CLB-08766; No. of pages: 7; 4C: Clinical Biochemistry xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
F
O
a
8
a r t i c l e
9 10 11 12 13
Article history: Received 8 February 2014 Received in revised form 20 May 2014 Accepted 21 May 2014 Available online xxxx
14 15 16 17 18 19
Keywords: Glomerular filtration rate Cystatin C Creatinine Chronic kidney disease Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)
Department of Laboratory Medicine, Peking Union Medical College Hospital, Chinese Academic Medical Science and Peking Union Medical College, Beijing, P. R. China Department of Nephrology, Peking Union Medical College Hospital, Chinese Academic Medical Science and Peking Union Medical College, Beijing, P. R. China
i n f o
a b s t r a c t
Objectives: The newly developed glomerular filtration rate (GFR)-estimating equations developed by the CKD-EPI Collaboration and Feng et al. (2013) that are based on standardized serum cystatin C (ScysC), combined/not combined with serum creatinine (Scr), require further validation in China. We compared the performance of four new equations (CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations) with the CKD-EPI creatinine equation (CKD-EPIcr) in adult Chinese chronic kidney disease (CKD) patients to clarify their clinical application. Design and Methods: GFR was measured using the dual plasma sampling 99mTc-DTPA method (mGFR) in 252 adult CKD patients enrolled from four centres. Scr and ScysC were measured by standardized assays in a central laboratory. Each equation's performance was assessed using bias, precision, accuracy, agreement, and correct classification of the CKD stage. Results: The measured GFR was 46 [25–83] mL/min per 1.73 m2. The CKD-EPIcys, CKD-EPIcr-cys and Fengcys equations provided significantly higher accuracy (P15: 38.9%, 39.7%, and 38.9%) than the CKD-EPIcr equation (29.8%). The CKD-EPIcr-cys and Fengcr-cys equations presented higher precision (IQR of the difference, 16.4 and 17.3 mL/min per 1.73 m2, respectively) and narrower acceptable limits in Bland–Altman analysis (56.6 and 50.8 mL/min per 1.73 m2, respectively) than single marker-based equations. The CKD-EPIcr-cys equation achieved the highest overall correct proportion (61.5%) in classification of CKD stages. Conclusions: Combining ScysC and Scr measurements for GFR estimation improves diagnostic performance. The Scr–ScysC equation showed better performance than equations based on either marker alone. The CKD-EPIcr-cys equation showed the best performance for GFR estimation in Chinese adult CKD patients. © 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
21 22 23 24 25 26 27 28 29 30 31 20 32
in China revealed that overall prevalence of CKD in China is 10.8%, and the number of CKD patients has reached approximately 119.5 million [3]. Early recognition and diagnosis of CKD is crucial to its timely treatment that can delay its progression and prevent CKD-related cardiovascular and metabolic disorders. Glomerular filtration rate (GFR), the best overall index to reflect kidney function, is central to the diagnosis and classification of CKD [4]. Currently, the ‘gold standard’ for GFR determination is to measure the clearance of exogenous substances, such as inulin, iohexol, 51Cr-EDTA, 99mTc-DTPA, and 125I-iothalamate. These techniques are time-consuming, labour-intensive, expensive, and require the administration of substances that make them incompatible with routine monitoring in clinical practice [5]. To measure GFR conveniently, certain equations based on serum creatinine (Scr) and demographic characteristics have been developed. The Cockcroft–Gault (CG) equation [6] and the abbreviated Modification of Diet in Renal Disease (MDRD) equation [7] were recommended for use by the Kidney Disease Outcomes Quality Initiative (K/DOQI)
49
R
E
b
R O
6 7
P
5
Xiuzhi Guo a, Yan Qin b, Ke Zheng b, Mengchun Gong b, Jie Wu a, Weiling Shou a, Xinqi Cheng a, Liangyu Xia a, Ermu Xu a, Xuemei Li b,⁎, Ling Qiu a,⁎⁎
D
4Q1
E
3
T
2
Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese adult chronic kidney disease patients
C
1
44 42 41 43
N C O
R
39 40
Introduction
46
Chronic kidney disease (CKD) has become a major health problem worldwide. The prevalence of CKD in USA is reported to be 13% [1]; in Europe, it ranges from 5% to 35% [2]. The latest epidemiological survey
U
45
47 48
33 34 35 36 37 38
Abbreviations: eGFR, estimated glomerular filtration rate; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; CKD, chronic kidney disease; MDRD, Modification of Diet in Renal Disease; Scr, serum creatinine; ScysC, serum cystatin C; CysC, cystatin C; DTPA, diethylene triamine pentacetate acid; K/DOQI, Kidney Disease Outcomes Quality Initiative; PETIA, particle-enhanced turbidimetric immunoassay; PENIA, particleenhanced nephelometric immunoassay. ⁎ Correspondence to: X. Li, Department of Nephrology, Peking Union Medical College Hospital, Chinese Academic Medical Science and Peking Union Medical College, Beijing 100730, P.R. China. ⁎⁎ Correspondence to: L. Qiu, Department of Laboratory Medicine, Peking Union Medical College Hospital, Chinese Academic Medical Science and Peking Union Medical College, Beijing 100730, P.R. China. Fax: +86 10 69159743. E-mail addresses: [email protected] (X. Li), [email protected] (L. Qiu).
http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060 0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
125
Overall, 252 adult CKD patients (aged 18–90 years) were enrolled in four major medical centres (North China, Beijing; East China, Shanghai; Central China, Changsha; and Northeast China, Dalian) from September 2007 to December 2010. The study was performed in accordance with
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
126 127 128
C
89 90
E
87 88
R
85 86
R
83 84
O
81 82
C
79 80
N
77 78
U
75 76
Measurement of GFR
141
F
Participant selection
73 74
The 99mTc-DTPA plasma clearance rate was used to measure GFR (mGFR) using the two-sample method [24]. The staff from the four hospitals participating underwent training for these procedures before the initiation of the study. 99mTc-DTPA 296 MBq was injected into the elbow median cubital vein via an intravenous bolus injection. Blood was collected, and radioactivity measurements (P1 and P2) were performed at 2 h (T1) and 4 h (T2), respectively. GFR was corrected for the standard body surface area by multiplying the measured value by 1.73 and dividing it by the patient's body surface area, derived from the Du Bois formula [25]. GFR was calculated as follows:
O
124
72
R O
Materials and methods
70 71
129 130
P
123
68 69
the declaration of the Ethics Review Board for Human Studies of Peking Union Medical College Hospital. Written informed consent was provided by all the participants after education with regard to the potential benefits, risks, and study procedures. The diagnosis of CKD was according to the criteria provided by K/DOQI guidelines [4]. Patients were excluded from the study if any of the following conditions were present: (a) acute kidney injury; (b) receiving haemodialysis or peritoneal dialysis; (c) general oedema, pleural effusion, ascites, or severe heart failure; (d) severe malnutrition, absence of limbs, or ketoacidosis; (e) receiving cimetidine or trimethoprim; (f) received glucocorticoid therapy in the previous 3 months; (g) hyperthyroidism or hypothyroidism; or (h) leukaemia or cancer.
131 132 133 134 135 136 137 138 139 140
142 143 144 145 146 147 148 149 150 151
GFR ¼ fD ln ðP1 =P2 Þ=ðT2 −T1 Þg expf½ðT1 lnP2 Þ−ðT2 lnP1 Þ=ðT2 −T1 Þg 1:73=BSA 2 0:425 0:725 height BSA m ¼ 0:007184 body weight
wherein D is the radioactive count for the injected drugs, T1 and T2 are the first and second blood collection times from the contralateral arm following the intravenous bolus injection of 99mTc-DTPA, respectively, and while P1 and P2 are the radioactive counts in blood plasma at T1 and T2, respectively. The units of weight and height were kg and cm, respectively. The Brochner-Mortensen method [26] was used for correcting for the systematic error of the slope–intercept technique. The corrected GFR was calculated as follows:
T
122
guidelines in 2002. However, subsequent validation studies demonstrated that neither equation provided satisfactory results in various patient populations [8,9]. To minimise such limitations, including imprecision and systematic underestimation of measured GFR with the MDRD equation, a Chronic Kidney Disease Epidemiology Collaboration creatinine equation (CKD-EPIcr equation) was developed based on standardized Scr [10]. Accumulating evidences demonstrated that the CKD-EPIcr equation performed better than the CG and MDRD equations and could be applicable in clinical nephrology [11,12]. In 2012, the K/DOQI guidelines recommended the use of the CKD-EPIcr equation to report estimated GFR (eGFR) in adults determined using serum creatinine levels, as measured by an assay calibrated to the isotope dilution mass spectrometry reference method [13]. Recent studies have suggested that the CKD-EPIcr equation may be the most appropriate creatinine-based equation for determining GFR in Chinese CKD patients [14–17]. However, the accuracy of creatinine-based equations is not satisfactory because the Scr concentration is easily affected by factors other than GFR [18]. Under such circumstances, the use of cystatin C (CysC) as an alternative marker has received great attention. CysC is a cysteine proteinase inhibitor with a molecular weight of 13 kDa that is produced by all nucleated cells at a constant rate and is considered to be close to the ‘ideal’ endogenous marker: it is freely filtered by the glomerulus and catabolized in the proximal tubular epithelial cells without being secreted [19]. Unlike Scr, CysC is not be easily affected by gender or muscle mass; its concentration in serum/plasma depends on the GFR [19]. As it has been reported to be possibly superior to Scr in GFR estimation, several equations based on serum cystatin C (ScysC) have been proposed between 2000 and 2010 [20]. However, the measurement of CysC varies across centres owing to the lack of international standardization, and thus, these equations were not widely used. In the fall of 2010, certified CysC reference material (ERM-DA471/IFCC) for calibrating laboratory assays was released [21], resulting in the measurement of CysC in different laboratories becoming comparable and traceable. This provides a foundation for the evaluation of equations in various populations. In 2012, Inker et al. reported two equations for estimated GFR [22], one based on standardized ScysC values (CKD-EPIcys equation) and the other based on standardized ScysC combined with standardized Scr values (CKD-EPIcr-cys equation). As reported, the CKD-EPIcr-cys equation was accurate compared to the equations based on either marker alone [22]. However, participants in the development study were mostly of western origin; therefore, it is crucial to validate the performance of the equations in ethnically diverse groups. Almost simultaneously, two GFR equations based on standardized ScysC (Fengcys equation) and combined with Scr (Fengcr-cys equation) were also developed using a population of 788 Chinese CKD patients [23]. Under these circumstances, several issues need to be resolved in China, as follows: (i) Are equations based on standardized ScysC better than those based on Scr? (ii) Are Scr–ScysC combined equations better than equations based on either marker alone? (iii) Among CKD-EPIcr, CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations, which is the optimal equation for the Chinese adult CKD population? In this study, we assessed the performance of the four newly developed equations in Chinese adult CKD populations and compared their performance against the CKD-EPIcr equation, which is presently considered as the best choice for estimating GFR in China [14–17].
D
66 67
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx
E
2
mGFR ¼ 0:990778 GFR−0:001218 GFR
153 154 155 156 157 158 159 160
2
162
Estimation of GFR Estimated GFR (eGFR) values were calculated separately using the CKD-EPIcr, CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations, and their corresponding results were labelled as eGFRCKD-EPIcr, eGFRCKD-EPIcys, eGFRCKD-EPIcr-cys, eGFRFengcys, and eGFRFengcr-cys, respectively, as presented in detail in Table 1.
