Role of TNFRSF17 and GPRC5D Structural and Point Mutations in Resistance to Targeted Immunotherapies in Multiple Myeloma (MM)

Adaptive T cell therapy using chimeric antigen receptor (CAR) T cells and T cell engagers (TCE) targeting BCMA and GPRC5D have demonstrated very encouraging responses in heavily pre-treated multiple myeloma (MM). Recent studies interrogating T cells repertoires and fitness correlated resistance to these immune based therapies with exhausted T cell states. While in lymphoma, antigen escape with CD19 loss is observed in up to 30% of patients with acquired resistance to adaptive therapies, in MM TNFRSF17 or GPRC5D structural mutations are thought to be less prevalent in recipients of BCMA or GPRC5D-targeting CARs or TCEs . Bi-allelic loss (n=2) and a missense truncating mutation (n=1) of TNFRSF17 with CAR T cells, as well as one case of biallelic loss in TNFRSF17 with TCE are reported to date with no reported cases of bi-allelic loss of GPRC5D. Further studies of the role TNFRSF17 or GPRC5D structural or point mutations to the immune escape to these adaptive therapies are needed.

Serial BM aspirates were collected from patients treated with BCMA or GPRC5D-targeting TCE or CARs prior to therapy and at relapse. MM cells were profiled with single cell mRNA (scRNAseq), single cell copy number (scCNVseq) using the 10x Genomics GemCode system and whole genome sequencing (WGS). Sequenced reads were aligned to hg38. Single cell reads were processed with CellRanger and downstream analyses with Seurat R package.

In order to define the prevalence of TNFRSF17 or GRPC5D structural variants and mutations in MM, we performed WGS (n=19, 14 pre-CAR/TCE, 5 post-CAR/TCE) as well as scCNVseq (n=33, 16 RRMM, 7 pre-CAR/TCE, 8 post-CAR/TCE) on sorted CD138+ cells from relapsed MM patients. WGS (100x coverage) identified at baseline a subclonal deletion at the TNFRSF17 gene locus in only 1 patient; this small subclone subsequently drove the progression post BCMA CAR T cell therapy with an acquired biallelic loss of TNFRSF17. Mono-allelic large and focal deletions of GPRC5D (chr.12p) were detected by WGS in 7 cases (5 deletions, 2 copy neutral LOH). scCNVseq studies identified more frequent subclonal monoallelic copy number losses of TNFRSF17 in 11.2% of the cells (27884 MM cells) prior to anti-BCMA therapy exposure with the deletions ranging from 200 kb to 2.3 Mb. A homozygous deletion of TNFRSF17 was noted in 2 patients at the time of disease progression, one post CAR T and the other post BCMAxCD3 TCE. Structural mutations at the GPRC5D locus were also identified by scCNVseq with monoallelic and biallelic deletions in 11.9% and 2.7% of the cells, respectively. Clonal biallelic loss of GPRC5D post GPRC5DxCD3 TCE was detected in one patient who harbored a monoallelic loss in 79.2% of the cells pre-TCE.

With regard to point mutations, in one patient progressing from a durable CR on BCMA TCE, we observed an acquired mono-allelic loss of TNFRSF17 (scCNV) coupled with de novo mutation in TNFRSF17 extracellular domain c.81G>C [p.(R27P)]. This point R27P mutation is predicted to be damaging to BCMA (Polyphen) with disruption of several H-bonds between R27 and C24, Q25, S29, S30 and T32 of BCMA extracellular aa residues. Importantly, molecular dynamic simulations of WT vs. mutant R27P using publicly deposited crystal structural of BCMA bound to anti-BCMA Ab (J22.9-xi) (PBD ID 4ZFO) demonstrated that this mutation disrupts all H-bonds between BCMA and the light chain of chimeric mouse/human anti-BCMA Ab (J22.9-xi). Functionally, primary CD138+ cells with mutant BCMA [p.(R27P)] as well as K562 cells transduced with mutant BCMA [p.(R27P)] were completely resistant to BCMAxCD3 cytotoxicity with teclistamab and elranatamab compared to cells with WT BCMA.

While structural alterations of TNFRSF17 and GPRC5D remain relatively rare, we have identified biallelic loss of TNFRSF17 post BCMA CAR T and BCMAxCD3 TCE therapy and herein reported the first case of biallelic loss of GPRC5D post GPRC5DxCD3 TCE treatment. Importantly, we also identified a monoallelic loss of TNFRSF17 (del16p) coupled with BCMA extracellular domain point mutation [p.(R27P)] that confers resistance to two BCMAxCD3 TCEs in MM. Therefore, in addition to structural alterations (biallelic loss of TNFRSF17 or GPRC5D), antigen escape resulting from non-truncating missense point mutations in BCMA extracellular domain also mediate resistance to targeted-immunotherapies.

Neri:Janssen: Consultancy, Honoraria; BMS: Consultancy, Honoraria; Pfizer: Consultancy, Honoraria; Sanofi-Aventis: Consultancy, Honoraria. Landgren:Janssen: Honoraria, Other: Independent Data Monitoring Committee (IDMC) member for clinical trials, Research Funding; Merck & Co., Inc.: Other: Independent Data Monitoring Committee (IDMC) member for clinical trials; Leukemia & Lymphoma Society: Research Funding; MMRF: Honoraria; Aptitude Health: Honoraria; Rising Tide Foundation: Research Funding; Tow Foundation: Research Funding; Riney Foundation: Research Funding; NCI/NIH: Research Funding; Theradex: Other: Independent Data Monitoring Committee (IDMC) member for clinical trials; Pfizer Inc: Consultancy; Amgen: Honoraria, Research Funding. Bahlis:Forus: Consultancy, Honoraria; Takeda: Consultancy; Karyopharm Therapeutics: Consultancy, Honoraria; Celgene: Consultancy, Honoraria; GSK: Consultancy, Other; Sanofi: Consultancy, Honoraria; Genentech: Consultancy; Amgen: Consultancy, Honoraria; Pfizer: Consultancy, Honoraria, Research Funding; AbbVie: Consultancy, Honoraria; Janssen: Consultancy, Honoraria, Research Funding.