抗PD‑1/PD‑L1用于消化道肿瘤患者的疗效评估模型的构建及评价

Construction and evaluation of the model assessing the efficacy of anti-PD-1/PD-L1 treatment for gastrointestinal cancer

目的探讨消化道肿瘤患者抗程序性死亡受体-1/程序性死亡受体配体-1(PD-1/PD-L1)治疗消化道肿瘤疗效的影响因素,并建立列线图预测模型。方法回顾性分析2018年1月至2022年12月在我院接受免疫治疗或免疫联合化疗治疗的147例消化道肿瘤患者的临床资料。采用受试者工作特征曲线分析连续性变量的最佳截断值;采用单因素和多因素Cox比例风险回归模型进行无进展生存期(PFS)预测因素的筛选;使用KaplanMeier法和Log-rank检验分析不同因素对患者PFS的影响;将多因素Cox回归中影响患者PFS的因素作为自变量构建列线图模型,并评价预测模型的效能;通过X-tile软件对列线图预测模型进行风险分层。结果单因素Cox回归分析显示,美国东部肿瘤协作组体力状况评分(ECOG PS)、中性粒细胞/淋巴细胞(NLR)、血小板/淋巴细胞(PLR)、白蛋白(ALB)、肌酐清除率(Ccr)、免疫治疗期间激素/质子泵抑制剂(PPI)/抗菌药物(ATB)暴露史及免疫相关不良反应(irAE)均是影响消化道肿瘤患者抗PD-1/PD-L1治疗后PFS的因素。多因素Cox回归分析显示,ECOG PS、免疫治疗期间激素/ATB暴露史是消化道肿瘤患者抗PD-1/PD-L1治疗后PFS的独立预测因素。构建包含以上独立预测因素的列线图预测模型,模型内部验证一致性指数(C-index)为0.814,90 d及180 d PFS的受试者工作特征曲线中AUC分别为0.9332、0.9325,校准曲线斜率接近1,临床决策分析曲线显示该列线图临床实用性良好,低风险患者的PFS明显优于中、高风险组。结论ECOG PS评分、免疫治疗期间激素暴露史和抗菌药物暴露史是影响消化道肿瘤患者抗PD-1/PD-L1治疗后PFS的独立预测因素,本研究所构建的列线图模型为预测消化道肿瘤患者免疫治疗疗效提供参考依据。

Objective To explore the influencing factors of the efficacy of anti-PD-1/PD-L1 treatment for gastrointestinal cancer patients,and to establish a nomogram prediction model.Methods The retrospective analysis encompassed clinical data from 147 patients diagnosed with gastrointestinal cancer who underwent immunotherapy or a combination of immunotherapy with chemotherapy in our hospital from January 2018 to December 2022.Receiver operator characteristic curve analysis was employed to determine the optimal cut-off values for continuous variables.Univariate and multivariate Cox regression models were utilized to identify predictive factors for progression-free survival(PFS).The impact of various factors on PFS was assessed using the Kaplan-Meier method and Log-rank test.The factors affecting PFS in multivariate Cox regression analysis were employed as independent variables to establish nomogram model,and the efficiency of the prediction model was evaluated.X-tile software was used to make risk stratification of the nomogram prediction model.Results Univariate Cox regression analysis showed that Eastern Cooperative Oncology Group performance status(ECOG PS)score,neutrophil-to-lymphocyte,blood platelet-to-lymphocyte,albumin,creatinine clearance rate,steroid/proton pump inhibitor/antibacterial drug exposure during immunotherapy and immune-related adverse events were the factors affecting PFS in patients with gastrointestinal cancer after anti-PD-1/PD-L1 treatment.Multivariate Cox regression analysis showed that Eastern Cooperative Oncology Group performance status score and steroid/antibacterial drug exposure during immunotherapy were independent predictors of PFS in patients with gastrointestinal cancer after anti-PD-1/PD-L1 treatment.The Concordance index(C-index)of the internal validation of the nomogram prediction model was 0.814.The AUC of the receiver operator characteristic curve of 90 d and 180 d PFS was 0.9332 and 0.9325.The slope of the calibration curve was close to 1.Clinical decision analysis curve showed that the clinical practicality of this nomogram was good.The PFS of low-risk patients was significantly better than that of middle-risk and high-risk groups.Conclusion ECOG PS,history of steroid or antibiotic exposure during immunotherapy are independent predictors of PFS in patients with gastrointestinal cancer after anti-PD-1/PD-L1 treatment.The nomogram model constructed in this study provides a reference for predicting the efficacy of immunotherapy in patients with gastrointestinal cancer.