Tumor-Specific and Tumor-Agnostic Molecular Signatures Associated With Response to Immune Checkpoint Inhibitors

PURPOSE Next-generation sequencing (NGS) testing is being incorporated into routine standard of care for patients with cancer. Immune checkpoint inhibitors (CPIs) are approved for use in both tumor-specific and tumor-agnostic indications. We sought to determine tumor type-specific or tumor-agnostic correlations between mutations detected by NGS and response to CPIs. MATERIALS AND METHODS A retrospective analysis of 26,004 patient records with NGS data available was conducted. Time to treatment failure and overall survival analyses were performed. Hazard ratios and associated statistics were computed in the R programming language. The study was considered exempt from internal review board review and data were considered nonhuman subjects. RESULTS Response to CPIs varied between tumor types with melanoma and lung cancer performing relatively better on CPIs than other tumor types. Within tumor types, response to CPIs was stratified by mutations in specific genes. Tumor-agnostic markers including high tumor mutation burden and microsatellite instability-high were also associated with longer time to treatment failure on CPIs. Importantly, within the high tumor mutation burden and microsatellite instability-high groups, mutations in individual genes correlate with response to CPIs. CONCLUSION The results from commercial NGS panels may be used to stratify patients for response to CPIs. In tumors where CPIs show relatively low efficacy, there may be distinct patient populations-based on gene mutation status-that are predicted to have better response to CPIs. Likewise, there may be distinct patient populations who do relatively worse on CPIs within tumor types known to respond well to CPIs.