Zhao Dan, Li Wei, Zhang Xiaomei
Department of Endocrinology, Peking University International Hospital, Beijing, China.
Department of Gastrointestinal Surgery, Peking University International Hospital, Beijing, China.
Gland Surg. 2024 Apr 29;13(4):528-539. doi: 10.21037/gs-23-478. Epub 2024 Apr 22.
The incidence of papillary thyroid cancer (PTC) has increased dramatically, and it is susceptible to cervical lymph node metastasis (LNM), predominantly in the ipsilateral cervical central lymph node metastasis (CLNM). Ipsilateral cervical CLNM affects patients' surgical options and survival rates. In this study, we integrated multiple factors to establish a nomogram-based preoperative prediction model of ipsilateral cervical CLNM in PTC.
Data were retrospectively collected from 609 patients with PTC admitted to Peking University International Hospital, all of whom underwent ipsilateral cervical lymph node dissection. They were randomly divided into a modeling set and validation set in the ratio of 7:3. Binary logistic regression was used to analyze independent risk factors for ipsilateral cervical CLNM in PTC and to construct a nomogram model. The performance of nomogram CLNM prediction was evaluated by the receiver operating characteristic (ROC) curve and calibration curve.
Binary Logistic Regression showed that age, history of osteoporosis, complicated by Hashimoto's thyroiditis, enlarged lymph nodes in the central neck, and extrathyroidal extension were risk factors for ipsilateral cervical CLNM. Combining these five independent risk factors, a nomogram prediction model was developed. In the modeling set, the area under the curve (AUC) of the nomogram ROC was 0.782 [95% confidence interval (CI): 0.730-0.833], and the sensitivity and specificity of the model were 0.761 and 0.763, respectively, with a well-calibrated curve fit. Moreover, the model presented better discrimination than any of the independent risk factors. The nomogram performed well in the validation set (AUC 0.753; 95% CI: 0.648-0.858).
A non-invasive, and accurate nomogram prediction model for ipsilateral cervical CLNM of PTC was established. It can help physicians identify patients with a high risk of ipsilateral cervical CLNM of PTC preoperative for individualized treatment.
甲状腺乳头状癌（PTC）的发病率急剧上升，且易发生颈部淋巴结转移（LNM），主要是同侧颈部中央区淋巴结转移（CLNM）。同侧颈部CLNM会影响患者的手术选择和生存率。在本研究中，我们综合多种因素建立了基于列线图的PTC同侧颈部CLNM术前预测模型。
回顾性收集北京大学国际医院收治的609例PTC患者的数据，所有患者均接受了同侧颈部淋巴结清扫术。他们以7:3的比例随机分为建模组和验证组。采用二元逻辑回归分析PTC同侧颈部CLNM的独立危险因素，并构建列线图模型。通过受试者工作特征（ROC）曲线和校准曲线评估列线图CLNM预测的性能。
二元逻辑回归显示，年龄、骨质疏松病史、合并桥本甲状腺炎、中央区颈部淋巴结肿大和甲状腺外侵犯是同侧颈部CLNM的危险因素。结合这五个独立危险因素，建立了列线图预测模型。在建模组中，列线图ROC曲线下面积（AUC）为0.782 [95%置信区间（CI）：0.730 - 0.833]，模型的灵敏度和特异度分别为0.761和0.763，曲线拟合校准良好。此外，该模型的辨别能力优于任何一个独立危险因素。列线图在验证组中表现良好（AUC 0.753；95% CI：0.648 - 0.858）。
建立了一种非侵入性、准确的PTC同侧颈部CLNM列线图预测模型。它可以帮助医生在术前识别PTC同侧颈部CLNM高危患者，以便进行个体化治疗。
