Journal of Pediatric Urology (2014) 10, 262e267
Assessment of cystatin C and cystatin C-based GFR formulas in reflux nephropathy Sevgi Yavuz*, Ali Anarat, Aysun K. Bayazıt Division of Pediatric Nephrology, Cukurova University School of Medicine, 01330 Adana, Turkey Received 29 April 2013; accepted 22 August 2013 Available online 15 September 2013
KEYWORDS Cystatin C; Glomerular filtration rate; Reflux nephropathy; Vesicoureteral reflux
Abstract Objective: Early identification of reflux nephropathy (RN) could reduce the frequency of chronic kidney disease (CKD) caused by vesicoureteral reflux (VUR). We aimed to assess whether cystatin C has value for determining RN in children with VUR. Materials and methods: Ninety-three children with VUR were classified into two groups according to the presence of renal parenchymal scarring (RS). Patients with RS were divided into three subgroups according to scar grade. Serum cystatin C, serum creatinine (Scr) and urine creatinine were measured. eGFR values of the patients were calculated with Scr-based, cystatin C-based and combined formulas. Results: Cystatin C was significantly higher in patients with RS than patients without RS and declined in parallel with grade of RS (p Z 0.01). Scr was not significant in patients with and without RS. It was only significant between mild and severe scar subgroups (p < 0.05). All eGFR values were lower in RS (þ) patients compared with RS () patients. All eGFR equations were negatively correlated with grade of RS (p < 0.05). Conclusion: Cystatin C could be a useful marker for identifying the risk and severity of RN in patients with VUR. Renal functions could be more accurately determined with Scr-cystatin C combined eGFR equations. ª 2013 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.
Introduction Reflux nephropathy (RN) is an undesirable complication of vesicoureteral reflux (VUR) and is usually identified as renal parenchymal lesions on dimercaptosuccinic acid (DMSA)
* Corresponding author. Tel.: þ90 322 3386060; fax: þ90 322 3386901. E-mail address: [email protected] (S. Yavuz).
scan. Post-infection injury or congenital renal dysplasia shares the features of RN [1,2]. Whether caused by acute pyelonephritis or dysplasia, some patients with VUR progress to RN and some develop chronic kidney disease (CKD). Male gender, high grade and bilateral VUR, high serum creatinine (Scr) or low estimated glomerular filtration rate (eGFR) at admission, proteinuria and hypertension are more likely to increase the risk of CKD in VUR [3,4]. However, which patients are candidates for RN and/or CKD is still unclear.
1477-5131/$36 ª 2013 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jpurol.2013.08.010
Assessment of cystatin C and GFR formulas
263
Cystatin C, a cystine protease inhibitor, has been proposed as an alternative to Scr and presents in all human body fluids and nucleated cells. Cystatin C fulfills the criteria of the ideal filtration marker as it is produced at a constant rate, freely filtered without tubular secretion and completely catabolized in the proximal tubule. Unlike Scr, cystatin C seems to be independent from body weight and height, age (after 1 year) or nutritional status. Clinical studies have shown that cystatin C is superior to Scr for the detection of mild to moderate chronic kidney dysfunction and acute kidney injury [5,6]. More recently, cystatin Cbased prediction equations were introduced to be as good as Scr-based formulas in children [7,8]. However, few cystatin C studies have been conducted in the VUR and RN area [9e11]. In this prospective study, we aimed to assess whether cystatin C might have value for determining RN. Additionally, Scr, cystatin C and combined Scr-cystatin C-based eGFR equations were compared in refluxing children with and without renal parenchymal scarring (RS).
Assays
Materials and methods
The presence of VUR was determined by VCUG and graded from I to V according to the International Reflux Study Committee system [15]. RS was identified by 99mTc-DMSA performed 3e6 months after UTI. Reduced uptake with loss of normal renal contour on the follow-up DMSA scintigraphy was defined as RS [16]. The kidney was divided into 12 segments to quantify the extent of RS. Renal lesions were classified according to the DMSA grading system of Ziessman and Majd: none (no segments affected), mild (1-2 segments affected), moderate (2-3 segments affected) and severe (4 segments affected) [17].
Study population Ninety-three children with primary VUR were enrolled in our prospective study in Cukurova University Hospital between 2009 and 2011. The exclusion criteria were additional renal malformations or systemic disease, stage III CKD (eGFR < 30 ml/min/1.73 m2), history of a recent urinary tract infection (UTI) within 4 weeks, reflux surgery or spontaneous resolution and incapability of 24-h urine sampling. Blood urea nitrogen (BUN), Scr and cystatin C; 24 h urine for protein (Up) and creatinine (Ucr) were measured in all patients. All children underwent 99mTcdimercaptosuccinic acid (99mTc-DMSA) scanning and voiding cystourethrogram (VCUG) imaging before the onset of the trial. Patients were divided into two groups according to presence of RS. Then, patients with RS were divided into three subgroups as mild, moderate and severe based on the grade of RS.
Table 1
BUN, Scr, Ucr and Up were concurrently measured using turbidimetric assay in a Beckman Coulter AU analyzer (Beckman Coulter, California, United States). Serum cystatin C concentrations were determined by ELISA using the human Cystatin C ELISA kit (BioVendor, Heidelberg, Germany). Intra-assay and inter-assay variations of the test were 2.5 and 6.9%, respectively.
Assessment of eGFR Estimated GFR values of the all group were calculated according to Scr-based formulas [creatinine clearance (Ccr), new Schwartz], cystatin C-based equations (Zappitelli1, Filler) and combined Scr-cystatin C formulas [Zappitelli2, the New Chronic Kidney Disease in children (Ckid) equations] (Table 1) [12e14].
