Fig. 5.

Potential application of immune checkpoint blockade in the ACT-S&HR-P subtype. A–C Representative gene set enrichment analysis plot showing upregulated interferon gamma signalling (A), cancer immunotherapy by PD-L blockade (B) and natural killer cell–mediated cytotoxicity (C) in the ACT-S&HR-P (sensitive to ACT and HR proficiency) subtype versus the other subtypes. NES, normalized enrichment score. D Pathway activity in ACT-S&HR-P patients and other patients. The highlighted pathways are indicated as being focused on in this research. E In the ACT-sensitive group, immune cell activity scores were higher in the HR-proficient samples. The dots depict the mean difference of immune cell activity scores in HR-deficient samples compared to HR-proficient samples, and the lines show the 95% confidence interval (CI) for the difference. P < 0.05 was considered significant (red colour), Wilcoxon rank-sum test. See Figure S6 for the full name of immune cell types. F, G Relative number of immune-stimulatory cells (F) and immune-suppressive cells (G) in the four TNBC subtypes calculated using the CIBERSORT algorithm. H, I The expression of immune stimulatory molecules (H) and immune checkpoint genes (I) in four TNBC subtypes. ***P < 0.001, **P < 0.01, *P < 0.05