目的 探索预后营养指数(PNI)与营养风险筛查 2002(NRS 2002)对肺癌患者预后的预测效能。 方法 根据纳入排 除标准选取 2021 年 6 月至 2023 年 6 月在同济大学附属上海市肺科医院胸外科收治的 652 例肺癌患者临床资料进行回顾性 分析,在患者术后进行为期 14 个月随访,根据患者预后情况分为两组,即生存组( n = 587),死亡组( n = 65),比较两组基本情 况,利用单因素初步选择用于候选因子,将有统计学差异的变量进行 Cox 比例风险回归模型分析,建立预测模型,绘制受试者 操作特征曲线(ROC 曲线)检测预测模型效能。 同时,还需利用 Kaplan-Meier 生存模型分析,不同 PNI、NRS 2002 评分值与生 存关系。 结果 652 例肺癌患者经 14 个月随访,无进展生存 587 例,死亡/ 复发 65 例。 与生存组比较,死亡组患者年龄 66 ~ 70 岁、存在饮酒史、TNM 分期Ⅲ期、PNI<45、治疗前 NRS 2002 评分≥3 分更高,差异存在统计学意义(P<0. 05)。 经 Cox 比例风险 回归模型分析,TNM 分期、PNI<45、治疗前 NRS 2002 评分≥3 分是肺癌患者预后的影响因素(P<0. 05),基于上述因素建立预 测模型,利用 Bootstrap 法内验证,ROC 曲线下面积为 0. 745,95%CI = 0. 680 ~ 0. 810,此次模型可较好预测患者预后情况。 同 时,经 Kaplan-Meier 生存分析,PNI≥45 患者生存时间长于 PNI<45 者,死亡率低于 PNI<45 者;NRS 2002<3 分患者生存时间 长于 NRS 2002≥3 分者,死亡率低于 NRS 2002≥3 分者,差异存在统计学意义(P<0. 05)。 结论 肺癌患者预后受到 TNM 分 期、NRS 2002 评分值、PNI 评分影响,通过构建预测模型,能够及时了解患者术前营养状态,识别高危状态,并进行相应针对性 干预,以期改善预后。

Objective Exploring the predictive value of prognostic nutritional index PNI and nutritional risk screening 2002 NRS2002 in predicting the prognosis of lung cancer patients using Cox proportional hazards model analysis. Method According to the inclusion and exclusion criteria a retrospective analysis was conducted on the clinical data of 652 lung cancer patients admitted to Shanghai Pulmonary Hospital from June 2021 to June 2023. Patients were followed up for 14 months after surgery and divided into two groups based on their prognosis Survival group n = 587 and death group n = 65 . The basic situation of the two groups was compared and a single factor was used to preliminarily select candidate factors. Cox proportional hazards analysis was conducted on variables with statistical differences to establish a prediction model. Receiver operating characteristic ROC curves were drawn to test the effectiveness of the prediction model. Meanwhile it is necessary to use the Kaplan Meier survival model to analyze the relationship between different PNI and NRS 2002 scores and survival. Result After 14 months of follow-up 587 patients with 652 lung cancer patients had no progression and 65 patients died or relapsed. Compared with the survival group the death group had a higher age of 66-70 years a history of alcohol consumption TNM stage Ⅲ PNI <45 and a higher NRS 2002 score of ≥3 before treatment which was statistically significant P<0. 05 . According to Cox proportional hazards analysis TNM staging PNI<45、A NRS 2002 score of≥3 before treatment is a prognostic factor for lung cancer patients P<0. 05 . Based on these factors a prediction model was established and validated using Bootstrap method. The ROC curve area was 0. 745 and the 95%CI was 0. 680-0. 810. This model can effectively predict the prognosis of patients. Meanwhile Kaplan Meier survival analysis showed that patients with PNI≥45 had longer survival times and lower mortality rates than those with PNI <45 The survival time of patients with NRS 2002<3 points is longer than that of patients with NRS 2002 ≥3 points and the mortality rate is lower than that of patients with NRS 2002≥3 points with statistical significance P<0. 05 . Conclusion The prognosis of lung cancer patients is influenced by TNM staging NRS 2002 score and PNI score. By constructing a predictive model it is possible to timely understand the preoperative nutritional status of patients identify high-risk states and carry out targeted interventions to improve prognosis.