163
Analytical methods
168
On the day of 99mTc-DTPA GFR measurement, a fasting blood sample was collected, centrifuged, and stored at −80 °C. General information of the patients, including age, gender, height, and weight, were documented. All the serum samples were transferred in the frozen state to the central laboratory at the Department of Laboratory Medicine, Peking Union Medical College Hospital, stored and tested via a standard procedure. Scr was measured by an isotope-dilution mass spectrometrytraceable enzymatic method (Roche-Hitachi P-Module instrument with Roche Creatininase Plus assay; Hoffman-La Roche, Basel, Switzerland). Accuracy of the creatinine assay was assessed using NIST SRM 967 I and II (National Institute of Standards and Technology Standard Reference Material). Bias for the creatinine assay with respect to NIST SRM 967 I and II was −0.75% and 1.96%, respectively. During the study period, the coefficients of variation were 0.7% and 0.6% at creatinine concentrations
169
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
164 165 166 167
170 171 172 173 174 175 176 177 178 179 180 181 182
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx t1:1 t1:2
Table 1 Equations for determining the estimated glomerular filtration rate.
t1:3
Name
Year
Equation
t1:4 t1:5 t1:6 t1:7 t1:8
CKD-EPIcr CKD-EPIcys CKD-EPIcr-cys Fengcys Fengcr-cys
2009 2012 2012 2013 2013
eGFR eGFR eGFR eGFR eGFR
= = = = =
141 × min(Scr/κ,1)α × max(Scr/κ,1)−1.209 × 0.993age × 1.018 (if female) × 1.159 (if black) 133 × min(ScysC/0.8, 1)−0.499 × max (ScysC/0.8, 1)−1.328 × 0.996age × 0.932 (if female) 135 × min(Scr/κ,1)α × max(Scr/κ,1)−0.601 × min(ScysC/0.8, 1)−0.375× max(ScysC/0.8, 1)−0.711 × 0.995age × 0.969 (if female) × 1.08 (if black) 78.64 × ScysC-0.964 173.9 × ScysC-0.725 × Scr-0.184 × age-0.193 × 0.89 (if female)
Abbreviations and definitions: eGFR, estimated GFR; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; CKD-EPIcr, CKD-EPI creatinine equation; CKD-EPIcys, CKD-EPI cystatin C equation [22]; CKD-EPIcr-cys, CKD-EPI creatinine-cystatin C equation [22]; Fengcys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC; Fengcr-cys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC in combination with Scr; Scr, serum creatinine; ScysC, serum cystatin C; Note: eGFR is presented in mL/min per 1.73 m2; Scr is shown as mg/L, ScysC is shown as mg/L, age is in years; In the CKD-EPIcr equation, κ is 0.7 for female and 0.9 for male, α is –0.329 for female and –0.411 for male, min indicates a minimum of Scr/κ or 1, and max indicates a maximum of Scr/κ or 1; In the CKD-EPIcys equation, min indicates the minimum of ScysC/0.8 or 1, and max indicates a maximum of ScysC/0.8 or 1; In the CKD-EPIcr-cys equation, min indicates a minimum of Scr/κ or 1, and ScysC/0.8 or 1, and max indicates a maximum of Cr/κ or 1 and ScysC/0.8 or 1.
Statistical analysis
191 192
217
SPSS19.0 (SPSS Inc., USA) and MedCalc (Medcalc software, Belgium) were used for data analysis. Equation performance was assessed by bias, precision, accuracy, and agreement, as recommended by the National Kidney Foundation guidelines [4]. Bias was expressed as the median of the difference between eGFR and mGFR, and precision was expressed as the interquartile range (IQR) for the difference. Accuracy was expressed as the percentage of individuals with eGFR within 15% (P15), 30% (P30), or 50% (P50) of mGFR. Differences between the eGFR values as calculated with the equations and mGFR were compared by Wilcoxon's signed-rank test and analysed by Passing-Bablok regression analysis. Spearman correlation was used to describe the relationship between eGFRs and mGFR. The P15, P30, and P50 values for the four equations were compared with those for the CKD-EPIcr equation by McNemar's test. The Bland–Altman statistical method was used to assess the agreement between mGFR and eGFR. The area between the regression line of difference and the zero difference line shows the degree of deviation of the eGFR equation from the mGFR [27]. Acceptable tolerance for the difference between mGFR and eGFR was defined as 60 mL/min per 1.73 m2. Kappa statistics were used to evaluate the agreement between CKD stage classification based on mGFR and those by eGFRs calculated using the different equations. The kappa value was interpreted as follows: poor (b 0.20), fair (0.21–0.40), moderate (0.41–0.60), good (0.61–0.80), and very good agreement (0.81–1.0). mGFR was categorized according to K/DOQI [4] (i.e. N90, CKD 1 stage; 60–89, CKD 2 stage; 30–59, CKD 3 stage; 15–29, CKD 4 stage; and b15 mL/min per 1.73 m2, CKD 5 stage). P b 0.05 was considered statistically significant.
218
Results
219
Participant characteristics
220
A total of 252 Chinese CKD participants were enrolled in the study, with 138 male patients and a median age of 43 [IQR, 33–59] years. The clinical characteristics of the participants are shown in Table 2. The concentrations of Scr and ScysC in this study were 1.3 [IQR, 0.9–3.3] mg/dL and 1.6 [IQR, 0.9–3.0] mg/L, respectively, and the mGFR was 46 [IQR, 25–83] mL/min per 1.73 m2.
203 204 205 206 207 208 209 210 211 212 213 214 215 216
221 222 223 224 225
C
E
R
201 202
R
199 200
N C O
197 198
U
195 196
T
190
193 194
The overall comparison between the calculated eGFRs and mGFR is shown in Table 3. All of the eGFRs correlated well with mGFR, and the CKD-EPIcr-cys equation showed the highest Spearman correlation coefficient (0.941), followed by the Fengcr-cys equation (0.939) (Table 3). Further, Passing-Bablok analysis revealed that eGFR values of CKDEPIcr, CKD-EPIcr-cys and Fengcys equations had no marked linear deviation from the mGFR (p N 0.05, Fig. 1). The Fengcr-cys equation did not illustrate a significant proportional difference [the 95% CI of slopes (b) include 1, Fig. 1E]; however, it revealed a large positive constant difference [intercept (a) = 7.301]. The CKD-EPIcys equation showed no significant constant error [the 95% CI of intercepts (a) include 0, Fig. 1-B] and a slight proportional difference. Significant differences in the proportional errors and the constant errors were noted for all the other three equations. Further comparisons of total bias, precision, and accuracy in subgroups of mGFR less than or greater than 60 mL/min per 1.73 m2 are shown in Table 4. The CKD-EPIcr-cys and Fengcr-cys equations provided better precision (IQR of the difference, 16.4 and 17.3 mL/min per 1.73 m2, respectively) as compared with the single marker-based equations. The medians of the difference of the CKD-EPIcys equation and CKDEPIcr-cys equations (-0.2 and 0.5 mL/min per 1.73 m2, respectively) were less than those of the other equations. The CKD-EPIcys, CKD-EPIcr-cys, and Fengcys equations provided significantly higher accuracy than the
R O
188 189
242
P
186 187
Assessment of performance
of 1.0 mg/dL and 6.2 mg/dL, respectively. ScysC was measured by a standardized particle-enhanced immunonephelometric assay on the Siemens Dade Behring Nephelometer with N Latex CysC reagent in 2011, which was traceable to ERM-DA471/IFCC, the international certified reference material for CysC. The coefficients of variation were 2.7% and 2.5% at CysC concentrations of 0.9 mg/L and 1.8 mg/L, respectively, during the study period.
D
184 185
E
183
O
F
t1:9 t1:10 t1:11 t1:12 t1:13 t1:14 t1:15
3
243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 t2:1 t2:2
Table 2 Baseline participant characteristics. Characteristics
Values (n = 252)
t2:3
Age (years, median [IQR]) Males, n (%) Weight (kg, median [IQR]) Height (m, median [IQR]) Serum creatinine (mg/dL, median [IQR]) Serum cystatin C (mg/L, median [IQR]) Measured GFR (mL/min per 1.73 m2, median [IQR]) Causes of CKD, n (%) Primary or secondary renal glomerular disease Hypertensive nephropathy Diabetic nephropathy Chronic tubulointerstitial nephropathy Obstructive nephropathy Cystic nephropathy Other or unclear CKD staging results, n (%) ≥90 60–89 30–59 15–29 b15
43 [33, 59] 138 (54.8) 65.0 [56.5, 75.0] 1.66 [1.60, 173] 1.3 [0.9, 3.3] 1.6 [0.9, 3.0] 46 [25, 83]
t2:4 t2:5 t2:6 t2:7 t2:8 t2:9 t2:10 t2:11 t2:12 t2:13 t2:14 t2:15 t2:16 t2:17 t2:18 t2:19 t2:20 t2:21 t2:22 t2:23 t2:24
166 (65.9) 26 (10.3) 40 (15.9) 10 (4.0) 2 (0.8) 2 (0.8) 6 (2.4) 51 (20.2) 50 (19.8) 69 (27.4) 45 (17.6) 37 (14.7)
Note: Data are shown as median [IQR] or n (%). IQR, interquartile range; GFR, glomerular filtration rate; CKD, chronic kidney disease. Serum creatinine in mg/dL to μmol/L, ×88.4. CKD staging results based on mGFR.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
t2:25 t2:26 t2:27
4 t3:1 t3:2
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx
Table 3 Overall comparison between eGFR and mGFR: Correlation, Bland–Altman analyses (n = 252).
t3:3
mean ± SD
correlation
t3:4 t3:5 t3:6 t3:7 t3:8 t3:9 t3:10
mGFR eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
54 60 55 57 58 61
± ± ± ± ± ±
35 43 39 41 32 34
Bland–Altman analysis
r
p-value
mean differences
area⁎ (arbitrary unit)
95% AL
0.935 0.924 0.941 0.913 0.939
b0.001 b0.001 b0.001 b0.001 b0.001
5.7 0.8 2.4 3.1 6.2
1178 402 761 470 726
-27.1–38.4 -29.1–30.6 -25.9–30.7 -26.8–33.1 -19.2–31.6
t3:11 t3:12
Note: Units are mL/min per 1.73 m2 for mean ± SD, Mean differences, 95% AL; Spearman correlation of eGFRs against mGFR were made; r is Spearman's correlation coefficient between eGFRs and mGFR; ⁎the area between the Bland–Altman regression line and the zero difference line; arbitrary unit, i.e., (mL/min per 1.73 m2)2; 95% AL denotes 95% agreement limits.
266 267
CKD-EPIcr equation (P15: 38.9%, 39.7%, and 38.9% vs 29.8%, McNemar's test, p b 0.05). The accuracy of the CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations showed no statistically significant differences as compared with each other (P30: 64.3%, 62.3%, 61.9%, and 62.7%, respectively, McNemar's test, p N 0.05). Table 5 shows the percentages of patients that were classified in the correct CKD stages by the eGFRs from the different equations. Kappa statistics were used to evaluate the agreement between classification of CKD stages by mGFR and that by eGFRs calculated by different equations. Higher kappa values indicated better agreement in CKD staging as calculated by mGFR. The CKD-EPIcr-cys equation achieved the best agreement with a kappa value of 0.520 and an overall correct classification proportion of 61.5%. In CKD stages 2–4, the proportions of correct classification by the Fengcys and Fengcr-cys equations were significantly higher than those by the other equations; however, in CKD stages 1 and 5, these proportions were lower, particularly in CKD stage 5 (Table 5). The Bland–Altman plots were used to evaluate the bias (mean of difference) and the degree of agreement between eGFRs and mGFR (Fig. 2). Both the slopes of the regression of differences against averages by the CKD-EPIcys and Fengcr-cys equations were nearly equal to zero, which indicated no proportional bias. However, the CKD-EPIcys equation demonstrated a smaller mean of difference, while the Fengcr-cys equation showed a larger constant positive difference than the other equations (mean difference, 0.8 and 6.2 mL/min per 1.73 m2, respectively).
