Zhang Fan, Ma Xin, Li Huizi, Guo Gang, Li Pengfei, Li Hongzhao, Gu Liangyou, Li Xintao, Chen Luyao, Zhang Xu
Department of Urology, State Key Laboratory of Kidney Diseases, Chinese PLA General Hospital, Beijing, PR China.
Department of Nutrition, The Rockets Army General Hospital, Beijing, PR China.
Urol Oncol. 2017 Jun;35(6):392-400. doi: 10.1016/j.urolonc.2017.01.004. Epub 2017 Feb 8.
To investigate the predictive and prognostic values of the logistic regression model based on the serum amino acid levels.
The study enrolled 42 patients with clear cell renal cell carcinoma (ccRCC) and 66 matched healthy people admitted to PLA General Hospital from April 2015 to June 2015. Serum samples from the 2 groups were analyzed by isobaric tags for relative and absolute quantitation-liquid chromatography-tandem mass spectrometry (iTRAQ-LC-MS/MS) method. Variables of the 2 groups were compared by Student's t-test or chi-square test. The prediction model was constructed by logistic regression analysis. Oncological outcomes were evaluated by Kaplan-Meier survival analysis and Cox regression analysis.
Pathological diagnosis confirmed that all patients had ccRCC. No significant differences were found in the distribution of age, sex, or body mass index between patients with ccRCC and matched healthy people. A total of 32 amino acids were quantitatively analyzed, and serum levels of 23 amino acids were significantly different between the 2 groups. A predictive logistic regression model containing histidine, glutamine, 1-methyl histidine, and norvaline was constructed. Receiver operating characteristic (ROC) curves for all patients with ccRCC and controls, patients with T1 ccRCC and controls, patients with Fuhrman grade 1 to 2 ccRCC and controls, patients with T2 ccRCC and controls, and patients with Fuhrman grade 3 ccRCC and controls were drawn based on the model. The area under the curve values of the 5 receiver operating characteristic curves were 0.878, 0.885, 0.890, 0.830, and 0.788, respectively. Patients with higher model scores (>2) had significantly poorer progression-free survival (PFS) than patients with lower model scores (≤2). Multivariable Cox regression analysis showed that logistic regression model score was an independent predictor of PFS.
Serum amino acid levels are a potential predictive biomarker for distinguishing patients with ccRCC, especially early T-stage and low Fuhrman grade patients, from healthy people. Serum amino acid levels can also be used in the prognostic evaluation of patients with ccRCC.
探讨基于血清氨基酸水平的逻辑回归模型的预测价值和预后价值。
本研究纳入了2015年4月至2015年6月在中国人民解放军总医院收治的42例透明细胞肾细胞癌（ccRCC）患者和66例匹配的健康人。采用等压标签相对与绝对定量-液相色谱-串联质谱法（iTRAQ-LC-MS/MS）分析两组的血清样本。两组变量采用Student's t检验或卡方检验进行比较。通过逻辑回归分析构建预测模型。采用Kaplan-Meier生存分析和Cox回归分析评估肿瘤学结局。
病理诊断证实所有患者均为ccRCC。ccRCC患者与匹配的健康人在年龄、性别或体重指数分布上未发现显著差异。共对32种氨基酸进行了定量分析，两组间23种氨基酸的血清水平存在显著差异。构建了一个包含组氨酸、谷氨酰胺、1-甲基组氨酸和正缬氨酸的预测逻辑回归模型。基于该模型绘制了所有ccRCC患者与对照组、T1期ccRCC患者与对照组、Fuhrman 1至2级ccRCC患者与对照组、T2期ccRCC患者与对照组以及Fuhrman 3级ccRCC患者与对照组的受试者工作特征（ROC）曲线。5条ROC曲线的曲线下面积值分别为0.878、0.885、0.890、0.830和0.788。模型评分较高（>2）的患者无进展生存期（PFS）明显低于模型评分较低（≤2）的患者。多变量Cox回归分析表明，逻辑回归模型评分是PFS的独立预测因子。
血清氨基酸水平是区分ccRCC患者，尤其是早期T分期和低Fuhrman分级患者与健康人的潜在预测生物标志物。血清氨基酸水平也可用于ccRCC患者的预后评估。
