Dr. Turajlic began by emphasizing that RCC is uniquely immunogenic among common solid tumors. This immunogenicity, defined as the ability to induce an adaptive immune response, is driven by the tumor’s antigenicity. This antigenicity is driven by cancer germline antigens, mutated peptides (neoantigens), oncogenic viral antigens, and lineage-restricted antigens. Characterization of this immunogenicity, in addition to a role as a biomarker for response to immunotherapy treatment (e.g. tumor mutational burden, TMB), also offers the potential as a novel therapeutic target for vaccines, antigen-directed T cell therapy, and neoantigen-directed T cell therapy.
Taking a step back, Dr. Turajlic highlighted the data supporting those immunotherapy responses are linked to the expansion of tumor antigen-driven T-cells. This data is largely derived from the ADAPTeR study in which patients underwent longitudinal collection of fresh samples during the course of treatment with nivolumab.
This trial eventually enrolled 15 patients of whom 11 had a partial response or stable disease and 4 of whom had progressive disease. Two patients underwent cytoreductive nephrectomy, provided a greater volume of tissue for analysis. The authors undertook an array of translational analyses on specimens derived from biopsy prior to treatment, biopsy following treatment (week 9) or nephrectomy, and autopsy sample (in 3 individuals).
In particular, the authors had an interest in T-cell receptor sequencing as the interaction at the T cell receptor (TCR) is necessary for immune response. Thus, they considered a variety of analyses looking at diversity/clonality, expansion, antigen specificity, and dynamic changes over time. They employed single-cell ribonucleic acid (RNA) sequencing.
They first demonstrated higher clonality of intratumoral TCRs, compared to those found in the blood of the same patients. Further, comparing patients who responded and those who didn’t respond to therapy, they identified that, at baseline, those who responded had higher intratumoral TCR clonality.
They then examined the treatment effect by comparing TCR sequences from prior to therapy to those at 9 weeks after nivolumab treatment. They found, interestingly, that TCR sequences were more likely to be maintained in responders, suggesting persistent antigen stimulation.
Among those patients in whom autopsy was performed, this allowed for sampling of all sites and tracking over TCR dynamics over time. These longitudinal analyses demonstrated that expanded TCR clones that were detected prior to therapy persist only at sites where there is tumor response.
The authors then used IgG4 antibody labeling to identify drug-bound cells. In an n=1 vs n=1 comparison of responder and non-responder, they found that drug binding induces clonal CD8+ T cell expansion. Notably, T cell expansion tracks with drug binding and with a response.
Then, they sought to assess cellular functionality by examining for Granzyme-B, a marker of cytotoxicity. At a single-cell sequencing resolution, the authors found that hyperexpanded CD8 T cells in the responder have upregulation of granzyme B. Thus, it is important to consider not just the quantity and spatial location of T cells, but also their state. These data suggest that they retain their cytotoxic ability despite exhaustion.
Given all of this, Dr. Turajlic then discussed the nature of the antigenic stimulus driving this response, with a focus on neoantigens. DNA mutation leads to a mutant peptide. This peptide then binds to MHC and if it is sufficiently different from “self”, it escapes central tolerance and initiates an immune response.
In kidney cancer, this is somewhat complex as tumor mutational burden, correlated with neoantigen load, does not correlate with response to immunotherapy. While assessed in other data sets, this finding was confirmed in Dr. Turajlic’s own analysis of the ADAPTeR cohort. However, this does not mean that individual mutations are not immunogenic, as evidenced by historical examples of spontaneous regression of RCC on the basis of T cell responses triggered by specific RCC antigens.
Further, non-coding epitopes such as human endogenous retroviruses (HERVs) may account for up to 80% of all tumor-specific antigens. Expression of these HERVs is silenced in normal somatic tissue but is de-repressed in disease states, facilitating transcription. HERV signatures associate with T-cell infiltration and cytolytic activity, as well as with nivolumab response. However, in the ADAPTeR trial, a previous HERV signature did not correlate with response to nivolumab.
Notably, the majority of HERVs are derived from immune cells, not from the tumor cells themselves. Thus, prior correlations between HERV expression and response may reflect an indirect assessment of immune infiltration. In the ADAPTeR cohort, they found that ccRCC specific HERVs were highly expressed, at baseline, in patients who did not respond to nivolumab. These tumors are less infiltrated by immune cells, thus, correlating with tumor purity. However, as with mutations discussed above, this does not mean that individual HERVs are not immunogenic.
In conclusion, Dr. Turajlic emphasized that pre-existing expanded T cells are maintained in responders to nivolumab. In non-responders, there is less evidence of this expansion. The nature of the underlying antigenic stimulus remains to be ascertained.
Presented by: Samra Turajlic, BA, MBBS, MRCP, Ph.D. is a Consultant Medical Oncologist at The Royal Marsden in London in the UK, specializing in the treatment of kidney cancer and melanoma. She is a Cancer Research UK Clinician Scientist at the Francis Crick Institute and Associate Honorary Faculty at the Institute of Cancer Research.
Written by: Christopher J.D. Wallis, Urologic Oncology Fellow, Vanderbilt University Medical Center Contact: @WallisCJD on Twitter during the 2021 European International Kidney Cancer Symposium (EIKCS), April 23-24, 2021