286 287 288 289 290
F
O
R O
P
E
D
Although the implementation of eGFR has resulted in the identification of large numbers of patients with undiagnosed CKD, an optimal equation for estimating the GFR in different populations has not been validated yet. The standardization of Cr measurement has greatly promoted the development and application of Scr-based eGFR equations [28]. After the standardization of ScysC measurement, equations based on standardized ScysC have attracted attention and have been validated in different populations. Masson et al. validated the CKDEPIcys and CKD-EPIcr-cys equations in kidney transplant recipients and found that both of these were better than the CKD-EPIcr equation [29]. Horio et al. studied the accuracy of these equations in the Japanese population and suggested that the CKD-EPIcys equation could be used in patients of different races without modification [30]. In China, Liu et al. assessed the CKD-EPIcr-cys equation in a cohort of elderly
T
284 285
C
282 283
E
280 281
299
R
278 279
Discussion
R
276 277
O
274 275
C
272 273
N
270 271
291 292
U
268 269
In the estimation of GFR consistency, two combined equations (CKD-EPIcr-cys and Fengcr-cys) were both better than the CKD-EPIcr, CKD-EPI cys, and Fengcys equations alone (limits of agreement = 50.8 and 56.6 vs 65.5, 59.7, and 59.9 mL/min per 1.73 m2, respectively). When the area between the regression line of difference and the zero difference line was compared, the CKD-EPIcys equation gave the smallest deviation (Table 1, 402 arbitrary units), followed by the Fengcys equation (Table 1, 470 arbitrary units).
Fig. 1. Passing-Bablok regression analysis comparing the differences between eGFR by (A) CKD-EPIcr, (B) CKD-EPIcys, (C) CKD-EPIcr-cys, (D) Fengcys, and (E) Fengcr-cys and mGFR. Note: All data are presented in mL/min per 1.73 m2; a, 95% confidence interval for the intercept; b, 95% confidence interval for the slope; solid lines, regression line; dashed lines, 95% confidence intervals of the regression line; dotted lines, identity line (x = y). Abbreviations and definitions: CKD-EPIcr, CKD-EPI creatinine equation [10]; CKD-EPIcys, CKD-EPI cystatin C equation [22]; CKD-EPIcr-cys, CKD-EPI creatinine-cystatin C equation [22]; Fengcys equation, newly developed equation by Feng et al.(2013) [23] based on standardized ScysC; Fengcr-cys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC in combination with Scr.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
293 294 295 296 297 298
300 301 302 303 304 305 306 307 308 309 310 311 312 313
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx Table 4 Comparison of median bias, precision, and accuracy of equations in subgroups when mGFR b 60 mL/min per 1.73 m2 or mGFR ≥ 60 mL/min per 1.73 m2. mGFR ≥ 60 mL/min per 1.73 m2 (n = 101)
Bias: Median difference, i.e. eGFR – mGFRa eGFRCKD-EPIcr 2.4; p b 0.001 eGFRCKD-EPIcys -0.2; p = 0.9 eGFRCKD-EPIcr-cys 0.5; p = 0.2 eGFRFengcys 3.3; p b 0.001 eGFRFengcr-cys 5.6; p b 0.001
-1.7; p = 0.7 -0.5; p b 0.001 -2.7; p b 0.001 8.2; p b 0.001 7.9; p b 0.001
14.6; p b 0.001 1.2; p b 0.001 4.9; p b 0.001 -3.9; p = 0.1 1.6; p b 0.001
Precision: IQR of the difference eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
22.1 18.4 16.4 19.0 17.3
15.2 14.6 13.8 15.7 13.1
22.3 25.6 21.0 19.4 22.0
Accuracy: P15b (%) eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
29.8 38.9; p 39.7; p 38.9; p 36.1; p
= = = =
0.03 0.02 0.03 0.1
22.5 27.8; p 27.2; p 23.2; p 20.5; p
= = = =
0.3 0.4 0.9 0.5
Accuracy: P30b (%) eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
59.1 64.3; p 62.3; p 61.9; p 62.7; p
= = = =
0.2 0.5 0.5 0.4
47.7 51.7; p 49.7; p 45.7; p 45.0; p
= = = =
0.5 0.7 0.7 0.6
Accuracy: P50b (%) eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
79.0 83.3; p 83.7; p 79.0; p 86.6; p
= = = =
0.2 0.2 1.0 0.7
69.5 73.5; p 74.8; p 64.9; p 67.5; p
= = = =
0.4 0.3 0.4 0.7
40.6 55.4; p 58.4; p 62.4; p 59.4; p
= = = =
0.04 0.01 0.002 0.008
76.2 83.2; p 81.2; p 86.1; p 89.1; p
= = = =
0.2 0.4 0.07 0.01
O R O P
t4:4 t4:5 t4:6 t4:7 t4:8 t4:9 t4:10 t4:11 t4:12 t4:13 t4:14 t4:15 t4:16 t4:17 t4:18 t4:19 t4:20 t4:21 t4:22 t4:23 t4:24 t4:25 t4:26 t4:27 t4:28 t4:29 t4:30 t4:31 t4:32 t4:33 t4:34 t4:35 t4:36 t4:37
Overall (n = 252)
93.1 98.0; p = 0.2 97.0; p = 0.2 100; p = 0.02 100; p = 0.02
D
Variable
F
mGFR b 60 mL/min per 1.73 m2 (n = 151)
t4:3
E
t4:1 t4:2
5
Note: mGFR is presented in mL/min per 1.73 m2. GFR, glomerular filtration rate; eGFR, estimated GFR; IQR, interquartile range; mGFR, measured GFR; P30 (P15, P50), percentage of estimates within 30% (15%, 50%) of mGFR. a Wilcoxon's signed-rank test was used to compare the bias of each of the eGFRs against mGFR. b McNemar's test was used to compare P15, P30, and P50 values of the eGFRs against the CKD-EPIcr equation.
314
330
Chinese participants, and found that the equation was more suitable for the elderly Chinese patients than the Scr-based equations [31]. However, Liu's study involved only elderly individuals, and GFR was measured using the renal dynamic imaging method, which provided a non-reliable GFR that cannot be considered a ‘gold standard’ [32]. Feng et al. developed two equations based on standardized ScysC in Chinese CKD patients in 2013 [23]; however, further external validation is necessary for its wide application. In the present study, we observed that the combination of ScysC with Cr for calculating the eGFR truly improved the equations' performance. The four new equations, i.e., the CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations, all presented higher precision and accuracy than the CKD-EPIcr equation. These results were similar to findings in western populations, as reported by the CKD-EPI Collaboration [22]. Reclassification of CKD patients using the more accurate GFR-estimating tools, especially those based on ScysC, has recently become an important research focus. In this study, it appeared that stage misclassification was
t5:1 t5:2
Table 5 Correct classification proportion of CKD using eGFRs and mGFR [n (%)].
320 321 322 323 324 325 326 327 328 329
C
E
R
R
318 319
N C O
317
U
315 316
T
t4:38 t4:39 t4:40 t4:41
t5:3 t5:4 t5:5 t5:6 t5:7 t5:8 t5:9 t5:10 t5:11 t5:12
mGFR⁎ eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
only reduced by the CKD-EPIcr-cys equation compared to the CKD-EPIcr equation (Table 5). Further analysis of the data revealed that this was because the CKD-EPIcr equation underestimated GFR in individuals with a low GFR and significantly overestimated GFR in individuals with a high GFR (Fig. 2); therefore, among the five equations, the CKD-EPIcr equation gave the highest proportions of correct CKD stage 1 and 5 classifications. For CKD stages 2–4, the Fengcys and Fengcr-cys equations provided significantly higher proportions of correct classifications than the other equations; however, these two equations overestimated GFR in patients with a low mGFR (Fig. 2), resulting in a large number of patients with CKD stage 5 misclassified as having CKD stage 4. Overall, the performance of the two ScysC-based equations (CKD-EPIcys and Fengcys) was superior to that of the CKD-EPIcr equation, and two combined equations (CKD-EPIcr-cys and Fengcr-cys equations) were better than the single marker–based equations. The performances of the Fengcys and Fengcr-cys equations was also compared with those of the CKD-EPIcys and CKD-EPIcr-cys equations,
CKD 1 stage
CKD 2 stage
CKD 3 stage
CKD 4 stage
CKD 5 stage
Total
Kappa
p-Value
51 46 (90.2)★ 40(78.4) 45 (88.0) 31 (61.8) 41 (80.1)
50 20 (40.0) 23 (46.0) 28 (56.0) 35 (70.0) 32 (64.0)
69 34 (49.3) 34 (49.3) 33 (47.8) 39 (56.5) 41 (59.4)
45 20 (44.4) 21 (46.7) 18 (40.0) 28 (62.2) 24 (53.3)
37 31 (83.8) 30 (81.1) 31 (83.8) 6 (16.2) 6 (16.2)
252 151 (59.9) 148 (58.7) 155 (61.5) 139 (55.2) 144 (57.1)
0.500 0.484 0.520 0.430 0.455
b0.001 b0.001 b0.001 b0.001 b0.001
Kappa analysis was used to evaluate the agreement between the stages classified according to mGFR and each eGFR-classified CKD stage. ⁎ CKD staging results based on mGFR according to the recommendations by K/DOQI. ★ Number of patients (%) correct classified.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx
R O
O
F
6
359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384
E
Fengcys and Fengcr-cys equations in this study may be due to these factors. The K/DOQI criteria require the acceptable level of 30% accuracy (P30) to be at least 70% [4]. However, none of the five equations in the study accurately fulfilled this criterion, with P30 ranging from 59.1% to 64.3% and a much lower accuracy for the subgroup of mGFR values b 60 mL/min per 1.73 m2 (P30, 45.0–51.7%). Low accuracy has also been reported in Chinese populations [14–17,31]. Possible reasons for these discrepancies are as follows. First, the applicability of equations may be not general enough, at least at present. Differences in race, physical fitness, diet, and living habits can influence the level of SCr [18]. Thus, the accuracy of equations in different populations may have varied. Further studies, for example, with considering a racial coefficient, may improve the accuracy of the CKD-EPIcr-cys equation. Second, differences in the method of GFR measurement may be a contributory factor. Unlike renal clearance of 125I-iothalamate, used when developing the CKD-EPI equations, Chinese studies generally employ 99mTc-DTPA dual plasma clearance as reference method for GFR [14,15,23]. Systematic bias cannot be avoided when two distinct methods are used to measure GFR [35]. Third, the clearance of dual plasma sampling 99mTc-DTPA method was used in our study. The first and second samples were obtained at 2 h and 4 h, respectively, for all patients. The use of a narrow time interval (2 h and 4 h) for patients with renal failure may reduce the accuracy of the plasma clearance of 99mTc-DTPA. It is worth mentioning that although ScysC was less affected by diet, race, and muscle mass than Scr, factors such as diabetes [36], thyroid function [37], immunosuppressant therapy [38], and smoking [39] may influence ScysC levels. Unfortunately, subjects with some of these factors were not completely excluded from the study (see exclusion criteria). The respective weightage of these confounding factors has not been thoroughly investigated to date, and clinicians should be prudent regarding the results in cases where these factors are present. Our study has some limitations. First, 99mTc-DTPA dual plasma clearance was used as reference method, in contrast to renal clearance of 125I-iothalamate that used in the development of the CKD-EPI equations. This would invariably introduce systematic bias. Second, the sampling times were 2 h and 4 h in our study, and the interval
T
C
E
R
358
R
356 357
O
354 355
C
352 353
N
350 351
