非远处转移性甲状腺乳头状癌131I疗效预测模型的分析

目的 分析影响非远处转移性甲状腺乳头状癌¹³¹I治疗疗效的危险因素,构建疗效预测模型。方法 回顾性分析我院2016年1月—2020年12月期间行甲状腺癌切除术和¹³¹I治疗的422例患者临床资料,采用多因素logistic回归分析筛选影响¹³¹I治疗疗效的独立危险因素,构建疗效预测模型,并对模型进行评价和验证。结果 选取75%的患者作为训练集(n=319),25%的患者作为验证集(n=103),训练集经多因素logistic回归分析后,非远处转移性甲状腺乳头状癌¹³¹I治疗疗效不满意的独立危险因素为BRAF^V600E突变阳性、sTg/TSH≥0.05 ng/μIU、sTg/TgAb≥0.60 ng/IU、病灶最大径≥1.05 cm、淋巴结分期N1b、淋巴结转移率≥34.58%(P＜0.05)。构建诺莫图模型,其在训练集和验证集中的ROC曲线下面积分别为0.90(95% CI:0.87~0.94,P＜0.001)和0.86(95% CI:0.78~0.94,P＜0.001)。结论 本研究基于非远处转移性甲状腺乳头状癌¹³¹I治疗疗效不满意的危险因素构建的疗效预测模型具有良好的预测效能。

Analysis of 131I efficacy prediction model for non-distant metastatic papillary thyroid carcinoma

Objective The objective of this study was to analyze the risk factors affecting the curative effect of ¹³¹I treatment for non-distant metastatic papillary thyroid carcinoma(PTC),and construct a predictive model for curative effect. Methods The clinical data of 422 patients who underwent thyroid cancer resection and ¹³¹I treatment in our hospital from January 2016 to December 2020 were retrospectively analyzed.Multivariate logistic regression analysis was used to screen independent risk factors affecting the efficacy of ¹³¹I treatment,and the efficacy predictive model was constructed as well as its evaluation and validation. Results Selecting 75% patients as training set(n=319),25% patients as validation set(n=103).After multivariate logistic regression analysis in the training set,the independent risk factors of unsatisfactory ¹³¹I treatment for non-distant metastatic PTC were positive BRAF^V600E mutation,sTg/TSH≥0.05 ng/μIU,sTg/TgAb≥0.60 ng/IU,the largest diameter of lesions≥1.05 cm,the lymph node staging N1b and the lymphatic metastasis rate≥34.58%(P＜0.05).A nomogram model was constructed,and the areas under the ROC curve in the training set and validation set were 0.90(95% CI:0.87-0.94,P＜0.001)and 0.86(95% CI:0.78-0.94,P＜0.001),respectively. Conclusion The efficacy predictive model constructed in this study based on the risk factors of unsatisfactory ¹³¹I treatment efficacy in non-distant metastatic PTC has good predictive performance.