Figure 1. Integrin β3 expression is increased in breast cancer cells after chemotherapy.

A. Flow cytometry of integrin β3 expression in PyMT-BO1 and 4T1 cells harvested 48 hours after overnight treatment with DTX. One of three independent experiments, each with n = 2 biological replicates per treatment group. One-way ANOVA with test for linear trend; ****, P<0.0001. B. Representative X-ray radiographs of tibiofemoral joints from vehicle and DTX-treated mice bearing either PyMT-BO1 or 4T1 metastases established by i.c. injection. Scale bar = 1.25mm. Red arrows indicate areas of significant bone erosion. n = 8–9 mice per group. C. Ex vivo flow cytometry of live, GFP+ PyMT-BO1 cells harvested from established bone metastases treated with either vehicle or DTX (5mg/kg i.v.). n = 7–8 mice per group. Two-tailed unpaired Welch’s t test; ****, P<0.0001. D. Design of tissue microarray with 38 primary TNBC biopsies obtained after neoadjuvant chemotherapy. Summary of β3-stratified patient demographics (right). E. Integrin β3 IHC in human TNBC patients after chemotherapy. Representative images of low and high tumor β3 staining (left), summary of scoring (right, see Materials and Methods). Scale bars = 100 μm (20x) or 50 μm (40x). F. Kaplan-Meier analysis of β3-stratified recurrence-free survival in patients from tissue microarray. Swimmer’s plot of individual time to recurrence (left), Kaplan-Meier curves and statistics (right). Hazard ratio (HR) and confidence intervals determined by Cox proportional hazards model; significance determined by log rank test. G. Kaplan-Meier analysis of β3-stratified recurrence-free survival in 315 high-risk TNBC patients receiving any chemotherapy obtained from publicly available microarray data. Hazard ratio (HR) and confidence intervals determined by Cox proportional hazards model; significance determined by log rank test.