Abstract
BACKGROUND: Lymph node metastasis is an important predictive factor for the prognosis of breast cancer. Bone morphogenetic protein (BMP) and hepatocyte growth factor (HGF), and its receptor MET, are involved in metastasis. However, their predictive potential in the prediction of lymph node metastasis from breast cancer has not been evaluated to date. The aim of this study is to evaluate the implication of HGF and BMPs in breast cancer lymphatic metastasis. METHODS: The association between the ligands and receptors of BMP, the regulators of HGF, and the modulators of MET were detected in Cardiff clinical breast cancer cohort and The Cancer Genome Atlas (TCGA) breast cancer ribonucleic acid (RNA) sequencing database. Predictive model of HGF and BMPs for nodal metastasis was evaluated by binary logistic regression and receiver operating characteristic (ROC) curve. RESULTS: BMP-2 expression was upregulated but BMP-7 and matriptase-2 expression were downregulated in patients with nodal metastases. MET, matriptase-1, BMP-15, HAI-1, and matriptase-2 were correlated with the lymphangiogenesis markers. Lymphatic metastasis was positively with MET, matriptase-1 and BMP-15 but was negatively with matriptase-2, BMP-3, and HAI-1. ROC curve analysis showed the six factors with nodal status had a significant area under the curve values (0.657, P=0.001). The integrated signature could effectively predict lymph nodes involvement (P=0.006, hazard ratio =2.929). CONCLUSIONS: The aberrant expression of HGF/MET and BMPs is related to lymphatic metastasis in breast cancer. Integrated expression level of MET/BMP-15/matriptase-1 and the inversed HAI-1/BMP-3/matriptase-2 establishes a predictive model for lymph node involvement.