Imaging studies
Statistical analysis Statistical analysis was done using SPSS version 13.0 for Windows. Initially, assumptions of normality and homogeneity of variance were checked. Descriptive results were expressed as mean standard deviation (SD) for variables with a normal distribution. The t test for independent
Formulas used for the estimation of glomerular filtration rate. Formula
With plasma creatinine Creatinine clearance (Ccr) (ml/min/1.73 m2) New Schwartz (ml/min/1.73 m2) [12] With serum cystatine C Zappitelli1 (ml/min/1.73 m2) [13] Filler (ml/min/1.73 m2) [14] Combined Zappitelli2 (ml/min/1.73 m2) [13] New Ckid (ml/min/1.73 m2) [12]
eGFR Z [Ucr (mg/dl)/Scr (mg/dl)] [urine volume (ml)/1440 (min)] [1.73/body surface area (m2)] eGFR Z 0.413 [height (cm)/Scr (mg/dl)] eGFR Z 75.94 [cystatin C (mg/l)1.17] eGFR Z 91.62 [cystatin C (mg/l)1.123] eGFR Z [507.76 e0.003 height(m)]/[cystatin C (mg/l)0.635 sCr (mmol/l)0.547] eGFR Z 39.1 [height (m)/Scr (mg/dl)]0.516 [1.8/cystatin C (mg/l)0.294] [30/BUN (mg/dl)0.169] [1.099]male [height (m)/1.4]0.188
Scr, serum creatinine; Ucr, urine creatinine; eGFR, estimated glomerular filtration rate; Ccr, creatinine clearance; Ckid, chronic kidney diseases in children.
264 Table 2
S. Yavuz et al. Baseline characteristics of study groups.
Characteristics
RS () group (n Z 36)
Age (year) 8.44 3.35 Gender, n (%) Male 14 (38.9) Female 22 (61.1) Weight 30.94 13.8 Height 131.68 21.1 Grade of VUR, n (%) I-II 19 (52.8) III 14 (38.9) IV-V 3 (8.3)
RS (þ) group (n Z 57)
p
9.09 3.46
NS
21 (36.8) 36 (63.2) 31.15 14.1 130.1 16.6
NS
14 (24.5) 34 (59.6) 9 (15.9)
0.02
NS NS
VUR, vesicoureteral reflux; RS, renal parenchymal scarring; n, number of children; NS, not significant.
samples was used to analyze the differences between groups. For variables with a skewed distribution, descriptive results were presented as median and ranges. The ManneWhitney U and KolmogoroveSmirnov tests were used to assess the differences between groups. One-way ANOVA and/or KruskalleWallis tests were used to compare the results between subgroups. Categorical variables were expressed as proportions and compared with results from chi-square tests. Correlations among variables were assessed using the Pearson rank coefficient. For all tests, a value of p < 0.05 was considered statistically significant. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of cystatin C were calculated. A receiver operating characteristic (ROC) curve was constructed to determine the cut-off value of cystatin C for the diagnosis of RN. The project was approved by the ethics committee of Cukurova University Hospital and in all cases written informed consent was provided by at least one of the parents of patients.
Results The current study involved 93 children: 58 girls and 35 boys. Patients were divided into two groups as RS () (n Z 36) and RS (þ) (n Z 57) groups. Anthropometric measurements, age and gender distribution were similar in the two groups. VUR grade was higher in RS (þ) patients compared with RS () ones (p Z 0.02). Patient characteristics are given in Table 2. Laboratory test results and eGFR measurements of RS () and RS (þ) groups are presented in Table 3. BUN, serum cystatin C and Up/Ucr concentrations were significantly higher in RS (þ) than RS () children (p Z 0.04, p Z 0.01 and p Z 0.01, respectively). Scr levels were not significant in patients with and without RS. Scr-based, cystatin Cbased and combined eGFR formulas were all significantly lower in the RS (þ) group than the RS () group (p Z 0.01). Children with RS were subdivided into three groups: mild, moderate and severe scar groups. Comparisons of laboratory data of subgroups are shown in Table 4. BUN, Scr, Ccr and Up/Ucr were significantly different only between mild and severe scar groups (p < 0.05). Serum
Table 3 Comparison of patients with and without renal parenchymal scarring (RS). Parameter
RS () group (n Z 36)
RS (þ) group (n Z 57)
p
BUN (mg/dl) Scr (mg/dl) Up/Ucr (mg/mg) Cystatin C (mg/l) eGFR Ccr New Schwartz Zappitelli1 Filler Zappitelli2 New Ckid
10.39 0.44 0.15 1.08
7.6 0.4 0.2 0.58
15.96 0.69 0.35 1.49
14.7 0.7 0.5 0.81
0.04 NS 0.01 0.01
188.3 153.9 80.4 96.6 118.1 115.5
62.8 50.5 24.4 28.3 34 28.6
125 113.6 61.6 74.7 87.3 87.8
59.7 46.2 29.5 34.4 39.2 32
0.01 0.01 0.01 0.01 0.01 0.01
RS, renal parenchymal scarring; BUN, blood urea nitrogen; Scr, serum creatinine; Up, urine protein; Ucr, creatinine; Ccr, creatinine clearance; eGFR, estimated glomerular filtration rate; Ckid, chronic kidney diseases in children; NS, not significant. Values are given as mean SD (standard deviation). eGFR formulas in ml/min/1.73 m2.
cystatin C was significantly different in three scar subgroups and significantly decreased according to severity of scar grade (p < 0.05). Serum cystatin C and Scr levels in comparison with scarring are shown in Fig. 1. Ccr was only different significantly between mild and severe scar subgroups. Other eGFR equations (New Schwartz, Zappitelli1, Filler, Zappitelli2 and new Ckid) were significantly different among three subgroups (p < 0.05). All eGFR equations were negatively correlated with grade of RS (p < 0.05). The most significant association was observed for new Ckid and Zappitelli2 formula (p < 0.05, r Z 0.58 and r Z 0.54, respectively). A ROC/area under the curve (AUC) curve was drawn by plotting the sensitivity versus the specificity for different cut-off levels of serum cystatin C for the diagnosis of RN (Fig. 2). When a cut-off value of 1.0 mg/l was used, the ROC/AUC analysis demonstrated a sensitivity of 77% and specificity of 50% [AUC Z 0.71, likelihood ratio (LR): 2.94]. The most significant association was observed with combined formulas: new Ckid and Zappitelli2 (p < 0.05, r Z 0.58 and r Z 0.54, respectively).