respectively. The Fengcr-cys and CKD-EPIcr-cys equations provided similar overall precision (IQR of the difference, 17.3 vs 16.4) and accuracy (P30, 64.3% vs 62.7%). In the Bland–Altman analysis, the Fengcr-cys equation presented narrower acceptable limits than the CKD-EPIcr-cys equation (50.8 vs 56.6 mL/min per 1.73 m2), and no significant difference in proportional error against mGFR was noted for the Fengcr-cys equation by Passing-Bablok analysis. However, the Fengcr-cys equation misclassified a large number of patients with CKD stage 5 as having CKD stage 4. Therefore, although the Fengcr-cys equation provided significantly higher proportions of correct CKD stage 2–4 classifications than the CKD-EPIcr-cys equation, its overall correct classification proportion remained lower than that of the CKD-EPIcr-cys equation. A similar situation was also observed for the Fengcys and CKD-EPIcys equations. Why were the performances of the two Chinese-based (Fengcys and Fengcr-cys) equations not better than the CKE-EPI equations, and why were these not as good as they were at the time of development? First, the CKD-EPIcr-cys and CKD-EPIcys equations provide different coefficients for different levels of Cr and CysC; however, Feng's equations were not stratified by the Cr and CysC levels. This may result in a high degree of compliance in the intermediate region, along with large deviations in the high and low regions of GFR. Therefore, further improvements are needed for the two Feng's equations. Second, the method for CysC measurement differed in the two studies. The Fengcys and Fengcr-cys equations were developed on the basis of a particle-enhanced turbidimetric immunoassay (PETIA) method for CysC measurement, whereas CysC was measured by using the particle-enhanced nephelometric immunoassay (PENIA) method in the development of the CKD-EPI equations and in our study. Although both methods are traceable to ERM-DA471/IFCC, differences cannot be avoided because two distinct methods were used; CysC values by the PETIA method have been reported to be 27.5% higher than those by the PENIA method [33]. Third, we used the Brochner-Mortensen correction rather than the Chantler correction used by Feng et al. to calculate the systematic error of the slope–intercept technique, as recommended by a guideline [34]. The Brochner-Mortensen technique is preferred as it addresses the variation in percentage error with GFR, whereas the Chantler technique does not adequately correct for the variation in percentage error with GFR [34]. The poor performance observed for the
U
348 349
D
P
Fig. 2. Bland–Altman plots comparing the agreement between eGFR by (A) CKD-EPIcr, (B) CKD-EPIcys, (C) CKD-EPIcr-cys, (D) Fengcys, and (E) Fengcr-cys and mGFR. Note: Full line, mean difference between two methods; dashed line, ±1.96 SD difference against mean; the regression line of differences versus averages and the 95% confidence interval are also presented. All data are presented in mL/min per 1.73 m2. Abbreviations and definitions: CKD-EPIcr, CKD-EPI creatinine equation [10]; CKD-EPIcys, CKD-EPI cystatin C equation [22]; CKD-EPIcr-cys, CKD-EPI creatinine-cystatin C equation [22]; Fengcys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC; Fengcr-cys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC in combination with Scr.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx
References
440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 561
[1] Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, et al. Prevalence of chronic kidney disease in the United States. JAMA 2007;298:2038–47. [2] Zhang QL, Rothenbacher D. Prevalence of chronic kidney disease in populationbased studies: Systematic review. BMC Public Health 2008;8:117–30. http:// dx.doi.org/10.1186/1471-2458-8-117. [3] Zhang L, Wang F, Wang L, Wang W, Liu B, Liu J, et al. Prevalence of chronic kidney disease in China: a cross-sectional survey. Lancet 2012;379:815–22. http:// dx.doi.org/10.1016/S0140-6736(12)60033-6. [4] N.K. Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39:S1–S266. [5] Florkowski CM, Chew-Harris JS. Methods of estimating GFR-different equations including CKD-EPI. Clin Biochem Rev 2011;32:75–9. [6] Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31–41. [7] Levey AS, Greene T, Kusek JW, Beck GJ, MDRD Study Group. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 2000;11:155A. [8] Poggio ED, Nef PC, Wang X, Greene T, Van Lente F, Dennis VW, et al. Performance of the Cockcroft-Gault and modification of diet in renal disease equations in estimating GFR in ill hospitalized patients. Am J Kidney Dis 2005;46:242–52. [9] Ibrahim H, Mondress M, Tello A, Fan Y, Koopmeiners J, Thomas W. An alternative formula to the Cockcroft-Gault and the modification of diet in renal diseases formulas in predicting GFR in individuals with type 1 diabetes. J Am Soc Nephrol 2005;16:1051–60. [10] Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, Feldman HI, et al. CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–12. [11] Stevens LA, Schmid CH, Greene T, Zhang YL, Beck GJ, Froissart M, et al. Comparative performance of the CKD Epidemiology Collaboration (CKD-EPI) and the Modification of Diet in Renal Disease (MDRD) study equations for estimating GFR levels above 60 mL/min/1.73 m2. Am J Kidney Dis 2010;56:486–95. http://dx.doi.org/ 10.1053/j.ajkd.2010.03.026. [12] Michels WM, Grootendorst DC, Verduijn M, Elliott EG, Dekker FW, Krediet RT. Performance of the Cockcroft-Gault, MDRD, and new CKD-EPI formulas in relation to GFR, age, and body size. Clin J Am Soc Nephrol 2010;5:1003–9. [13] Stevens PE, Levin Adeera, the Kidney Disease: Improving Global Outcomes Chronic Kidney Disease Guideline Development Work Group Members. Evaluation and management of chronic kidney disease: synopsis of the kidney disease: Improving Global Outcomes 2012 Clinical Practice Guideline. Ann Intern Med 2013;158:825–30. [14] Kong X, Ma Y, Chen J, Luo Q, Yu X, Li Y, et al. Evaluation of the Chronic Kidney Disease Epidemiology Collaboration equation for estimating glomerular filtration rate in the Chinese population. Nephrol Dial Transplant 2013;28:641–51. [15] Xie P, Huang JM, Lin HY, Wu WJ, Pan LP. CDK-EPI equation may be the most proper formula based on creatinine in determining glomerular filtration rate in Chinese patients with chronic kidney disease. Int Urol Nephrol 2013;45:1057–64. http:// dx.doi.org/10.1007/s11255-012-0325-7.
F
439
C
E
R
R
N C O
U
431 432
O
437 438
This study was supported by the Beijing Municipal Science and Technology Commission (D09050704310901) and Sichuan Provincial Department of Science and Technology (2009sz0066). We thank the members and staff of our research group for their cooperative work.
429 430
R O
435 436
428
P
Acknowledgments
426 427
[16] Li JT, Xun C, Cui CL, Wang HF, Wu YT, Yun AH, et al. Relative performance of two equations for estimation of glomerular filtration rate in a Chinese population having chronic kidney disease. Chin Med J (Engl) 2012;125:599–603. [17] Pei XH, He J, Liu Q, Zhu B, Bao LH, Yan CJ, et al. Evaluation of serum creatinine- and CysC-based equations for the estimation of glomerular filtration rate in a Chinese population. Scand J Urol Nephrol 2012;46:223–31. http://dx.doi.org/10.3109/ 00365599.2012.660985. [18] Perrone RD, Madias NE, Levey AS. Serum creatinine as an index of renal function: new insights into old concepts. Clin Chem 1992;38:1933–53. [19] Filler G, Bökenkamp A, Hofmann W, Le Bricon T, Martínez-Brú C, Grubb A. Cystatin C as a marker of GFR-history, indications, and future research. Clin Biochem 2005;38:1–8. [20] Harman G, Akbari A, Hiremath S, White CA, Ramsay T, Kokolo MB, et al. Accuracy of cystatin C-based estimates of glomerular filtration rate in kidney transplant recipients: a systematic review. Nephrol Dial Transplant 2013;28:741–57. http:// dx.doi.org/10.1093/ndt/gfs498. [21] Grubb A, Blirup-Jensen S, Lindström V, Schmidt C, Althaus H, Zegers I. First certified reference material for cystatin C in human serum ERM-DA471/IFCC. Clin Chem Lab Med 2010;48:1619–21. http://dx.doi.org/10.1515/CCLM.2010.318. [22] Inker LA, Schmid CH, Tighiouart H, Eckfeldt JH, Feldman HI, Greene T, et al. Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med 2012;367:20–9. http://dx.doi.org/10.1056/NEJMoa1114248. [23] Feng JF, Qiu L, Zhang L, Li X, Yang Y, Zeng P, et al. Multicenter study of creatinineand/or cystatin C-based equations for estimation of glomerular filtration rates in Chinese patients with chronic kidney disease. PLoS One 2013;8:e57240. http:// dx.doi.org/10.1371/journal.pone.0057240. [24] Fleming JS, Zivanovic MA, Blake GM, Burniston M, Cosgriff PS, British Nuclear Medicine Society. Guidelines for the measurement of glomerular filtration rate using plasma sampling. Nucl Med Commun 2004;25:759–69. [25] Du BD, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. Nutrition 1989;5:303–11. [26] Bröchner-Mortensen J. A simple method for the determination of glomerular filtration rate. Scand J Clin Lab Invest 1972;30:271–4. [27] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10. [28] Myers GL, Miller WG, Coresh J, Fleming J, Greenberg N, Greene T, et al. Recommendations for improving serum creatinine measurement: a report from the Laboratory Working Group of the National Kidney Disease Education Program. Clin Chem 2006;52:5–18. [29] Masson I, Maillard N, Tack I, Thibaudin L, Dubourg L, Delanaye P, et al. GFR estimation using standardized cystatin C in kidney transplant recipients. Am J Kidney Dis 2013;61:279–84. http://dx.doi.org/10.1053/j.ajkd.2012.09.010. [30] Horio M, Imai E, Yasuda Y, Watanabe T, Matsuo S. GFR estimation using standardized serum cystatin C in Japan. Am J Kidney Dis 2013;61:197–203. http://dx.doi.org/ 10.1053/j.ajkd.2012.07.007. [31] Liu X, Ma H, Huang H, Wang C, Tang H, Li M, et al. Is the Chronic Kidney Disease Epidemiology Collaboration creatinine-cystatin C equation useful for glomerular filtration rate estimation in the elderly? Clin Interv Aging 2013;8:1387–91. http:// dx.doi.org/10.2147/CIA.S52774. [32] Xie P, Huang JM, Liu XM, Wu WJ, Pan LP, Lin HY. 99mTc-DTPA renal dynamic imaging method may be unsuitable to be used as the reference method in investigating the validity of CDK-EPI Equation for determining Glomerular Filtration Rate. PLoS One 2013;8:e62328. [33] Voskoboev NV, Larson TS, Rule AD, Lieske JC. Analytic and clinical validation of a standardized cystatin C particle enhanced turbidimetric assay (PETIA) to estimate glomerular filtration rate. Clin Chem Lab Med 2012;50:1591–6. http://dx.doi.org/ 10.1515/cclm-2012-0063. [34] Fleming JS, Zivanovic MA, Blake GM, Burniston M, Cosgriff PS, British Nuclear Medicine Society. Guidelines for the measurement of glomerular filtration rate using plasma sampling. Nucl Med Commun 2004;25:759–69. [35] Dai SS, Yasuda Y, Zhang CL, Horio M, Zuo L, Wang HY. Evaluation of GFR measurement method as an explanation for differences among GFR estimation equations. Am J Kidney Dis 2011;58:496–8. http://dx.doi.org/10.1053/j.ajkd.2011.05.016. [36] Stevens LA, Schmid CH, Greene T, Li L, Beck GJ, Joffe MM, et al. Factors other than glomerular filtration rate affect serum cystatin C levels. Kidney Int 2009;75:652–60. http://dx.doi.org/10.1038/ki.2008.638. [37] Kotajima N, Yanagawa Y, Aoki T, Tsunekawa K, Morimura T, Ogiwara T, et al. Influence of thyroid hormones and transforming growth factor-β1 on cystatin C concentrations. J Int Med Res 2010;38:1365–73. [38] Rule AD, Bergstralh EJ, Slezak JM, Bergert J, Larson TS. Glomerular filtration rate estimated by cystatin C among different clinical presentations. Kidney Int 2006;69:399–405. [39] Sjöström PA, Jones IL, Tidman MA. Cystatin C as a filtration marker-haemodialysis patients expose its strengths and limitations. Scand J Clin Lab Invest 2009;69:65–72. http://dx.doi.org/10.1080/00365510802326469.
D
434
424 425
T
433
may be too narrow for patients with renal failure, thus reducing accuracy. In summary, our data suggest that the addition of CysC to the equations for estimating GFR improved the equations' performance. The performances of two CysC-based equations (CKD-EPIcys and Fengcys equations) were superior to that of the CKD-EPIcr equation. The two combined equations CKD-EPIcr-cys and Fengcr-cys were better than those based on either marker alone. The CKD-EPIcr-cys equation performed the best among all the equations for Chinese adult CKD patients; however, its accuracy still does not fulfil the requirements of the K/DOQI criteria. Further developments are needed to improve the performance of these equations.
