ORIGINAL ARTICLE

Urinary proteins in children with urinary tract infection

Lena Andersson & Iulian Preda & Mirjana Hahn-Zoric &

Lars Å. Hanson & Ulf Jodal & Rune Sixt &Lars Barregard & Sverker Hansson

Received: 19 June 2008 /Revised: 4 March 2009 /Accepted: 6 March 2009 /Published online: 8 April 2009# IPNA 2009

Abstract The aim of this study was to test our hypothesisthat the urinary excretion of C-reactive protein (CRP),alpha 1-microglobulin (A1M), retinol-binding protein(RBP) and Clara cell protein (CC16) is increased inchildren with urinary tract infection (UTI) and relates torenal damage as measured by acute dimercaptosuccinic acid(DMSA) scintigraphy. Fifty-two children <2 years of agewith UTI were enrolled in the study, 44 of whom werefebrile. The control group consisted of 23 patients withnon-UTI infection and elevated serum CRP (s-CRP) levels.Thirty-six patients had abnormal DMSA uptake, classifiedas mild, moderate or severe damage (DMSA class 1, 2, 3,respectively). There was a significant association betweenDMSA class and the excretion of urinary RBP (u-RBP) andu-CC16. There was also a significant difference in u-CRPlevels between children with UTI and control children withnon-UTI infections, although u-CRP excretion was notsignificantly correlated to DMSA class. In conclusion, theurinary excretion of the low-molecular-weight proteinsRBP and CC16 showed a strong association with uptakedefects on renal DMSA scans. The urinary level of CRP

seems to distinguish between children with UTI and otherfebrile conditions. A combination of these biomarkers maybe useful in the clinical assessment of children with UTI.

Keywords Alpha 1-microglobulin . Biomarker .

Clara cell protein . C-reactive protein . Scintigraphy .

Retinol-binding protein

Introduction

Renal involvement during urinary tract infection (upperUTI) causes parenchymal inflammation that can be visual-ized by scintigraphy [1]. This is an invasive and expensivetechnique that is not always available, and a search for non-invasive methods is therefore relevant. Several studies havemonitored proteinuria in UTI patients [2], but most of theseinvolved the use of non-specific reagent strips or measuredmicroalbuminuria; few have investigated low-molecular-weight proteins (LMWPs) with specific immunoassays.Sandberg et al. showed that in an adult population, thelevels of LMWPs, such as alpha 1-microglobulin (A1M),beta 2-microglobulin (B2M) and retinol-binding protein(RBP), increased during acute pyelonephritis comparedwith cystitis and asymptomatic bacteriuria; however,patients with fever of non-renal origin also had increasedurine levels of LMWPs [3]. In children, microalbuminuriaand N-acetyl-beta-glucosaminidase (NAG) level have beenfound to increase in patients with acute pyelonephritiscompared to those with cystitis and they were higher thanthose in a non-renal fever control group. Beta 2-microglobulindid not differ significantly between the groups [4–6]. Childrenwith glomerular disease had increased excretion of LMWPs,such as A1M, Clara cell protein (CC16) (also known as‘urine protein 1’), RBP and B2M [7]. In children with UTI,

Pediatr Nephrol (2009) 24:1533–1538DOI 10.1007/s00467-009-1173-2

L. Andersson (*) : L. BarregardDepartment of Occupational and Environmental Medicine,Sahlgrenska University Hospital and Academy,University of Gothenburg,P.O. Box 414, 405 30 Gothenburg, Swedene-mail: lena.andersson@amm.gu.se

I. Preda :U. Jodal :R. Sixt : S. HanssonPediatric Uro-Nephrologic Centre,Queen Silvia Children’s Hospital, University of Gothenburg,Gothenburg, Sweden

M. Hahn-Zoric : L. Å. HansonDepartment of Clinical Immunology, Sahlgrenska UniversityHospital and Academy, University of Gothenburg,Gothenburg, Sweden

aberrations in acute 99mtechnetium dimercaptosuccinic acid(DMSA) scintigraphy were shown to relate to the urinaryexcretion of A1M [8], but not to B2M or NAG [4, 5].However, an association between NAG excretion andDMSA uptake defects was shown in infants <1 year of age[6].