Discussion The results of the current study demonstrate that cystatin C was significantly elevated in refluxing children with RN in parallel with grade of RS. However, Scr was not distinguishable between RS () and RS (þ) patients and did not significantly change unless severe scarring appeared. Although Scr-based, cystatin C-based and combined eGFR equations all declined in the presence of RN, combined new Ckid and Zappitelli2 formulas seem to be the most accurate for assumption of severity of RN. Recently, there has been enthusiasm for the use of serum cystatin C in various conditions; either in renal or
Assessment of cystatin C and GFR formulas Table 4
265
Comparison of patients with renal parenchymal scarring according to the scar grade.
Parameter BUN (mg/dl) Scr (mg/dl) Up/Ucr (mg/mg) Cystatin C (mg/l) eGFR Ccr New Schwartz Zappitelli1 Filler Zappitelli2 New Ckid
Mild scar (n Z 32)
Moderate scar (n Z 12)
Severe scar (n Z 13)
p
9 0.41 0.1 1.12
(5e29) (0.25e0.96) (0.05e0.7) (0.58e2.79)
12 0.47 0.25 1.38
(6e64) (0.32e0.99) (0.04e1.49) (0.8e2.04)
22 1.0 0.52 1.97
(7e70) (0.29e4.1) (0.07e2.61) (0.94e4.32)
c
142.8 135.0 66.5 80.3 97.5 106.7
(45.9e284.2) (52.1e193.3) (22.9e143.6) (28.9e168.9) (33.6e182.2) (40.4e149.2)
118.1 124.6 52.1 63.8 73.2 87.4
(57.5e167.6) (61.5e157.2) (32.9e97.2) (41.1e116.1) (46.3e117.5) (52.3e119.3)
74.5 61.9 34.4 42.8 44.8 49.9
(8.9e234) (12.3e163.8) (13.7e81.6) (17.7e98.2) (11.5e126.7) (13.7e118.7)
c
c c a,b,c
a,b,c a,b,c a,b,c a,b,c a,b,c
Values are given as median (range). eGFR values in ml/min/1.73 m2. a Significantly differ in mild and moderate scar groups (p 0.05). b Significantly differ in moderate and severe scar groups (p 0.05). c Significantly differ in mild and severe scar groups (p 0.05).
extra-renal disorders. However, data in VUR and RN are restricted. The present study was the first in which the value of cystatin C was assessed as an indicator of RN in pure VUR patients. RS (þ) patients exhibited higher cystatin C levels versus RS () ones. Our findings suggested that the cut-off of 1.0 mg/l cystatin C had adequate sensitivity and specificity for diagnosing RN. Of note, cystatin C showed a rising tendency with grade of RS. On the other hand, Scr was not significantly affected with presence of RS and merely increased in severe RS (þ) children. Our results are in agreement with previous reports regarding cystatin C to be a more sensitive indicator of mild reductions of renal function than Scr [18e20]. On the contrary, in the paper by Islekel et al. examining cystatin C in children with post-pyelonephritic scarring, no association was found between RS and cystatin C levels. This discrepancy could be because of the small sample size of Islekel’s study consisting of children with unilateral RS only
[9]. In another recent report regarding long-term follow-up of patients after UTI, Hannula et al. found that serum cystatin C levels were within normal ranges among all participants with/without VUR and RS. However, ultrasonography was used to detect renal scars and their results did not reflect actual RN [11]. No data are evident in the literature concerning renal functions with cystatin C-based eGFR equations either in VUR or RN. In the present study, Scr, cystatin C and combined Scr-cystatin C-based eGFR equations were lower in RS (þ) patients. All eGFR values except Ccr were decreased in parallel with the severity of RS. Furthermore, eGFR equations were all negatively correlated with RS and the most significant correlation was observed for combined ones: new Ckid and Zappitelli2 formula. Few researchers used Scr-based eGFR equations among patients with RN and VUR. Lahdes-Vasama et al. examined outcomes of kidneys in patients treated for VUR during
Figure 1 Box plot graphs demonstrating serum creatinine and cystatin C levels in patients with and without renal parenchymal scarring. Serum creatinine levels were significantly different between mild and severe scar groups (p 0.05). Serum cystatin C levels were significantly different between all patient groups (p 0.05).
266
S. Yavuz et al. Scr-cystatin C combined eGFR equations, particularly in the presence of mild and moderate RS. Further controlled studies could provide more evidence on the utility of cystatin C in RN in larger populations.
Conflict of interest None.
Funding None.
References
Figure 2 ROC curve showing the sensitivity and specificity of serum cystatin C for diagnosing reflux nephropathy (RN) at different cut-off values. PPV, positive predictive value; NPV, negative predictive value; AUC, area under the curve.
childhood. In contrast to our study, unilateral scarred and unscarred patients had similar eGFR values (based on the Cockroft-Gault formula). In the same study, bilateral scarred patients had the lowest eGFR values, in agreement with the observations from our study [21]. Likewise, a recent cohort indicated that eGFR levels (based on the Schwartz equation) were lower in bilateral RN than in unilateral RN, but not significant between the unilateral and absent RN groups [22]. Moreover, a Spanish group reported decreased eGFR values (based on the Schwartz formula) in patients with RS versus those without. However, sensitivity of eGFR found to be low for detecting loss of renal parenchyma [23]. The common point of these three reports was inefficacy of identifying mild and moderate RN with traditional eGFR methods. Based on the data from our study, cystatin C and Scr-cystatin combined eGFR formulas could fill this gap if the advantages of cystatin C such as an easy performance, no relation of age, gender and body composition, and tendency to increase in the blind area of Scr (mildly reduced GFR of 70e90 ml/min) are considered. A limitation of our study was not using a reference eGFR method, such as inulin or 99 Tc DTPA, to compare accuracy of our different eGFR results with such a reference. However, these reference methods are invasive, hard to perform and nuclear medicine GFR scans involve radiation exposure. Additionally, the count of RS (þ) patients were greater than RS () ones. Another limitation was the lack of a control group. Studying cystatin C in healthy children might be useful when considering reduction of the possible effect of VUR on kidney function.
