Nomograms to Predict Individual Potential Benefit from Targeted Therapy for patients with lymph node positive Luminal B (HER2-) breast cancer

Abstract

Abstract Purpose To construct nomograms combining clinicopathological characteristics, bone metastases (BM), viscera metastases (VM) and cancer-related deaths (CRD) to predict the higher-risk patients with lymph node positive (N+) Luminal B (HER2-) breast cancer (BC). Method Kaplan-Meier survival analysis, Venn diagram, Bar charts were used to describe the data for subsequent metastasis and CRD. Thirty-six clinicopathological characteristics were extracted in univariate and multivariate Cox regression analyses to develop nomogram to predict potential risk of BM, VM and CRD among patients with N+ Luminal B (HER2-) BC. The calibration plots, concordance index and receiver operating characteristics (ROC) analysis were applied to determine the nomogram accuracy. Result The median age of 8139 patients was 51 years, with a median follow-up of 124 months (4–216 months). There was no statistical difference between the metachronous primary bilateral BC and synchronous primary bilateral BC. The number-peak period of patients with subsequent BM was the third year, VM was the 4th year and CRD was the 6th year (range second–6th year, 4th–6th year and third–8th year, respectively). BM, VM and CRD nomograms showed outstanding performance and discriminative ability (C-index 0.69, 0.68 and 0.71, respectively). The calibration curves and ROC curves analysis demonstrated the considerable clinical usefulness of the combined nomogram. Three clinical examples showed results differences in optimal period who had similar pathological stage. Conclusion The developed nomogram model consisting of time-event-dependent clinicopathological characteristics could reliable in predicting BM, VM and CRD probability of patients with N+ Luminal B (HER2-) BC.