Role of urinary N-acetyl-beta-D-glucosaminidase in predicting the prognosis of antenatal hydronephrosis

Antenatal hydronephrosis (AH) is a commonly observed finding in prenatal ultrasonography. It is observed in about 0.5% to 1% of pregnancies after the second trimester [1, 2]. Among the various causes of hydronephrosis, the most common one is ureteropelvic junction (UPJ) obstruction [3]. Congenital obstructive nephropathy caused by UPJ obstruction is an important cause of permanent renal insufficiency in infants and children [4]. Most infants with AH do not require additional treatment other than close follow-up [5]. However, immediate surgery is required for infants whose renal function is greatly reduced due to UPJ obstruction [6].

To date, the most reliable methods for the assessment of obstructive nephropathy are renal scintigraphy modalities such as ^99mTc-mercaptoacetyltriglycine (MAG3), ^99mTc-diethylenetriaminepentaacetic acid (DTPA) scans [5, 7]. However, it is difficult to define a true obtrusion that requires immediate intervention based on these imaging methods alone [8]. Moreover, the limitations of renal scintigraphy are that it is expensive, involves radiation exposure, and requires intravenous injections and sedation. For the above reasons, other diagnostic methods that are relatively less invasive but sensitive to obstructive nephropathy have been investigated, and a urine biomarker emerged as a good alternative [9, 10].

Theoretically, urinary biomarkers are known to be able to diagnose renal damage caused by obstruction at an early stage [11]. N-acetyl-beta-D-glucosaminidase (NAG) has been proved to be a sensitive marker for acute tubular damage [12]. NAG is a lysosomal enzyme that is abundantly present in the cells of the proximal tubule. NAG has a relative high molecular weight that does not permit its filtration through the glomerular basal membrane [13]. Thus, urinary NAG originates primarily from the proximal tubule, and increased urinary concentration is a consequence of renal tubular damage.

Conventional methods may not be sufficient to decide on surgery in infants with severe hydronephrosis detected in the antenatal period. Considering this situation, this study aimed to investigate whether urinary NAG level can play an important role in determining the prognosis of AH, especially in determining surgery in severe hydronephrosis.

This study was approved by the Institutional Review Board (IRB) of Seoul National University Hospital (IRB number: H-2403-067-1519). Informed consent was waived owing to the retrospective nature of the study design. The study was performed in accordance with applicable laws and regulations, good clinical practices, and ethical principles, as described in the Declaration of Helsinki.

Following the approval of IRB, we retrospectively analyzed infants diagnosed with AH from January 2019 to December 2021 at our hospital. AH was defined as an anteroposterior pelvic diameter (APD) of >7 mm after the 36th gestational week on an ultrasonography. Initial postnatal ultrasonography was performed 3–7 days or 4–8 weeks after birth depending on the severity of AH. The severity of hydronephrosis was assessed according to the Society of Fetal Urology (SFU) criteria and is presented as grade 1 to 4. Among infants under 2 years old diagnosed with AH, only unilateral hydronephrosis with an SFU grade of 3 or higher on initial ultrasonography was included in this study. Those with bilateral grade 3 or higher hydronephrosis and other urinary abnormalities such as single kidney, duplex system, megaureter, posterior urethral valve, and confirmed vesicoureteral reflux were excluded from the study (Fig. 1).

Fig. 1
Study flow diagram. HN, hydronephrosis; PUV, posterior urethral valve; VUR, vesicoureteral reflux.

• Click for larger image
• Download as PowerPoint slide

In all patients, ultrasonography was performed at intervals of 1 to 6 months depending on the severity of hydronephrosis. Changes in APD and parenchymal thickness (PT) were measured with ultrasonography. Patients who had persistent significant hydronephrosis were evaluated by diuretic renal scintigraphy with ^99mTc-MAG3 for assessment of the presence of significant obstruction and the severity of renal function impairment. A differential renal function (DRF) of less than 40% was considered functional impairment of the affected renal unit.

We performed surgery in patients whose DRF decreased less than 40%, who had progressively worsening hydronephrosis on ultrasonography, or who had recurrent febrile urinary tract infections. The worsening of hydronephrosis was determined by an increase in APD or a decrease in PT within the same SFU grade. In this study, patients were categorized into two groups: group 1 (laparoscopic pyeloplasty [LP] group) included infants who had severe hydronephrosis and underwent surgical treatment; group 2 (active surveillance [AS] group) included infants with less severe hydronephrosis with preserved renal function who did not need any intervention. In the surgery group, laparoscopic dismembered pyeloplasty (Anderson–Hynes technique) was performed in all patients with a single surgeon. A ureteral stent was inserted while performing LP, and was removed 4 weeks after surgery. There were no complications related with operation and ureteral stent in all patients.

All patients were assessed for urinary levels of NAG and Cr. In clinical practice, whenever a patient came to the hospital for an ultrasonography, we performed a urinalysis and urinary NAG sampling and then normalized the results to urinary Cr values to adjust the daily changes in the urinary excretion of markers. The values were expressed in IU/g Cr. We compared and analyzed the results immediately before surgery in the LP group and the results when hydronephrosis was most severe in the AS group during follow-up of at least 1 year.

