The tumor immune microenvironment (TIME) shows significant heterogeneity in primary clear cell renal cell carcinoma (ccRCC). As TIME heterogeneity between primary and paired metastatic tumors of ccRCC is less understood, we characterized and compared the TIME of primary ccRCC with paired asynchronous metastases. We analyzed patients who developed ccRCC recurrence post radical nephrectomy and had both primary and metastatic treatment-naïve tissue available. Capture whole-transcriptome sequencing was performed on formalin-fixed paraffin-embedded (FFPE) specimens using the Illumina platform. Differential gene expression (DGE) analysis and gene set enrichment analysis (GSE) was performed using R packages limma and fgsea respectively. TIME deconvolution was quantified using CIBERSORT, an in-silico flow cytometry tool. In aggregate, 42 tumor samples from 19 patients (19 primary tumors with 23 matched metastases) were analyzed. Metastatic sites included lung (n = 6), bone (n = 6), adrenal (n = 4), liver (n = 2), lymph node (n = 2), and soft tissue (n = 3). In unsupervised hierarchical clustering, primary tumors clustered together and not with their matched metastatic tumor. Of the immune cells assayed, primary tumors displayed greater Tregs than their matched (and unmatched) metastases (p < 0.001). Among metastatic sites, bone had high levels of EMT activity compared to their matched primary tumors and lung metastatic tumors were enriched in E2F targets. We demonstrate differences in pathway enrichment and immune cell populations in primary ccRCC and their matched metastases, including a higher infiltration of immunosuppressive T regulatory cells in the tumor immune microenvironment of primary ccRCC. Metastatic tumors not only differed from their paired primary tumors but also differed in gene expression, gene set enrichment, and immune cell composition between metastatic tissue sites.
Cotta et al. (Wed,) studied this question.