目的: 探讨预后营养指数(prognostic nutrition index, PNI)对非转移性肾细胞癌术后患者预后的意义，并将PNI与中性粒细胞与淋巴细胞比值(neutrophil to lymphocyte ratio, NLR)、血小板与淋巴细胞比值(platelet to lymphocyte ratio, PLR)、全身免疫炎症指数(systemic immune inflammation index, SII)等血液学指标进行比较。方法: 回顾性收集2010年1月至2012年12月于北京大学第一医院泌尿外科接受手术治疗的328例非转移性肾细胞癌患者的临床病理资料。采用受试者工作特征(receiver operating characteristic, ROC)曲线分析各个血液学指标的敏感度、特异度，根据最大约登指数(Youden index)确定其最佳截断值。采用Kaplan-Meier法绘制术后生存曲线，Cox回归模型分析PNI与总生存期(overall survival, OS)、无病生存期(disease-free survival, DFS)的相关性。结果: 根据ROC曲线最大约登指数得出PNI最佳截断值为47.3。低水平的PNI与患者高龄、低体重指数、更高的肿瘤病理T分期相关(P < 0.05)。Kaplan-Meier单因素分析显示，较低的PNI与较差的OS和DFS均显著相关(P < 0.05)，此外，高龄、低体重指数、肿瘤坏死、更高的肿瘤病理T分期及Fuhrman分级均与较差的OS显著相关(P < 0.05)。Cox多因素分析结果表明，4种血液学指标中，只有PNI不论作为连续变量(HR=0.9，95%CI=0.828~0.978，P=0.013)还是分类变量(HR=2.397，95%CI=1.061~5.418，P=0.036)都是影响OS的独立因素。结论: 非转移性肾细胞癌患者术前低PNI是术后高病理T分期的重要预测指标，同时也是术后OS、DFS不佳的独立危险因素。PNI作为预测肾细胞癌患者的预后指标优于其他血清学指标。

Objective: To evaluate the implications of the prognostic nutrition index (PNI) in non-metastatic renal cell carcinoma (RCC) patients treated with surgery and to compare it with other hematological biomarkers, including neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), and systemic immune inflammation index (SII). Methods: A cohort of 328 non-metastatic RCC patients who received surgical treatment between 2010 and 2012 at Peking University First Hospital was analyzed retrospectively. Receiver operating characteristic (ROC) curve analysis was used to determine the optimal cutoff values of the hematological biomarkers. The Youden index was maximum for PNI was value of 47.3. So we divided the patients into two groups (PNI≤ 47. 3 and >47. 3) for further analysis. Categorical variables [age, gender, body mass index (BMI), surgery type, histological subtype, necrosis, pathological T stage and tumor grade] were compared using the Chi-square test and Student' s t test. The association of the biomarkers with overall survival (OS) and disease-free survival (DFS) was analyzed using Kaplan-Meier methods with log-rank test, followed by multivariate Cox proportional hazards model. Results: According to the maximum Youden index of ROC curve, the best cut-off value of PNI is 47. 3. Low level of PNI was significantly associated with older age, lower BMI and higher tumor pathological T stage (P < 0.05). Kaplan-Meier univariate analysis showed that lower PNI was significantly correlated with poor OS and DFS (P < 0.05). In addition, older age, lower BMI, tumor necrosis, higher tumor pathological T stage and Fuhrman grade were significantly correlated with poor OS (P < 0.05). Cox multivariate analysis showed that among the four hematological indexes, only PNI was an independent factor significantly associated with OS, whether as a continuous variable (HR=0.9, 95%CI=0.828-0.978, P=0.013) or a classified variable (HR=2.397, 95%CI=1.061-5.418, P=0.036). Conclusion: Low PNI was a significant predictor for advanced pathological T stage, decreased OS, or DFS in non-metastatic RCC patients treated with surgery. In addition, PNI was superior to the other hematological biomar-kers as a useful tool for predicting prognosis of RCC in our study. It should be externally validated in future research before the PNI can be used widely as a predictor of RCC patients undergoing nephrectomy.