C-reactive protein (CRP) is a direct and quantitativemarker of the acute-phase reaction, and there is a significantassociation between the severity of inflammation at firstUTI, as measured by serum CRP (s-CRP), and permanentrenal damage [9, 10]. It has been recently demonstrated thatCRP can also be produced locally in the kidney [11].Therefore, studies of the association between CRP in urine(u-CRP) and renal damage are of interest. To the best of ourknowledge, such studies have not been previously reported.

Clara cell protein is an LMWP that has been used lessoften than A1M and RBP as a marker of renal damage [12].It is an anti-inflammatory protein that is mainly producedby the Clara cells in the bronchioli and is excreted into theurine.

The aim of our study was to test the hypothesis that theurinary excretion of CRP, A1M, RBP and CC16 isincreased in children with UTI and related to renal damageas measured by acute DMSA scintigraphy. We report theresults in children with acute UTI and control patients withsevere infections in organs other than the urinary tract.

Patients and methods

Children eligible for the study were <2 years of age, withfirst-time, community-acquired symptomatic UTI (withfever, failure to thrive, poor weight gain, irritability,vomiting) and who were attending the hospital emergencyroom when the study team was available. Children withknown urogenital or anorectal malformations or neurolog-ical diseases were excluded. The diagnosis of UTI requiredbacteriuria with growth of any number of bacteria in theurine from a suprapubic bladder aspiration, or ≥100,000colony-forming units in urine from two midstream or bagsamples. None of the children with bag samples haddiarrhoea. Investigations, treatment and further manage-ment were according to the hospital guidelines [10]. A spotsample for analysis of urinary proteins was obtained within3 days of diagnosis. The highest temperature was recordedas well as the highest s-CRP concentration.

The patient group enrolled in the study comprised 52Caucasian children <2 years of age with first-time UTI, ofwhom 26 were male infants (median age 0.2 years, range12 days–1.0 year) and 26 were female (median age0.9 years, range 1.5 months–1.9 years). A fever of ≥38.5°C was recorded in 44 (85%) children. Non-febrile childrenhad (in addition to bacteriuria) other symptoms (failure to

thrive, poor weight gain, irritability or vomiting) and weregenerally <3 months of age. The bacteriuria was caused byEscherichia coli (47 cases, 90%), Klebsiella (three),coagulase-negative staphylococci (one) and Proteus (one).

Ultrasound examination of the kidneys and bladder wasperformed within 1–2 days and showed dilatation of theurinary tract compatible with distal ureteral stenosis in fivecases and pelvo-ureteral stenosis in one. Voiding cystour-ethrography was performed within 1–2 months in 48patients and showed vesicoureteral reflux in seven children,four of whom had dilatation.

A DMSA scan was performed within 1 month afteradmission in all cases (median 5 days). The scan wasperformed in accordance with the guidelines of thePaediatric Committee of the European Association ofNuclear Medicine [13]. Thus, three images, one posteriorand two posterior oblique, were acquired 2–3 h afterinjection of DMSA in a dose scaled down from an adultdose of 70 MBq according to body surface area, with aminimum dose of 15 MBq. A kidney without uptake defectand relative (split) function ≥45% was classified as normal(class 0); a kidney with reduced or absent uptake in one ormore areas or relative function <45% was considered to beabnormal, i.e. upper UTI. The extent of kidney damage wasgraded semi-quantitatively (class 1: uptake abnormal withrelative function ≥45%; class 2: relative function 40–44%irrespective of uptake; class 3: relative function <40%irrespective of uptake). In the case of bilateral involvement,the kidneys were individually classified according to theextent of the damage. In the analysis, the kidney with the morepronounced involvement was used to characterize the patient.

Controls were 23 Caucasian children (median age1.8 years, range 0.2–2.8 years) with an acute infection inan organ other than the urinary tract, who in addition to anegative urine culture also had elevated s-CRP (>20 mg/L).Of these, five had X-ray-verified pneumonia, ten had upperrespiratory tract infection and eight had other infections.The control children were collected on two differentoccasions since 14 control children were added 3 yearslater to enlarge this group.

