Table_1_A Novel Prognostic Nomogram for Predicting Survival of Hormone Receptor-Positive and HER2 Negative Advanced Breast Cancer Among the Han-Population.docx

PurposeTo develop a nomogram model to predict overall survival in HR+/HER2- subtype advanced breast cancer.
MethodsA total of 3,577 ABC (advanced breast cancer) patients from 21 hospitals in China were involved in this study from January 2012 to December 2014. From all ABC patients, 1,671 HR+/HER2- ABC patients were extracted and enrolled in our study. A nomogram was built based on univariable and multivariable Cox regression analyses, identifying independent predictors. The discriminatory and predictive capacities of the nomogram were assessed using the ROC (receiver operating characteristic) curve and calibration plots.
ResultsUnivariable and multivariable analysis found that ER (estrogen receptor) status, MFIs (metastatic-free intervals), first-line therapy options, the number of metastatic sites, and whether local therapy for metastatic sites was chosen, were significantly related to overall survival (all P < 0.05). These variables were incorporated into a nomogram to predict the 2- year, 3-year, and 5-year OS (overall survival) of ABC patients. The AUC (the area under the curve) of the nomogram was 0.748 (95% CI (confidence interval):0.693-0.804) for 5-year OS in the training cohort and 0.732 (95% CI: 0.676-0.789) for the validation cohort. The calibration curves revealed good consistency between actual survival and nomogram prediction in the training and validation cohorts. Additionally, the nomogram showed an excellent ability to stratify patients into different risk cohorts.
ConclusionWe established a nomogram that provided a more straightforward predictive model for the outcome of HR+/HER2- ABC subtype patients and, to some extent, assisted physicians in making the personalized therapeutic option.

本研究旨在构建一列线图（nomogram）模型，用于预测HR+/HER2-亚型晚期乳腺癌患者的总体生存期（overall survival, OS）。方法 本研究纳入2012年1月至2014年12月期间，来自中国21家医院的共计3577例晚期乳腺癌（advanced breast cancer, ABC）患者。从中筛选出1671例HR+/HER2-型晚期乳腺癌患者纳入本研究。基于单因素及多因素Cox回归分析，本研究构建了一列线图模型，以识别独立预后危险因素。采用受试者工作特征（receiver operating characteristic, ROC）曲线及校准曲线（calibration plots）评估该列线图的区分度与预测效能。结果 单因素及多因素分析显示，雌激素受体（estrogen receptor, ER）状态、无转移间隔（metastatic-free intervals, MFI）、一线治疗方案、转移灶数量以及是否针对转移灶实施局部治疗，均与总体生存期显著相关（所有P<0.05）。将上述变量纳入列线图模型，用于预测晚期乳腺癌患者的2年、3年及5年总体生存期。训练队列中，该列线图预测5年总体生存期的曲线下面积（area under the curve, AUC）为0.748（95%置信区间（confidence interval, CI）：0.693~0.804）；验证队列中该值为0.732（95%CI：0.676~0.789）。校准曲线结果显示，在训练队列与验证队列中，实际生存期与列线图预测结果均具有良好的一致性。此外，该列线图可有效将患者划分为不同风险分层队列。结论 本研究构建的列线图模型可为HR+/HER2-型晚期乳腺癌患者的预后评估提供更为直观的预测工具，在一定程度上可辅助临床医师制定个体化治疗方案。