Extended Data Fig. 7: CXCL10-associated CD8 + T cell infiltration and correlation with response to anti-PD-1 therapy in cancer patient cohorts.
a-c, Positive regulation of leucocyte chemotaxis (a), positive regulation of leucocyte migration (b), and chemokine activity (c) were enriched in the responsive tumours versus non-responsive tumours from a melanoma patient cohort treated with anti-PD-1, using the GSEA analysis (n = 118). d, Heatmap for the core enrichment genes in the pathway of positive regulation of leucocyte chemotaxis (n = 118). e, Distribution of CD8 + T cell fraction in the tumours from nonresponders and responders treated with anti-PD-1 in the Liu and the Riaz et al. cohorts (n = 141). f, Multivariate Cox proportional hazard analysis of the correlation of tumor-infiltrating immune cells with overall survival in the TCGA-COAD cohort, and the results were visualized as a forest plot. Red squares represented hazard ratio (n = 444). g, Partial Spearman’s correlation of CXCL10 mRNA expression and relative abundance of CD8 + T cell in the CIBERSORT analysis in the TCGA-COAD cohort. The correlation was adjusted by tumor purity, since most immune cell types were negatively correlated with tumor purity. The shade represented the confidence intervals of the regression estimate. h-i, Spearman’s correlation of CXCL10 expression and immune (h) and ESTIMATE (i) scores using the ESTIMATE algorithm in the TCGA-COAD cohort (n = 457). The shade represented the 95% confidence intervals of the regression estimate. Data were presented as the mean ± s.e.m. Each dot represented a biological replicate. Statistical analysis was performed using unpaired two-tailed Student’s t-test (e), log-rank (Mantel–Cox) tests (f), or Fisher’s exact test (g-i).