Conclusions Our study demonstrated that cystatin C could be a useful marker for identifying the risk of RN in patients with VUR. Renal functions might be more accurately determined with
[1] Cendron M. Reflux nephropathy. J Pediatr Urol 2008;4: 414e21. [2] Smith EA. Pyelonephritis, renal scarring, and reflux nephropathy: a pediatric urologist’s perspective. Pediatr Radiol 2008; 38(1):76e82. [3] Gargollo PC, Diamond DA. Therapy insight: what nephrologists need to know about primary vesicoureteral reflux. Nat Clin Pract Nephrol 2007;3(10):551e63. [4] Brakeman P. Vesicoureteral reflux, reflux nephropathy, and end-stage renal disease. Adv Urol 2008. 508949. [5] Inker LA, Okparavero A. Cystatin C as a marker of glomerular filtration rate: prospects and limitations. Curr Opin Nephrol Hypertens 2011;20:631e9. [6] Herget-Rosenthal S, Bo ¨kenkamp A, Hofmann W. How to estimate GFR-serum creatinine, serum cystatin C or equations? Clin Biochem 2007;40(3e4):153e61. [7] Andersen TB, Eskild-Jensen A, Frøkiær J, BrøchnerMortensen J. Measuring glomerular filtration rate in children; can cystatin C replace established methods? A review. Pediatr Nephrol 2009;24:929e41. [8] Weinert LS, Camargo EG, Soares AA, Silveiro SP. Glomerular filtration rate estimation: performance of serum cystatin Cbased prediction equations. Clin Chem Lab Med 2011;49(11): 1761e71. [9] Islekel H, Soylu A, Altun Z, Yis U, Turkmen M, Kavukcu S. Serum and urine cystatin c levels in children with postpyelonephritic renal scarring: a pilot study. Int Urol Nephrol 2007;39(4):1241e50. [10] Lo `pez Pereira P, Moreno Valle JA, Espinosa L, Alonso Dorrego JM, Burgos Lucena L, Martı´nez Urrutia MJ, et al. Enterocystoplasty in children with neuropathic bladders: longterm follow-up. J Pediatr Urol 2008;4(1):27e31. [11] Hannula A, Perhomaa M, Venhola M, Pokka T, Renko M, Uhari M. Long-term follow-up of patients after childhood urinary tract infection. Int Urol Nephrol 2012;44(3):817e27. [12] Schwartz GJ, Munoz A, Schneider MF, Mak RH, Kaskel F, Warady BA, et al. New equations to estimate GFR in children with CKD. J Am Soc Nephrol 2009;20(3):629e37. [13] Zappitelli M, Parvex P, Joseph L, Paradis G, Grey V, Lau S, et al. Derivation and validation of cystatin C-based prediction equations for GFR in children. Am J Kidney Dis 2006;48(2):221e30. [14] Filler G, Lepage N. Should the Schwartz formula for estimation of GFR be replaced by cystatin c formula? Pediatr Nephrol 2003;18(10):981e5. [15] Lebowitz RL, Olbing H, Parkkulainen KV, Smellie JM, Tamminen-Mo ¨bius TE. International system of radiographic grading of vesicoureteral reflux. International Reflux Study in Children. Pediatr Radiol 1985;15(2):105e9. [16] Rushton HG. The evaluation of acute pyelonephritis and renal scarring with technetium 99m-dimercaptosuccinic acid renal
Assessment of cystatin C and GFR formulas
[17]
[18]
[19]
[20]
scintigraphy: evolving concepts and future directions. Pediatr Nephrol 1997;11:108e20. Ziessman HA, Majd M. Importance of methodology on 99m technetium dimercapto-succinic acid scintigraphic image quality: imaging pilot study for RIVUR (Randomized intervention for children with vesicoureteral reflux) multicenter investigation. J Urol 2009;182:272e9. Astor BC, Shaikh S, Chaudhry M. Associations of endogenous markers of kidney function with outcomes: more and less than glomerular filtration rate. Curr Opin Nephrol Hypertens 2013. Mar 15 [epub ahead of print]. Dharnidharka VR, Kwon C, Stevens G. Serum cystatin C is superior to serum creatinine as a marker of kidney function: a meta-analysis. Am J Kidney Dis 2002;40(2):221e6. Roos JF, Doust J, Tett SE, Kirkpatrick CM. Diagnostic accuracy of cystatin C compared to serum creatinine for the estimation
267 of renal dysfunction in adults and children-a meta-analysis. Clin Biochem 2007;40:383e91. [21] Lahdes-Vasama T, Niskanen K, Ro ¨nnholm K. Outcome of kidneys in patients treated for vesicoureteral reflux (VUR) during childhood. Nephrol Dial Transplant 2006;21(9): 2491e7. [22] Lambert HJ, Stewart A, Gullett AM, Cordell HJ, Malcolm S, Feather SA, et al., UK VUR Study Group. Primary, nonsyndromic vesicoureteral reflux and nephropathy in sibling pairs: a United Kingdom cohort for a DNA bank. Clin J Am Soc Nephrol 2011;6(4):760e6. [23] Garcia-Nieto VM, Afonso-Coderch M, Garcia-Rodriguez VE, Monge-Zamorano M, Hernandez-Gonzalez MJ, Luis-Yanes MI. Diagnostic efficiency and quality indexes of several markers of renal function for detecting the loss of parenchyma in paediatric patients. Nephrologia 2012;32(4):486e93.