E
422 423
7
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 Q2 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
F
O
a
8
a r t i c l e
9 10 11 12 13
Article history: Received 8 February 2014 Received in revised form 20 May 2014 Accepted 21 May 2014 Available online xxxx
14 15 16 17 18 19
Keywords: Glomerular filtration rate Cystatin C Creatinine Chronic kidney disease Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)
Department of Laboratory Medicine, Peking Union Medical College Hospital, Chinese Academic Medical Science and Peking Union Medical College, Beijing, P. R. China Department of Nephrology, Peking Union Medical College Hospital, Chinese Academic Medical Science and Peking Union Medical College, Beijing, P. R. China
i n f o
a b s t r a c t
Objectives: The newly developed glomerular filtration rate (GFR)-estimating equations developed by the CKD-EPI Collaboration and Feng et al. (2013) that are based on standardized serum cystatin C (ScysC), combined/not combined with serum creatinine (Scr), require further validation in China. We compared the performance of four new equations (CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations) with the CKD-EPI creatinine equation (CKD-EPIcr) in adult Chinese chronic kidney disease (CKD) patients to clarify their clinical application. Design and Methods: GFR was measured using the dual plasma sampling 99mTc-DTPA method (mGFR) in 252 adult CKD patients enrolled from four centres. Scr and ScysC were measured by standardized assays in a central laboratory. Each equation's performance was assessed using bias, precision, accuracy, agreement, and correct classification of the CKD stage. Results: The measured GFR was 46 [25–83] mL/min per 1.73 m2. The CKD-EPIcys, CKD-EPIcr-cys and Fengcys equations provided significantly higher accuracy (P15: 38.9%, 39.7%, and 38.9%) than the CKD-EPIcr equation (29.8%). The CKD-EPIcr-cys and Fengcr-cys equations presented higher precision (IQR of the difference, 16.4 and 17.3 mL/min per 1.73 m2, respectively) and narrower acceptable limits in Bland–Altman analysis (56.6 and 50.8 mL/min per 1.73 m2, respectively) than single marker-based equations. The CKD-EPIcr-cys equation achieved the highest overall correct proportion (61.5%) in classification of CKD stages. Conclusions: Combining ScysC and Scr measurements for GFR estimation improves diagnostic performance. The Scr–ScysC equation showed better performance than equations based on either marker alone. The CKD-EPIcr-cys equation showed the best performance for GFR estimation in Chinese adult CKD patients. © 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
21 22 23 24 25 26 27 28 29 30 31 20 32
in China revealed that overall prevalence of CKD in China is 10.8%, and the number of CKD patients has reached approximately 119.5 million [3]. Early recognition and diagnosis of CKD is crucial to its timely treatment that can delay its progression and prevent CKD-related cardiovascular and metabolic disorders. Glomerular filtration rate (GFR), the best overall index to reflect kidney function, is central to the diagnosis and classification of CKD [4]. Currently, the ‘gold standard’ for GFR determination is to measure the clearance of exogenous substances, such as inulin, iohexol, 51Cr-EDTA, 99mTc-DTPA, and 125I-iothalamate. These techniques are time-consuming, labour-intensive, expensive, and require the administration of substances that make them incompatible with routine monitoring in clinical practice [5]. To measure GFR conveniently, certain equations based on serum creatinine (Scr) and demographic characteristics have been developed. The Cockcroft–Gault (CG) equation [6] and the abbreviated Modification of Diet in Renal Disease (MDRD) equation [7] were recommended for use by the Kidney Disease Outcomes Quality Initiative (K/DOQI)
49
R
E
b
R O
6 7
P
5
Xiuzhi Guo a, Yan Qin b, Ke Zheng b, Mengchun Gong b, Jie Wu a, Weiling Shou a, Xinqi Cheng a, Liangyu Xia a, Ermu Xu a, Xuemei Li b,⁎, Ling Qiu a,⁎⁎
D
4Q1
E
3
T
2
Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese adult chronic kidney disease patients
C
1
44 42 41 43
N C O
R
39 40
Introduction
46
Chronic kidney disease (CKD) has become a major health problem worldwide. The prevalence of CKD in USA is reported to be 13% [1]; in Europe, it ranges from 5% to 35% [2]. The latest epidemiological survey
U
45
47 48
33 34 35 36 37 38
Abbreviations: eGFR, estimated glomerular filtration rate; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; CKD, chronic kidney disease; MDRD, Modification of Diet in Renal Disease; Scr, serum creatinine; ScysC, serum cystatin C; CysC, cystatin C; DTPA, diethylene triamine pentacetate acid; K/DOQI, Kidney Disease Outcomes Quality Initiative; PETIA, particle-enhanced turbidimetric immunoassay; PENIA, particleenhanced nephelometric immunoassay. ⁎ Correspondence to: X. Li, Department of Nephrology, Peking Union Medical College Hospital, Chinese Academic Medical Science and Peking Union Medical College, Beijing 100730, P.R. China. ⁎⁎ Correspondence to: L. Qiu, Department of Laboratory Medicine, Peking Union Medical College Hospital, Chinese Academic Medical Science and Peking Union Medical College, Beijing 100730, P.R. China. Fax: +86 10 69159743. E-mail addresses: [email protected] (X. Li), [email protected] (L. Qiu).
http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060 0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
125
Overall, 252 adult CKD patients (aged 18–90 years) were enrolled in four major medical centres (North China, Beijing; East China, Shanghai; Central China, Changsha; and Northeast China, Dalian) from September 2007 to December 2010. The study was performed in accordance with
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
126 127 128
C
89 90
E
87 88
R
85 86
R
83 84
O
81 82
C
79 80
N
77 78
U
75 76
Measurement of GFR
141
F
Participant selection
73 74
The 99mTc-DTPA plasma clearance rate was used to measure GFR (mGFR) using the two-sample method [24]. The staff from the four hospitals participating underwent training for these procedures before the initiation of the study. 99mTc-DTPA 296 MBq was injected into the elbow median cubital vein via an intravenous bolus injection. Blood was collected, and radioactivity measurements (P1 and P2) were performed at 2 h (T1) and 4 h (T2), respectively. GFR was corrected for the standard body surface area by multiplying the measured value by 1.73 and dividing it by the patient's body surface area, derived from the Du Bois formula [25]. GFR was calculated as follows:
O
124
72
R O
Materials and methods
70 71
129 130
P
123
68 69
the declaration of the Ethics Review Board for Human Studies of Peking Union Medical College Hospital. Written informed consent was provided by all the participants after education with regard to the potential benefits, risks, and study procedures. The diagnosis of CKD was according to the criteria provided by K/DOQI guidelines [4]. Patients were excluded from the study if any of the following conditions were present: (a) acute kidney injury; (b) receiving haemodialysis or peritoneal dialysis; (c) general oedema, pleural effusion, ascites, or severe heart failure; (d) severe malnutrition, absence of limbs, or ketoacidosis; (e) receiving cimetidine or trimethoprim; (f) received glucocorticoid therapy in the previous 3 months; (g) hyperthyroidism or hypothyroidism; or (h) leukaemia or cancer.
131 132 133 134 135 136 137 138 139 140
142 143 144 145 146 147 148 149 150 151
GFR ¼ fD ln ðP1 =P2 Þ=ðT2 −T1 Þg expf½ðT1 lnP2 Þ−ðT2 lnP1 Þ=ðT2 −T1 Þg 1:73=BSA 2 0:425 0:725 height BSA m ¼ 0:007184 body weight
wherein D is the radioactive count for the injected drugs, T1 and T2 are the first and second blood collection times from the contralateral arm following the intravenous bolus injection of 99mTc-DTPA, respectively, and while P1 and P2 are the radioactive counts in blood plasma at T1 and T2, respectively. The units of weight and height were kg and cm, respectively. The Brochner-Mortensen method [26] was used for correcting for the systematic error of the slope–intercept technique. The corrected GFR was calculated as follows:
T
122
guidelines in 2002. However, subsequent validation studies demonstrated that neither equation provided satisfactory results in various patient populations [8,9]. To minimise such limitations, including imprecision and systematic underestimation of measured GFR with the MDRD equation, a Chronic Kidney Disease Epidemiology Collaboration creatinine equation (CKD-EPIcr equation) was developed based on standardized Scr [10]. Accumulating evidences demonstrated that the CKD-EPIcr equation performed better than the CG and MDRD equations and could be applicable in clinical nephrology [11,12]. In 2012, the K/DOQI guidelines recommended the use of the CKD-EPIcr equation to report estimated GFR (eGFR) in adults determined using serum creatinine levels, as measured by an assay calibrated to the isotope dilution mass spectrometry reference method [13]. Recent studies have suggested that the CKD-EPIcr equation may be the most appropriate creatinine-based equation for determining GFR in Chinese CKD patients [14–17]. However, the accuracy of creatinine-based equations is not satisfactory because the Scr concentration is easily affected by factors other than GFR [18]. Under such circumstances, the use of cystatin C (CysC) as an alternative marker has received great attention. CysC is a cysteine proteinase inhibitor with a molecular weight of 13 kDa that is produced by all nucleated cells at a constant rate and is considered to be close to the ‘ideal’ endogenous marker: it is freely filtered by the glomerulus and catabolized in the proximal tubular epithelial cells without being secreted [19]. Unlike Scr, CysC is not be easily affected by gender or muscle mass; its concentration in serum/plasma depends on the GFR [19]. As it has been reported to be possibly superior to Scr in GFR estimation, several equations based on serum cystatin C (ScysC) have been proposed between 2000 and 2010 [20]. However, the measurement of CysC varies across centres owing to the lack of international standardization, and thus, these equations were not widely used. In the fall of 2010, certified CysC reference material (ERM-DA471/IFCC) for calibrating laboratory assays was released [21], resulting in the measurement of CysC in different laboratories becoming comparable and traceable. This provides a foundation for the evaluation of equations in various populations. In 2012, Inker et al. reported two equations for estimated GFR [22], one based on standardized ScysC values (CKD-EPIcys equation) and the other based on standardized ScysC combined with standardized Scr values (CKD-EPIcr-cys equation). As reported, the CKD-EPIcr-cys equation was accurate compared to the equations based on either marker alone [22]. However, participants in the development study were mostly of western origin; therefore, it is crucial to validate the performance of the equations in ethnically diverse groups. Almost simultaneously, two GFR equations based on standardized ScysC (Fengcys equation) and combined with Scr (Fengcr-cys equation) were also developed using a population of 788 Chinese CKD patients [23]. Under these circumstances, several issues need to be resolved in China, as follows: (i) Are equations based on standardized ScysC better than those based on Scr? (ii) Are Scr–ScysC combined equations better than equations based on either marker alone? (iii) Among CKD-EPIcr, CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations, which is the optimal equation for the Chinese adult CKD population? In this study, we assessed the performance of the four newly developed equations in Chinese adult CKD populations and compared their performance against the CKD-EPIcr equation, which is presently considered as the best choice for estimating GFR in China [14–17].
D
66 67
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx
E
2
mGFR ¼ 0:990778 GFR−0:001218 GFR
153 154 155 156 157 158 159 160
2
162
Estimation of GFR Estimated GFR (eGFR) values were calculated separately using the CKD-EPIcr, CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations, and their corresponding results were labelled as eGFRCKD-EPIcr, eGFRCKD-EPIcys, eGFRCKD-EPIcr-cys, eGFRFengcys, and eGFRFengcr-cys, respectively, as presented in detail in Table 1.
