食管癌根治术后喉返神经损伤风险预测模型构建  

Construction of risk prediction model for recurrent laryngeal nerve injury after radical resection of esophageal cancer

作　　者：刘宝兴 马海波 刘兴雨 邢文群 LIU Baoxing;MA Haibo;LIU Xingyu;XING Wenqun(Department of Thoracic Surgery,the Affiliated Cancer Hospital of Zhengzhou University,Henan Cancer Hospital,Zhengzhou,Henan 450008,China)

机构地区：[1]郑州大学附属肿瘤医院河南省肿瘤医院胸外科,河南郑州450008

出　　处：《中华实用诊断与治疗杂志》2024年第4期367-371,共5页Journal of Chinese Practical Diagnosis and Therapy

基　　金：河南省医学科技攻关计划省部共建重点项目(202002019)。

摘　　要：目的分析食管癌患者行食管癌根治术后发生喉返神经损伤的影响因素,并构建列线图模型。方法2017年1月—2022年12月河南省肿瘤医院诊治食管癌患者508例,均行食管癌根治术,根据术后是否发生喉返神经损伤分为损伤组43例和未损伤组465例。记录患者临床资料,包括年龄、性别、肿瘤直径、肿瘤位置、组织病理分型、T分期、N分期、临床分期、术前放疗及化疗情况、淋巴结清扫数目、喉返神经旁淋巴结清扫数目。采用lasso回归筛选预测食管癌患者行食管癌根治术后发生喉返神经损伤的影响因素;根据影响因素构建预测食管癌患者行食管癌根治术后发生喉返神经损伤的列线图模型;分别绘制决策曲线和临床影响曲线,评估列线图模型的预测效能。结果损伤组年龄[(70.91±10.41)岁]大于未损伤组[(62.07±14.83)岁](t=3.820,P<0.001),肿瘤直径[(58.60±21.62)mm]长于未损伤组[(42.78±26.07)mm](t=3.858,P<0.001),肿瘤位于上段食管、T分期≥T2、N分期≥N1、临床分期Ⅲ期、术前放疗及化疗比率(86.0%、81.4%、34.9%、34.9%、58.1%、81.4%)均高于未损伤组(17.0%、50.8%、11.2%、7.7%、10.5%、8.2%)(P<0.05),淋巴结清扫数目[(11.53±10.35)个]、喉返神经旁淋巴结清扫数目[(4.09±3.22)个]均多于未损伤组[(9.03±6.43)、(1.97±1.02)个](t=2.294,P=0.022;t=9.909,P<0.001),性别比例、组织病理分型与未损伤组比较差异均无统计学意义(P>0.05)。lasso回归筛选预测食管癌患者行食管癌根治术后发生喉返神经损伤的影响因素为肿瘤直径、肿瘤位置、术前放疗、术前化疗、喉返神经旁淋巴结清扫数目。决策曲线及临床影响曲线显示,列线图模型预测食管癌患者行食管癌根治术后发生喉返神经损伤的净收益率较高,临床实用性较强。结论肿瘤直径长、肿瘤位于食管上段、术前行放疗及化疗、喉返神经旁淋巴结清扫数目多的食管癌患者行食管癌根�Objective To analyze the influencing factors of recurrent laryngeal nerve injury in patients with esophageal cancer after radical resection,and to construct a nomogram model.Methods From January 2017 to December 2022,508 patients with esophageal cancer in Henan Cancer Hospital were divided into injury group(n=43)and no-injury group(n=465)according to the presence or absence of recurrent laryngeal nerve injury after surgery.The clinical data were recorded,including the age,gender,diameter of tumor,tumor location,histopathological classification,T stage,N stage,clinical stage,preoperative radiotherapy and chemotherapy,number of lymph node dissection,and number of recurrent laryngeal nerve lymph node dissection.Lasso regression was used to screen and predict the influencing factors of recurrent laryngeal nerve injury after radical resection of esophageal cancer.According to the influencing factors,a nomogram model was constructed to predict recurrent laryngeal nerve injury after radical resection of esophageal cancer.The decision curve and clinical impact curve were drawn to evaluate the predictive efficiency of the nomogram model.Results The patients were older in injury group[(70.91±10.41)years]than no-injury group[(62.07±14.83)years](t=3.820,P<0.001),and the tumor diameter was longer in injury group[(58.60±21.62)mm]than that in no-injury group[(42.78±26.07)mm](t=3.858,P<0.001).The proportions of patients with tumor located in the upper esophagus,≥T2 stage,≥N1stage,clinical stageⅢ,and preoperative radiotherapy and chemotherapy were higher in injury group(86.0%,81.4%,34.9%,34.9%,58.1%,81.4%)than those in no-injury group(17.0%,50.8%,11.2%,7.7%,10.5%,8.2%)(P<0.05).The numbers of lymph node dissection and recurrent laryngeal nerve lymph node dissection were more in injury group(11.53±10.35,4.09±3.22)than those in no-injury group(9.03±6.43,1.97±1.02)(t=2.294,P=0.022;t=9.909,P<0.001).There were no significant differences in the gender ratio and histopathological classification between two groups(P>0.05).La

关 键 词：食管癌 食管癌根治术 喉返神经损伤 lasso回归 列线图 

分 类 号：R735.1[医药卫生—肿瘤]