胃肠胰神经内分泌肿瘤肝转移患者预后列线图的建立

Establishment of prognostic nomogram for patients with gastrointestinal pancreatic neuroendocrine tumors with liver metastasis

目的 分析胃肠胰神经内分泌肿瘤(GEP-NEN)肝转移患者的预后影响因素,并建立有效的列线图模型。方法 收集美国国立癌症研究所监测、流行病学和最终结果(SEER)数据库2010—2018年GEP-NEN肝转移患者的临床信息,基于COX比例风险回归模型筛选的预后独立影响因素构建预测患者1、3、5年癌症特异性生存率(CSS)的列线图,并评估该模型的区分度及校准度。结果 共纳入GEP-NEN肝转移患者1 154例,其中建模组808例,验证组346例。COX多因素分析显示,年龄≥59岁、分化程度低分化和未分化、N1和N2分期为预后不良的独立危险因素[HR(95%CI)=1.84(1.51~2.24)、4.80(3.59~6.43)、3.29(2.30~4.70)、1.37(1.09~1.72)、3.16(1.91~5.22)],原发性肿瘤位于小肠、胰腺及行原发灶手术为其独立保护因素[HR(95%CI)=0.38(0.23~0.63),0.66(0.44~0.99),0.46(0.35~0.60)]。建模组和验证组的一致性指数分别为0.805(95%CI 0.755~0.846)和0.814(95%CI 0.733~0.875),2组1、3、5年特异性生存率的受试者工作特征曲线下面积分别为0.900、0.893、0.853和0.879、0.879、0.856,校准曲线表明模型预测的生存率与实际生存率之间具有良好的一致性。结论 年龄<59岁、小肠和胰腺原发肿瘤、分化程度好、N0分期、行原发灶手术的患者预后较好,由以上因素建立的列线图模型对GEP-NEN肝转移患者的预后具有良好的预测能力。

Objective To analyze the prognostic factors of patients with gastrointestinal pancreatic neuroendocrine tumor(GEP-NEN) with liver metastasis, and establish an effective nomogram model. Methods Collect the clinical information of GEP-NEN patients with liver metastasis from 2010 to 2018 in the National Cancer Institute Surveillance, Epidemiology and End Results(SEER) database. Based on the independent prognostic factors screened by COX proportional hazard regression model, an nomograph was constructed to predict the 1, 3, and 5 year cancer specific survival(CSS) of patients, and the differentiation and calibration of the model were evaluated. Results 1 154 patients with liver metastasis from GEP-NEN were enrolled, including 808 in the modeling group and 346 in the validation group. COX multivariate analysis showed that age ≥ 59 years, poorly differentiated and undifferentiated, N1 and N2 stage were independent risk factors for prognosis [HR(95%CI)=1.84(1.51-2.24), 4.80(3.59-6.43), 3.29(2.30-4.70), 1.37(1.09-1.72), 3.16(1.91-5.22)]. Primary tumors located in the small intestine and pancreas and primary focus surgery were independent protective factors [HR(95%CI)=0.38(0.23-0.63),0.66(0.44-0.99),0.46(0.35-0.60)].The consistency indexes of the modeling group and the validation group were 0.805(95%CI 0.755-0.846) and 0.814(95%CI 0.733-0.875), respectively. The area under the working curve of the subjects with 1, 3, and 5-year specific survival rates in the two groups were 0.900, 0.893, 0.853, and 0.879, 0.879, 0.856, respectively. The predicted survival rate of the model was in good consistency with the actual survival rate. Conclusion The patients with age<59 years, primary tumors of small intestine and pancreas, good differentiation, N0 staging, and primary focus surgery had a good prognosis. The nomogram model established by the above factors had a good predictive ability for the prognosis of patients with GEP-NEN liver metastasis.