Continuous variables were expressed as mean±standard deviation. Statistical analysis of the differences between groups with normal distributions was determined by using the t-test or Fisher’s exact test to compare the two groups. Nonparametric tests such as the Mann–Whitney test were used for variables that were not distributed normally. A receiver operating characteristic (ROC) curve was drawn and the sensitivity and specificity of the different cut-off points for APD, PT and urinary NAG/Cr were determined. The best cut-off point was chosen according to the ROC curve. A p-value of <0.05 was considered statistically significant. All statistical analyses were performed using IBM SPSS Statistics software version 25.0.0.2 (IBM Corp.).

Twenty-one infants underwent LP for UPJ obstruction, and 14 infants underwent AS. Out of a total of 35 patients, there were 30 boys. The mean age at surgery in the LP group was 3.7 months (range, 1.7–7.5 months). Meanwhile, in the AS group, the mean age when hydronephrosis was most severe was 5.2 months (range, 0.5–21.5 months). Age, laterality and gender difference did not differ between the two groups. In the LP group, the mean value of APD at surgery was 30.0 mm (range, 15.0–49.0 mm) and the mean value of PT was 2.6 mm (range, 1.2–3.7 mm), and these showed a statistically significant difference from the AS group. The mean value of DRF in the LP group was 46.7% (range, 26.9%–55.9%). Most patients in the AS group did not undergo MAG3 diuretic renal scintigraphy. The LP group had a significantly higher mean urinary NAG/Cr ratio at the time of operation than that when hydronephrosis was most severe in the AS group (26.1 IU/g, p=0.003) (Table 1).

Table 1
Comparison of demographic and imaging findings, and urine NAG/Cr ratios between patients diagnosed with antenatal hydronephrosis who underwent surgery and those who did not

• Click for larger image
• Click for full table
• Download as Excel file

When comparing the preoperative and postoperative 1 year results, SFU and APD decreased and PT increased significantly in all patients. The urinary NAG/Cr ratio also decreased statistically significantly from 26.1 to 10.4 IU/g (p=0.023). However, in the case of mean DRF, there was no significant change from 46.7 to 46.8 at postoperative 1 year. This is because patients whose DRF was significantly reduced before surgery showed no significant change after surgery, and in some patients, DRF actually decreased after surgery (Table 2). This underscores the inability to recover renal function once irreversibly damaged, highlighting the importance of resolving obstruction before renal function deteriorates. Additionally, as DRF is a relative function, the growth rate of the unaffected kidney after surgery may exceed that of the affected side, leading to further DRF decrease on the affected side.

Table 2
Changes in preoperative and postoperative imaging study findings and urine NAG/Cr ratios in patients who underwent laparoscopic pyeloplasty

• Click for larger image
• Click for full table
• Download as Excel file

The sensitivity and specificity of the APD, PT, and urinary NAG/Cr ratio were calculated from the ROC curves (Fig. 2). The LP group and AS group were compared to find the cut-off value to distinguish infants with severe hydronephrosis from those without an indication for surgical intervention. For comparison of the two groups, the cut-off values were measured as follows: 21.0 mm (sensitivity, 95.0%; specificity, 93.3%) for APD, 3.7 mm (sensitivity, 100.0%; specificity, 86.7%) for PT, and 15.5 IU/g (sensitivity, 85.0%; specificity, 93.3%) for NAG/Cr ratio (Table 3).

Fig. 2
Receiver operating characteristic curve to decide surgery (differentiate laparoscopic pyeloplasty group from active surveillance group). (A) For anteroposterior diameter, area under the curve (AUC)=0.972. (B) For parenchymal thickness, AUC=0.983. (C) For N-acetyl-beta-D-glucosaminidase to creatinine ratio, AUC=0.955.

• Click for larger image
• Download as PowerPoint slide

Table 3
Cutoff value, sensitivity, specificity, PPV, NPV and AUC of APD, PT, and NAG/Cr ratios for decision of surgery in infants with antenatal hydronephrosis

• Click for larger image
• Click for full table
• Download as Excel file

Urinary tract obstruction in the perinatal period poses a significant challenge for clinicians due to the risk of permanent renal damage, necessitating surgical intervention within the first year of life [6, 14]. Traditionally, the assessment of renal function, particularly DRF, relies on nuclear scintigraphy. However, these radioisotope studies are relatively invasive. Recognizing the need for less invasive yet sensitive diagnostic methods for obstructive nephropathy, researchers have explored alternative approaches, with urine biomarkers emerging as promising alternatives [11, 15]. Urine biomarkers offer several advantages: they are non-invasive; collecting urine samples is minimally invasive compared to imaging studies, making it suitable for repeated measurements and monitoring. They are sensitive to renal injury; biomarkers reflect specific pathophysiological processes occurring in the kidneys, which provides insights into renal function and injury. Finally, early detection is possible; biomarkers may detect renal injury at earlier stages than imaging studies, allowing for timely intervention to prevent irreversible damage. The cellular effects of urinary obstruction involve leukocyte migration and interstitial infiltration, vasogenic effects and ischemia, and cellular death and apoptosis. Specific biomarkers have been identified for each of these pathways. Among these, NAG is involved in the cellular stage [16].

