摘要目的:探讨胃癌患者术前外周血系统免疫炎症指数（SII）、血小板与淋巴细胞比值（PLR）、中性粒细胞与淋巴细胞比值（NLR）预测淋巴结转移的价值。方法:回顾性病例队列研究。回顾性分析2021年10月至2022年12月在江苏省肿瘤医院行手术治疗的319例胃癌患者的临床病理资料，其中150例经术后病理检查证实无淋巴结转移（无转移组），169例有淋巴结转移（转移组）。比较两组患者术前SII、PLR和NLR。绘制受试者工作特征曲线，计算曲线下面积（AUC），确定各指标判断淋巴结转移的最佳临界值。采用logistic回归分析胃癌淋巴结转移的影响因素。根据多因素logistic回归结果建立胃癌淋巴结转移的预测模型，使用R4.2.1软件构建列线图模型，采用C统计量与Hosmer-Lemeshow检验评价模型的预测效能与拟合度。结果:150例无转移患者中男性105例，女性45例，年龄[ M（ Q1， Q3）]60岁（54岁，70岁）；169例转移患者中男性124例，女性45例，年龄64岁（56岁，70岁）；转移组患者术前SII、PLR、NLR水平均较无转移患者高，差异均有统计学意义（ Z值分别为-4.58、-4.14、-3.79，均 P<0.05）。SII、PLR和NLR的AUC分别为0.649、0.634和0.623，最佳临界值分别为710.24、182.21和1.72。多因素logistic回归分析结果提示SII>710.24（ OR=2.051，95% CI：1.006~4.184， P=0.048）、PLR>182.21（ OR=1.950，95% CI：1.011~3.760， P=0.046）、NLR>1.72（ OR=1.976，95% CI：1.116~3.500， P=0.020）、癌胚抗原（CEA）>2.57 ng/ml（ OR=2.179，95% CI：1.290~3.679， P=0.004）、糖类抗原724（CA724）>1.83 U/ml（ OR=2.365，95% CI：1.420~3.938， P=0.001）和糖类抗原199（CA199）>14.85 U/ml（ OR=1.965，95% CI：1.129~3.420， P=0.017）是胃癌淋巴结转移的独立危险因素。预测模型：logit（P）=-1.730+0.718×SII+0.668×PLR+0.681×NLR+0.779×CEA+0.861×CA724+0.676×CA199，预报正确率71.47%，C统计量0.762（95% CI：0.710~0.815， P<0.01）。模型预测值与实际观测值间差异无统计学意义（ χ2=3.84， P=0.871）。 结论:术前监测外周血SII、PLR、NLR水平预测胃癌患者发生淋巴结转移有一定价值。

abstractsObjective:To investigate the predictive value of preoperative peripheral blood systemic immune-inflammation index (SII), platelet-to-lymphocyte ratio (PLR) and neutrophil-to-lymphocyte ratio (NLR) for lymph node metastasis in gastric cancer patients.Methods:A retrospective case cohort study was performed. The clinicopathological data of 319 gastric cancer patients who underwent surgical resection in Jiangsu Cancer Hospital from October 2021 to December 2022 were retrospectively analyzed. Among them, 150 cases were postoperatively pathologically confirmed as non-lymph node (non-metastasis group) and 169 cases were confirmed as lymph node (metastasis group). Preoperative SII, PLR, and NLR of patients in both groups were compared. The receiver operating characteristic (ROC) curve was drawn and the area under the curve (AUC) was calculated to determine the optimal cut-off value of all hematological indicators to judge lymph node metastasis. Logistic regression was used to analyze the factors influencing the lymph node metastasis of gastric cancer. According to multivariate logistic regression result, the predictive model for lymph node metastasis in gastric cancer was established. R4.2.1 software was used to establish a nomogram model. C statistics and Hosmer-Lemeshow test were used to evaluate the predictive performance and the fitness of the model.Results:Among 150 non-metastatic patients, 105 cases were male and 45 cases were female, the age [ M ( Q1, Q3)] was 60 years (54 years, 70 years); among 169 metastatic patients, 124 cases were male and 45 cases were female, the age was 64 years (56 years, 70 years). Compared with non-metastasis group, patients in metastasis group had higher levels of preoperative SII, PLR and NLR, and the differences were statistically significant ( Z = -4.58, -4.14, -3.79, all P < 0.05). The AUC values of SII, PLR, and NLR were 0.649, 0.634 and 0.623, respectively, with optimal cut-off values of 710.24, 182.21 and 1.72. Multivariate logistic analysis showed that SII > 710.24 ( OR = 2.051, 95% CI: 1.006-4.184, P = 0.048), PLR > 182.21 ( OR = 1.950, 95% CI: 1.011-3.760, P = 0.046), NLR > 1.72 ( OR =1.976, 95% CI: 1.116-3.500, P = 0.020), carcinoembryonic antigen (CEA) > 2.57 ng/ml ( OR =2.179, 95% CI: 1.290-3.679, P = 0.004), carbohydrate antigen 724 (CA724) > 1.83 U/ml ( OR =2.365, 95% CI: 1.420-3.938, P = 0.001), and carbohydrate antigen 199 (CA199) > 14.85 U/ml ( OR =1.965, 95% CI: 1.129-3.420, P = 0.017) were independent risk factors for lymph node metastasis in gastric cancer. The predictive model of gastric cancer patients with lymph node metastasis: logit (P) = -1.730+0.718×SII+0.668×PLR+0.681×NLR+0.779×CEA+0.861×CA724+0.676×CA199. The prediction accuracy rate was 71.47% and C statistics was 0.762 (95% CI: 0.710-0.815, P < 0.01). There was no statistically significant difference between the model predictive value and the actual observed value ( χ2 = 3.84, P = 0.871). Conclusions:Preoperative monitoring of peripheral blood SII, PLR and NLR levels has a certain value in predicting lymph node metastasis in patients with gastric cancer.