Increased SNX20 and PD-L1 Levels Can Predict the Clinical Response to PD-1 Inhibitors in Lung Adenocarcinoma.

PURPOSE

Programmed death ligand 1 (PD-L1) is widely used for predicting immune checkpoint inhibitors but has a limited effect on predicting clinical response. The aim of this study was to examine the prognostic value and PD-1 inhibitor therapeutic efficiency of SNX20 in lung adenocarcinoma.

METHODS

We evaluated the mRNA and protein expression levels of SNX20 and PD-L1 and confirmed their predictive role in clinical response to anti-PD-1 therapy in 56 patients with advanced, refractory lung adenocarcinoma treated with PD-1 inhibitors. The expression of SNX family in different cancer types and the relationship between SNX20 and immune cells were evaluated in TCGA. The protein expression levels of SNX20, PD-L1 in 56 lung adenocarcinoma tissues were evaluated by immunohistochemistry.

RESULTS

SNX20 mRNA expression has the strongest relationship with CD8a of the sorting nexin (SNX) family in lung adenocarcinoma and is strongly correlated with immune infiltration levels in 30 cancer types, especially in lung adenocarcinoma. A positive correlation between SNX20 and PD-L1 was found based on immunohistochemical data (Pearson's r=0.3731 and p=0.0466). SNX20 and PD-L1 were also observed to have a significant positive correlation at the mRNA level. According to the receiver operating characteristic (ROC) curve, the best expression differentiation score of SNX20 and PD-L1 between responder versus non-responders in patients with lung adenocarcinoma using PD-1 inhibitors is 5. In univariate logistic regression analysis, both SNX20 (odds ratio [OR]=3.778, p=0.019) and PD-L1 (OR=5.727, p=0.004) expression levels are significant predictors of clinical response in the PD-1 inhibitor responder group, and SNX20 (OR=3.575, p=0.038) and PD-L1 (OR=5.484, p=0.007) are also predictors of the response to PD-1 inhibitors in the multivariate analysis. High SNX20/high PD-L1 expression group had longer overall survival than patients with high SNX20/low PD-L1 expression group or low SNX20/high PD-L1 expression group (p=0.013) and patients with low SNX20/low PD-L1 expression group (p=0.01).

CONCLUSION

SNX20 expression can be a promising predictor for therapeutic decision-making and treatment response assessment regarding PD-1 inhibitors, and special attention is required for the subgroup of patients with lung adenocarcinoma whose tumors express both high SNX20 and PD-L1.