Clinical Medical College, Yangzhou University, Yangzhou, Jiangsu Province, China.
Department of Breast Surgery, Northern Jiangsu People's Hospital, Clinical Medical College of Yangzhou University, Yangzhou, Jiangsu Province, China.
J Cancer Res Clin Oncol. 2023 Nov;149(16):14721-14730. doi: 10.1007/s00432-023-05283-z. Epub 2023 Aug 16.
The lymph node (LN) status is a crucial prognostic factor for breast cancer (BC) patients. Our study aimed to compare the predictive capabilities of three different LN staging systems in node-positive BC patients and develop nomograms to predict overall survival (OS).
We enrolled 71,213 eligible patients from the SEER database, and 667 cases from our hospital were used for external validation. All patients were divided into two groups based on the number of removed lymph nodes (RLNs). The prognostic abilities of pN stage, positive LN ratio (LNR), and log odds of positive LN (LODDS) were compared using the C-indexes and AUC values. LASSO regression was performed to identify significant factors associated with prognosis and develop corresponding nomogram models.
Our study found that LNR had superior predictive performance compared to pN and LODDS among patients with RLNs < 10, while pN performed better in patients with RLNs ≥ 10. In the training set, the nomogram models exhibited excellent clinical applicability, as evidenced by the C-indexes, ROC curves, calibration plots, and decision curve analysis curves. Moreover, the nomogram classification accurately differentiated risk subgroups and improved discrimination. These results were externally validated in the validation cohort.
Physicians should select different LN staging systems based on the number of RLNs. Our novel nomograms demonstrated excellent performance in both internal and external validations, which may assist in patient counseling and guide treatment decision-making.
淋巴结（LN）状态是乳腺癌（BC）患者的一个重要预后因素。我们的研究旨在比较三种不同的 LN 分期系统在阳性淋巴结 BC 患者中的预测能力，并开发列线图来预测总生存（OS）。
我们从 SEER 数据库中纳入了 71213 名合格患者，从我们医院纳入了 667 例患者进行外部验证。所有患者均根据切除的淋巴结（RLNs）数量分为两组。使用 C 指数和 AUC 值比较 pN 分期、阳性 LN 比（LNR）和阳性 LN 的对数优势比（LODDS）的预后能力。使用 LASSO 回归识别与预后相关的显著因素，并开发相应的列线图模型。
我们的研究发现，在 RLNs<10 的患者中，LNR 与 pN 和 LODDS 相比具有更好的预测性能，而在 RLNs≥10 的患者中，pN 表现更好。在训练集中，列线图模型表现出良好的临床适用性，表现在 C 指数、ROC 曲线、校准图和决策曲线分析曲线。此外，列线图分类能够准确地区分风险亚组并提高区分度。这些结果在验证队列中得到了外部验证。
医生应根据 RLNs 的数量选择不同的 LN 分期系统。我们的新列线图在内部和外部验证中均表现出良好的性能，可能有助于患者咨询和指导治疗决策。
