Pediatric Nephrology

, Volume 26, Issue 8, pp 1263–1268

Urinary MMP-9/NGAL complex in children with acute cystitis

Authors

    • Department of PediatricsBakirkoy Dr. Sadi Konuk Training and Research Hospital
  • Esra Sevketoglu
    • Department of PediatricsBakirkoy Dr. Sadi Konuk Training and Research Hospital
  • Asuman Gedikbasi
    • Department of BiochemistryBakirkoy Dr. Sadi Konuk Training and Research Hospital
  • Alev Yilmaz
    • Department of Pediatric NephrologyBakirkoy Maternity and Children’s Hospital
  • Aysel Kiyak
    • Department of Pediatric NephrologyBakirkoy Maternity and Children’s Hospital
  • Mehmet Mulazimoglu
    • Department of Nuclear MedicineOkmeydani Training and Research Hospital
  • Gonul Aydogan
    • Department of PediatricsBakirkoy Maternity and Children’s Hospital
  • Tevfik Ozpacaci
    • Department of Nuclear MedicineOkmeydani Training and Research Hospital
Original Article

DOI: 10.1007/s00467-011-1856-3

Cite this article as:
Hatipoglu, S., Sevketoglu, E., Gedikbasi, A. et al. Pediatr Nephrol (2011) 26: 1263. doi:10.1007/s00467-011-1856-3

Abstract

The matrix metalloproteinase-9 (MMP-9) and neutrophil gelatinase associated lipocalin (NGAL) are shown to increase in an inflammatory situation. Based on our previous reports that NGAL can be detected in the urine of children with urinary tract infection (UTI), we also asked whether MMP-9/NGAL complex could be detected in the urine of children with UTI. This multicenter, prospective study was conducted between October 2009 and October 2010. Seventy-one patients with symptomatic culture proven UTI, 37 asymptomatic children with contaminated urine and 37 healthy children were recruited. Mean uMMP-9/NGAL/Cr levels were significantly higher in the UTI group than in the control group (p < 0.0001). According to ROC analysis, the optimal cut-off level was 0.08 ng/mg to predict UTI. Using a cut-off value, sensitivity and specificity were 98.6 and 97.3%, respectively. The mean levels of uMMP-9/NGAL/cr in the UTI group were also significantly higher than those in the contamination group (p < 0.0001). There was no statistically significant difference between contamination group and the control group (p = 0.21). The mean uMMP-9/NGAL/Cr in the UTI group were significantly higher before treatment than after treatment (p < 0.0001). The area under the curve was 0.997 (SE: 0.002, 95% CI: 0.993 to 1.001) for uMMP-9/NGAL/Cr. Urinary MMP-9/NGAL/Cr level was also correlated with positive urine nitrite test, positive urine leukocyte esterase reaction and renal scarring (p = 0.0001, p = 0.0001, p = 0.04, respectively) whereas was not correlated to leukocytosis and positive CRP level in serum. Urine MMP-9/NGAL/cr can be used as a diagnostic biomarker for UTI in children. Identification of NGAL-MMP-9/cr levels in the urine of suspected UTI patients may also be useful to differentiate between contamination and infection and for monitoring of treatment response in children.

Keywords

Urinary tract infectionChildrenCytokinesBacterial infections

Introduction

Urinary tract infection (UTI) is one of the most common infections of childhood. It is known to be associated with long-term consequences including renal scarring, hypertension, and chronic renal failure [1]. A urine culture result is the definitive standard for diagnosis of a UTI, but difficulty in specimen collection and interpretation of inadequately collected specimens may contribute to its misdiagnosis in children [2]. Furthermore, contamination of the urine sample is another problem in children that can cause false-positive results and lead to unnecessary treatment.

Matrix metalloproteinases (MMPs) are zinc-dependent endopeptidases known for their role in tissue remodeling through the degradation of extracellular matrix components. Typically, MMP expression and activity can be increased in any tissue injury and inflammatory disease process [3]. MMPs also promote re-epithelialization and restoration of epithelial barriers to bacteria or other invading pathogens. Several proinflammatory cytokines produced in response to bacterial infections, including TNF-α, IL-1, and granulocyte-macrophage colony-stimulating factor (GM-CSF), have been shown to upregulate monocyte and macrophage MMP production in vitro [4, 5]. Studies also have shown that bacterial products including lipopolysaccharide, phospholipase C, and chlamydial heat shock proteins induced MMP production in cell culture [69]. Systemic Escherichia coli infection, acute Lyme neuroborreliosis, and pneumococcal meningitis lead to secretion of significant amounts of MMP-9 [1012].

