直肠癌三孔腹腔镜直肠前切除术中经自然腔道标本取出失败的预测模型

Construction of a predictive model for failure of natural orifice specimen extraction during three-port laparoscopic anterior resection for rectal cancer

摘要:

目的 构建直肠癌三孔腹腔镜直肠前切除术中经自然腔道标本取出（NOSE）失败的预测模型。方法 选取2021年3月—2023年3月苏州市立医院收治的123例行三孔腹腔镜直肠前切除术的直肠癌患者，根据NOSE是否成功分为成功组（39例）与失败组（84例）。比较两组临床资料，进行多因素一般Logistic回归分析，并构建预测模型和受试者工作特征（ROC）曲线。结果 失败组体质量指数>25 kg/m²、肿瘤梗阻率、肿瘤距肛缘距离>5 cm、肿瘤最大直径>5 cm、标本直径/肛管直径>0.8、直肠系膜脂肪厚度>3 cm占比均高于成功组（P <0.05）；多因素一般Logistic分析结果表明：发生肿瘤梗阻［O^R=8.442（95% CI：0.007，1.800）］、肿瘤距肛缘距离>5 cm［O^R=6.965（95% CI：0.003，1.937）］、肿瘤最大直径>5 cm［O^R=4.681（95% CI：0.005，1.583）］、标本直径/肛管直径>0.8［O^R=5.064（95% CI：0.003，1.721）］、肠系膜脂肪厚度>3 cm［O^R=4.524（95% CI：0.005，1.562）］均为直肠癌三孔腹腔镜直肠前切除术中NOSE失败的危险因素（P <0.05）；ROC曲线结果表明，联合肿瘤梗阻、肿瘤距肛缘距离>5 cm、肿瘤最大直径>5 cm、标本直径/肛管直径>0.8、肠系膜脂肪厚度>3 cm联合预测直肠癌三孔腹腔镜直肠前切除术中NOSE失败的AUC为0.825，敏感性为89.3%（95% CI：0.806，0.950），特异性为64.1%（95% CI：0.472，0.788）。结论 肿瘤梗阻、肿瘤距肛缘距离>5 cm、肿瘤最大直径>5 cm、标本直径/肛管直径>0.8、直肠系膜脂肪厚度>3 cm是直肠癌三孔腹腔镜直肠前切除术中NOSE失败的危险因素，且联合预测效能较好。

Abstract:

Objective To establish a predictive model for failure of natural orifice specimen extraction (NOSE) during three-port laparoscopic anterior resection for rectal cancer.Methods A total of 123 patients with rectal cancer undergoing three-port laparoscopic anterior resection at Suzhou Municipal Hospital between March 2021 and March 2023 were enrolled. Patients were divided into a successful group (n = 39) and a failed group (n = 84) based on NOSE outcome. Clinical data were compared between groups, followed by multivariable logistic regression analysis to construct predictive models and receiver operating characteristic (ROC) curves.Results The failure group exhibited significantly higher prevalence of body mass index >25 kg/m2, tumour obstruction rate, tumour distance from anal margin >5 cm, tumour maximum diameter >5 cm, specimen diameter/ anal canal diameter >0.8 cm, and rectal mesenteric fat thickness >3 cm were higher in the failure group than in the success group (P <0.05). Multivariate logistic regression analysis indicated that tumour obstruction [O^R = 8.442 (95% CI: 0.007, 1.800), tumour distance from anal margin >5 cm [O^R = 6.965 (95% CI: 0.003, 1.937) ], tumour maximum diameter >5 cm [O^R = 4.681 (95% CI: 0.005, 1.583) ], specimen diameter/ anal canal diameter >0.8 cm [O^R = 5.064 (95% CI: 0.003, 1.721) ], mesenteric fat thickness >3 cm [O^R = 4.524 (95% CI: 0.005, 1.562) ] were all risk factors for NOSE failure during three-port laparoscopic anterior resection for rectal cancer (P < 0.05). ROC curve results indicated that the combined use of tumour obstruction, tumour distance from anal margin >5 cm, tumour maximum diameter >5 cm, specimen diameter/ anal canal diameter >0.8, and mesenteric fat thickness >3 cm yielded an AUC of 0.825, sensitivity of 89.3% (95% CI: 0.806, 0.950), and specificity of 64.1% (95% CI: 0.472, 0.788).Conclusion Tumour obstruction, tumour distance from anal margin >5 cm, tumour maximum diameter >5 cm, specimen diameter/anal canal diameter >0.8, and rectal mesenteric fat thickness >3 cm constitute risk factors for NOSE failure during three-port laparoscopic anterior rectal resection for rectal cancer, demonstrating favourable combined predictive efficacy.