Figure 4 from Inferring Characteristics of the Tumor Immune Microenvironment of Patients with HNSCC from Single-Cell Transcriptomics of Peripheral Blood

Exhausted T-cell signature in the TME can be inferred from the blood and the inferred exhausted T-cell score can further predict immunotherapy response in HNSCC. A, Correlation between the predicted and the measured exhausted T-cell signature scores in TME in the matched blood/tumor training dataset (n = 25) and validation dataset (n = 5). B, ROC curve for predicting ICB RECIST response (n = 16) and pathologic response (n = 26) by the predicted exhausted T-cell signature scores. C, OS and PFS analyses of ICB-treated patients in exhausted-high (exhausted score > quantile 50%) vs. exhausted-low (exhausted score ≤ quantile 50%) tumor groups using the predicted exhausted T-cell signature scores. D–F are the same as A–C, respectively, except that the predicted exhausted T-cell signature scores in the TME are replaced by the measured exhausted T-cell signature scores in the blood. r, Pearson correlation coefficient; ρ, Spearman correlation coefficient. FPR, false-positive rate; TPR, true-positive rate.