Serum CRP was determined by a nephelometric method(NycoCard; Axis-Shield PoC AS, Oslo, Norway); concen-trations >20 mg/L were considered to be compatible withrenal inflammation. Urinary CRP, A1M, RBP and CC16were measured by commercial high-sensitivity enzyme-linked immunosorbent assays (ELISAs) (ImmundiagnostikAG, Bensheim, Germany, and BioVendor, LaboratoryMedicine, Brno, Czech Republic) using polyclonal rabbitantibodies against human proteins coated onto the wells ofthe micro-titre plates [14, 15]. The levels of u-CRP wereanalysed 1–2 days after diagnosis, while u-A1M, u-RBPand u-CC16 levels were analysed in samples that had beenstored for 3 years at −20°C, with the exception of the

1534 Pediatr Nephrol (2009) 24:1533–1538

samples from the 14 control children, which had beencollected later. These 14 samples were frozen at −20°C andanalysed within 1 month. Urine samples were diluted 1:5for CRP, 1:20 for A1M, 1:10 for RBP and 1:25 for CC16.Urinary creatinine was measured using the Jaffe technique,which is a kinetic method (Cobas Mira; Roche Diagnostics,Rotkreuz, Switzerland). The excretion of each urinaryprotein was expressed as the ratio between the concentra-tion of the respective protein and urinary creatinine.

Data analyses were performed using the SAS software,release 9.1 (SAS Institute, Cary, NC). Since the distribu-tions of urinary proteins were not normal, group compar-isons were performed using the non-parametric Kruskal–Wallis test and correlations between variables using Spear-man’s rank correlation coefficient. For values below thelimit of detection (LoD), the calculated value of the LoDdivided by the square root of 2 was used in the statisticalcalculations [16]. Stepwise logistic regression (using thecumulative logit model) was used for multivariate purposes.A p value<0.05 was considered statistically significant.

The study was approved by the Ethics Committee of theUniversity of Gothenburg (Ö118–02). The informed con-sent of parents was obtained.

Results

The DMSA scintigraphy images showed abnormal uptakein 36 patients (69%). The urinary excretion of LMWPs inthe control children and in children with UTI, stratified byDMSA class, is shown in Table 1 and Figs. 1 and 2.

In children with UTI (n=52), the acute phase proteins-CRP was positively correlated with temperature and all ofthe other proteins (u-CRP, u-A1M, u-RBP and u-CC16).Urinary CRP and u-CC16 were positively correlated withall other proteins but not with temperature. Urinary A1Mand u-RBP were not significantly associated with temper-ature but with each other.

Urinary CRP was significantly higher in children with UTIthan in control children (p<0.001; Fig. 1), but it was notassociated with DMSA class (p=0.23).

Urinary RBP and u-CC16 were significantly higher inchildren with UTI than in control children (p=0.005 and p=0.04, respectively). Furthermore, children without scintigraphicsigns of renal involvement (DMSA class 0) had significantlylower urinary protein excretion than those with such signs(DMSA classes 1–3) in terms of u-CC16 (p<0.001) and u-RBP (p =0.008) but not for u-A1M (p=0.25). There was alsoan association between DMSA class and s-CRP (p <0.001).

Urinary CC16, u-RBP and s-CRP and temperature wereincluded in a stepwise logistic regression model, and u-CC16 and s-CRP were found to be independent predictorsof the degree of renal involvement, according to DMSAscintigraphy (p<0.001 and p<0.016, respectively).

Discussion

In our study, the urinary excretion of CRP increased inchildren with UTI, but there was no clear associationbetween u-CRP and the extent of renal damage, as assessedby DMSA scintigraphy. Urinary CC16, u-RBP and s-CRP,

Table 1 Highest recorded temperature, serum C-reactive protein (CRP), urinary CRP, u-alpha 1-microglobulin, u-retinol-binding protein and u-Claracell protein in children with acute urinary tract infection (UTI) and in control children with febrile non-UTI infections

Parameters Controls (n=23) DMSA class 0(n=16)

DMSA class 1(n =12)

DMSA class 2(n=15)

DMSA class 3(n=9)

Temperaturea (°C) 40.0 (38.6–41.0) 39.2 (37.0–40.3) 40.0 (37.0–40.4) 40.0 (39.0–41.0) 39.3 (38.2–41.0)

s-CRPb (mg/L) 96 (21–243) 42 (<LoD–142) 110 (<LoD–247) 130 (64–300) 169 (92–224)

u-CRP (μg/g creatinine) 27 (<LoD–177) 42 (<LoD–5916) 308 (3–8000) 88 (7–5407) 511 (55–3114)

u-A1M (mg/g creatinine) 4.9 (0.6–35) 3.1 (1.1–23) 2.6 (0.4–18) 6.6 (2.0–51) 10.5 (<LoD–108)

u-RBPc (μg/g creatinine) 110 (<LoD–2935) 141 (<LoD–1975) 289 (10–1981) 386 (<LoD–7620) 2502 (172–7792)

u-CC16d (μg/g creatinine) 6.3 (<LoD–303) 5.7 (<LoD–177) 97 (<LoD–384) 105 (1–476) 488 (91–904)