Neutrophil gelatinase associated lipocalin (NGAL) is a member of the lipocalin family that has been implicated in a variety of processes including inflammation, apoptosis, organogenesis, and is a promising biomarker for acute renal failure [13, 14]. NGAL biosynthesis is stimulated upon bacterial invasion; secreted NGAL then limits bacterial growth by sequestrating the iron-laden siderophore [15]. Neutrophil gelatinase associated lipocalin exists as monomer, homodimer and also occurs as a complex with human neutrophil gelatinase B (MMP-9) [16]. NGAL was first identified as a 25-kDa protein co-purified with MMP-9. The 125-kDa urinary MMP activity was identified as a complex of MMP-9 and NGAL. The MMP-9/NGAL complex (125 kDa) is one of several MMPs forms with high molecular weights, which independently serve as a predictor of the metastatic phenotype in patients with a variety of cancers [17]. Yan Li et al. demonstrated that NGAL is capable of protecting MMP-9 from auto-degradation and resulted in a significant increase in MMP-9 enzymatic activity in vitro. They suggest a potential regulatory role for NGAL in modulating MMP-9 activity. The elevated expression of NGAL from infiltrating tumor cells or from inflammatory cells might result in an increase in the local concentration of MMP-9 [18]. In our previous study, we demonstrated that urinary NGAL increased in children with UTI [19], we therefore hypothesized that NGAL may be involved in UTI via its interaction with MMP-9. Based on our previous reports that NGAL can be detected in the urine of children with UTI, we also asked whether MMP-9/NGAL complex could be detected in the urine of children with UTI and whether it might be predictive of disease status.

Material and methods

This multicenter prospective study was conducted at three government training and research hospitals in Istanbul over a 1-year period between October 2009 and October 2010. Seventy-one patients with symptomatic culture proven UTI (UTI group), 37 patients with asymptomatic contamination (contamination group) and 37 healthy children (control group) were enrolled in the study. This study was approved by the local ethics committee and written informed consent was obtained from all parents or legal guardians.

Urine samples were obtained by catheter from children with symptoms suggesting UTI including any of the following: fever, vomiting, dysuria, abdominal pain, and voiding frequency. Urinary tract infection was diagnosed by significant bacteriuria (≥10,000 CFU/ml) on urine culture [1] in symptomatic children and classified as UTI group. All children in the UTI group were treated as standard UTI protocol.

In patients suffering from recurrent UTI who have been followed-up at pediatric nephrology outpatient clinics, urine samples for culture and urinalysis were obtained at their routine visit by collecting bag or midstream urine. A contaminated urine culture was defined as a culture growing a single organism with <105 CFU/ml or a mix of two or more organisms. [20] In these patients, who were asymptomatic and had bacterial growth on routine culture, repeated urine sample for culture was obtained by catheter. If there was no growth at repeated urine sample, these asymptomatic patients were classified as the contamination group. No treatment was given to these patients.

Healthy children with no symptoms of UTI, urine sample for culture and urinalysis were obtained by collecting bag or midstream urine. These patients, whose urine cultures were negative, were classified as the control group.

Urine culture, urinalysis, serum urea and creatinine, C-reactive protein (CRP) and serum white blood cell (WBC) count measurements were obtained in groups at the time of presentation. Urinalysis including leukocyte esterase reaction, nitrite test, and microscopic analysis of urine were performed by Iris IQ 200 full automatic urine analyzer.