Assessment of cystatin C and cystatin C-based GFR formulas in reflux nephropathy Sevgi Yavuz*, Ali Anarat, Aysun K. Bayazıt Division of Pediatric Nephrology, Cukurova University School of Medicine, 01330 Adana, Turkey Received 29 April 2013; accepted 22 August 2013 Available online 15 September 2013
KEYWORDS Cystatin C; Glomerular filtration rate; Reflux nephropathy; Vesicoureteral reflux
Abstract Objective: Early identification of reflux nephropathy (RN) could reduce the frequency of chronic kidney disease (CKD) caused by vesicoureteral reflux (VUR). We aimed to assess whether cystatin C has value for determining RN in children with VUR. Materials and methods: Ninety-three children with VUR were classified into two groups according to the presence of renal parenchymal scarring (RS). Patients with RS were divided into three subgroups according to scar grade. Serum cystatin C, serum creatinine (Scr) and urine creatinine were measured. eGFR values of the patients were calculated with Scr-based, cystatin C-based and combined formulas. Results: Cystatin C was significantly higher in patients with RS than patients without RS and declined in parallel with grade of RS (p Z 0.01). Scr was not significant in patients with and without RS. It was only significant between mild and severe scar subgroups (p < 0.05). All eGFR values were lower in RS (þ) patients compared with RS () patients. All eGFR equations were negatively correlated with grade of RS (p < 0.05). Conclusion: Cystatin C could be a useful marker for identifying the risk and severity of RN in patients with VUR. Renal functions could be more accurately determined with Scr-cystatin C combined eGFR equations. ª 2013 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.
Introduction Reflux nephropathy (RN) is an undesirable complication of vesicoureteral reflux (VUR) and is usually identified as renal parenchymal lesions on dimercaptosuccinic acid (DMSA)
* Corresponding author. Tel.: þ90 322 3386060; fax: þ90 322 3386901. E-mail address: [email protected] (S. Yavuz).
scan. Post-infection injury or congenital renal dysplasia shares the features of RN [1,2]. Whether caused by acute pyelonephritis or dysplasia, some patients with VUR progress to RN and some develop chronic kidney disease (CKD). Male gender, high grade and bilateral VUR, high serum creatinine (Scr) or low estimated glomerular filtration rate (eGFR) at admission, proteinuria and hypertension are more likely to increase the risk of CKD in VUR [3,4]. However, which patients are candidates for RN and/or CKD is still unclear.
1477-5131/$36 ª 2013 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jpurol.2013.08.010
Assessment of cystatin C and GFR formulas
263
Cystatin C, a cystine protease inhibitor, has been proposed as an alternative to Scr and presents in all human body fluids and nucleated cells. Cystatin C fulfills the criteria of the ideal filtration marker as it is produced at a constant rate, freely filtered without tubular secretion and completely catabolized in the proximal tubule. Unlike Scr, cystatin C seems to be independent from body weight and height, age (after 1 year) or nutritional status. Clinical studies have shown that cystatin C is superior to Scr for the detection of mild to moderate chronic kidney dysfunction and acute kidney injury [5,6]. More recently, cystatin Cbased prediction equations were introduced to be as good as Scr-based formulas in children [7,8]. However, few cystatin C studies have been conducted in the VUR and RN area [9e11]. In this prospective study, we aimed to assess whether cystatin C might have value for determining RN. Additionally, Scr, cystatin C and combined Scr-cystatin C-based eGFR equations were compared in refluxing children with and without renal parenchymal scarring (RS).
Assays
Materials and methods
The presence of VUR was determined by VCUG and graded from I to V according to the International Reflux Study Committee system [15]. RS was identified by 99mTc-DMSA performed 3e6 months after UTI. Reduced uptake with loss of normal renal contour on the follow-up DMSA scintigraphy was defined as RS [16]. The kidney was divided into 12 segments to quantify the extent of RS. Renal lesions were classified according to the DMSA grading system of Ziessman and Majd: none (no segments affected), mild (1-2 segments affected), moderate (2-3 segments affected) and severe (4 segments affected) [17].
Study population Ninety-three children with primary VUR were enrolled in our prospective study in Cukurova University Hospital between 2009 and 2011. The exclusion criteria were additional renal malformations or systemic disease, stage III CKD (eGFR < 30 ml/min/1.73 m2), history of a recent urinary tract infection (UTI) within 4 weeks, reflux surgery or spontaneous resolution and incapability of 24-h urine sampling. Blood urea nitrogen (BUN), Scr and cystatin C; 24 h urine for protein (Up) and creatinine (Ucr) were measured in all patients. All children underwent 99mTcdimercaptosuccinic acid (99mTc-DMSA) scanning and voiding cystourethrogram (VCUG) imaging before the onset of the trial. Patients were divided into two groups according to presence of RS. Then, patients with RS were divided into three subgroups as mild, moderate and severe based on the grade of RS.
Table 1
BUN, Scr, Ucr and Up were concurrently measured using turbidimetric assay in a Beckman Coulter AU analyzer (Beckman Coulter, California, United States). Serum cystatin C concentrations were determined by ELISA using the human Cystatin C ELISA kit (BioVendor, Heidelberg, Germany). Intra-assay and inter-assay variations of the test were 2.5 and 6.9%, respectively.
Assessment of eGFR Estimated GFR values of the all group were calculated according to Scr-based formulas [creatinine clearance (Ccr), new Schwartz], cystatin C-based equations (Zappitelli1, Filler) and combined Scr-cystatin C formulas [Zappitelli2, the New Chronic Kidney Disease in children (Ckid) equations] (Table 1) [12e14].