Values are presented as the median with the range given in parenthesis

s-CRP, Serum CRP; u-CRP, urinary CRP; u-AIM, u-alpha 1-microglobulin; u-RBP, u-retinol-binding protein; u-CC16, u-Clara cell protein;DMSA, dimercaptosuccinic acid; LoD, limit of detection

Children with UTI are stratified by degree of renal damage, as assessed by DMSA scintigraphy (DMSA class 0–3). See caption to Fig. 1 fordefinition of classes.a p=0.04 for association between DMSA class and temperature (Kruskal–Wallis test)b p=0.0003 for association between DMSA class and s-CRP (Kruskal–Wallis test)c p=0.002 for association between DMSA class and u-RBP (Kruskal–Wallis test)d p<0.0001 for association between DMSA class and u-CC16 (Kruskal–Wallis test)

Pediatr Nephrol (2009) 24:1533–1538 1535

on the other hand, were significantly associated with theextent of uptake defects, based on the DMSA scans.Children with uptake defects at the time of an acute DMSAscan are mostly regarded as having an inflammatoryprocess of the renal parenchyma, i.e. acute pyelonephritis.An alternative explanation is that the uptake defects depict

areas of fibrous tissue that were either antenatally acquiredor caused by a previous renal infection (scarring). Theremay have been a number of patients with this type of renaldamage in the study group, but the low age of the children(median 2 months in the male infants) and previous historymakes it unlikely that the proportion of patients with

Fig. 2 Levels of alpha1-microglobulin (u-A1M)(a), retinol-binding protein(u-RBP) (b) and Clara cell pro-tein (u-CC16) (c) in patientswith acute UTI, categorizedinto DMSA classes 0–3(see caption to Fig. 1), and incontrols. Median values and 10,25, 75 and 90 percentilesare given

Fig. 1 Levels of serum C-reactive protein (s-CRP) (a) and urinaryC-reactive protein (u-CRP) (b) in patients with acute urinary tractinfection (UTI), categorized into dimercaptosuccinic acid (DMSA)scintigraphy classes 0–3, and in controls. Median values and 10, 25,

75 and 90 percentiles are given. DMSA classes: 0 normal kidneyfunction, 1 uptake abnormal with relative function ≥45%, 2 relativefunction 40–44% irrespective of uptake, 3 relative function <40%irrespective of uptake

1536 Pediatr Nephrol (2009) 24:1533–1538

previous infections was large. Whatever the uptake defectsrepresent, they depict areas of decreased renal functionwhich may be clinically important to detect.

C-reactive protein is formed mainly in the liver byhepatocytes, but studies carried out in recent years haveshown that it is also produced at other sites, such ascardiovascular locations, adipose tissue and the kidney [11,17–19]. Liver synthesis is triggered by cytokines involvedin the inflammatory process, with s-CRP increasingexponentially, doubling every 8–9 h. The elimination rateof CRP is constant, and since the local production at othersites is small, the concentration of s-CRP is regulated byliver synthesis. The biological half-life of CRP is estimatedto be 13–16 h [20].

The relatively large size of CRP, 106 kDa (comparedwith, for example, 65 kDa for albumin), indicates that theincreased urinary excretion is the result of the localproduction and secretion of this protein from kidney tissuerather than from glomerular filtration of s-CRP. In patientswith secondary amyloidosis, s-CRP is lost into the urineonly in massive proteinuria, indicating that for increasedu-CRP by this mechanism, advanced glomerular injury isrequired [21]. It is unlikely that s-CRP (or fragments of it)was the source of the u-CRP in the UTI children of ourstudy, since all of the non-UTI control children hadelevated s-CRP (>20 mg/L) while u-CRP was barelydetectable (Fig. 1).

Little is known about the kinetics of renal production and theexcretion of CRP in children with UTI. To assess the clinicalvalue of u-CRP in UTI, there is a need for studies with repeateddeterminations of u-CRP during the clinical course of UTI.

