[Relationship between preoperative inflammatory indexes and prognosis of patients with rectal cancer and establishment of prognostic nomogram prediction model].

Objective: To compare the prognostic evaluation value of preoperative neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), and systemic immune-inflammation index (SII) in rectal cancer patients. Nomogram survival prediction model based on inflammatory markers was constructed. Methods: The clinical and survival data of 585 patients with rectal cancer who underwent radical resection in the First Affiliated Hospital of Xi'an Jiao tong University from January 2013 to December 2016 were retrospectively analyzed. The optimal cut-off values of NLR, PLR, LMR, and SII were determined by the receiver operating characteristic (ROC) curve. The relationship between different NLR, PLR, LMR and SII levels and the clinic pathological characteristics of the rectal cancer patients were compared. Cox proportional risk model was used for univariate and multivariate regression analysis. Nomogram prediction models of overall survival (OS) and disease-free survival (DFS) of patients with rectal cancer were established by the R Language software. The internal validation and accuracy of the nomograms were determined by the calculation of concordance index (C-index). Calibration curve was used to evaluate nomograms' efficiency. Results: The optimal cut-off values of preoperative NLR, PLR, LMR and SII of OS for rectal cancer patients were 2.44, 134.88, 4.70 and 354.18, respectively. There was statistically significant difference in tumor differentiation degree between the low NLR group and the high NLR group (P<0.05), and there were statistically significant differences in T stage, N stage, TNM stage, tumor differentiation degree and preoperative carcinoembryonic antigen (CEA) level between the low PLR group and the high PLR group (P<0.05). There was statistically significant difference in tumor differentiation degree between the low LMR group and the high LMR group (P<0.05), and there were statistically significant differences in T stage, N stage, TNM stage, tumor differentiation degree and preoperative CEA level between the low SII group and the high SII group (P<0.05). The multivariate Cox regression analysis showed that the age (HR=2.221, 95%CI: 1.526-3.231), TNM stage (Ⅲ grade: HR=4.425, 95%CI: 1.848-10.596), grade of differentiation (HR=1.630, 95%CI: 1.074-2.474), SII level (HR=2.949, 95%CI: 1.799-4.835), and postoperative chemoradiotherapy (HR=2.123, 95%CI: 1.506-2.992) were independent risk factors for the OS of patients with rectal cancer. The age (HR=2.107, 95%CI: 1.535-2.893), TNM stage (Ⅲ grade, HR=2.850, 95%CI: 1.430-5.680), grade of differentiation (HR=1.681, 95%CI: 1.150-2.457), SII level (HR=2.309, 95%CI: 1.546-3.447), and postoperative chemoradiotherapy (HR=1.837, 95%CI: 1.369-2.464) were independent risk factors of the DFS of patients with rectal cancer. According to the OS and DFS nomograms predict models of rectal cancer patients established by multivariate COX regression analysis, the C-index were 0.786 and 0.746, respectively. The calibration curve of the nomograms showed high consistence of predict and actual curves. Conclusions: Preoperative NLR, PLR, LMR and SII levels are all correlated with the prognosis of rectal cancer patients, and the SII level is an independent prognostic risk factor for patients with rectal cancer. Preoperative SII level can complement with the age, TNM stage, differentiation degree and postoperative adjuvant chemoradiotherapy to accurately predict the prognosis of rectal cancer patients, which can provide reference and help for clinical decision.目的： 探讨术前中性粒细胞与淋巴细胞比值(NLR)、血小板与淋巴细胞比值(PLR)、淋巴细胞与单核细胞比值(LMR)和系统免疫炎症指数(SII)等炎症指标对直肠癌患者预后的预测价值，并构建基于炎症指标的生存预测模型。 方法： 收集2013年1月至2016年12月于西安交通大学第一附属医院行根治性切除的585例直肠癌患者的临床和生存资料，采用受试者工作特征(ROC)曲线确定术前NLR、PLR、LMR和SII的最佳界值，比较不同NLR、PLR、LMR和SII水平患者的临床病理特征。采用Kaplan-Meier法进行生存分析，采用Cox比例风险模型进行预后影响因素的单因素和多因素回归分析。应用R软件构建直肠癌患者的总生存和无病生存列线图预测模型，计算C指数评估模型的准确性，绘制校准图评价模型的预测效能。 结果： 术前NLR、PLR、LMR和SII水平预测直肠癌患者术后总生存的最佳界值分别为2.44、134.88、4.70和354.18。低NLR组与高NLR组患者的肿瘤分化程度差异有统计学意义(P<0.05)，低PLR组与高PLR组患者的T分期、N分期、TNM分期、肿瘤分化程度和术前癌胚抗原(CEA)差异有统计学意义(均P<0.05)，低LMR组与高LMR组患者的肿瘤分化程度有统计学意义(P<0.05)，低SII组与高SII组患者的T分期、N分期、TNM分期、肿瘤分化程度和术前CEA差异有统计学意义(均P<0.05)。多因素Cox回归分析显示，年龄(HR＝2.221，95%CI为1.526~3.231)、TNM分期(Ⅲ期：HR＝4.425，95%CI为1.848~10.596)、肿瘤分化程度(HR＝1.630，95%CI为1.074~2.474)、SII(HR＝2.949，95%CI为1.799~4.835)及术后辅助放化疗(HR＝2.123，95%CI为1.506~2.992)为直肠癌患者总生存的独立影响因素，年龄(HR＝2.107，95%CI为1.535~2.893)、TNM分期(Ⅲ期：HR＝2.850，95%CI为1.430~5.680)、肿瘤分化程度(HR＝1.681，95%CI为1.150~2.457)、SII(HR＝2.309，95%CI为1.546~3.447)及术后辅助放化疗(HR＝1.837，95%CI为1.369~2.464)为直肠癌患者无病生存的独立影响因素。根据多因素Cox回归分析构建直肠癌患者总生存和无病生存的列线图预测模型，其C指数分别为0.786和0.746，校准曲线显示预测曲线和真实曲线拟合度较好。 结论： 术前NLR、PLR、LMR和SII水平均与直肠癌患者的预后有关，其中SII具有独立预测价值。术前SII水平可与年龄、TNM分期、肿瘤分化程度以及术后辅助放化疗联合准确预测直肠癌患者的预后，为临床决策提供参考和帮助。.