Gene signature of antigen processing and presentation machinery predicts response to checkpoint blockade in non-small cell lung cancer (NSCLC) and melanoma

BACKGROUND: Limited data exist on the role of alterations in HLA Class I antigen processing and presentation machinery in mediating response to immune checkpoint blockade (ICB). METHODS: This retrospective cohort study analyzed transcriptional profiles from pre-treatment tumor samples of 51 chemotherapy-refractory advanced non-small cell lung cancer (NSCLC) patients and two independent melanoma cohorts treated with ICB. An antigen processing machinery (APM) score was generated utilizing eight genes associated with APM (B2M, CALR, NLRC5, PSMB9, PSME1, PSME3, RFX5, and HSP90AB1). Associations were made for therapeutic response, progression-free survival (PFS) and overall survival (OS). RESULTS: In NSCLC, the APM score was significantly higher in responders compared with non-responders (p=0.0001). An APM score above the median value for the cohort was associated with improved PFS (HR 0.34 (0.18 to 0.64), p=0.001) and OS (HR 0.44 (0.23 to 0.83), p=0.006). The APM score was correlated with an inflammation score based on the established T-cell-inflamed resistance gene expression profile (Pearson’s r=0.58, p<0.0001). However, the APM score better predicted response to ICB relative to the inflammation score with area under a receiving operating characteristics curve of 0.84 and 0.70 for PFS and OS, respectively. In a cohort of 14 high-risk resectable stage III/IV melanoma patients treated with neoadjuvant anti-PD1 ICB, a higher APM score was associated with improved disease-free survival (HR: 0.08 (0.01 to 0.50), p=0.0065). In an additional independent melanoma cohort of 27 metastatic patients treated with ICB, a higher APM score was associated with improved OS (HR 0.29 (0.09 to 0.89), p=0.044). CONCLUSION: Our data demonstrate that defects in antigen presentation may be an important feature in predicting outcomes to ICB in both lung cancer and melanoma.