We found a strong association between the extent ofuptake defects, as evidenced on DMSA scans, and theexcretion of u-CC16 and u-RBP. Results from previousstudies indicate that tubular proteinuria is common inpyelonephritis and, occasionally, also albuminuria, asreviewed by Carter et al. [2]. Studies on u-B2M, anotherLMWP, have shown, however, that this molecule has a lowability to discriminate between pyelonephritis and cystitisin children [4, 5]. In contrast, we have shown in our studythat there was a significantly higher excretion of the othertubular proteins, especially u-CC16 and u-RBP, in childrenwith UTI and with renal uptake defects visualized onDMSA scans. Consequently, these proteins are potentialbiomarkers for the prediction of renal involvement inchildren with UTI. In order to quantitatively assess thesensitivity and specificity of these tubular markers, optimalcut-off limits are needed, based on studies in larger groupsof children. There is also a need for data on normal levelsof these LMWPs in infants, although data on RBP insomewhat older children are available. Dillon et al. reporteda value of 27 μg/g creatinine in 183 healthy children aged2–16 years [22].

Urinary CRP, which is not a LMWP, was not clearlyassociated with DMSA class, but its level was significantlyincreased in UTI patients compared with non-renal fevercontrols (Fig. 1). This finding is similar to those of Chiou etal., who observed that albumin excretion was increased inUTI patients compared with controls with fever not causedby UTI [4]. Although we obtained a statistically significantdifference between our two groups, it should be noted thatour groups were small. Therefore, our findings must bevalidated in a larger study. If they are confirmed, a combinationof several urinary biomarkers may be useful in the evaluation offebrile children—u-CRP or u-RBP for differentiating betweenUTI and other conditions, and u-RBP or u-CC16 for thescreening of renal defects. Such a screening procedure wouldbe non-invasive and relatively rapid since u-CRP, u-RBP and u-CC16 can be analysed with commercially available ELISA kits.

The limitations of our study are that only one urinesample was collected from each child and that this samplewas not taken at a fixed time after presentation with feveror diagnosis of UTI. Furthermore, the lack of data on thekinetics of the urinary proteins in children with UTI and theprolonged storage of samples before analysis emphasizethe need for extended studies. According to the literature,the concentrations of LMWPs (such as RBP, CC16 andA1M) in urine tend to decrease with long-term storagewithout preservatives [14–15, 23, 24]. From our ownunpublished data, however, CC16 seems to be rather stableeven after 3 years of storage. It may be assumed that theprotein concentrations in our study have decreased some-what after 3 years of storage, but this is independent ofDMSA class and, therefore, cannot explain our findings.The true difference in protein concentrations betweencontrol children and children with UTI may even beunderestimated since the samples from 14 of the controlchildren were analysed within 1 month of collection, whenthere would be no expected loss of LMWPs.

In summary, the LMWPs u-CC16 and u-RBP showed astrong association with uptake defects visualized in renalDMSA scans. The levels of u-RBP and especially u-CRPwere significantly higher in children with UTI compared tochildren with fever of a non-UTI origin, although thegroups were small. If our findings are confirmed in largerstudies, a combination of these biomarkers may be useful inthe clinical assessment of children with UTI.

References

1. Majd M, Rushton HG (1992) Renal cortical scintigraphy in thediagnosis of acute pyelonephritis. Semin Nucl Med 22:98–111

2. Carter J, Tomson C, Stevens P, Lamb E (2006) Does urinary tractinfection cause proteinuria or microalbuminuria? Nephrol DialTransplant 21:3031–3037

Pediatr Nephrol (2009) 24:1533–1538 1537

3. Sandberg T, Cooper EH, Lidin-Janson G, Yu H (1985) Fever andproximal tubular function in acute pyelonephritis. Nephron 41:39–44

4. Chiou Y, Chiu N, Chen M, Cheng H (2001) Role of β 2-microglobulinuria and microalbuminuria in pediatric febrile urinarytract infection. Acta Paediatr Taiwan 42:84–89

5. Jantausch BA, Rifai N, Getson P, Akram S, Majd M, WiedermanBL (1994) Urinary N-acetyl-beta-glucosaminidase and beta-2-microglobulin in the diagnosis of urinary tract infection in febrileinfants. Pediatr Infect Dis J 13:294–299