163
Analytical methods
168
On the day of 99mTc-DTPA GFR measurement, a fasting blood sample was collected, centrifuged, and stored at −80 °C. General information of the patients, including age, gender, height, and weight, were documented. All the serum samples were transferred in the frozen state to the central laboratory at the Department of Laboratory Medicine, Peking Union Medical College Hospital, stored and tested via a standard procedure. Scr was measured by an isotope-dilution mass spectrometrytraceable enzymatic method (Roche-Hitachi P-Module instrument with Roche Creatininase Plus assay; Hoffman-La Roche, Basel, Switzerland). Accuracy of the creatinine assay was assessed using NIST SRM 967 I and II (National Institute of Standards and Technology Standard Reference Material). Bias for the creatinine assay with respect to NIST SRM 967 I and II was −0.75% and 1.96%, respectively. During the study period, the coefficients of variation were 0.7% and 0.6% at creatinine concentrations
169
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
164 165 166 167
170 171 172 173 174 175 176 177 178 179 180 181 182
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx t1:1 t1:2
Table 1 Equations for determining the estimated glomerular filtration rate.
t1:3
Name
Year
Equation
t1:4 t1:5 t1:6 t1:7 t1:8
CKD-EPIcr CKD-EPIcys CKD-EPIcr-cys Fengcys Fengcr-cys
2009 2012 2012 2013 2013
eGFR eGFR eGFR eGFR eGFR
= = = = =
141 × min(Scr/κ,1)α × max(Scr/κ,1)−1.209 × 0.993age × 1.018 (if female) × 1.159 (if black) 133 × min(ScysC/0.8, 1)−0.499 × max (ScysC/0.8, 1)−1.328 × 0.996age × 0.932 (if female) 135 × min(Scr/κ,1)α × max(Scr/κ,1)−0.601 × min(ScysC/0.8, 1)−0.375× max(ScysC/0.8, 1)−0.711 × 0.995age × 0.969 (if female) × 1.08 (if black) 78.64 × ScysC-0.964 173.9 × ScysC-0.725 × Scr-0.184 × age-0.193 × 0.89 (if female)
Abbreviations and definitions: eGFR, estimated GFR; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; CKD-EPIcr, CKD-EPI creatinine equation; CKD-EPIcys, CKD-EPI cystatin C equation [22]; CKD-EPIcr-cys, CKD-EPI creatinine-cystatin C equation [22]; Fengcys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC; Fengcr-cys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC in combination with Scr; Scr, serum creatinine; ScysC, serum cystatin C; Note: eGFR is presented in mL/min per 1.73 m2; Scr is shown as mg/L, ScysC is shown as mg/L, age is in years; In the CKD-EPIcr equation, κ is 0.7 for female and 0.9 for male, α is –0.329 for female and –0.411 for male, min indicates a minimum of Scr/κ or 1, and max indicates a maximum of Scr/κ or 1; In the CKD-EPIcys equation, min indicates the minimum of ScysC/0.8 or 1, and max indicates a maximum of ScysC/0.8 or 1; In the CKD-EPIcr-cys equation, min indicates a minimum of Scr/κ or 1, and ScysC/0.8 or 1, and max indicates a maximum of Cr/κ or 1 and ScysC/0.8 or 1.
Statistical analysis
191 192
217
SPSS19.0 (SPSS Inc., USA) and MedCalc (Medcalc software, Belgium) were used for data analysis. Equation performance was assessed by bias, precision, accuracy, and agreement, as recommended by the National Kidney Foundation guidelines [4]. Bias was expressed as the median of the difference between eGFR and mGFR, and precision was expressed as the interquartile range (IQR) for the difference. Accuracy was expressed as the percentage of individuals with eGFR within 15% (P15), 30% (P30), or 50% (P50) of mGFR. Differences between the eGFR values as calculated with the equations and mGFR were compared by Wilcoxon's signed-rank test and analysed by Passing-Bablok regression analysis. Spearman correlation was used to describe the relationship between eGFRs and mGFR. The P15, P30, and P50 values for the four equations were compared with those for the CKD-EPIcr equation by McNemar's test. The Bland–Altman statistical method was used to assess the agreement between mGFR and eGFR. The area between the regression line of difference and the zero difference line shows the degree of deviation of the eGFR equation from the mGFR [27]. Acceptable tolerance for the difference between mGFR and eGFR was defined as 60 mL/min per 1.73 m2. Kappa statistics were used to evaluate the agreement between CKD stage classification based on mGFR and those by eGFRs calculated using the different equations. The kappa value was interpreted as follows: poor (b 0.20), fair (0.21–0.40), moderate (0.41–0.60), good (0.61–0.80), and very good agreement (0.81–1.0). mGFR was categorized according to K/DOQI [4] (i.e. N90, CKD 1 stage; 60–89, CKD 2 stage; 30–59, CKD 3 stage; 15–29, CKD 4 stage; and b15 mL/min per 1.73 m2, CKD 5 stage). P b 0.05 was considered statistically significant.
218
Results
219
Participant characteristics
220
A total of 252 Chinese CKD participants were enrolled in the study, with 138 male patients and a median age of 43 [IQR, 33–59] years. The clinical characteristics of the participants are shown in Table 2. The concentrations of Scr and ScysC in this study were 1.3 [IQR, 0.9–3.3] mg/dL and 1.6 [IQR, 0.9–3.0] mg/L, respectively, and the mGFR was 46 [IQR, 25–83] mL/min per 1.73 m2.
203 204 205 206 207 208 209 210 211 212 213 214 215 216
221 222 223 224 225
C
E
R
201 202
R
199 200
N C O
197 198
U
195 196
T
190
193 194
The overall comparison between the calculated eGFRs and mGFR is shown in Table 3. All of the eGFRs correlated well with mGFR, and the CKD-EPIcr-cys equation showed the highest Spearman correlation coefficient (0.941), followed by the Fengcr-cys equation (0.939) (Table 3). Further, Passing-Bablok analysis revealed that eGFR values of CKDEPIcr, CKD-EPIcr-cys and Fengcys equations had no marked linear deviation from the mGFR (p N 0.05, Fig. 1). The Fengcr-cys equation did not illustrate a significant proportional difference [the 95% CI of slopes (b) include 1, Fig. 1E]; however, it revealed a large positive constant difference [intercept (a) = 7.301]. The CKD-EPIcys equation showed no significant constant error [the 95% CI of intercepts (a) include 0, Fig. 1-B] and a slight proportional difference. Significant differences in the proportional errors and the constant errors were noted for all the other three equations. Further comparisons of total bias, precision, and accuracy in subgroups of mGFR less than or greater than 60 mL/min per 1.73 m2 are shown in Table 4. The CKD-EPIcr-cys and Fengcr-cys equations provided better precision (IQR of the difference, 16.4 and 17.3 mL/min per 1.73 m2, respectively) as compared with the single marker-based equations. The medians of the difference of the CKD-EPIcys equation and CKDEPIcr-cys equations (-0.2 and 0.5 mL/min per 1.73 m2, respectively) were less than those of the other equations. The CKD-EPIcys, CKD-EPIcr-cys, and Fengcys equations provided significantly higher accuracy than the
R O
188 189
242
P
186 187
Assessment of performance
of 1.0 mg/dL and 6.2 mg/dL, respectively. ScysC was measured by a standardized particle-enhanced immunonephelometric assay on the Siemens Dade Behring Nephelometer with N Latex CysC reagent in 2011, which was traceable to ERM-DA471/IFCC, the international certified reference material for CysC. The coefficients of variation were 2.7% and 2.5% at CysC concentrations of 0.9 mg/L and 1.8 mg/L, respectively, during the study period.
D
184 185
E
183
O
F
t1:9 t1:10 t1:11 t1:12 t1:13 t1:14 t1:15
3
243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 t2:1 t2:2
Table 2 Baseline participant characteristics. Characteristics
Values (n = 252)
t2:3
Age (years, median [IQR]) Males, n (%) Weight (kg, median [IQR]) Height (m, median [IQR]) Serum creatinine (mg/dL, median [IQR]) Serum cystatin C (mg/L, median [IQR]) Measured GFR (mL/min per 1.73 m2, median [IQR]) Causes of CKD, n (%) Primary or secondary renal glomerular disease Hypertensive nephropathy Diabetic nephropathy Chronic tubulointerstitial nephropathy Obstructive nephropathy Cystic nephropathy Other or unclear CKD staging results, n (%) ≥90 60–89 30–59 15–29 b15
43 [33, 59] 138 (54.8) 65.0 [56.5, 75.0] 1.66 [1.60, 173] 1.3 [0.9, 3.3] 1.6 [0.9, 3.0] 46 [25, 83]
t2:4 t2:5 t2:6 t2:7 t2:8 t2:9 t2:10 t2:11 t2:12 t2:13 t2:14 t2:15 t2:16 t2:17 t2:18 t2:19 t2:20 t2:21 t2:22 t2:23 t2:24
166 (65.9) 26 (10.3) 40 (15.9) 10 (4.0) 2 (0.8) 2 (0.8) 6 (2.4) 51 (20.2) 50 (19.8) 69 (27.4) 45 (17.6) 37 (14.7)
Note: Data are shown as median [IQR] or n (%). IQR, interquartile range; GFR, glomerular filtration rate; CKD, chronic kidney disease. Serum creatinine in mg/dL to μmol/L, ×88.4. CKD staging results based on mGFR.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
t2:25 t2:26 t2:27
4 t3:1 t3:2
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx
Table 3 Overall comparison between eGFR and mGFR: Correlation, Bland–Altman analyses (n = 252).
t3:3
mean ± SD
correlation
t3:4 t3:5 t3:6 t3:7 t3:8 t3:9 t3:10
mGFR eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
54 60 55 57 58 61
± ± ± ± ± ±
35 43 39 41 32 34
Bland–Altman analysis
r
p-value
mean differences
area⁎ (arbitrary unit)
95% AL
0.935 0.924 0.941 0.913 0.939
b0.001 b0.001 b0.001 b0.001 b0.001
5.7 0.8 2.4 3.1 6.2
1178 402 761 470 726
-27.1–38.4 -29.1–30.6 -25.9–30.7 -26.8–33.1 -19.2–31.6
t3:11 t3:12
Note: Units are mL/min per 1.73 m2 for mean ± SD, Mean differences, 95% AL; Spearman correlation of eGFRs against mGFR were made; r is Spearman's correlation coefficient between eGFRs and mGFR; ⁎the area between the Bland–Altman regression line and the zero difference line; arbitrary unit, i.e., (mL/min per 1.73 m2)2; 95% AL denotes 95% agreement limits.
266 267
CKD-EPIcr equation (P15: 38.9%, 39.7%, and 38.9% vs 29.8%, McNemar's test, p b 0.05). The accuracy of the CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations showed no statistically significant differences as compared with each other (P30: 64.3%, 62.3%, 61.9%, and 62.7%, respectively, McNemar's test, p N 0.05). Table 5 shows the percentages of patients that were classified in the correct CKD stages by the eGFRs from the different equations. Kappa statistics were used to evaluate the agreement between classification of CKD stages by mGFR and that by eGFRs calculated by different equations. Higher kappa values indicated better agreement in CKD staging as calculated by mGFR. The CKD-EPIcr-cys equation achieved the best agreement with a kappa value of 0.520 and an overall correct classification proportion of 61.5%. In CKD stages 2–4, the proportions of correct classification by the Fengcys and Fengcr-cys equations were significantly higher than those by the other equations; however, in CKD stages 1 and 5, these proportions were lower, particularly in CKD stage 5 (Table 5). The Bland–Altman plots were used to evaluate the bias (mean of difference) and the degree of agreement between eGFRs and mGFR (Fig. 2). Both the slopes of the regression of differences against averages by the CKD-EPIcys and Fengcr-cys equations were nearly equal to zero, which indicated no proportional bias. However, the CKD-EPIcys equation demonstrated a smaller mean of difference, while the Fengcr-cys equation showed a larger constant positive difference than the other equations (mean difference, 0.8 and 6.2 mL/min per 1.73 m2, respectively).
286 287 288 289 290
F
O
R O
P
E
D
Although the implementation of eGFR has resulted in the identification of large numbers of patients with undiagnosed CKD, an optimal equation for estimating the GFR in different populations has not been validated yet. The standardization of Cr measurement has greatly promoted the development and application of Scr-based eGFR equations [28]. After the standardization of ScysC measurement, equations based on standardized ScysC have attracted attention and have been validated in different populations. Masson et al. validated the CKDEPIcys and CKD-EPIcr-cys equations in kidney transplant recipients and found that both of these were better than the CKD-EPIcr equation [29]. Horio et al. studied the accuracy of these equations in the Japanese population and suggested that the CKD-EPIcys equation could be used in patients of different races without modification [30]. In China, Liu et al. assessed the CKD-EPIcr-cys equation in a cohort of elderly
T
284 285
C
282 283
E
280 281
299
R
278 279
Discussion
R
276 277
O
274 275
C
272 273
N
270 271
291 292
U
268 269
In the estimation of GFR consistency, two combined equations (CKD-EPIcr-cys and Fengcr-cys) were both better than the CKD-EPIcr, CKD-EPI cys, and Fengcys equations alone (limits of agreement = 50.8 and 56.6 vs 65.5, 59.7, and 59.9 mL/min per 1.73 m2, respectively). When the area between the regression line of difference and the zero difference line was compared, the CKD-EPIcys equation gave the smallest deviation (Table 1, 402 arbitrary units), followed by the Fengcys equation (Table 1, 470 arbitrary units).
