starzinski

Multiple myeloma is characterized by frequent clinical relapses after conventional therapy. Recently, chimeric antigen receptor (CAR) T cells targeting B-cell maturation antigen (BCMA) has been established as a treatment option for patients with relapsed or refractory disease. However, although >70% of patients initially respond to this treatment, clinical relapse and disease progression occur in most cases. Recent studies showed persistent expression of BCMA at the time of relapse, indicating that immune-intrinsic mechanisms may contribute to this resistance. Although there were no preexisting T-cell features associated with clinical outcomes, we found that patients with a durable response to CAR T-cell treatment had greater persistence of their CAR T cells than patients with transient clinical responses. They also possessed a significantly higher proportion of CD8+ T-effector memory cells. In contrast, patients with short-lived responses to treatment have increased frequencies of cytotoxic CD4+ CAR T cells. These cells expand in vivo early after infusion but express exhaustion markers (hepatitis A virus cellular receptor 2 [HAVCR2] and T-cell immunoglobulin and mucin domain-containing-3 [TIGIT]) and remain polyclonal. Finally, we demonstrate that nonclassical monocytes are enriched in the myeloma niche and may induce CAR T-cell dysfunction through mechanisms that include transforming growth factor β. These findings shed new light on the role of cytotoxic CD4+ T cells in disease progression after CAR T-cell therapy.

Sotigalimab induces activation and trafficking of immune cells within the blood. A and B, Heat maps of cell frequency and cell surface marker expression as analyzed by CyTOF for T and NK cells (A) and antigen-presenting cells (B; pre n = 6, post n = 6). The count column shows the number of cells in each cluster. The “frequency signif” columns indicates statistical significance for differences in cell frequency of which there were none. Red and blue in the protein columns indicate significant up- or down-regulation of expression for the indicated cell cluster. C, UMAP plots of PBMCs analyzed by CITEseq showing cell annotation (top) and abundance of cells by timepoint (bottom; pre n = 6, post n = 6). D, Quantification of cell frequency as percentage of all immune cells by timepoint from the CITEseq data (pre n = 6, post n = 6). *, P < 0.05; **, P < 0.01, unmarked comparisons are not significant. E, Expression of cell surface proteins indicative of cell trafficking and activation via CITEseq for CD8+ and CD4+ T cells, and adhesion molecules in CD14+ monocytes, for pre- and post-sotigalimab (pre n = 6, post n = 6). For scRNAseq data in E, all comparisons are significant with an adjusted P < 0.05.

Patient clinical characteristics and sample availability.

Supplementary Table 4 from Neoadjuvant CD40 Agonism Remodels the Tumor Immune Microenvironment in Locally Advanced Esophageal/Gastroesophageal Junction Cancer

Treatment with sotigalimab increases immune infiltration into the TME, including T cells and myeloid cells. A, Patients were treated with sotigalimab at the initiation of the study (a), at the beginning of week 3 (b) and at week 8 (c). Chemotherapy and radiation were initiated after the second dose of sotigalimab (b). Tissue biopsies and blood samples were collected at timepoints pre-treatment, at week 3, and at surgery. Blood samples were used to assess changes to the peripheral immune system using CyTOF (1) and scRNAseq (3). Fixed tissue samples were sectioned, stained, and analyzed using MIBI (2). The tumor tissue was also dissociated into a single-cell suspension and analyzed by scRNAseq (3). B, Images from MIBI analysis from samples taken pre- and post-sotigalimab treatment for patient 04, showing markers for DNA, keratin (tumor cells) and CD45 for immune cells. The square at the center of the image indicates the 5x zoom image depicting DNA, keratin, CD3 and CD68 (representative image: Patient 04). C and D, The density of each immune cell type per tumor area (C) and the frequency out of total immune cells (D) is shown for each paired sample pre-sotigalimab (pre) and post-sotigalimab (post; n = 4). E, Uniform Manifold Approximation and Projection (UMAP) plot demonstrates all immune cell types within samples analyzed by scRNAseq (top). The abundance of cells within the UMAP is shown by density heat map for pre and post samples (bottom; pre n = 3, post n = 4). F and G, Distances between each tumor cell and the nearest immune cell are depicted (F) and quantified (G; n = 4). All statistical changes were measured by t test. *, P ≤ 0.05.

Sotigalimab results in immunomodulation and apoptosis of tumor cells. A, Tumor cells in the TME were characterized using MIBI analysis of keratin, HLA-I, HLA-II, PD-L1, and CD40 (representative image: patient 04). B, Quantification of tumor cell density by MIBI (pre n = 5, post n = 4). C–E, Expression of the phenotypic markers was also quantified in the MIBI analysis for HLA I and II (C), PD-L1 (D), and CD40 (E; pre n = 5, post n = 4). F–H, Differential expression by timepoint using scRNAseq of CD274 (F), HLA I and II genes (G), and apoptosis genes (H) on tumor cells. For scRNAseq data, comparisons denoted as significant have adjusted P < 0.05, and those without an asterisk are not significant.