多灶性甲状腺乳头状癌发生颈部转移性淋巴结的危险因素及其预测模型构建

Risk factors of cervical metastatic lymph nodes in multifocal papillary thyroid carcinoma and its prediction model construction

  • 摘要:
    目的 分析多灶性甲状腺乳头状癌(MPTC)患者发生中央区颈部淋巴结转移(CLNM)和颈侧区淋巴结转移(LLNM)的危险因素,并建立预测模型。
    方法 选取93例MPTC患者作为研究对象。根据是否发生CLNM分为CLNM组(n=42)和无CLNM组(n=51); 根据是否发生LLNM分为LLNM组(n=31)和无LLNM组(n=62)。分别分析MPTC患者发生CLNM和LLNM的危险因素,并构建预测模型。通过受试者工作特征(ROC)曲线评估模型的预测效能。
    结果 93例MPTC患者CLNM的发生率为45.16%, LLNM的发生率为33.33%。男性、年龄≤45岁、最大肿瘤直径>10 mm、被膜侵犯是MPTC患者发生CLNM的危险因素(P < 0.05); 最大肿瘤直径>10 mm、被膜侵犯和有CLNM是MTPC发生LLNM的危险因素(P < 0.05)。MPTC患者发生CLNM模型的ROC曲线的曲线下面积(AUC)为0.846(95%CI: 0.749~0.947), 灵敏度为78.21%, 特异度为86.07%; MPTC患者发生LLNM模型的AUC为0.848(95%CI: 0.749~0.947), 灵敏度为76.38%, 特异度为82.46%。
    结论 男性、年龄≤45岁、最大肿瘤直径>10 mm、被膜侵犯与CLNM和LLNM均有关, LLNM发生与有CLNM有关。构建的CLNM和LLNM预测模型,模型拟合度良好,具有一定的预测价值。

     

    Abstract:
    Objective To analyze the risk factors of central cervical lymph node metastasis (CLNM) and lateral cervical lymph node metastasis (LLNM) in patients with multifocal papillary thyroid cancer (MPTC), and establish prediction models.
    Methods A total of 93 MPTC patients were selected as the study objects. They were divided into CLNM group (n=42) and non-CLNM group (n=51) according to the occurrence of CLNM; they were divided into LLNM group (n=31) and non-LLNM group (n=62) according to whether LLNM occurred. Risk factors for CLNM and LLNM in MPTC patients were analyzed, and prediction models were constructed. The predictive efficacy of the model was evaluated by receiver operating characteristic (ROC) curve.
    Results The incidence of CLNM and LLNM was 45.16% and 33.33% in 93 MPTC patients. Male, age≤45 years, maximum tumor diameter>10 mm and membrane invasion were the risk factors for CLNM in MPTC patients (P < 0.05); the maximum tumor diameter>10 mm, membrane invasion and CLNM were the risk factors for LLNM in MTPC (P < 0.05). The area under ROC curve (AUC) of CLNM model in MPTC patients was 0.846 (95%CI, 0.749 to 0.947), the sensitivity was 78.21%, and the specificity was 86.07%; the AUC of LLNM model in MPTC patients was 0.848 (95%CI, 0.749 to 0.947), the sensitivity was 76.38%, and the specificity was 82.46%.
    Conclusion The male, age≤45 years, maximum tumor diameter>10 mm and capsule invasion are associated with both CLNM and LLNM, and LLNM is associated with the presence of CLNM. The constructed CLNM and LLNM prediction models have good fit and certain prediction value.

     

/

返回文章
返回