治疗前炎症指标构建模型预测卵巢癌患者术后生存  

Model Based on Preoperative Inflammatory Markers to Predict the Overall Survival in Patients with Ovarian Cancer

作　　者：李希 林敏 马犇[1] 顾正田[1] 康佳丽[1] LI Xi;LIN Min;MA Ben;GU Zheng-tian;KANG Jia-li(Department of Obstetrics and Gynecology,Guangzhou First People's Hospital,Guangzhou 510180,China)

机构地区：[1]广州市第一人民医院妇产科,广东广州510180

出　　处：《中山大学学报（医学科学版）》2021年第6期937-943,共7页Journal of Sun Yat-Sen University:Medical Sciences

基　　金：广州市科技计划项目(201707010075)。

摘　　要：【目的】探讨治疗前炎性指标在卵巢癌患者预后中的作用,建立卵巢癌患者的便捷、准确的生存预测模型。【方法】回顾性分析2013年1月至2018年12月在广州市第一人民医院就诊的145例卵巢癌患者及同期60例卵巢良性肿瘤患者的临床、病理及随访资料。根据随访截止日期时患者的生存情况,将患者分成生存组(48例)和死亡组(97例),比较两组患者治疗前外周血中性粒细胞/淋巴细胞比值(NLR)、血小板/淋巴细胞比值(PLR)、单核细胞与淋巴细胞计数比值(MLR)、单核细胞与淋巴细胞计数比值(NAR)水平,采用Cox回归分析明确影响卵巢癌患者生存的因素,用R软件建立列线图生存预测模型,并进行内部验证。【结果】NLR、PLR、MLR、NAR在卵巢癌患者中的中位数(四分位数间距)分别为2.98(1.85,5.55)、156.17(110.38,194.66)、0.28(0.21,0.42)、0.17(0.13,0.28);四个指标在卵巢良性肿瘤患者中的中位数(四分位数间距)分别为1.94(1.49,3.65)、125.00(109.38,180.66)、0.09(0.04,0.19)、0.14(0.11,0.17)。卵巢癌患者四个指标均高于卵巢良性肿瘤患者,差异具有统计学差异(P<0.05);卵巢癌生存亚组在年龄、FIGO分期、治疗前血清白蛋白水平(37.22±4.86)、NLR(3.12±2.00)、PLR(143.61±54.18)、MLR(0.30±0.17)、NAR[0.14(0.11,0.18)]与死亡亚组的年龄、FIGO分期、治疗前血清白蛋白水平(34.33±3.73)、NLR(5.38±2.29)、PLR(185.76±64.04)、MLR(0.41±0.19)、NAR[0.56(0.19,0.61)]存在统计学差异(P<0.05);多因素分析显示病理分型、腹水、淋巴结转移、NLR、PLR、NAR是卵巢癌患者5年生存的独立预测因素。列线图预测术后生存模型的C-index为0.865(0.792,0.921),校准预测曲线和理想曲线拟合良好。【结论】病理分型、腹水、淋巴结转移、治疗前外周血的NLR、PLR、NAR值是卵巢癌患者预后的独立预测因素,以此为基础构建的列线图模型对卵巢癌患者术后生存情况有良好的预测价值。【Objective】To investigate the value of preoperative inflammatory markers in patients with ovarian cancer.【Methods】The clinical data of 145 cases of ovarian cancer and 60 cases of benign ovarian tumor were selected from January 2013 to December 2018 in Guangzhou First People’s Hospital.The patients were divided into survival group(n=48)and death group(n=97)according to follow-up results.The inflammatory markers such as neutrophil/lymphocyte ratio(NLR),platelet/lymphocyte ratio(PLR),monocyte to lymphocyte ratio(MLR),neutrophil to albumin ratio(NAR)were compared between the two groups.Univariate analysis and Cox multivariate analysis were used to analyze the factors affecting the survival.R software was used to establish the survival prediction nomogram model.【Results】The value of NLR[2.98(1.85,5.55)],PLR[156.17(110.38,194.66)],MLR[0.28(0.21,0.42)]and NAR[0.17(0.13,0.28)]in patients with ovarian cancer were higher than NLR[1.94(1.49,3.65)],PLR[125.00(109.38,180.66)],MLR[0.09(0.04,0.19)]and NAR[0.14(0.11,0.17)]in the control group(all P<0.05).There were significant differences in age,FIGO stage,pre-treatment albumin(37.22±4.86),NLR(3.12±2.00),PLR(143.61±54.18),MLR(0.30±0.17)and NAR[0.14(0.11,0.18)]of survival group comparing with the age,FIGO stage,pre-treatment albumin(34.33±3.73)、NLR(5.38±2.29)、PLR(185.76±64.04)、MLR(0.41±0.19)、NAR[0.56(0.19,0.61)]in the death group(all P<0.05).Multivariate analysis showed pathological type,ascites,lymph node metastasis,NLR,PLR and NAR were independent risk factors for 5-year survival of patients with ovarian cancer.The C-index of the nomogram model was 0.865[95%CI(0.792,0.921)],and the calibration prediction curve and ideal curve fit well.【Conclusion】NLR,PLR and NAR were independent risk factors for the prognosis of patients with ovarian cancer.The nomogram model based on them had potential value in predicting the postoperative survival.

关 键 词：卵巢癌 治疗前炎症指标 预后 列线图 

分 类 号：R711[医药卫生—妇产科学]