基于彩色多普勒超声参数及CAR、SII构建宫颈癌淋巴结转移的风险预测模型  

Construction of a risk prediction model for lymph node metastasis of cervical cancer based on color Doppler ultrasound parameters,C-reactive protein albumin ratio,and SII

作　　者：张姣 冯玉珂 张琳颖 ZHANG Jiao;FENG Yu-ke;ZHANG Lin-ying(Department of Ultrasound Diagnosis,Zhengzhou Yihe Hospital,Zhengzhou 450000,Henan,CHINA)

机构地区：[1]郑州颐和医院超声诊断科,河南郑州450000

出　　处：《海南医学》2024年第18期2666-2671,共6页Hainan Medical Journal

基　　金：2021年度河南省医学科技攻关计划联合共建项目(编号:LHGJ20210801)。

摘　　要：目的构建基于彩色多普勒超声参数及C-反应蛋白白蛋白比(CAR)、全身免疫炎症指数(SII)的宫颈癌淋巴结转移的风险预测模型。方法回顾性选取2018年1月至2024年3月期间郑州颐和医院收治的152例宫颈癌患者作为研究对象,按照有无淋巴结转移分为淋巴结转移组61例和非淋巴结转移组91例,比较两组患者的临床特征、彩色多普勒超声参数及CAR、SII,采用二元Logisitic回归分析宫颈癌淋巴结转移的危险因素,并构建风险预测模型,采用受试者工作特征(ROC)曲线分析预测模型对宫颈癌淋巴结转移的预测效能。结果淋巴结转移组患者的肿瘤大小<2 cm占比、临床分期Ⅱa占比、浸润深度浅1/3占比、搏动指数(PI)、阻力指数(RI)与非淋巴结转移组比较更低,差异均有统计学意义(P<0.05),而淋巴结转移组患者的肿瘤大小>4 cm占比、临床分期Ⅱb占比、浸润深度深1/3占比、收缩期峰值流速(PSV)、CAR、SII与非淋巴结转移组比较更高,差异均有统计学意义(P<0.05)。二元Logisic回归分析结果显示,浸润深度(深1/3)、PSV、CAR、SII均是宫颈癌淋巴结转移的危险因素(P<0.05),而PI、RI是宫颈癌淋巴结转移的保护因素(P<0.05)。风险预测模型=2.227×浸润深度(深1/3)+0.285×PSV+(-24.753)×PI+(-17.186)×RI+16.530×CAR+0.008×SII,Hosmer-Lemeshow拟合度检验显示,χ^(2)=6.999,P=0.5379,模型拟合优度较好,ROC分析结果显示,风险预测模型预测宫颈癌淋巴结转移的曲线下面积(AUC)为0.960,约登指数为0.802,敏感度、特异度分别为96.7%、83.5%,95%CI为0.932~0.987,实际应用准确性为89.5%。结论基于彩色多普勒超声参数及CAR、SII构建的宫颈癌淋巴结转移风险预测模型的预测价值较高。Objective To construct the risk prediction model for lymph node metastasis of cervical cancer based on color Doppler ultrasound parameters,Creactive protein albumin ratio(CAR),and systemic immune inflammation index(SII).Methods A retrospective study was conducted on 152 cervical cancer patients admitted to Zhengzhou Yihe Hospital from January 2018 to March 2024.The patients were divided into a lymph node metastasis group of 61 cases and a nonlymph node metastasis group of 91 cases based on the presence or absence of lymph node metastasis.The clinical characteristics,color Doppler ultrasound parameters,CAR,and SII were compared between the two groups.Binary logistic regression was used to analyze the risk factors for lymph node metastasis of cervical cancer,and a risk prediction model was constructed.The receiver operating characteristic(ROC)curve was used to analyze the predictive performance of the prediction model for lymph node metastasis of cervical cancer.Results The proportion of patients with tumor size<2 cm,clinical stageⅡa,infiltration depth of shallow 1/3,pulsatility index(PI),and resistance index(RI)were significantly lower in the lymph node metastasis group than in the nonlymph node metastasis group(P<0.05).However,the proportions of patients with tumor size>4 cm,clinical stageⅡb,infiltration depth of deep 1/3,peak systolic velocity(PSV),CAR,and SII were significantly higher in the lymph node metastasis group than in the nonlymph node metastasis group(P<0.05).The binary logistic regression analysis showed that infiltration depth of deep 1/3,PSV,CAR,and SII were all risk factors for lymph node metastasis of cervical cancer(P<0.05),while PI and RI were protective factors for lymph node metastasis of cervical cancer(P<0.05).The risk prediction model was 2.227×infiltration depth(deep 1/3)+0.285×PSV+(24.753)×PI+(17.186)×RI+16.530×CAR+0.008×SII.The HosmerLemeshow fit test showedχ^(2)=6.999,P=0.5379,indicating good model goodness of fit.The ROC analysis results showed that the area under the curve

关 键 词：宫颈癌 淋巴结转移 彩色多普勒超声 C-反应蛋白白蛋白比 全身免疫炎症指数 预测模型 

分 类 号：R737.33[医药卫生—肿瘤]