Prenatal urinary matrix metalloproteinase profiling as a potential diagnostic tool in fetal obstructive uropathy

Urinary extracellular matrix proteins as predictors of the severity of ureteropelvic junction obstruction in children

INTRODUCTION

Ureteropelvic junction obstruction (UPJO) particularly represents a challenge in regard to management, as not all hydronephrosis (HN) represent a kidney-damaging state. Urinary biomarkers have been proposed as noninvasive tools. Extracellular matrix (ECM) proteins are associated with tissue fibrosis in the setting of UPJO and is poorly explored.

OBJECTIVE

To investigate whether urinary ECM proteins are useful to discriminate the severity of urinary obstruction on unilateral UPJO.

STUDY DESIGN

Children with unilateral UPJO were prospective enrolled. Urinary (u) matrix metalloproteinases (MMP-1,-2,-9 and TIMP-1,-2) as well as clinical characteristics were measured in the following groups: 26 children with obstructive HN at initial diagnosis and after six months of dismembered pyeloplasty; 22 children with non-obstructive HN at diagnosis and after six months of observation; 26 children without any urinary tract condition, as the control group. Results were assessed statistically using for homogenous groups, a one-way analysis of variances (ANOVA) and for nonparametric groups, Mann-Whitney test or Kruskal-Wallis test was performed. ROC curves were performed.

RESULTS

Baseline samples demonstrated a higher concentration of uMMP-1/Cr, uMMP-2/Cr, u-TIMP-1/Cr and u-TIMP-2/Cr in obstructive HN group; uMMP-9/Cr levels were higher in non-obstructive HN group and all studied biomarkers had lower concentrations for the control group. On follow-up, for the obstructive HN group, urinary concentration of uTIMP-1/Cr and uTIMP-2/Cr decreased, and uMMP-1/Cr, uMMP-2/Cr and uMMP-9/Cr increased when comparing preoperative to postoperative values. In the non-obstructive HN group, all proteins analyzed were stable after six months of observation. ROC curves analysis showed a promising diagnostic profile for the detection of obstructive HN for uTIMP-1/Cr (area under the curve -AUC-; of 0.692), uTIMP-2/Cr (AUC of 0.678) and for uMMP-2/Cr (AUC of 0.655).

DISCUSSION

The severity of kidney obstruction could correlate with the urinary ECM proteins concentration in this study. This is concordant with prior studies demonstrating that a disruption of the balance of accumulation/degradation of the ECM proteins occur on obstructive uropathy. Limitations of our study include the older age of our patients and that these markers had no influence at all on the surgical decision.

CONCLUSION

We demonstrate that obstructive HN have significantly higher uMMP-2, uTIMP-1 and uTIMP-2 concentrations. Particularly, uTIMP-2 levels were correlated to severity of obstruction and therefore, it might be a useful urinary biomarker to correctly allocate children with HN between surgical management vs follow-up. After pyeloplasty, uTIMP-1 and uTIMP-2 presented a progressive decrease postoperatively, which is also highly desirable for urinary markers.

The role of urinary TIMP1 and MMP9 levels in predicting vesicoureteral reflux in neonates with antenatal hydronephrosis

BACKGROUND

Antenatal hydronephrosis (AH) is commonly found on evaluation of pregnant women and 20-30 % of neonates have vesicoureteral reflux (VUR). In order to diagnose VUR, we required invasive testing and exposure of the neonate to radiation. The concentrations of a matrix metalloproteinase, MMP9, and its inhibitor TIMP1, were analyzed in hydronephrotic newborns with VUR and were compared to those without reflux.

METHODS

The neonates with a history of prenatal hydronephrosis were enrolled in two groups based on imaging study results, the neonates with VUR and without VUR. Neonates with a normal prenatal history and postnatal ultrasound were placed in a third group. We measured the random urinary levels of MMP9, TIMP1, and creatinine, their cut-off values and the MMP9/Cr and MMP9/TIMP1/Cr ratio was calculated, and an ROC curve was drawn.

RESULTS

Sixty-nine neonates were enrolled in three groups; 27 patients (20 male, seven female) with AH and VUR were in group 1, 23 neonates (19 male, four female) without VUR were placed in group 2, and 19 (15 male, four female) acted as controls in group 3. The differences between the three groups and the normal and total hydronephrotic groups were statistically significant for MMP9, the MMP9/Cr, MMP9/TIMP1, and MMP9/TIMP1/Cr ratios. The urinary TIMP1 and TIMP1/Cr ratios were not significantly different between the groups. A cut-off value of MMP9 was measured as 358.5 ng/ml (sensitivity [sens] 74 %, specificity [spec] 78 %) and was used to compare groups 1 and 2. For groups 2 and 3, this cut-off was 181.00 pg/ml (sens 91 %, spec 89 %). The cut-off values measured for the MMP9/TIMP1 ratio were 30.32 (sens 70 %, spec 61 %) and 9.85 (sens 96 %, spec 89 %) to compare groups 1 and 2, and 2 and 3, respectively. We found no valuable cut-offs for the TIMP1 and TIMP1/Cr values. There was no difference between neonates with mild, moderate, and severe VUR according to urinary biomarker concentrations.

CONCLUSIONS

Evaluation of urinary levels of MMP9, or the MMP9/Cr, MMP9/TIMP1, or MMP9/TIMP1/Cr ratios may help us to differentiate the newborns with hydronephrosis and VUR from those without reflux.