Abstract

Background: Chromosome 3p genes (tumor suppressor VHL and chromatin modifiers BAP1, SETD2, PBRM1) are frequently altered in ccRCC and may have prognostic significance, though the mechanism by which they drive ccRCC pathogenesis is incompletely understood. Therapies with novel mechanisms, for example belzutifan targeting hypoxia via HIF, increase the relevance of these mechanisms.

Methods: The Tempus Lens Platform (Tempus AI, Chicago, IL) was used to identify a cohort of de-identified ccRCC cases. Data was analyzed in Tempus Workspaces using the tempusverse suite of R packages. GSEA was applied to differential RNAseq expression data using the msigdbr (“hallmarks”) and clusterprofiler R packages. GSEA pathways are presented as normalized enrichment score with adjusted Q-value.

Results: We identified 166 ccRCC pts with solid tumor DNA and wtRNAseq genomic profiling, excluding biopsies from distant metastatic sites. Median age at diagnosis was 57 (IQR 47.2-66.0). The cohort was 63% male (n=101). Within the cohort, VHL SNV/indels were present in 52% (n=86), PBRM1 in 24% (n=39), BAP1 in 7% (n=12), and SETD2 in 10% (n=17). Copy number changes and structural alterations were not included. As expected, BAP1 and PBRM1 mutations were mutually exclusive (Fisher’s p = 0.037). VHLmut tumors demonstrated a clear hypoxic phenotype (hypoxia 1.50, q 2.27e-02, G2M checkpoint -2.37, q 4.00e-09). The other 3p mut tumors showed distinct phenotypes. BAP1mut was notable for an “immune-hot” signature (allograft rejection 2.83, q <7.37e-10, IFN-g response 2.49, q<7.37e-10, E2F targets 2.33 q<7.37e-10). PBRM1mut tumors were “immune-cold” with hallmarks of metabolic dysregulation (allograft rejection -2.13, q 1.96e-7, inflammatory response -1.92, q 1.22e-5, fatty acid metabolism 1.54, q 6.00e-3 ). Interestingly, SETD2mut tumors were notably distinct from VHLmut tumors, with downregulation of hypoxia, NFKB, and p53, and upregulation of cell cycle hallmarks, necessitating further investigation (hypoxia -2.15, q 1.57e-7, TNFa signaling -2.43, q 3.05e-9, G2M cell cycle 1.96, q 3.82e-6, p53 pathway -1.73, q 5.25e-4).

Conclusions: New treatments are emerging via rational design efforts enabled by a more detailed understanding of ccRCC pathogenesis. Here we leveraged real-world genomic data to characterize the landscape of 4 common, related genetic alterations known to be pivotal to ccRCC biology. Non-VHL altered patients were overrepresented in our cohort relative to prior studies, though only SNV/indels were captured here. We observed that non-VHL cases frequently have alterations in other chromosome 3p genes, which may drive ccRCC development, progression, and resistance via cellular processes distinct from the classic VHL-HIF-hypoxia axis. As HIF directed therapies become more widely utilized, chromosome 3p alterations have the potential to provide additional predictive and prognostic value to ccRCC patients.

VIEW THE PUBLICATION