Clinical Value of Surgery and Establishment a Predictive Model for Breast Cancer Patients with Bone-Only Metastasis.

Background: Whether to perform surgery on breast cancer with initial bone metastasis is heatedly debated. The aim of this study was to assess the efficacy of surgery to prolong survival time towards breast cancer patients with bone-only metastasis, and create a prognostic nomogram to predict long-term survival.

Methods: Patients diagnosed with bone-only metastatic breast cancer from 2010 to 2015 were included in this study. National Cancer Database (NCDB) was used to obtain patients' data.

Results: A total of 7751 patients were included for final analysis, among which 3114 (40.2%) patients had received specialized breast surgery and 4637 (59.8%) had not. Patients who had undergone surgery showed a superior overall survival (OS) compared to patients without surgery (multivariate: HR (hazard ratios) = 0.58; 95% CI (confidence interval) [0.54-0.62]; p < 0.001). Moreover, survival benefit from surgery was also true in almost all subgroups. Prognostic nomogram to individually predict patients' long-term survival rate exhibited an acceptable predictive capability, with a C-index around 0.7 both in training and validation cohorts. Clinicopathological factors included in the nomogram could discriminate patients into subgroup with different prognoses.

Conclusions: Our data suggests that surgery may enhance OS in patients with initial bone-only metastatic breast cancer. Additionally, the created nomogram showed an acceptable could predict patients' long term survival.