6. Linné T, Fituri O, Escobar-Billing R, Karlsson A, Wikstad I,Aperia A, Tullus K (1994) Functional parameters and 99mtechne-tium-dimercaptosuccinic acid scan in acute pyelonephritis. PediatrNephrol 8:694–699

7. Tomlinson PA, Dalton RN, Hartley B, Haycock GB, Chantler C(1997) Low molecular weight protein excretion in glomerulardisease: a comparative analysis. Pediatr Nephrol 11:285–290

8. Everaert K, Raes A, Hoebeke P, Samijn W, Delanghe J, Van deWiele C, Van de Walle J (1998) Combined use of urinary α1-microglobulin and 99mTc DMSA scintigraphy in the diagnosis andfollow-up of acute pyelonephritis and cystitis in children. Eur Urol34:486–491

9. Stokland E, Hellström M, Jacobsson B, Jodal U, Sixt R (1998)Evaluation of DMSA scintigraphy and urography in assessingboth acute and permanent renal damage in children. Acta Radiol39:447–452

10. Swerkersson S, Jodal U, Sixt R, Stokland E, Hansson S (2007)Relationship among vesicoureteral reflux, urinary tract infectionand renal damage in children. J Urol 178:647–651

11. Jabs WJ, Lögering BA, Gerke P, Kreft B, Wolber EM, KlingerMHF, Fricke L, Steinhoff J (2003) The kidney as a second site ofhuman CRP formation in vivo. Eur J Immunol 33:152–161

12. Penders J, Delanghe JR (2004) Alpha 1-microglobulin: clinicallaboratory aspects and applications. Clin Chem Acta 346:107–118

13. Piepsz A, Colarinha P, Gordon I, Hahn K, Olivier P, Roca I, Sixt R,Van Velzen J (2001) Guidelines for 99mTc-DMSA scintigraphy inchildren. Eur J Nucl Med 28:BP37–41

14. Andersson L, Haraldsson B, Johansson C, Barregard L (2008)Methodological issues on the use of urinary alpha-1-microglobulinin epidemiological studies. Nephrol Dial Transplant 23:1252–1256

15. Andersson L, Lundberg PA, Barregard L (2007) Methodologicalaspects onmeasurement of Clara cell protein in urine as a biomarker forairway toxicity, compared with serum levels. J Appl Toxicol 27:60–66

16. Hornung RW, Reed LD (1990) Estimation of average concentra-tion in the presence of nondetectable values. Appl Occup EnvironHyg 5:46–51

17. Wilson AM, Ryan MC, Boyle AJ (2006) The novel role of CRP incardiovascular disease: risk marker or pathogen. Int J Cardiol106:291–297

18. Yasojima K, Schwab C, McGeer EG, McGeer PL (2001)Generation of C-reactive protein and complement components inatherosclerotic plaques. Am J Pathol 158:1039–1051

19. Ouchi N, Kihara S, Funahashi T, Nakamura T, Nishida M,Kumada M, Okamoto Y, Ohashi K, Nagaretani H, Kishida K,Nishizawa H, Maeda N, Kobayashi H, Hiraoka H, Matsuzawa Y(2003) Reciprocal association of C-reactive protein with adiponec-tin in blood stream and adipose tissue. Circulation 107:671–674

20. Foglar C, Lindsey RW (1998) C-reactive protein in orthopedics.Orthopedics 21:687–691

21. Laiho K, Tiitinen S, Teppo AM, Kauppi M, Kaarela K (1998)Serum C-reactive protein is rarely lost in patients with secondaryamyloidosis and proteinuria. Clin Rheumatol 17:234–235

22. Dillon SC, Taylor GM, Shah V (1998) Diagnostic value of urinaryretinol-binding protein in childhood nephritic syndrome. PediatrNephrol 12:643–647

23. Schultz CJ, Dalton RN, Turner C, Neil HA, Dunger DB (2000)Freezing method affects the concentration and variability of urineproteins and the interpretation of data on microalbuminuria.Diabet Med 17:7–14

24. Tencer J, Thysell H, Andersson K, Grubb A (1997) Long-termstability of albumin, protein HC, immunoglobulin G, kappa- andlambda-chain immunoreactivity, orosomucoid and alpha 1-antitrypsin in urine stored at −20°C. Scand J Urol Nephrol 31:67–71

1538 Pediatr Nephrol (2009) 24:1533–1538