Figure 3.

Quiescence-Associated Genes are Correlated with Late Metastatic Recurrence. (A) Comparison of integrated expression scores for quiescence-associated genes in patients with oestrogen receptor-positive (ER+) and oestrogen receptor-negative (ER−) tumours within data sets (GSE2034, GSE2603 and GSE12276) comprising the CB572 cohort (Wilcoxon rank-sum test, ** = p ≤ 0.01, *** = p ≤ 0.001). (B) Distant metastasis-free survival probability over time in patients within the CB572 cohort with ER− tumours integrated signature scores for expression of quiescence-associated signature genes (n = 186, log-rank test). (C) Distant metastasis-free survival probability over time in patients within the CB572 cohort with ER+ tumours stratified by integrated signature scores for expression of quiescence-associated signature genes (n = 371, log-rank test). (D) Cox proportional hazards survival analysis of the quiescence-associated signature genes represented in primary breast tumours of patients in the CB572 composite data set (n = 572). Patients were stratified at all possible cut-off points (n−1) for all signature genes. For each gene, the log₂-transformed cumulative hazard ratio for all cut-off points significantly (p ≤ 0.05) associated with late distant metastatic recurrence and early distant metastatic recurrence (<5 and ≥5 years after primary tumour diagnosis, respectively) normalised to total possible cut-off points are shown. Quiescence-associated signature genes were significantly associated with late metastatic recurrence (Wilcoxon rank-sum test, **** = p ≤ 0.0001).