糖尿病患者基于胱抑素C肾小球滤过率估算公式的建立及评价

Establishment and assessment of glomerular filtration rate estimation equation based on cystatin C for diabetes mellitus patients

目的建立2型糖尿病患者胱抑素C肾小球滤过率(GFR)估算公式,并对其预测肾功能状态的适用性进行评价。方法收集856例2型糖尿病患者相关资料,检测血清肌酐(SCr)、半胱氨酸蛋白酶抑制剂C(Cys C)及糖化血红蛋白(Hb A1c)浓度,同时行同位素99m锝-二乙三胺五乙酸(99mTc-DTPA)肾动态显像,得出同位素GFR(i GFR),并以i GFR作为GFR参考标准。从856例患者中随机选取660例用于建立糖尿病人群估算公式,其余196例作为验证组,采用配对t检验、Bland-Altman一致性检验、χ2检验、Spearman相关分析、ROC曲线等统计学方法对自建公式与简化肾脏病膳食改良试验(MDRD)公式、慢性肾脏病流行病学协助组Cys C公式和SCr/Cys C联合公式(CKD-EPICys C、CKD-EPISCr/Cys C)估算的GFR(e GFR)进行比较。结果糖尿病患者基于Cys C建立的公式为87.2×Cys C-1-4.6。在验证组中,自建公式估算e GFR与i GFR比较,其偏差、一致性及符合率均显著优于MDRD和CKD-EPISCr/Cys C公式,偏差和符合率优于CKD-EPICys C公式。自建公式和CKD-EPICys C公式对肾功能轻中度下降的糖尿病患者进行慢性肾脏病正确分期的百分比显著高于MDRD和CKD-EPISCr/Cys C公式。自建公式和CKD-EPICys C公式在不同血糖控制水平患者间的偏差和符合率比较,差异无统计学意义(P>0.05)。结论自建Cys C公式在估算不同肾功能水平和不同血糖控制水平的糖尿病患者GFR时均表现出较好适用性。

Objective To establish the glomerular fdtration rate （GFR） estimation equation based on cystatin C （Cys C） for type 2 diabetes mellitus patients and assess the applicability for predicting the renal function. Methods Clinical data of 856 type 2 diabetes mellitus patients were collected and levels of serum creatinine （SCr）, Gys C, and plasma glycosylated hemoglobin concentration （HbAlc） were detected. The 99mTc- diethylenetriamine pentaacetic acid-deoxyglucose （99mTc-DTPA） renal dynamic imaging was conducted to obtain the isotopic GFR （iGFR）, which was used as the reference of GFR. Among 856 patients, 660 of them were randomly selected for establishing the estimation equation and other 196 patients belonged to the validation group. The GFR values （eGFR） estimated by self-established equation, modification of diet in renal disease （MDRD）, chronic kidney disease epidemiology collaboration Cys G equation, and Cys C/SCr combined equation （CKD-EPIcysc, CKD-EPIscr/cysc） were compared by statistical methods, including paired t test, Bland and Altman procedures, X2 test, Spearman correlation analysis, and receiver operating characteristic （ROC） curves. Results The self-established equation based on Cys C for type 2 diabetes mellitus patients was 87. 2 × Gys C -1 -4.6. For the validation group, compared with iGFR, the bias, consistency, and coincidence rate of eGFR estimated by the self-established equation were better than those of eGFRs estimated by MDRD and CKD-EPIscr/cys c equations, and the bias and coincidence rate of eGFR estimated by the self-established equation were better than those of eGFR estimated by CKD-EPIcy,c equation. The percentages of correct staging of chronic kidney disease for diabetes mellitus patients with mildly to moderately decreased renal function by the self-established equation and CKD-EPIcy, c equation were significantly higher than those by MDRD and CKD- EPIscr/cysc functions. The differences of bias and coincidence rate between the self-established and CKD-EPIcysc equations among patients with different blood glucose controlled levels were not statistically significant （P〉0, 05）. Conclusion The self-established Cys C equation is applicable for estimating the GFR of diabetes mellitus patients with different renal function levels and different blood glucose controlled levels.