Bevc Sebastjan, Hojs Nina, Hojs Radovan, Ekart Robert, Gorenjak Maksimiljan, Puklavec Ludvik
Department of Nephrology, Clinic for Internal Medicine, University Clinical Centre Maribor, Maribor, Slovenia.
Faculty of Medicine, University of Maribor, Maribor, Slovenia.
Ther Apher Dial. 2017 Apr;21(2):126-132. doi: 10.1111/1744-9987.12523. Epub 2017 Mar 10.
Estimating glomerular filtration rate (GFR) in elderly patients is a problem, since they are poorly represented in studies developing GFR equations. Serum cystatin C is a better indicator of GFR than serum creatinine in elderly patients. Therefore the aim of our study was to compare frequently used serum cystatin C based GFR equations with a gold standard ( CrEDTA clearance) in elderly chronic kidney disease (CKD) patients. 106 adult Caucasian patients, older than 65 years (58 women, 48 men; mean age 72.5 years), were included. In each patient CrEDTA clearance, serum creatinine (IDMS traceable method) and serum cystatin C (immunonephelometric method) were determined. GFR was estimated using the Simple cystatin C, CKD-EPI cystatin C, CKD-EPI creatinine-cystatin C and BIS2 equation. Mean serum creatinine of our patients was 141.4 ± 41.5 μmol/L, mean serum cystatin C 1.79 ± 0.6 mg/L, mean CrEDTA clearance was 52.2 ± 15.9 mL/min per 1.73 m . Statistically significant correlations between CrEDTA clearance and all formulas were found (P < 0.0001). In the receiver operating characteristic (ROC) curve analysis (cut-off for GFR 45 mL/min per 1.73 m ) no significant differences in diagnostic accuracy between all the before mentioned equations were found. Bland-Altman analysis for the same cut-off showed that CKD-EPI creatinine-cystatin C and BIS2 equation underestimated and CKD-EPI cystatin C and Simple cystatin C equation overestimated measured GFR. All equations lacked precision. Analysis of ability to correctly predict patient's GFR below or above 45 mL/min per 1.73 m showed similar ability for all equations (P = 0.24-0.89). All equations are equally accurate for estimating GFR in elderly Caucasian CKD patients. For daily practice Simple cystatin C equation is most practical.
估算老年患者的肾小球滤过率（GFR）是个难题，因为在制定GFR方程的研究中，老年患者的代表性不足。在老年患者中，血清胱抑素C比血清肌酐是更好的GFR指标。因此，我们研究的目的是在老年慢性肾脏病（CKD）患者中，将常用的基于血清胱抑素C的GFR方程与金标准（铬标记乙二胺四乙酸清除率）进行比较。纳入了106名年龄超过65岁的成年白种人患者（58名女性，48名男性；平均年龄72.5岁）。测定了每位患者的铬标记乙二胺四乙酸清除率、血清肌酐（可溯源至国际计量局的方法）和血清胱抑素C（免疫比浊法）。使用简单胱抑素C方程、慢性肾脏病流行病学协作组（CKD-EPI）胱抑素C方程、CKD-EPI肌酐-胱抑素C方程和BIS2方程估算GFR。我们患者的平均血清肌酐为141.4±41.5μmol/L，平均血清胱抑素C为1.79±0.6mg/L，平均铬标记乙二胺四乙酸清除率为每1.73平方米52.2±15.9毫升/分钟。发现铬标记乙二胺四乙酸清除率与所有公式之间存在统计学显著相关性（P<0.0001）。在受试者工作特征（ROC）曲线分析中（GFR的截断值为每1.73平方米45毫升/分钟），未发现上述所有方程在诊断准确性上有显著差异。针对相同截断值的布兰德-奥特曼分析表明，CKD-EPI肌酐-胱抑素C方程和BIS2方程低估了实测GFR，而CKD-EPI胱抑素C方程和简单胱抑素C方程高估了实测GFR。所有方程都缺乏精确性。对正确预测患者每1.73平方米低于或高于45毫升/分钟的GFR能力的分析表明，所有方程的能力相似（P=0.24-0.89）。在估算老年白种人CKD患者的GFR方面，所有方程同样准确。对于日常实践而言，简单胱抑素C方程最为实用。