Random urine samples were obtained for measurement of MMP-9/NGAL and urine creatinine from UTI group before and after treatment. Single sample was obtained from contamination and control groups (no treatment was given to these groups). Urine samples immediately were centrifuged at 4°C for 15 min at 13,000 × g. Aliquots of urine supernatant were stored at –80°C for assaying. MMP-9/NGAL amount in urine was determined using the Quantikine Human MMP-9/NGAL complex enzyme-linked immunosorbent assay (ELISA) kit (Cat no: DM9L20), purchased from R&D Systems, Inc. (MN, USA) following the manufacturer’s instructions. Quantikine Human MMP-9/NGAL complex ELISA kit is an in vitro enzyme-linked immunosorbent assay for the quantitative measurement of human MMP-9/NGAL complex in urine. It contains natural human MMP-9/NGAL as the standard. The antibodies were raised against recombinant human MMP-9 and recombinant human NGAL. The mean recovery of natural human MMP-9/NGAL complex, from extraction to assay was 93% for urine. The intra-assay and inter-assay variations were 2.9 and 6.5%, respectively. This assay does not cross-react with either MMP9 or NGAL individually. Minimum detectable dose of MMP-9/NGAL complex ranged from 0.002–0.058 ng/ml. Results were expressed as pg/mg urinary creatinine.

Forty-seven children in the UTI group showed symptoms and laboratory data suggesting pyelonephritis including any of the following: fever, loin pain, high CRP level, was investigated by 99mTechnetium-dimercaptosuccinic acid (DMSA) scan within 7 days in an attempt to distinguish acute pyelonephritis from cystitis. The children in UTI group with normal scintigraphy were considered as having lower UTI. Scan features to suggest acute changes included focal, multifocal, or diffusely decreased or absent cortical uptake without cortical volume loss, in which the renal cortical contour remains intact considered as acute pyelonephritis. When scintigraphy revealed renal parenchymal lesions, a second scintigraphy was performed at least 6 months after the initial scan to evaluate the progression of renal lesions. On the follow-up scan we considered every persisting lesion as a scar in the absence of recurrent UTI [21]. Diagnosis of acute pyelonephritis was confirmed only in patients with totally or partially reversible lesions on follow-up DMSA scan.

Statistical calculations were performed with NCSS 2007 program for Windows. Besides standard descriptive statistical calculations (mean, standard deviation, median, and geometric mean), Kruskal–Wallis test was used in the comparison of groups, post hoc Dunn’s multiple comparison test was utilized in the comparison of subgroups, Mann–Whitney U test was used in the comparison of two groups, and Chi-square test was performed during the evaluation qualitative data. The results were evaluated within a 95% confidence interval. Receiver operating curve (ROC) analysis was performed to determine sensitivity and specificity of different cut-off points for uMMP-9/NGAL/Cr for the prediction of UTI. The most appropriate cut-off point was chosen according to ROC analysis and the area under the curve (AUC) was calculated. Statistical significance level was established at p < 0.05.

Results

The UTI group consisted of 71 children (15 male, 56 female) and the mean age was 7.23 ± 3.95 years. The contamination group consisted of 37 children (20 male, 17 female) and the mean age was 2.24 ± 2.98 years. Table 1 shows bacterial etiology of groups. The control group consisted of 37 children (13 male, 24 female) and the mean age was 6.37 ± 3.54 years. Serum urea and creatinine levels were normal in all groups. In the UTI group, DMSA scan was performed in 47 patients, 13 had renal scarring. Pyelonephritis was not detected in any patients in the UTI groups. In the contamination group, six patients had a DMSA scan performed within 3 months of which all were normal. Table 2 shows complaints of the patients at the time of presentation.
Table 1

Bacterial etiology of contamination and UTI* groups

Bacterial etiology

Contamination group, n (%)

UTIa group, n (%)

Escherichia coli

35

56

Klebsiella pneumoniae

3

7

Pseudomonas aeruginosa

5

5

Proteus mirabilis

1

Enterococcus

5

2

Staphylococcus

9

 

aUTI Urinary tract infection

Table 2

Complaints of the patients at the time of presentation

UTI group (n = 71)

Number (%)

Fever

20 (28.2)

Loin pain

1 (1.4)

Abdominal pain

15 (21.1)

Nausea/vomiting

19 (26.8)

Hematuria

1 (1.4)

Agitation

1 (1.4)

Dysuria

6 (8.5)

Voiding frequency

8 (11.3)

Urinary incontinence

7 (9.9)

Foul odor in urine

2 (2.8)

