基于临床特征和肿瘤标志物构建预测舌鳞状细胞癌（tongue squamous cell carcinoma，TSCC）术后复发的Nomogram模型，并验证其预测效能。

选择2019年1月—2021年1月于郑州大学第一附属医院行手术治疗的297例TSCC患者为研究对象。收集患者临床资料并检测其血管内皮生长因子（vascular endothelial growth factor，VEGF）、催乳素（prolactin，PRL）和组织多肽特异性抗原（tissue polypeptide specific antigen，TPS）水平。所有患者随访至2024年1月，最终288例完成随访。根据复发情况将患者分为复发组（n=71）与未复发组（n=217）。采用Cox比例风险模型分析TSCC术后复发的独立影响因素，基于筛选出的影响因素应用R软件构建Nomogram模型，通过受试者工作特征（receiver operating characteristic，ROC）曲线和校准曲线评估模型的预测效能及校准度。

288例患者中位随访时间为48（41，53）个月，术后复发率为24.65%（71/288）。多因素Cox比例风险回归模型分析显示，有吸烟史[风险比（hazard ratio，HR）=1.685，95%置信区间（confidence interval，CI）：1.164~2.441]、TNM分期为Ⅲ期（HR=1.848，95%CI：1.319~2.589）、浸润型生长（HR=1.935，95%CI：1.247~3.001）、未分化和低分化（HR=1.725，95%CI：1.163~2.557）、VEGF（HR=1.870，95%CI：1.345~2.599）、PRL（HR=1.664，95%CI：1.228~2.254）及TPS（HR=1.480，95%CI：1.127~1.943）均是TSCC患者术后复发的独立危险因素（均P<0.01）。基于上述7个因素构建的Nomogram模型预测复发的ROC曲线下面积（area under the curve，AUC）为0.938（95% CI：0.904~0.963），灵敏度为87.32%，特异度为89.86%；Hosmer-Lemeshow拟合优度检验显示模型拟合度良好（P>0.05），校准曲线显示预测值与实际观测值一致性良好。Nomogram模型的AUC高于第8版美国癌症联合委员会（American Joint Committee on Cancer，AJCC）TNM分期系统（AUC=0.757，95% CI：0.703~0.805，P<0.05）。

吸烟史、TNM分期为Ⅲ期、浸润型生长、未分化和低分化及VEGF、PRL、TPS均是TSCC患者术后复发的独立危险因素。基于上述临床特征和肿瘤标志物构建的Nomogram模型具有良好的预测效能，可为TSCC术后复发风险的个体化评估提供参考。

To construct and validate a Nomogram model for predicting postoperative recurrence of tongue squamous cell carcinoma (TSCC) based on clinical characteristics and tumor markers.

A total of 297 patients with TSCC who underwent surgical treatment at the First Affiliated Hospital of Zhengzhou University from January 2019 to January 2021 were enrolled in this study. Clinical data were collected, and serum levels of vascular endothelial growth factor (VEGF), prolactin (PRL), and tissue polypeptide specific antigen (TPS) were measured. All patients were followed up until January 2024, and 288 completed follow-up. According to recurrence status, patients were divided into a recurrence group (n=71) and a non-recurrence group (n=217). Multivariate Cox regression analysis was performed to identify independent risk factors for postoperative recurrence. A Nomogram model was constructed using R software based on these factors, and its predictive performance and calibration were evaluated using receiver operating characteristic (ROC) and calibration curves.

The median follow-up time of the 288 patients was 48 (41, 53) months, and the postoperative recurrence rate was 24.65% (71/288). Multivariate Cox regression analysis showed that smoking history [hazard ratio (HR)=1.685, 95% confidence interval (CI): 1.164–2.441], TNM stageⅢ (HR=1.848, 95% CI: 1.319–2.589), infiltrative growth pattern (HR=1.935, 95% CI: 1.247–3.001), undifferentiated and poorly differentiated status (HR=1.725, 95% CI: 1.163–2.557), VEGF (HR=1.870, 95% CI: 1.345–2.599), PRL (HR=1.664, 95% CI: 1.228–2.254), and TPS (HR=1.480, 95% CI: 1.127–1.943) were independent risk factors for postoperative recurrence (all P<0.01). The area under the ROC curve (AUC) of the nomogram model was 0.938 (95% CI: 0.904–0.963), with a sensitivity of 87.32% and a specificity of 89.86%. The Hosmer-Lemeshow goodness-of-fit test indicated good model fit (P>0.05), and the calibration curve demonstrated good agreement between predicted and observed outcomes. The Nomogram model showed a significantly higher AUC than the 8th edition of American Joint Committee on Cancer （AJCC） TNM staging system (AUC=0.757, 95% CI: 0.703–0.805, P<0.05).

Smoking history, TNM stageⅢ, infiltrative growth pattern, undifferentiated/poorly differentiated status, and VEGF, PRL, and TPS are independent risk factors for postoperative recurrence of TSCC. The Nomogram model based on these clinical characteristics and tumor markers demonstrated good predictive performance and may provide a reference for individualized assessment of postoperative recurrence risk in patients with TSCC.