PDL1 high expression without TP53, KEAP1 and EPHA5 mutations could better predict survival for patients with NSCLC receiving atezolizumab

•ERBB4, EP300, PREX2, SLIT2, EPHB1 and IGF2R mutations were high frequency mutations in patients with PDL1 high.•The patients with EGFR, SMARCA4, EPHA5, FAT1, STK11, TET2 mutations were more likely to be seen in negative PDL1 expression group.•PDL1 high expression without TP53, KEAP1 and EPHA5 mutations could better predict survival for NSCLC patients receiving atezolizumab. Non-small cell lung cancer (NSCLC) patients with high expression of PDL1 are more likely to benefit from atezolizumab. There are no relevant research focusing on the relationship between the PDL1 expression and clinical variables and gene mutation types among NSCLC patients. NSCLC patients with confirmed PDL1 expression and gene mutation information from OAK study were included in our study. Logistic regression proportional model was applied to analyze the risk factors on PDL1 high expression. The biomarker evaluable population (BEP) was screened to analyze the gene mutation informaion among these patients. High frequency gene mutations were screened based on different PDL1 expressions. Moreover, the log rank test was applied to analyze the overall survival (OS) difference based on different gene mutation types. A total of 838 patients with NSCLC were included in our study. White patients are more likely to have PDL1 ≥ 1% (P = 0.004). ERBB4, EP300, PREX2, SLIT2, EPHB1 and IGF2R mutations were high frequency mutations in patients with high PDL1 expression, and the patients with EGFR, SMARCA4, EPHA5, FAT1, STK11, TET2 mutations were more likely to be seen in negative PDL1 expression group. Worse survival could be found in patients with KEAP1 (P