Fig. 1. Passing-Bablok regression analysis comparing the differences between eGFR by (A) CKD-EPIcr, (B) CKD-EPIcys, (C) CKD-EPIcr-cys, (D) Fengcys, and (E) Fengcr-cys and mGFR. Note: All data are presented in mL/min per 1.73 m2; a, 95% confidence interval for the intercept; b, 95% confidence interval for the slope; solid lines, regression line; dashed lines, 95% confidence intervals of the regression line; dotted lines, identity line (x = y). Abbreviations and definitions: CKD-EPIcr, CKD-EPI creatinine equation [10]; CKD-EPIcys, CKD-EPI cystatin C equation [22]; CKD-EPIcr-cys, CKD-EPI creatinine-cystatin C equation [22]; Fengcys equation, newly developed equation by Feng et al.(2013) [23] based on standardized ScysC; Fengcr-cys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC in combination with Scr.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
293 294 295 296 297 298
300 301 302 303 304 305 306 307 308 309 310 311 312 313
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx Table 4 Comparison of median bias, precision, and accuracy of equations in subgroups when mGFR b 60 mL/min per 1.73 m2 or mGFR ≥ 60 mL/min per 1.73 m2. mGFR ≥ 60 mL/min per 1.73 m2 (n = 101)
Bias: Median difference, i.e. eGFR – mGFRa eGFRCKD-EPIcr 2.4; p b 0.001 eGFRCKD-EPIcys -0.2; p = 0.9 eGFRCKD-EPIcr-cys 0.5; p = 0.2 eGFRFengcys 3.3; p b 0.001 eGFRFengcr-cys 5.6; p b 0.001
-1.7; p = 0.7 -0.5; p b 0.001 -2.7; p b 0.001 8.2; p b 0.001 7.9; p b 0.001
14.6; p b 0.001 1.2; p b 0.001 4.9; p b 0.001 -3.9; p = 0.1 1.6; p b 0.001
Precision: IQR of the difference eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
22.1 18.4 16.4 19.0 17.3
15.2 14.6 13.8 15.7 13.1
22.3 25.6 21.0 19.4 22.0
Accuracy: P15b (%) eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
29.8 38.9; p 39.7; p 38.9; p 36.1; p
= = = =
0.03 0.02 0.03 0.1
22.5 27.8; p 27.2; p 23.2; p 20.5; p
= = = =
0.3 0.4 0.9 0.5
Accuracy: P30b (%) eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
59.1 64.3; p 62.3; p 61.9; p 62.7; p
= = = =
0.2 0.5 0.5 0.4
47.7 51.7; p 49.7; p 45.7; p 45.0; p
= = = =
0.5 0.7 0.7 0.6
Accuracy: P50b (%) eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
79.0 83.3; p 83.7; p 79.0; p 86.6; p
= = = =
0.2 0.2 1.0 0.7
69.5 73.5; p 74.8; p 64.9; p 67.5; p
= = = =
0.4 0.3 0.4 0.7
40.6 55.4; p 58.4; p 62.4; p 59.4; p
= = = =
0.04 0.01 0.002 0.008
76.2 83.2; p 81.2; p 86.1; p 89.1; p
= = = =
0.2 0.4 0.07 0.01
O R O P
t4:4 t4:5 t4:6 t4:7 t4:8 t4:9 t4:10 t4:11 t4:12 t4:13 t4:14 t4:15 t4:16 t4:17 t4:18 t4:19 t4:20 t4:21 t4:22 t4:23 t4:24 t4:25 t4:26 t4:27 t4:28 t4:29 t4:30 t4:31 t4:32 t4:33 t4:34 t4:35 t4:36 t4:37
Overall (n = 252)
93.1 98.0; p = 0.2 97.0; p = 0.2 100; p = 0.02 100; p = 0.02
D
Variable
F
mGFR b 60 mL/min per 1.73 m2 (n = 151)
t4:3
E
t4:1 t4:2
5
Note: mGFR is presented in mL/min per 1.73 m2. GFR, glomerular filtration rate; eGFR, estimated GFR; IQR, interquartile range; mGFR, measured GFR; P30 (P15, P50), percentage of estimates within 30% (15%, 50%) of mGFR. a Wilcoxon's signed-rank test was used to compare the bias of each of the eGFRs against mGFR. b McNemar's test was used to compare P15, P30, and P50 values of the eGFRs against the CKD-EPIcr equation.
314
330
Chinese participants, and found that the equation was more suitable for the elderly Chinese patients than the Scr-based equations [31]. However, Liu's study involved only elderly individuals, and GFR was measured using the renal dynamic imaging method, which provided a non-reliable GFR that cannot be considered a ‘gold standard’ [32]. Feng et al. developed two equations based on standardized ScysC in Chinese CKD patients in 2013 [23]; however, further external validation is necessary for its wide application. In the present study, we observed that the combination of ScysC with Cr for calculating the eGFR truly improved the equations' performance. The four new equations, i.e., the CKD-EPIcys, CKD-EPIcr-cys, Fengcys, and Fengcr-cys equations, all presented higher precision and accuracy than the CKD-EPIcr equation. These results were similar to findings in western populations, as reported by the CKD-EPI Collaboration [22]. Reclassification of CKD patients using the more accurate GFR-estimating tools, especially those based on ScysC, has recently become an important research focus. In this study, it appeared that stage misclassification was
t5:1 t5:2
Table 5 Correct classification proportion of CKD using eGFRs and mGFR [n (%)].
320 321 322 323 324 325 326 327 328 329
C
E
R
R
318 319
N C O
317
U
315 316
T
t4:38 t4:39 t4:40 t4:41
t5:3 t5:4 t5:5 t5:6 t5:7 t5:8 t5:9 t5:10 t5:11 t5:12
mGFR⁎ eGFRCKD-EPIcr eGFRCKD-EPIcys eGFRCKD-EPIcr-cys eGFRFengcys eGFRFengcr-cys
only reduced by the CKD-EPIcr-cys equation compared to the CKD-EPIcr equation (Table 5). Further analysis of the data revealed that this was because the CKD-EPIcr equation underestimated GFR in individuals with a low GFR and significantly overestimated GFR in individuals with a high GFR (Fig. 2); therefore, among the five equations, the CKD-EPIcr equation gave the highest proportions of correct CKD stage 1 and 5 classifications. For CKD stages 2–4, the Fengcys and Fengcr-cys equations provided significantly higher proportions of correct classifications than the other equations; however, these two equations overestimated GFR in patients with a low mGFR (Fig. 2), resulting in a large number of patients with CKD stage 5 misclassified as having CKD stage 4. Overall, the performance of the two ScysC-based equations (CKD-EPIcys and Fengcys) was superior to that of the CKD-EPIcr equation, and two combined equations (CKD-EPIcr-cys and Fengcr-cys equations) were better than the single marker–based equations. The performances of the Fengcys and Fengcr-cys equations was also compared with those of the CKD-EPIcys and CKD-EPIcr-cys equations,
CKD 1 stage
CKD 2 stage
CKD 3 stage
CKD 4 stage
CKD 5 stage
Total
Kappa
p-Value
51 46 (90.2)★ 40(78.4) 45 (88.0) 31 (61.8) 41 (80.1)
50 20 (40.0) 23 (46.0) 28 (56.0) 35 (70.0) 32 (64.0)
69 34 (49.3) 34 (49.3) 33 (47.8) 39 (56.5) 41 (59.4)
45 20 (44.4) 21 (46.7) 18 (40.0) 28 (62.2) 24 (53.3)
37 31 (83.8) 30 (81.1) 31 (83.8) 6 (16.2) 6 (16.2)
252 151 (59.9) 148 (58.7) 155 (61.5) 139 (55.2) 144 (57.1)
0.500 0.484 0.520 0.430 0.455
b0.001 b0.001 b0.001 b0.001 b0.001
Kappa analysis was used to evaluate the agreement between the stages classified according to mGFR and each eGFR-classified CKD stage. ⁎ CKD staging results based on mGFR according to the recommendations by K/DOQI. ★ Number of patients (%) correct classified.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx
R O
O
F
6
359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384
E
Fengcys and Fengcr-cys equations in this study may be due to these factors. The K/DOQI criteria require the acceptable level of 30% accuracy (P30) to be at least 70% [4]. However, none of the five equations in the study accurately fulfilled this criterion, with P30 ranging from 59.1% to 64.3% and a much lower accuracy for the subgroup of mGFR values b 60 mL/min per 1.73 m2 (P30, 45.0–51.7%). Low accuracy has also been reported in Chinese populations [14–17,31]. Possible reasons for these discrepancies are as follows. First, the applicability of equations may be not general enough, at least at present. Differences in race, physical fitness, diet, and living habits can influence the level of SCr [18]. Thus, the accuracy of equations in different populations may have varied. Further studies, for example, with considering a racial coefficient, may improve the accuracy of the CKD-EPIcr-cys equation. Second, differences in the method of GFR measurement may be a contributory factor. Unlike renal clearance of 125I-iothalamate, used when developing the CKD-EPI equations, Chinese studies generally employ 99mTc-DTPA dual plasma clearance as reference method for GFR [14,15,23]. Systematic bias cannot be avoided when two distinct methods are used to measure GFR [35]. Third, the clearance of dual plasma sampling 99mTc-DTPA method was used in our study. The first and second samples were obtained at 2 h and 4 h, respectively, for all patients. The use of a narrow time interval (2 h and 4 h) for patients with renal failure may reduce the accuracy of the plasma clearance of 99mTc-DTPA. It is worth mentioning that although ScysC was less affected by diet, race, and muscle mass than Scr, factors such as diabetes [36], thyroid function [37], immunosuppressant therapy [38], and smoking [39] may influence ScysC levels. Unfortunately, subjects with some of these factors were not completely excluded from the study (see exclusion criteria). The respective weightage of these confounding factors has not been thoroughly investigated to date, and clinicians should be prudent regarding the results in cases where these factors are present. Our study has some limitations. First, 99mTc-DTPA dual plasma clearance was used as reference method, in contrast to renal clearance of 125I-iothalamate that used in the development of the CKD-EPI equations. This would invariably introduce systematic bias. Second, the sampling times were 2 h and 4 h in our study, and the interval
T
C
E
R
358
R
356 357
O
354 355
C
352 353
N
350 351
respectively. The Fengcr-cys and CKD-EPIcr-cys equations provided similar overall precision (IQR of the difference, 17.3 vs 16.4) and accuracy (P30, 64.3% vs 62.7%). In the Bland–Altman analysis, the Fengcr-cys equation presented narrower acceptable limits than the CKD-EPIcr-cys equation (50.8 vs 56.6 mL/min per 1.73 m2), and no significant difference in proportional error against mGFR was noted for the Fengcr-cys equation by Passing-Bablok analysis. However, the Fengcr-cys equation misclassified a large number of patients with CKD stage 5 as having CKD stage 4. Therefore, although the Fengcr-cys equation provided significantly higher proportions of correct CKD stage 2–4 classifications than the CKD-EPIcr-cys equation, its overall correct classification proportion remained lower than that of the CKD-EPIcr-cys equation. A similar situation was also observed for the Fengcys and CKD-EPIcys equations. Why were the performances of the two Chinese-based (Fengcys and Fengcr-cys) equations not better than the CKE-EPI equations, and why were these not as good as they were at the time of development? First, the CKD-EPIcr-cys and CKD-EPIcys equations provide different coefficients for different levels of Cr and CysC; however, Feng's equations were not stratified by the Cr and CysC levels. This may result in a high degree of compliance in the intermediate region, along with large deviations in the high and low regions of GFR. Therefore, further improvements are needed for the two Feng's equations. Second, the method for CysC measurement differed in the two studies. The Fengcys and Fengcr-cys equations were developed on the basis of a particle-enhanced turbidimetric immunoassay (PETIA) method for CysC measurement, whereas CysC was measured by using the particle-enhanced nephelometric immunoassay (PENIA) method in the development of the CKD-EPI equations and in our study. Although both methods are traceable to ERM-DA471/IFCC, differences cannot be avoided because two distinct methods were used; CysC values by the PETIA method have been reported to be 27.5% higher than those by the PENIA method [33]. Third, we used the Brochner-Mortensen correction rather than the Chantler correction used by Feng et al. to calculate the systematic error of the slope–intercept technique, as recommended by a guideline [34]. The Brochner-Mortensen technique is preferred as it addresses the variation in percentage error with GFR, whereas the Chantler technique does not adequately correct for the variation in percentage error with GFR [34]. The poor performance observed for the
U
348 349
D
P
Fig. 2. Bland–Altman plots comparing the agreement between eGFR by (A) CKD-EPIcr, (B) CKD-EPIcys, (C) CKD-EPIcr-cys, (D) Fengcys, and (E) Fengcr-cys and mGFR. Note: Full line, mean difference between two methods; dashed line, ±1.96 SD difference against mean; the regression line of differences versus averages and the 95% confidence interval are also presented. All data are presented in mL/min per 1.73 m2. Abbreviations and definitions: CKD-EPIcr, CKD-EPI creatinine equation [10]; CKD-EPIcys, CKD-EPI cystatin C equation [22]; CKD-EPIcr-cys, CKD-EPI creatinine-cystatin C equation [22]; Fengcys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC; Fengcr-cys equation, newly developed equation by Feng et al. (2013) [23] based on standardized ScysC in combination with Scr.