Huang et al. [17] investigated the clinical utility of urinary NAG levels in evaluating the degree of damage to hydronephrotic kidneys in children. They found that urinary NAG levels were significantly higher in hydronephrotic kidneys compared to control group. Moreover, there was a significant positive correlation between urinary NAG levels and pathologic grades of hydronephrosis, suggesting that NAG levels could serve as a useful biomarker for assessing renal damage in this population. Similarly, Skalova et al. [18] studied children with hydronephrosis diagnosed by abdominal ultrasonography, either prenatally or postnatally. They observed significantly higher urinary NAG/Cr ratios in patients with hydronephrosis compared to reference data. Interestingly, they found no correlation between the urinary NAG/Cr level and the severity of obstruction. On the other hand, our study showed that there was a significant difference in the urinary NAG/Cr ratios depending on the severity of hydronephrosis (LP group 26.1 IU/g vs. AS group 11.1 IU/g, p=0.003). Additionally, when obstruction was resolved after surgery, mean NAG/Cr ratios were found to significantly decrease (26.1 IU/g to 10.4 IU/g, p=0.023). The absence of a correlation between urine NAG/Cr ratios and the severity of hydronephrosis in the study by Skalova et al. [18] may be attributed to the inclusion of some patients over 2 years old in their study cohort. In contrast, our study specifically targeted infants under 2 years old at the time of admission to the hospital. By focusing on this younger age group, we aimed to minimize potential confounding factors related to age-related variations in renal function and biomarker levels. The susceptibility to renal function deterioration can be attributed to incomplete kidney maturation during early childhood, as documented by Lythgoe et al. [19], who suggested that renal maturation occurs predominantly within the first year after birth. Therefore, patients under 2 years old are vulnerable to renal injury caused by obstructive nephropathy, and urinary NAG/Cr ratios are expected to vary significantly depending on the severity of obstruction. Conversely, the prognosis for kidney recovery in infants under 2 years old following the resolution of UPJ obstruction is generally favorable, highlighting the importance of early detection and immediate intervention.

In addition to differences in the urinary NAG/Cr ratios, we also observed significant disparities in ultrasonographic findings such as APD and PT between the LP and AS groups (Table 1). This result is inevitable because the indications for surgery include not only a decrease in DRF but also worsening hydronephrosis on ultrasonography. These ultrasonographic parameters (APD and PT) are commonly used to assess the severity of hydronephrosis and provide valuable information about renal morphology. The significant differences in both urinary NAG/Cr ratio and ultrasonographic findings between the LP and AS groups underscore the complex interplay between biochemical markers and imaging modalities in the evaluation of hydronephrosis. While urine biomarkers such as NAG provide insights into renal tubular function and injury, ultrasonography remains essential for visualizing structural abnormalities and assessing the degree of obstruction. Integrating information from multiple diagnostic modalities enhances the accuracy of diagnosis and informs treatment decisions, particularly in cases where the severity of hydronephrosis may not be evident from imaging alone.

Our study has several limitations. First, we only compared urine NAG levels at a specific point in time between the LP and AS groups. Considering study on normal NAG levels in healthy children, it’s noted that NAG levels are highest in infants under 1 year old [20]. Thus, younger infants may exhibit higher normal levels due to renal tubule immaturity. Therefore, it’s more crucial to assess serial changes within a specific cohort rather than comparing two groups. The primary aim of this study was to assess if urine NAG levels could replace nuclear scintigraphy in evaluating renal function. However, a notable limitation is the lack of analysis regarding the correlation between urine NAG levels and renal function assessed by nuclear scintigraphy. Since nuclear scintigraphy was not conducted in AS group, determining the precise correlation between urine NAG and nuclear scintigraphy proves challenging. Nonetheless, in the LP group, despite the absence of significant changes in the average DRF before and after surgery, a notable decrease in urine NAG levels was observed. This suggests that urine biomarkers may offer greater sensitivity and accuracy compared to nuclear scintigraphy.

Additional limitation is the small sample size. A larger sample size would enhance the statistical power of our analysis and allow for more robust conclusions to be drawn. The last limitation to discuss is the variability in the timing of urinary NAG measurements among patients. Differences in urinary NAG concentration may arise depending on the feeding status, particularly in infants. Correcting for Cr levels may not entirely eliminate this variability, as urinary Cr concentrations can also fluctuate. Therefore, the accuracy of NAG measurements may be compromised, necessitating repeated assessments to improve reliability. Future prospective studies are necessary to address the limitations discussed above.

One of the biomarkers of acute renal injury, urinary NAG, was closely related to the degree of hydronephrosis. This biomarker offers the potential to supplement or even replace traditional imaging studies in the assessment of obstructive nephropathy. We suggest that the urinary NAG level may be useful to determine whether to perform surgery on the UPJ obstruction and to predict the prognosis.