Prior to the treatment, mean uMMP-9/NGAL/Cr levels were significantly higher in the UTI group than in the control group (0.375 ng/mg vs. 0.006 ng/mg, p < 0.0001, Table 3). According to ROC analysis, the optimal cut-off level was 0.08 ng/mg for uMMP-9/NGAL/Cr to predict UTI. Using a cut-off of 0.08 ng/mg, had a sensitivity of 98.6% and a specificity of 97.3% for detecting UTI. The positive and negative predictive values of this cut-off point were 97.2 and 98.6%, respectively. The mean levels of uMMP-9/NGAL/Cr in the UTI group were also significantly higher than those in the contamination group (0.375 ng/mg vs. 0.003 ng/mg p < 0.0001, Fig. 1, Table 3). There was no statistically significant difference between the contamination group and control group (p = 0.21, Table 3). The mean uMMP-9/NGAL/Cr in the UTI group were significantly higher before treatment than after treatment (0.375 ng/mg vs. 0.01 ng/mg p < 0.0001, Table 3, Fig. 2). The area under curve was found to be 0.997 (SE: 0.002, 95% CI: 0.993 to 1.001) by ROC analysis (Fig. 3).
Table 3

Comparison of the mean MMP9/NGAL/cr levels of groups

Mean MMP9/NGAL/cr (ng/mg) (SD)

 

Control

UTI

Contamination

p

Before treatment

0.006 (0.023)

0.375 (0.187)

0.003(0.016)

0.0001

After treatment

0.010 (0.026)

0.827

p

0.0001

Control/UTI p = 0.0001, Contamination/UTI p = 0.0001, Control/contamination p = 0.21, SD Standard deviation

https://static-content.springer.com/image/art%3A10.1007%2Fs00467-011-1856-3/MediaObjects/467_2011_1856_Fig1_HTML.gif
Fig. 1

Urine MMP9/NGAL/cr levels of groups

https://static-content.springer.com/image/art%3A10.1007%2Fs00467-011-1856-3/MediaObjects/467_2011_1856_Fig2_HTML.gif
Fig. 2

Urine MMP9/NGAL/cr levels of children before and after treatment

https://static-content.springer.com/image/art%3A10.1007%2Fs00467-011-1856-3/MediaObjects/467_2011_1856_Fig3_HTML.gif
Fig. 3

Receiver operating curve for uMMP-9/NGAL/cr, area under curve (AUC) = 0.997 (SE: 0.002, 95% CI: 0.993 to 1.001)

Urinary MMP-9/NGAL/Cr levels were also correlated with positive urine nitrite test, positive urine leukocyte esterase reaction and renal scarring (p < 0.05), whereas, were not correlated to leukocytosis and positive CRP level in serum (Table 4).
Table 4

Comparison of MMP-9/NGAL/Cr levels with laboratory parameters

 

Urine MMP9/NGAL/cr (ng/mg) mean (SD)

P

CRP

<5 mg/l, n = 37

0.423 (0.191)

0.1

>5 mg/l, n = 22

0.447 (0.667)

DMSA

Normal, n = 34

0.286 (0.215)

0.04*

Scar, n = 13

0.398 (0.140)

Urine leukocyte esterase

Negative, n = 56

0.071 (0.166)

0.0001*

Positive, n = 106

0.408 (0.444)

Leukocytosis

Negative, n = 56

0.374 (0.202)

0.74

Positive, n = 7

0.357(0.159)

Urine nitrite

Negative, n = 82

0.141(0.227)

0.0001*

Positive, n = 42

0.365 (0.163)

MMP-9/NGAL/Cr Matrix metalloproteinase-9/Neutrophil gelatinase associated lipocalin/creatinine; SD standard deviation; CRP C-reactive protein, DMSA dimercaptosuccinic acid

Discussion

Matrix metalloproteinase-9 is produced by neutrophils and macrophages after different stimuli, like lipopolysaccharides [22]. Furthermore, glomerular epithelial cells, mesangial cells, tubular cells, and fibroblasts have been shown to produce MMP-9 after stimulation by cytokines and growth factors [2325]. Chromek et al. [26] reported that significant increase of MMP-9/cr and its main inhibitor tissue inhibitor of metalloproteinases-1 (TIMP-1)/cr in the urine of 40 children with acute pyelonephritis and renal scarring. No data is available about uMMP-9 levels in children with lower UTI in their study. In our study, in larger group of children, we confirmed that uMMP-9/NGAL/Cr levels increased in lower UTI patients. However, we could not determine the cut-off point for differentiate upper and lower UTI because none of the patients had acute pyelonephritis. Urinary MMP-9/NGAL/Cr levels significantly increased in patients with renal scarring (p < 0.05). Chromek et al. also observed that urinary MMP-9 was produced mainly by leukocytes on the second day from inoculation of bacteria [26]. In our study, uMMP-9/NGAL/Cr increased in children with UTI, but did not increase in children with contaminated urine samples. One possible explanation of our findings can be that the production of MMP-9 from leukocytes depends on the cytokine stimulation.

