RHOA Deletion Downregulates CD19 and Promotes Dysfunctional Immune Microenvironments in CAR-T Resistant B-Cell Lymphoma

ABSTRACT CD19-directed chimeric antigen receptor (CAR)-T cells are breakthrough therapies for aggressive B-cell lymphomas, but less than half of patients achieve durable responses. We previously showed through whole-genome sequencing of tumors from CAR-T-treated patients that deletions of RHOA (3p21.31) are enriched in cases progressing after treatment. RHOA’s roles in resistance and pathogenesis are poorly defined, despite loss-of-function alterations that occur in ∼20% of newly diagnosed large B-cell lymphoma (LBCL) cases. We created RHOA-deficient LBCL systems and confirmed cell-intrinsic loss of response to CAR-19 in vitro and in vivo driven by CD19 downregulation. Impact on CD19, however, was variable and would not explain selection for RHOA deletion in newly diagnosed cases. We therefore created RHOA-deficient tumors in immunocompetent mice and found remarkable correlation with dysfunctional lymphoma microenvironment (LME) signatures in CAR-19 resistant patients. These LMEs are marked by a type 1-like immune infiltrate with terminally exhausted CD8 T cells, Th1-like CD4 cytotoxic lymphocytes (CTLs), and increased production of interferon gamma (IFNγ). RHOA-deficient tumor cells themselves have significantly impaired IFNγ responses, providing resistance to CD8 T cell clearance by way of diminished induction of major histocompatibility complex class I (MHC-I). These findings support a model that depletion of healthy effector populations by RHOA-deficient lymphoma is a key driver of immune dysfunction thwarting CAR-19 clinical responses. Overall, we describe for the first time how a single-gene alteration found recurrently in CAR-19-resistant LBCL contributes to treatment failures. Competing Interest Statement M.D.J. declares consultancy/advisory for Kite/Gilead and Novartis and research funding (Inst) from Kite/Gilead, Incyte, and Lilly. F.L.L reports a consulting or advisory role for A2, Adaptive, Adaptimmune, Allogene, Amgen, Bluebird Bio, Bristol Myers Squibb/Celgene, Caribou, ecoR1, Gerson Lehman Group, Iovance, Janssen, Kite a Gilead Company, Legend Biotech, Novartis; research funding from the US NIH National Cancer Institute (grants RO1CA276040, RO1CA244328, and P30CA076292), Leukemia and Lymphoma Society, Allogene, BMS, Novartis, 2Sevnty Bio, Kite, and Novartis; and patents, royalties, other intellectual property in the field of cellular immunotherapy. J.P.A. declares research funding from ADC Therapeutics, Genmab, AbbVie, and BeiGene and consultancy or advisory from ADC Therapeutics, Genentech, Genmab, AbbVie, Novartis, Lilly, and Regeneron. F.M. received honoraria from Medidata and Sanofi. J.Y.S. served on an advisory board for Kite. All other authors declare no competing interests. Footnotes Data Sharing Statement; Lead contact – requests for further information and resources should be directed to Dr. Jonathan Schatz (jschatz{at}miami.edu) Materials availability – Cell lines, plasmids, and other reagents are available from the lead contact with completed materials transfer agreement NGS data and code availability – PCAWG data available via the ICGC Data Portal (https://dcc.icgc.org/) and at EGA under study ID EGAS00001001692. The CAR19 WGS and RNA-seq data have been submitted to the dbGaP. scRNA-seq and RNA-seq from this study are in process to be submitted to GEO (GE, RRID: SCR_005012) and will resume upon full government operations. Article Category; Lymphoid Neoplasia The manuscript has been revised to critically expand upon the scRNA-seq experiments and the functional validation of these findings.