Imaging studies
Statistical analysis Statistical analysis was done using SPSS version 13.0 for Windows. Initially, assumptions of normality and homogeneity of variance were checked. Descriptive results were expressed as mean standard deviation (SD) for variables with a normal distribution. The t test for independent
Formulas used for the estimation of glomerular filtration rate. Formula
With plasma creatinine Creatinine clearance (Ccr) (ml/min/1.73 m2) New Schwartz (ml/min/1.73 m2) [12] With serum cystatine C Zappitelli1 (ml/min/1.73 m2) [13] Filler (ml/min/1.73 m2) [14] Combined Zappitelli2 (ml/min/1.73 m2) [13] New Ckid (ml/min/1.73 m2) [12]
eGFR Z [Ucr (mg/dl)/Scr (mg/dl)] [urine volume (ml)/1440 (min)] [1.73/body surface area (m2)] eGFR Z 0.413 [height (cm)/Scr (mg/dl)] eGFR Z 75.94 [cystatin C (mg/l)1.17] eGFR Z 91.62 [cystatin C (mg/l)1.123] eGFR Z [507.76 e0.003 height(m)]/[cystatin C (mg/l)0.635 sCr (mmol/l)0.547] eGFR Z 39.1 [height (m)/Scr (mg/dl)]0.516 [1.8/cystatin C (mg/l)0.294] [30/BUN (mg/dl)0.169] [1.099]male [height (m)/1.4]0.188
Scr, serum creatinine; Ucr, urine creatinine; eGFR, estimated glomerular filtration rate; Ccr, creatinine clearance; Ckid, chronic kidney diseases in children.
264 Table 2
S. Yavuz et al. Baseline characteristics of study groups.
Characteristics
RS () group (n Z 36)
Age (year) 8.44 3.35 Gender, n (%) Male 14 (38.9) Female 22 (61.1) Weight 30.94 13.8 Height 131.68 21.1 Grade of VUR, n (%) I-II 19 (52.8) III 14 (38.9) IV-V 3 (8.3)
RS (þ) group (n Z 57)
p
9.09 3.46
NS
21 (36.8) 36 (63.2) 31.15 14.1 130.1 16.6
NS
14 (24.5) 34 (59.6) 9 (15.9)
0.02
NS NS
VUR, vesicoureteral reflux; RS, renal parenchymal scarring; n, number of children; NS, not significant.
samples was used to analyze the differences between groups. For variables with a skewed distribution, descriptive results were presented as median and ranges. The ManneWhitney U and KolmogoroveSmirnov tests were used to assess the differences between groups. One-way ANOVA and/or KruskalleWallis tests were used to compare the results between subgroups. Categorical variables were expressed as proportions and compared with results from chi-square tests. Correlations among variables were assessed using the Pearson rank coefficient. For all tests, a value of p < 0.05 was considered statistically significant. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of cystatin C were calculated. A receiver operating characteristic (ROC) curve was constructed to determine the cut-off value of cystatin C for the diagnosis of RN. The project was approved by the ethics committee of Cukurova University Hospital and in all cases written informed consent was provided by at least one of the parents of patients.
Results The current study involved 93 children: 58 girls and 35 boys. Patients were divided into two groups as RS () (n Z 36) and RS (þ) (n Z 57) groups. Anthropometric measurements, age and gender distribution were similar in the two groups. VUR grade was higher in RS (þ) patients compared with RS () ones (p Z 0.02). Patient characteristics are given in Table 2. Laboratory test results and eGFR measurements of RS () and RS (þ) groups are presented in Table 3. BUN, serum cystatin C and Up/Ucr concentrations were significantly higher in RS (þ) than RS () children (p Z 0.04, p Z 0.01 and p Z 0.01, respectively). Scr levels were not significant in patients with and without RS. Scr-based, cystatin Cbased and combined eGFR formulas were all significantly lower in the RS (þ) group than the RS () group (p Z 0.01). Children with RS were subdivided into three groups: mild, moderate and severe scar groups. Comparisons of laboratory data of subgroups are shown in Table 4. BUN, Scr, Ccr and Up/Ucr were significantly different only between mild and severe scar groups (p < 0.05). Serum
Table 3 Comparison of patients with and without renal parenchymal scarring (RS). Parameter
RS () group (n Z 36)
RS (þ) group (n Z 57)
p
BUN (mg/dl) Scr (mg/dl) Up/Ucr (mg/mg) Cystatin C (mg/l) eGFR Ccr New Schwartz Zappitelli1 Filler Zappitelli2 New Ckid
10.39 0.44 0.15 1.08
7.6 0.4 0.2 0.58
15.96 0.69 0.35 1.49
14.7 0.7 0.5 0.81
0.04 NS 0.01 0.01
188.3 153.9 80.4 96.6 118.1 115.5
62.8 50.5 24.4 28.3 34 28.6
125 113.6 61.6 74.7 87.3 87.8
59.7 46.2 29.5 34.4 39.2 32
0.01 0.01 0.01 0.01 0.01 0.01
RS, renal parenchymal scarring; BUN, blood urea nitrogen; Scr, serum creatinine; Up, urine protein; Ucr, creatinine; Ccr, creatinine clearance; eGFR, estimated glomerular filtration rate; Ckid, chronic kidney diseases in children; NS, not significant. Values are given as mean SD (standard deviation). eGFR formulas in ml/min/1.73 m2.
cystatin C was significantly different in three scar subgroups and significantly decreased according to severity of scar grade (p < 0.05). Serum cystatin C and Scr levels in comparison with scarring are shown in Fig. 1. Ccr was only different significantly between mild and severe scar subgroups. Other eGFR equations (New Schwartz, Zappitelli1, Filler, Zappitelli2 and new Ckid) were significantly different among three subgroups (p < 0.05). All eGFR equations were negatively correlated with grade of RS (p < 0.05). The most significant association was observed for new Ckid and Zappitelli2 formula (p < 0.05, r Z 0.58 and r Z 0.54, respectively). A ROC/area under the curve (AUC) curve was drawn by plotting the sensitivity versus the specificity for different cut-off levels of serum cystatin C for the diagnosis of RN (Fig. 2). When a cut-off value of 1.0 mg/l was used, the ROC/AUC analysis demonstrated a sensitivity of 77% and specificity of 50% [AUC Z 0.71, likelihood ratio (LR): 2.94]. The most significant association was observed with combined formulas: new Ckid and Zappitelli2 (p < 0.05, r Z 0.58 and r Z 0.54, respectively).
