Fig. 2: Univariable prognostic models. | British Journal of Cancer

Fig. 2: Univariable prognostic models.

From: Pathway-based signatures predict patient outcome, chemotherapy benefit and synthetic lethal dependencies in invasive lobular breast cancer

Fig. 2

a Forest plot showing prognostic assessment of CPs in the combined validation cohort (HR, 95% CI are shown). Predicted risk scores for CPs in the combined validation cohort were dichotomised into low- and high-risk groups using the discovery cohort’s median risk score for the corresponding CP. CPs with significant prognostic association are left of the vertical reference line in blue section (FDR-adjusted P < 0.05). b Venn diagram of prognostic genes from the top 3 CPs. c Heatmap showing mRNA abundance profiles of prognostic genes from CP14 in combined validation cohorts. Row covariate indicates whether a gene was associated with poor or good outcome in univariable analysis in the discovery cohort. Column covariate indicate predicted risk group in the combined validation cohort. Hierarchical clustering was performed on both rows and columns using “euclidean” as the distance measure and “ward” as the agglomeration method. d Kaplan–Meier survival curves (overall survival (OS)) indicating prognostic assessment of CP14 stratified by chemotherapy response using a contemporary clinical cohort (SCAN-B). e–h Same as (c, d), showing mRNA abundance profiles of prognostic genes from CP20 and CP25 and their prognostic assessment stratified by chemotherapy.

Back to article page