Fig. 4

TEXscore predicts benefit from immune checkpoint blockers therapy. A ROC analyses suggested the TEXscore exerted inferior predictive sensitivity to anti-PD-L1 response compared with CD8 + T cells and GEP in IMvigor210 dataset. (AUC = 0.576, 0.628, 0.515, respectively). B The Kaplan–Meier survival analysis indicated a poor overall survival of high-TEXscore patients (high: red; low: blue) in the IMvigor210 dataset (P = 0.0026, Hazard ratio = 1.49, 95% CI = 1.15–1.93). C Rate of clinical response (complete response [CR]/partial response [PR] and stable disease [SD]/progressive disease [PD]) to anti–PD-L1 immunotherapy in high or low TEXscore groups in the IMvigor210 dataset (chi-square test, P = 0.02382). D Distribution of TEXscore in different clinical responses (CR/PR, SD/PD) in the IMvigor210 dataset (Chi-square test, P = 0.02382). E PD-L1 expression (TC level) were associated with TEXscore. P values are indicated. Tumor cells (TC) level was measured to evaluate PD-L1 expression on tumor cells. The specimens were scored as IHC TC0, TC1, or TC2 if < 1%, ≥ 1% but < 5%, or ≥ 5% of TC were PD-L1 positive, respectively. F GO enrichment analysis of the DEGs between high (blue) and low (red) TEXscore in IMvigor210 dataset. The x-axis indicates the log10 (P value). G KEGG enrichment analysis of the DEGs between high (blue) and low (red) TEXscore in the IMvigor210 dataset. The x-axis indicates the log10 (P value). H GSEA analyses displayed key pathways enriched in high (up) TEXscore group. Gene sets that are inferred to reflect key underlying biological processes are colored