T cell receptor-engineered T cells targeting a human endogenous retrovirus in kidney cancer.

Abstract

4542 Background: Previously, our group described a novel human endogenous retrovirus type E (CT-RCC HERV-E) that is selectively expressed in most clear cell renal cell carcinomas (ccRCC). Here, we report the characterization and preclinical testing of a T cell receptor (HERV-E TCR) that targets an HLA-A11-restricted 10-mer peptide antigen (CT-RCC-1) that is derived from CT-RCC HERV-E. Methods: From a patient with ccRCC who had prolonged immune-mediated tumor regression following an allogeneic hematopoietic stem cell transplant, we identified a CD8+ T cell clone with HLA-A11-restricted specificity for the CT-RCC-1 peptide. The TCR of this clone was encoded into a retroviral vector containing a truncated CD34 cassette as a selection marker. Gene-engineered T cells expressing the HERV-E TCR (HERV-E T cells) were characterized for tumor recognition, lysis, and cross-reactivity. In vivo anti-tumor activity was assessed in a xeno-murine model using a human subcutaneous ccRCC tumor graft. A good manufacturing practice (GMP)-compliant method to produce HERV-E T cells was implemented for a subsequent clinical trial in humans with metastatic ccRCC. Results: When transduced into human T cells, the HERV-E TCR showed CD8-dependent specific recognition and lysis of HERV-E-expressing ccRCC cells that were HLA-A11+. The extent of tumor cytotoxicity correlated with CT-RCC HERV-E mRNA expression levels and HLA-A11 surface density (r=0.82). Cross-reactivity assessment established that there were no naturally occurring peptides with substantial sequence identity to CT-RCC-1 that could serve as alternative targets for HERV-E T cells. Further, immunopeptidomics and in silico analyses provided strong evidence that presentation of the CT-RCC-1 epitope depends exclusively on the transcription of the CT-RCC HERV-E genomic region. In a murine model, human HERV-E T cells mediated regression of established human ccRCC tumor grafts, significantly prolonging animal survival compared to controls that either received non-transduced T cells or no T cells (median survival 50 days vs. 20 and 20 days, respectively; p<0.001). Finally, the GMP-compliant production of HERV-E T cells yielded pure populations (i.e., >90% TCR-transduced) of HERV-E T cells that were highly cytotoxic to ccRCC tumor cells. The expansion numbers of GMP-produced cells were sufficient for testing the safety of the adoptively infused T cells in a phase I clinical trial. Conclusions: Here, we provide the first data showing T cells armed with a HERV-E TCR acquire specific anti-tumor activity in vitro and in vivo against ccRCC cells. These preclinical data provide the foundation for an ongoing first-in-human phase I clinical trial evaluating the safety of HERV-E T cell infusions in patients with advanced ccRCC.