Discussion The results of the current study demonstrate that cystatin C was significantly elevated in refluxing children with RN in parallel with grade of RS. However, Scr was not distinguishable between RS () and RS (þ) patients and did not significantly change unless severe scarring appeared. Although Scr-based, cystatin C-based and combined eGFR equations all declined in the presence of RN, combined new Ckid and Zappitelli2 formulas seem to be the most accurate for assumption of severity of RN. Recently, there has been enthusiasm for the use of serum cystatin C in various conditions; either in renal or
Assessment of cystatin C and GFR formulas Table 4
265
Comparison of patients with renal parenchymal scarring according to the scar grade.
Parameter BUN (mg/dl) Scr (mg/dl) Up/Ucr (mg/mg) Cystatin C (mg/l) eGFR Ccr New Schwartz Zappitelli1 Filler Zappitelli2 New Ckid
Mild scar (n Z 32)
Moderate scar (n Z 12)
Severe scar (n Z 13)
p
9 0.41 0.1 1.12
(5e29) (0.25e0.96) (0.05e0.7) (0.58e2.79)
12 0.47 0.25 1.38
(6e64) (0.32e0.99) (0.04e1.49) (0.8e2.04)
22 1.0 0.52 1.97
(7e70) (0.29e4.1) (0.07e2.61) (0.94e4.32)
c
142.8 135.0 66.5 80.3 97.5 106.7
(45.9e284.2) (52.1e193.3) (22.9e143.6) (28.9e168.9) (33.6e182.2) (40.4e149.2)
118.1 124.6 52.1 63.8 73.2 87.4
(57.5e167.6) (61.5e157.2) (32.9e97.2) (41.1e116.1) (46.3e117.5) (52.3e119.3)
74.5 61.9 34.4 42.8 44.8 49.9
(8.9e234) (12.3e163.8) (13.7e81.6) (17.7e98.2) (11.5e126.7) (13.7e118.7)
c
c c a,b,c
a,b,c a,b,c a,b,c a,b,c a,b,c
Values are given as median (range). eGFR values in ml/min/1.73 m2. a Significantly differ in mild and moderate scar groups (p 0.05). b Significantly differ in moderate and severe scar groups (p 0.05). c Significantly differ in mild and severe scar groups (p 0.05).
extra-renal disorders. However, data in VUR and RN are restricted. The present study was the first in which the value of cystatin C was assessed as an indicator of RN in pure VUR patients. RS (þ) patients exhibited higher cystatin C levels versus RS () ones. Our findings suggested that the cut-off of 1.0 mg/l cystatin C had adequate sensitivity and specificity for diagnosing RN. Of note, cystatin C showed a rising tendency with grade of RS. On the other hand, Scr was not significantly affected with presence of RS and merely increased in severe RS (þ) children. Our results are in agreement with previous reports regarding cystatin C to be a more sensitive indicator of mild reductions of renal function than Scr [18e20]. On the contrary, in the paper by Islekel et al. examining cystatin C in children with post-pyelonephritic scarring, no association was found between RS and cystatin C levels. This discrepancy could be because of the small sample size of Islekel’s study consisting of children with unilateral RS only
[9]. In another recent report regarding long-term follow-up of patients after UTI, Hannula et al. found that serum cystatin C levels were within normal ranges among all participants with/without VUR and RS. However, ultrasonography was used to detect renal scars and their results did not reflect actual RN [11]. No data are evident in the literature concerning renal functions with cystatin C-based eGFR equations either in VUR or RN. In the present study, Scr, cystatin C and combined Scr-cystatin C-based eGFR equations were lower in RS (þ) patients. All eGFR values except Ccr were decreased in parallel with the severity of RS. Furthermore, eGFR equations were all negatively correlated with RS and the most significant correlation was observed for combined ones: new Ckid and Zappitelli2 formula. Few researchers used Scr-based eGFR equations among patients with RN and VUR. Lahdes-Vasama et al. examined outcomes of kidneys in patients treated for VUR during
Figure 1 Box plot graphs demonstrating serum creatinine and cystatin C levels in patients with and without renal parenchymal scarring. Serum creatinine levels were significantly different between mild and severe scar groups (p 0.05). Serum cystatin C levels were significantly different between all patient groups (p 0.05).
266
S. Yavuz et al. Scr-cystatin C combined eGFR equations, particularly in the presence of mild and moderate RS. Further controlled studies could provide more evidence on the utility of cystatin C in RN in larger populations.
Conflict of interest None.
Funding None.
References
Figure 2 ROC curve showing the sensitivity and specificity of serum cystatin C for diagnosing reflux nephropathy (RN) at different cut-off values. PPV, positive predictive value; NPV, negative predictive value; AUC, area under the curve.
childhood. In contrast to our study, unilateral scarred and unscarred patients had similar eGFR values (based on the Cockroft-Gault formula). In the same study, bilateral scarred patients had the lowest eGFR values, in agreement with the observations from our study [21]. Likewise, a recent cohort indicated that eGFR levels (based on the Schwartz equation) were lower in bilateral RN than in unilateral RN, but not significant between the unilateral and absent RN groups [22]. Moreover, a Spanish group reported decreased eGFR values (based on the Schwartz formula) in patients with RS versus those without. However, sensitivity of eGFR found to be low for detecting loss of renal parenchyma [23]. The common point of these three reports was inefficacy of identifying mild and moderate RN with traditional eGFR methods. Based on the data from our study, cystatin C and Scr-cystatin combined eGFR formulas could fill this gap if the advantages of cystatin C such as an easy performance, no relation of age, gender and body composition, and tendency to increase in the blind area of Scr (mildly reduced GFR of 70e90 ml/min) are considered. A limitation of our study was not using a reference eGFR method, such as inulin or 99 Tc DTPA, to compare accuracy of our different eGFR results with such a reference. However, these reference methods are invasive, hard to perform and nuclear medicine GFR scans involve radiation exposure. Additionally, the count of RS (þ) patients were greater than RS () ones. Another limitation was the lack of a control group. Studying cystatin C in healthy children might be useful when considering reduction of the possible effect of VUR on kidney function.
