Aneuploidy, also known as somatic copy number alterations (SCNAs), is widespread in human cancers and has been proposed to drive tumorigenesis. The relationship between SCNAs and the characteristic functional features or “hallmarks” of cancer is not well understood. Among these cancer hallmarks is immune evasion, which is accomplished by neoantigen editing, defects in antigen presentation and inhibition of tumor infiltration, and/or cytotoxic activities of immune cells. Whether and how tumor SCNA levels influence immune evasion is of particular interest as this information could potentially be used to improve the efficacy of immune checkpoint blockade, a therapy that has produced durable responses in a subset of cancer patients.

Understanding how SCNAs and mutation load affect tumor evolution, and through what mechanisms, is a key objective in cancer research. To explore the relationships between SCNA levels, tumor mutations, and cancer hallmarks, we examined data from 5255 tumor/normal samples representing 12 cancer types from The Cancer Genome Atlas project. We assigned each tumor an SCNA score and looked for correlations with the number and types of tumor mutations. We also compared the gene expression profiles of tumors with high versus low SCNA levels to identify differences in cellular signaling pathways.

First, we found that, for most tumors, there was a positive correlation between SCNA levels and the total number of mutations. Second, tumors harboring activating oncogenic mutations in the receptor tyrosine kinase–RAS–phosphatidylinositol 3-kinase pathway showed fewer SCNAs, a finding at odds with the hypothesis of oncogene-driven genomic instability. Third, we found that tumors with high levels of SCNAs showed elevated expression of cell cycle and cell proliferation markers (cell cycle signature) and reduced expression of markers for cytotoxic immune cell infiltrates (immune signature). The increased expression level of the cell cycle signature was primarily predicted by focal SCNAs, with a lesser contribution of arm and whole-chromosome SCNAs. In contrast, the lower expression level of the immune signature was primarily predicted by high levels of arm and whole-chromosome SCNAs. SCNA levels were a stronger predictor of markers of cytotoxic immune cell infiltration than tumor mutational load. Finally, through analysis of data from two published clinical trials of immunotherapy in melanoma patients, we found that high SCNA levels in tumors correlated with poorer survival of patients. The combination of the tumor SCNA score and the tumor mutational load was a better predictor of survival after immunotherapy than either biomarker alone.

We found that two hallmarks of cancer, cell proliferation and immune evasion, are predicted by distinct types of aneuploidy that likely act through distinct mechanisms. Proliferation markers mainly correlated with focal SCNAs, implying a mechanism related to the action of specific genes targeted by these SCNAs. Immune evasion markers mainly correlated with arm- and chromosome-level SCNAs, consistent with a mechanism related to general gene dosage imbalance rather than the action of specific genes. A retrospective analysis of melanoma patients treated with immune checkpoint blockade anti–CTLA-4 (cytotoxic T lymphocyte–associated protein 4) therapy revealed that high SCNA levels were associated with a poorer response, suggesting that tumor aneuploidy might be a useful biomarker for predicting which patients are most likely to benefit from this therapy.