摘要
目的 探讨食管鳞状细胞癌(ESCC)患者在接受根治性放化疗(CRT)过程中纵向炎症标志物与患者预后的关系,并联合各指标构建预后预测模型.方法 回顾性纳入两家医院接受CRT的ESCC患者.采用广义估计方程评估不同疗效患者炎症标志物[系统免疫炎症指数(SII)、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)]的纵向差异.根据最佳截断值将炎症标志物转换为二分类变量.通过Kaplan-Meier曲线分析不同时间点炎症标志物与患者总生存期(OS)和局部无复发生存期(LRFS)的关系.LASSO回归和多因素Cox比例风险回归分析独立危险因素并分别构建预测ESCC患者OS和LRFS的列线图,交叉验证模型的性能.结果 不同疗效患者与纵向SII、NLR和PLR存在显著交互效应(P＜0.05).CRT前和CRT后的炎症标志物,CRT第2周的SII、NLR与PLR和CRT第4周的PLR与LMR均与患者OS和LRFS显著相关(P＜0.05).在多变量生存分析中,CRT 2周和4周的PLR及CRT 4周和结束后的LMR是OS和LRFS的独立影响因素(P＜0.05).基于Cox回归分析P＜0.10的变量建立OS和LRFS预测模型,交叉验证后模型一致性指数分别为0.666和0.665.结论 除基线外,CRT过程中的纵向炎症标志物也可作为ESCC患者的可靠生物标志物.
Abstract
Objective To investigate the relationship between longitudinal inflammatory biomarkers and prognosis in esophageal squamous cell carcinoma(ESCC)patients receiving definitive chemoradiotherapy(CRT),and to construct prognostic models by combining various indicators.Methods ESCC patients who received CRT at two hospitals were retrospectively included.Generalized estimating equation was used to evaluate the longitudinal differences of inflammatory biomarkers[systemic immune-inflammatory index(SII),neutrophil-to-lymphocyte ratio(NLR),platelet-to-lymphocyte ratio(PLR),and lymphocyte-to-monocyte ratio(LMR)]among patients with different treatment responses.The inflammatory biomarkers were converted into categorical variables according to the optimal cut-off values.Kaplan-Meier curve was used to analyze the relationship between inflammatory biomarkers at different time points and overall survival(OS)and local recurrence-free survival(LRFS).LASSO regression and multivariable Cox proportional hazards regression were used to analyze independent risk factors and construct nomograms for predicting OS and LRFS in ESCC patients,respectively,to cross-validate the performance of the model.Results Patients with different treatment responses had significant interaction effects with longitudinal SII,NLR and PLR(P＜0.05).Inflammatory biomarkers before and after CRT,SII,NLR and PLR at week 2 of CRT,and PLR and LMR at week 4 of CRT were significantly correlated with OS and LRFS(P＜0.05).In multivariate survival analysis,PLR at week 2 and week 4 of CRT,and LMR at week 4 and after CRT were independent influencing factors for OS and LRFS(P＜0.05).Based on the variables in Cox regression analysis(P＜0.10),prediction models for OS and LRFS were established.After cross-validation,consistency indexes of the model were 0.666 and 0.665,respectively.Conclusion In addition to baseline,longitudinal inflammatory biomarkers during CRT can also serve as reliable biomarkers for ESCC patients.
维普
万方数据