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
X. Guo et al. / Clinical Biochemistry xxx (2014) xxx–xxx
References
440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 561
[1] Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, et al. Prevalence of chronic kidney disease in the United States. JAMA 2007;298:2038–47. [2] Zhang QL, Rothenbacher D. Prevalence of chronic kidney disease in populationbased studies: Systematic review. BMC Public Health 2008;8:117–30. http:// dx.doi.org/10.1186/1471-2458-8-117. [3] Zhang L, Wang F, Wang L, Wang W, Liu B, Liu J, et al. Prevalence of chronic kidney disease in China: a cross-sectional survey. Lancet 2012;379:815–22. http:// dx.doi.org/10.1016/S0140-6736(12)60033-6. [4] N.K. Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39:S1–S266. [5] Florkowski CM, Chew-Harris JS. Methods of estimating GFR-different equations including CKD-EPI. Clin Biochem Rev 2011;32:75–9. [6] Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31–41. [7] Levey AS, Greene T, Kusek JW, Beck GJ, MDRD Study Group. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 2000;11:155A. [8] Poggio ED, Nef PC, Wang X, Greene T, Van Lente F, Dennis VW, et al. Performance of the Cockcroft-Gault and modification of diet in renal disease equations in estimating GFR in ill hospitalized patients. Am J Kidney Dis 2005;46:242–52. [9] Ibrahim H, Mondress M, Tello A, Fan Y, Koopmeiners J, Thomas W. An alternative formula to the Cockcroft-Gault and the modification of diet in renal diseases formulas in predicting GFR in individuals with type 1 diabetes. J Am Soc Nephrol 2005;16:1051–60. [10] Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, Feldman HI, et al. CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–12. [11] Stevens LA, Schmid CH, Greene T, Zhang YL, Beck GJ, Froissart M, et al. Comparative performance of the CKD Epidemiology Collaboration (CKD-EPI) and the Modification of Diet in Renal Disease (MDRD) study equations for estimating GFR levels above 60 mL/min/1.73 m2. Am J Kidney Dis 2010;56:486–95. http://dx.doi.org/ 10.1053/j.ajkd.2010.03.026. [12] Michels WM, Grootendorst DC, Verduijn M, Elliott EG, Dekker FW, Krediet RT. Performance of the Cockcroft-Gault, MDRD, and new CKD-EPI formulas in relation to GFR, age, and body size. Clin J Am Soc Nephrol 2010;5:1003–9. [13] Stevens PE, Levin Adeera, the Kidney Disease: Improving Global Outcomes Chronic Kidney Disease Guideline Development Work Group Members. Evaluation and management of chronic kidney disease: synopsis of the kidney disease: Improving Global Outcomes 2012 Clinical Practice Guideline. Ann Intern Med 2013;158:825–30. [14] Kong X, Ma Y, Chen J, Luo Q, Yu X, Li Y, et al. Evaluation of the Chronic Kidney Disease Epidemiology Collaboration equation for estimating glomerular filtration rate in the Chinese population. Nephrol Dial Transplant 2013;28:641–51. [15] Xie P, Huang JM, Lin HY, Wu WJ, Pan LP. CDK-EPI equation may be the most proper formula based on creatinine in determining glomerular filtration rate in Chinese patients with chronic kidney disease. Int Urol Nephrol 2013;45:1057–64. http:// dx.doi.org/10.1007/s11255-012-0325-7.
F
439
C
E
R
R
N C O
U
431 432
O
437 438
This study was supported by the Beijing Municipal Science and Technology Commission (D09050704310901) and Sichuan Provincial Department of Science and Technology (2009sz0066). We thank the members and staff of our research group for their cooperative work.
429 430
R O
435 436
428
P
Acknowledgments
426 427
[16] Li JT, Xun C, Cui CL, Wang HF, Wu YT, Yun AH, et al. Relative performance of two equations for estimation of glomerular filtration rate in a Chinese population having chronic kidney disease. Chin Med J (Engl) 2012;125:599–603. [17] Pei XH, He J, Liu Q, Zhu B, Bao LH, Yan CJ, et al. Evaluation of serum creatinine- and CysC-based equations for the estimation of glomerular filtration rate in a Chinese population. Scand J Urol Nephrol 2012;46:223–31. http://dx.doi.org/10.3109/ 00365599.2012.660985. [18] Perrone RD, Madias NE, Levey AS. Serum creatinine as an index of renal function: new insights into old concepts. Clin Chem 1992;38:1933–53. [19] Filler G, Bökenkamp A, Hofmann W, Le Bricon T, Martínez-Brú C, Grubb A. Cystatin C as a marker of GFR-history, indications, and future research. Clin Biochem 2005;38:1–8. [20] Harman G, Akbari A, Hiremath S, White CA, Ramsay T, Kokolo MB, et al. Accuracy of cystatin C-based estimates of glomerular filtration rate in kidney transplant recipients: a systematic review. Nephrol Dial Transplant 2013;28:741–57. http:// dx.doi.org/10.1093/ndt/gfs498. [21] Grubb A, Blirup-Jensen S, Lindström V, Schmidt C, Althaus H, Zegers I. First certified reference material for cystatin C in human serum ERM-DA471/IFCC. Clin Chem Lab Med 2010;48:1619–21. http://dx.doi.org/10.1515/CCLM.2010.318. [22] Inker LA, Schmid CH, Tighiouart H, Eckfeldt JH, Feldman HI, Greene T, et al. Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med 2012;367:20–9. http://dx.doi.org/10.1056/NEJMoa1114248. [23] Feng JF, Qiu L, Zhang L, Li X, Yang Y, Zeng P, et al. Multicenter study of creatinineand/or cystatin C-based equations for estimation of glomerular filtration rates in Chinese patients with chronic kidney disease. PLoS One 2013;8:e57240. http:// dx.doi.org/10.1371/journal.pone.0057240. [24] Fleming JS, Zivanovic MA, Blake GM, Burniston M, Cosgriff PS, British Nuclear Medicine Society. Guidelines for the measurement of glomerular filtration rate using plasma sampling. Nucl Med Commun 2004;25:759–69. [25] Du BD, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. Nutrition 1989;5:303–11. [26] Bröchner-Mortensen J. A simple method for the determination of glomerular filtration rate. Scand J Clin Lab Invest 1972;30:271–4. [27] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10. [28] Myers GL, Miller WG, Coresh J, Fleming J, Greenberg N, Greene T, et al. Recommendations for improving serum creatinine measurement: a report from the Laboratory Working Group of the National Kidney Disease Education Program. Clin Chem 2006;52:5–18. [29] Masson I, Maillard N, Tack I, Thibaudin L, Dubourg L, Delanaye P, et al. GFR estimation using standardized cystatin C in kidney transplant recipients. Am J Kidney Dis 2013;61:279–84. http://dx.doi.org/10.1053/j.ajkd.2012.09.010. [30] Horio M, Imai E, Yasuda Y, Watanabe T, Matsuo S. GFR estimation using standardized serum cystatin C in Japan. Am J Kidney Dis 2013;61:197–203. http://dx.doi.org/ 10.1053/j.ajkd.2012.07.007. [31] Liu X, Ma H, Huang H, Wang C, Tang H, Li M, et al. Is the Chronic Kidney Disease Epidemiology Collaboration creatinine-cystatin C equation useful for glomerular filtration rate estimation in the elderly? Clin Interv Aging 2013;8:1387–91. http:// dx.doi.org/10.2147/CIA.S52774. [32] Xie P, Huang JM, Liu XM, Wu WJ, Pan LP, Lin HY. 99mTc-DTPA renal dynamic imaging method may be unsuitable to be used as the reference method in investigating the validity of CDK-EPI Equation for determining Glomerular Filtration Rate. PLoS One 2013;8:e62328. [33] Voskoboev NV, Larson TS, Rule AD, Lieske JC. Analytic and clinical validation of a standardized cystatin C particle enhanced turbidimetric assay (PETIA) to estimate glomerular filtration rate. Clin Chem Lab Med 2012;50:1591–6. http://dx.doi.org/ 10.1515/cclm-2012-0063. [34] Fleming JS, Zivanovic MA, Blake GM, Burniston M, Cosgriff PS, British Nuclear Medicine Society. Guidelines for the measurement of glomerular filtration rate using plasma sampling. Nucl Med Commun 2004;25:759–69. [35] Dai SS, Yasuda Y, Zhang CL, Horio M, Zuo L, Wang HY. Evaluation of GFR measurement method as an explanation for differences among GFR estimation equations. Am J Kidney Dis 2011;58:496–8. http://dx.doi.org/10.1053/j.ajkd.2011.05.016. [36] Stevens LA, Schmid CH, Greene T, Li L, Beck GJ, Joffe MM, et al. Factors other than glomerular filtration rate affect serum cystatin C levels. Kidney Int 2009;75:652–60. http://dx.doi.org/10.1038/ki.2008.638. [37] Kotajima N, Yanagawa Y, Aoki T, Tsunekawa K, Morimura T, Ogiwara T, et al. Influence of thyroid hormones and transforming growth factor-β1 on cystatin C concentrations. J Int Med Res 2010;38:1365–73. [38] Rule AD, Bergstralh EJ, Slezak JM, Bergert J, Larson TS. Glomerular filtration rate estimated by cystatin C among different clinical presentations. Kidney Int 2006;69:399–405. [39] Sjöström PA, Jones IL, Tidman MA. Cystatin C as a filtration marker-haemodialysis patients expose its strengths and limitations. Scand J Clin Lab Invest 2009;69:65–72. http://dx.doi.org/10.1080/00365510802326469.
D
434
424 425
T
433
may be too narrow for patients with renal failure, thus reducing accuracy. In summary, our data suggest that the addition of CysC to the equations for estimating GFR improved the equations' performance. The performances of two CysC-based equations (CKD-EPIcys and Fengcys equations) were superior to that of the CKD-EPIcr equation. The two combined equations CKD-EPIcr-cys and Fengcr-cys were better than those based on either marker alone. The CKD-EPIcr-cys equation performed the best among all the equations for Chinese adult CKD patients; however, its accuracy still does not fulfil the requirements of the K/DOQI criteria. Further developments are needed to improve the performance of these equations.
E
422 423
7
Please cite this article as: Guo X, et al, Improved glomerular filtration rate estimation using New equations combined with standardized cystatin C and creatinine in Chinese..., Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.060
486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 Q2 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560