Conclusions Our study demonstrated that cystatin C could be a useful marker for identifying the risk of RN in patients with VUR. Renal functions might be more accurately determined with
[1] Cendron M. Reflux nephropathy. J Pediatr Urol 2008;4: 414e21. [2] Smith EA. Pyelonephritis, renal scarring, and reflux nephropathy: a pediatric urologist’s perspective. Pediatr Radiol 2008; 38(1):76e82. [3] Gargollo PC, Diamond DA. Therapy insight: what nephrologists need to know about primary vesicoureteral reflux. Nat Clin Pract Nephrol 2007;3(10):551e63. [4] Brakeman P. Vesicoureteral reflux, reflux nephropathy, and end-stage renal disease. Adv Urol 2008. 508949. [5] Inker LA, Okparavero A. Cystatin C as a marker of glomerular filtration rate: prospects and limitations. Curr Opin Nephrol Hypertens 2011;20:631e9. [6] Herget-Rosenthal S, Bo ¨kenkamp A, Hofmann W. How to estimate GFR-serum creatinine, serum cystatin C or equations? Clin Biochem 2007;40(3e4):153e61. [7] Andersen TB, Eskild-Jensen A, Frøkiær J, BrøchnerMortensen J. Measuring glomerular filtration rate in children; can cystatin C replace established methods? A review. Pediatr Nephrol 2009;24:929e41. [8] Weinert LS, Camargo EG, Soares AA, Silveiro SP. Glomerular filtration rate estimation: performance of serum cystatin Cbased prediction equations. Clin Chem Lab Med 2011;49(11): 1761e71. [9] Islekel H, Soylu A, Altun Z, Yis U, Turkmen M, Kavukcu S. Serum and urine cystatin c levels in children with postpyelonephritic renal scarring: a pilot study. Int Urol Nephrol 2007;39(4):1241e50. [10] Lo `pez Pereira P, Moreno Valle JA, Espinosa L, Alonso Dorrego JM, Burgos Lucena L, Martı´nez Urrutia MJ, et al. Enterocystoplasty in children with neuropathic bladders: longterm follow-up. J Pediatr Urol 2008;4(1):27e31. [11] Hannula A, Perhomaa M, Venhola M, Pokka T, Renko M, Uhari M. Long-term follow-up of patients after childhood urinary tract infection. Int Urol Nephrol 2012;44(3):817e27. [12] Schwartz GJ, Munoz A, Schneider MF, Mak RH, Kaskel F, Warady BA, et al. New equations to estimate GFR in children with CKD. J Am Soc Nephrol 2009;20(3):629e37. [13] Zappitelli M, Parvex P, Joseph L, Paradis G, Grey V, Lau S, et al. Derivation and validation of cystatin C-based prediction equations for GFR in children. Am J Kidney Dis 2006;48(2):221e30. [14] Filler G, Lepage N. Should the Schwartz formula for estimation of GFR be replaced by cystatin c formula? Pediatr Nephrol 2003;18(10):981e5. [15] Lebowitz RL, Olbing H, Parkkulainen KV, Smellie JM, Tamminen-Mo ¨bius TE. International system of radiographic grading of vesicoureteral reflux. International Reflux Study in Children. Pediatr Radiol 1985;15(2):105e9. [16] Rushton HG. The evaluation of acute pyelonephritis and renal scarring with technetium 99m-dimercaptosuccinic acid renal
Assessment of cystatin C and GFR formulas
[17]
[18]
[19]
[20]
scintigraphy: evolving concepts and future directions. Pediatr Nephrol 1997;11:108e20. Ziessman HA, Majd M. Importance of methodology on 99m technetium dimercapto-succinic acid scintigraphic image quality: imaging pilot study for RIVUR (Randomized intervention for children with vesicoureteral reflux) multicenter investigation. J Urol 2009;182:272e9. Astor BC, Shaikh S, Chaudhry M. Associations of endogenous markers of kidney function with outcomes: more and less than glomerular filtration rate. Curr Opin Nephrol Hypertens 2013. Mar 15 [epub ahead of print]. Dharnidharka VR, Kwon C, Stevens G. Serum cystatin C is superior to serum creatinine as a marker of kidney function: a meta-analysis. Am J Kidney Dis 2002;40(2):221e6. Roos JF, Doust J, Tett SE, Kirkpatrick CM. Diagnostic accuracy of cystatin C compared to serum creatinine for the estimation
267 of renal dysfunction in adults and children-a meta-analysis. Clin Biochem 2007;40:383e91. [21] Lahdes-Vasama T, Niskanen K, Ro ¨nnholm K. Outcome of kidneys in patients treated for vesicoureteral reflux (VUR) during childhood. Nephrol Dial Transplant 2006;21(9): 2491e7. [22] Lambert HJ, Stewart A, Gullett AM, Cordell HJ, Malcolm S, Feather SA, et al., UK VUR Study Group. Primary, nonsyndromic vesicoureteral reflux and nephropathy in sibling pairs: a United Kingdom cohort for a DNA bank. Clin J Am Soc Nephrol 2011;6(4):760e6. [23] Garcia-Nieto VM, Afonso-Coderch M, Garcia-Rodriguez VE, Monge-Zamorano M, Hernandez-Gonzalez MJ, Luis-Yanes MI. Diagnostic efficiency and quality indexes of several markers of renal function for detecting the loss of parenchyma in paediatric patients. Nephrologia 2012;32(4):486e93.
