Association between PD1 mRNA and response to anti-PD1 monotherapy across multiple cancers.

3076 Background: In advanced cancer, the overall response rate (ORR) following anti-PD1 monotherapy is variable (0% to 50%). Here, we hypothesized that this observation is partly explained by different amounts of the drug target (i.e. PD1) in the tumor. Methods: RNA-seq data from 10,078 tumors representing 34 cancer-types were obtained from TCGA. The expression of PD1 and 566 immune-related genes/signatures were evaluated. Correlations between each gene/signature and ORRs reported in the literature were calculated. We included only studies of anti-PD1 monotherapy that enrolled at least 20 patients (pts) who were not selected for PDL1 expression. To translate the in-silico findings to the clinical setting, we analyzed the expression of PD1 mRNA using the nCounter platform in 694 formalin-fixed paraffin embedded (FFPE) tumor samples from 16 cancer-types. Finally, we evaluated FFPE-based PD1 mRNA from an independent dataset of 102 pts with advanced solid tumors treated with anti-PD1 monotherapy. Results: PD1 expression varied significantly across TCGA (p < 0.001) with 0% to 84% of tumors within a cancer-type being PD1-high (H) (defined as %ile 80). Interestingly, % of PD1-H tumors within a cancer-type were found highly correlated with ORRs reported in the literature (correlation coefficient [CC] = 0.91), suggesting that 83% of the differences in the ORR across cancer-types may be explained by the abundance of PD1. Lower CCs were identified with different PD1 cutoffs, other genes/signatures and tumor mutational burden. To translate these findings in clinical samples, the expression of PD1 was evaluated in 773 FFPE tumor samples across 17 cancer-types. Using the optimal cutoff (%ile 80), similar proportions of PD1-H tumors within each cancer-type were identified in our in-house FFPE-based cohort compared to TCGA (CC = 0.92). Finally, the optimal PD1 FFPE-based cutoff was found significantly associated with ORR (PD1-H 42.8% vs. PD1-low 17.6; P = 0.018) in 102 pts with advanced cancer treated with anti-PD1 monotherapy. Conclusions: Our findings highlight the strong relationship between PD1 mRNA expression and the activity of anti-PD1 therapies across multiple cancers.