The accurate diagnosis of a UTI is necessary to ensure appropriate therapy for infected children and to avoid unnecessary antibiotic therapy for non-infected children. Bacterial contamination of urine culture specimens is almost inevitable especially in children who are not toilet trained. Contaminated urine samples can have important consequences including erroneous diagnosis or delay in the administration of appropriate antibiotic therapy. In the acutely ill child, one of the recommended sampling methods is suprapubic bladder aspiration, but this is relatively invasive and the rate of complications increases with multiple attempts. Other non-invasive methods, such as clean catch or bag specimens, are difficult to obtain, time-consuming to explain, and associated with high rates of contamination by fecal bacteria on the perineal skin. In our previous study, we demonstrated that urinary NGAL is increased in children with UTI and that it might serve as an early biomarker of infection [19]. One of the limitations of that study is that no data was available whether urinary NGAL level increased in contaminated urine. In the present study, we demonstrated that uMMP-9/NGAL/Cr levels were significantly higher in children with UTI but not elevated in children with contaminated urine. We also observed that uMMP-9/NGAL/Cr levels were not significantly different in contaminated urine samples and healthy control group urine samples. We therefore conclude that, uMMP-9/NGAL/Cr level may serve as a differentiation marker between infection and contamination.

Regardless of the urine collection techniques in children, the problem remains regarding the time required for urine culture. In most cases, complete identification of the bacteria is not available for 48–72 h. Because of the known consequences of the untreated UTI, urinalysis is commonly used for diagnosis of UTI and for the decision of whether empiric antibiotic treatment should be started. However, diagnostic value of urinalysis is an issue of debate. The nitrite test is based on the fact that bacteria reduce nitrates to nitrites in the urine. False-negative results are obtained if the bacteria present in the urine does not produce nitrites. Another indicator of UTI is leukocyte esterase reaction that is released by lysed leukocytes [27]. Using this test alone to detect pyuria would result in a large number of false-positive and some false-negative results. Arinzon et al. [28] reported that positive leukocyte esterase test and nitrite test are not sufficiently reliable in the assessment of patients with UTI. Thus, there is a need for a more rapid, specific, and sensitive laboratory diagnostic test to help guide the clinician in the decision of whether to begin empiric antibiotic therapy. In our previous study, using a cut-off 30 ng/mg, sensitivity and specificity of uNGAL/cr were 98 and 76%, for diagnosis of UTI, respectively [19]. Another study showed that sensitivity and specificity were 65.4 and 94% for urine leukocyte esterase test and were 38.9 and 99.5% for urine nitrite test, respectively [29]. In that determined cut-off point, sensitivity of uNGAL/Cr (98%) was found to be higher than in leukocyte esterase or the nitrite test, whereas specificity of uNGAL/Cr (76%) was lower than those tests. In the present study, determined cut-off points for diagnosis of UTI (0.08 ng/mg), sensitivity, and specificity of uMMP-9/NGAL/cr were 98.6 and 97.3%, respectively. Compared to uNGAL/cr alone, uMMP-9/NGAL/cr had better specificity and also compared to the leukocyte esterase or the nitrite test, had better sensitivity for diagnosis UTI in children. In the present study, our findings suggest that uMMP-9/NGAL/cr is an excellent indicator for UTI in children, with high sensitivity, specificity, and AUC value. Using optimal cut-off values, treatment can be started before urine culture results are obtained.

Another result of our study is that the uMMP-9/NGAL/Cr level in children with UTI was significantly decreased after adequate antibiotic therapy. It might also be used for monitoring the effectiveness of antibiotic therapy.

In conclusion, to our knowledge, this is the first study demonstrating that uMMP-9/NGAL/Cr is increased in children with UTI and that may be used as diagnostic biomarker for UTI in children with normal renal function and to decide whether empirical antibiotic therapy should be started pending culture results. Identification of NGAL-MMP-9/cr levels in the urine of suspected UTI patients may also be useful to differentiate between contamination and infection and for monitoring of treatment response in children.

Copyright information

© IPNA 2011