Functional prediction of immune checkpoint inhibitor response in renal cell carcinoma using a zebrafish patient-derived xenograft platform

To predict individual patient responses to immune checkpoint inhibitors (ICIs), we developed a biologic assay for renal cell carcinoma (RCC) in which whole tumor cells (WTCs) isolated from patients are mixed ex vivo with nivolumab or control IgG and injected into zebrafish larvae (zPDX). ICI response is quantified as the percentage reduction in tumor size relative to control treatment. In zPDX treated with nivolumab, tumor shrinkage correlated with upregulated hIL-2 mRNA expression and was completely reversed by co-injection of a CD3-neutralizing antibody, indicating a T cell-mediated effect. Histological examination of whole-body zPDX treated with ICI revealed a significant increase in T cells and a marked decrease in vimentin + tumor cells. The zPDX assays were repeatable and reproducible within accepted precision thresholds. We performed zPDX assays for 21 cases, which clustered into three groups; High-, Low-, and Non-responders. zPDX response showed no significant correlation with immune-response gene signatures or percentage of tumor-infiltrating lymphocytes (TILs). However, gene-expression analysis comparing High-responders with Low-/Non-responders revealed strong associations between zPDX response and the expression of positive immune regulators. Non-responder WTCs retained the intrinsic ability of T cells to respond to TCR stimulation. In the four patients with available clinical data, zPDX responses were generally consistent with outcomes following ICI therapy, supporting the potential of our novel zPDX platform to predict ICI efficacy in patients with RCC and, potentially, other malignancies.