癌变·畸变·突变 ›› 2025, Vol. 37 ›› Issue (3): 234-241,261.doi: 10.3969/j.issn.1004-616x.2025.03.010

• 肿瘤防治 • 上一篇    

基于临床特征和血清学指标的舌鳞状细胞癌术后患者预后分析

蔡泽曼1, 郭海鹏2, 李慧儿3, 林逸伟3, 伍方财1, 康炜骠4   

  1. 1. 汕头大学医学院附属肿瘤医院放疗科, 广东汕头 515041;
    2. 汕头大学医学院附属肿瘤医院国际医疗部, 广东 汕头 515041;
    3. 汕头大学医学院附属肿瘤医院检验科, 广东 汕头 515041;
    4. 汕头大学医学院第二附属医院耳鼻喉科, 广东 汕头 515041
  • 收稿日期:2024-12-29 修回日期:2025-04-24 发布日期:2025-06-13
  • 通讯作者: 康炜骠
  • 作者简介:蔡泽曼,E-mail:13zmcai@stu.edu.cn。
  • 基金资助:
    汕头市科技计划(221115156492214);广东省科技创新战略专项(STKJ202209069)

A nomogram based on clinical characteristics and serological indicators for prognosis prediction of surgical tongue squamous cell carcinomas

CAI Zeman1, GUO Haipeng2, LI Huier3, LIN Yiwei3, WU Fangcai1, KANG Weipiao4   

  1. 1. Department of Radiation Oncology, Cancer Hospital of Shantou University Medical College, Shantou 515041;
    2. Department of International Medical Services, Cancer Hospital of Shantou University Medical College, Shantou 515041;
    3. Department of Clinical Laboratory Medicine, Cancer Hospital of Shantou University Medical College, Shantou 515041;
    4. Department of Otolaryngology, the Second Affiliated Hospital of Shantou University Medical College, Shantou 515041, Guangdong, China
  • Received:2024-12-29 Revised:2025-04-24 Published:2025-06-13

摘要: 目的:识别与预后相关的临床血清学指标,建立一种预测舌鳞状细胞癌(TSCC)术后患者预后情况的列线图模型。方法:收集169例在汕头大学医学院附属肿瘤医院接受手术治疗的TSCC患者和相应临床信息,随机分为训练组(118例)和验证组(51例)。在训练组中,通过单因素和多因素Cox回归分析筛选出与TSCC患者总生存期(OS)相关的预后影响因素并构建列线图模型,随后采用赤池信息量准则和贝叶斯信息量准则评估该模型的拟合优度;bootstrap检验的一致性指数和时间依赖的一致性指数评估该模型的区分能力;校正曲线分析评估该模型的预测能力;临床决策曲线分析评估该模型的决策获益率。上述方法同时在验证组中进行分析验证。结果:与TSCC患者OS相关的预后影响因子分别为pTNM分期、羟基丁酸脱氢酶、血液尿素氮、总胆红素、丙氨酸氨基转移酶和B因子,以此构建列线图模型。在训练组中,该列线图的赤池信息量准则和贝叶斯信息量准则分析数值为202.231和209.779,显著低于pTNM分期的216.526和217.784,表明该列线图具有更好的拟合优度。在训练组中,列线图bootstrap检验的一致性指数为0.814,高于pTNM分期的0.683,相同的结果也出现在验证组中,这表明列线图相较于pTNM分期具有更好的区分能力。临床决策曲线分析结果显示,列线图的临床决策曲线也显著高于pTNM分期,表明其具有更好的净获益。校正曲线分析结果显示训练组中列线图预测患者OS的结果与实际情况较为符合。Kaplan-Meier生存分析结果显示出列线图模型可较好地预测训练组中高低风险TSCC患者OS(P<0.01)。上述分析结果在验证组中可得到相应验证。结论:基于TSCC患者的临床特征和临床血清学指标构建并验证的列线图,可能是一种新型且实用的评估术后TSCC患者OS的工具。

关键词: 舌鳞状细胞癌, 列线图, pTNM分期, 预后, 手术

Abstract: OBJECTIVE: To identify prognostic serological indicators and to establish a nomogram model for predicting prognosis of tongue squamous cell carcinoma (TSCC) patients after surgery. METHODS: Patients with TSCC (n=169) who received surgical treatment in Cancer Hospital of Shantou University Medical College and their clinical information were collected. These patients were randomly divided into training group (118 patients) and verification group (51 patients). In the training group, the prognostic factors related to overall survival (OS) of TSCC patients were screened by univariate and multivariate Cox regression analysis, and a nomogram model was constructed. Akaike information and Bayesian information criteria were used to evaluate the goodness of fit of the nomogram. The bootstrap-corrected of consistency and time-dependent consistency indices were used to evaluate discrimination ability of the nomogram. The predictive ability of the nomogram was evaluated by calibration curve analysis. Decision curve analysis was used to evaluate the decision benefit rate of the model. The above methods were analyzed and verified in the verification group. RESULTS: The prognostic factors related to OS of TSCC patients were pTNM stage, hydroxybutyrate dehydrogenase,blood urea nitrogen,total bilirubin,alanine aminotransferase and factor B. The analysis results of Akaike information and Bayesian information criteria showed that the nomogram had better goodness of fit than pTNM stage. In the training group, the bootstrap-corrected of consistency index of the nomogram was 0.814,higher than that of the pTNM stage of 0.683. The same result also appeared in the verification group, indicating that the nomogram had better discrimination ability than the pTNM stage. The result of the decision curve analysis showed that, compared with pTNM stage, the nomogram had a better net benefit within a reasonable threshold probability range. The calibration curves of the nomogram showed good consistency between the probabilities and observed values in both the training and verification groups. The Kaplan-Meier survival analysis showed that the nomogram predicted the OS of high- and low-risk TSCC patients in both the training and the verification groups (P<0.01). CONCLUSION: Based on clinical characteristics and serological markers,the nomogram might be a novel tool for evaluating the survival probability of TSCC patients.

Key words: tongue squamous cell carcinoma, nomogram, pTNM stage, prognosis, surgery

中图分类号: