Outcome of Patients with Antenatally Diagnosed hydronephrosis with Respect to Postnatal Diagnosis and Need for Surgical Intervention.

Aims: This study aims to determine the etiology of antenatal hydronephrosis (ANH) and predict need for surgical intervention based on antenatal renal pelvis anteroposterior diameter (APD). Materials and
Methods: Combined prospective and retrospective study (2012-2018) of ANH cases with postnatal follow-up. Surgical intervention was correlated with the degree of hydronephrosis (HDN) and pelvis APD measured at the 2nd trimester, 3rd trimester, and postnatal 6-week follow-up.
Results: One hundred and sixty-five patients were studied with a total of 219 ANH units. Transient HDN was seen in 116 units. Surgical intervention included pyeloplasty (n = 76), sub ureteric dextranomer injection (n = 8) and nephrectomy (n = 1). Chances of requiring surgery based on the degree of HDN in 2nd and 3rd trimester respectively were mild: 11.32% and 9.52%, moderate: 34.21% and 37.03% and severe: 85.71% and 86.27%. The mean increase in APD between the 2nd and 3rd trimester (n = 50) was lesser in conservatively managed (3.548 ± 4.219 mm) than surgically managed (8.261 ± 5.857 mm) patients (P = 0.002). In another subset (n = 37), the mean increase in APD between the 2nd trimester and postnatal period was less in conservatively managed (1.432 ± 0.612 mm) (P = 0.088) than surgically managed patients (12.91 ± 3.247 mm) (P = 0.004). The area under the receiver operating characteristic (ROC) curve showed that an APD of 8.2 mm in the 2nd trimester and 12.85 mm in the 3rd trimester correlated with the requirement of postnatal surgery.
Conclusion: Apart from the degree of HDN, significant changes in APD between 2nd and 3rd trimester and cut off value suggested by the ROC curve will help during antenatal counseling with regard to requirement of postnatal surgery. Copyright:

INTRODUCTION

Hydronephrosis (HDN) is defined as aseptic dilatation of the renal collecting system. Structural fetal anomalies detected by antenatal ultrasonography (USG) are seen in approximately 1% of pregnancies. Genitourinary anomalies are seen in 20%–50% with 50% of these presenting as HDN.[123]

Prenatal diagnosis of HDN helps in early follow-up after birth and can reduce the risks of developing urinary tract infection (UTI), pyelonephritis, hypertension, growth failure, and even renal failure.[12] A parameter commonly used to classify the severity is anteroposterior diameter (APD) of pelvis which helps in prognostication during pregnancy.[4] While similar studies are available in the literature, the cutoff values suggesting surgical intervention after birth are variable.[5678] Often, the exact postnatal etiology is not known as the study is not conducted in a pediatric surgical center. This study aimed to specifically compare the pelvis diameter in conservatively and surgically managed patients in a pediatric surgical setting and add to the existing knowledge.

MATERIALS AND METHODS

This was a combined retrospective (2012–2016) and prospective (2017–2018) study of all cases of antenatal hydronephrosis (ANH) referred to the Pediatric Surgery Department of a tertiary care center. Data maintained prospectively in a standard format for all antenatal and urological cases in our unit for many years were retrieved. Even though part of the data is “retrospective” all data collection was prospectively collected and management done using the same protocol.

Only those patients whose antenatal and postnatal USGs were available and a final etiological diagnosis could be made were included. Patients who had antenatal USG done either in the 2nd or 3rd trimester or both were included. Antenatal unilateral and bilateral HDN as well as hydroureteronephrosis (HDUN) were included. Cases of intrauterine demise, where medical termination of pregnancy was advised for oligohydramnios or multiple congenital anomalies, and those lost to postnatal follow-up, were excluded. Posterior urethral valve (PUV) cases were excluded from final analysis for the following reasons: (a) There is a strong suspicion of PUV on antenatal USG in the presence of distended bladder/thickened bladder wall; bilateral hydro ureteronephrosis and oligohydramnios.(b) All cases need surgical intervention.(c) If detected early, there are higher chances of termination of pregnancy when prognosis is explained to parents.

Where antenatal counseling was first done in the 2nd trimester, we repeat the scan once in the third trimester. Those with bilateral HDN and findings suggestive of lower urinary tract obstruction were monitored more frequently at an interval of 4–6 weeks. Based on pelvis APD, the degree of ANH in the 2nd trimester was classified as mild: 4 mm to <7 mm; moderate: 7 mm to ≤10 mm and severe: >10 mm and that in the 3rd trimester as mild: 7 mm to <9 mm; moderate: 9 mm to ≤15 mm; and severe: >15 mm.[8]

USG was performed at the end of the 1st week, 6 weeks age, and 3 months after birth. Based on pelvis APD, postnatal HDN was categorized as: Mild (6 mm to <10 mm), moderate (10 mm to <15 mm) and severe (>15 mm). Micturating cystourethrogram (MCUG) was done at 6 weeks of life in unilateral or bilateral HDN with APD >10 mm or associated ureteric dilatation, or if there was a history of UTI. Ethylene dicysteine (EC) diuretic renography scan (F0 protocol) was done in all patients with moderate to severe HDN (APD >10 mm) at 4–6 weeks age. There is a standard protocol for performing MCU and EC scan in our institute and there was no change in this protocol during the entire period of study.

Patients were planned for pyeloplasty based on antenatal sonography, postnatal clinical findings, and imaging. When any of the postnatal imaging was equivocal, a wait and watch policy was applied and investigations repeated at 3–6-month interval. In USG KUB, the following parameters were considered for diagnosis of ureteropelvic junction obstruction (UPJO): Renal pelvis anteroposterior diameter (≥2 cms, or 1–2 cms with other parameters mentioned below), abrupt cut off at the UPJ, nonvisualization of ureter, moderate to severe calyceal separation, and reduced renal parenchymal thickness.

All patients underwent EC renal dynamic scan with frusemide being injected intravenously along with the tracer (F 0 protocol). A continuously rising renogram pattern with no clearance of the tracer during the initial 25-min dynamic study, with subsequent retention of tracer in the pelvicalyceal system in the pre- and post-void static images up to 3 h and in compromised renal cortical function a 5-h delayed image was taken as obstructed drainage. A fall of 5% differential function of the affected kidney during follow-up was also considered significant. Differential renal function more than 40% was not criteria for conservative management.

IVP was done in all cases to further confirm the diagnosis. The persistence of the contrast in the pelvicalyceal system 2 h after its injection and its further retention even 20 min after Lasix injection was the criteria used to diagnose UPJO.

In patients with vesicoureteral reflux (VUR) and ANH in general, the norm was to start uroprophylaxis soon after birth. All patients were managed conservatively initially with repeat MCUG done after 6 months. Patients with persistent grade 3–5 VUR were counseled for dextranomer hyaluronic acid sub ureteric injection in the presence of more than 2 breakthrough febrile UTIs in a 6-month period, dimercaptosuccinic acid (DMSA) scan showing pyelonephritic changes or scars, in the presence of a single kidney, or raised creatinine levels. Those who could not afford the cost of dextranomer injection were counseled for ureteric reimplantation at or after 1 year of age especially in the presence of grade IV-V reflux.

In cases of partial vesicoureteric junction obstruction (VUJO) and megaureter, no surgical intervention was considered if there was a reduction in dilatation on USG KUB, negative urine cultures and improvement in drainage on serial EC scans first done after 6 months and then at yearly intervals. Those with complete VUJO would be advised ureteric reimplantation. In patients with multicystic dysplastic kidney (MCDK) serial follow up with USG for cyst size, and blood pressure monitoring was done after initial DMSA scan.

For the study protocol, patients were followed up for a minimum of 3 months after birth or until the diagnosis was established. They were otherwise indefinitely followed up for the treatment purposes. The incidence of spontaneous resolution of HDN and the time frame in which this happened was studied. Change in APD of pelvis of the same fetus in second trimester, third trimester and postnatal 6 weeks was analyzed. The chances of requiring surgical intervention postnatally were compared with the severity of ANH in the second and third trimester.

Statistical analysis

The statistical analysis was carried out using Statistical Package for Social Sciences (SPSS Inc., Chicago, IL, USA version 15.0 for Windows). All quantitative variables were estimated using measures of central location (mean, median) and measures of dispersion (standard deviation and standard error). Normality of data was checked by measures of skewness and Kolmogorov–Smirnov tests of normality. For normally distributed data means were compared using Student's t-test for groups. For skewed data Mann–Whitney test was applied for group. Qualitative or categorical variables were described as frequencies and proportions and proportions compared using Chi-square or Fisher's exact test whichever was applicable. For time-related variables, repeated measure analysis of variance (ANOVA) was applied followed by one-way ANOVA for normally distributed data or Wilcoxon Signed-Rank test for skewed data. All statistical tests were two-sided and performed at a significance level of α =0.05. Mean APD of pelvis, paired correlation and paired differences were compared between conservatively and surgically managed patients where both second and third trimester scans were available. The area under receiver operating characteristic (ROC) curve which reflects the probability of dimensions of the APD pelvis and its correlation with postnatal surgery was calculated for the second and third trimester.

RESULTS

A total of 165 patients were studied (130 male; 35 females; M: F ratio, 3.7:1). The mean gestational age at diagnosis of ANH was 27.10 ± 5.950 weeks (17 weeks–40 weeks).

There were a total of 219 renal units for analysis with unilateral HDN in 111 and bilateral in 54 [Figure 1]. One hundred and sixteen units were categorized as transient HDN. Complete resolution of HDN was seen in 35 units which occurred in utero in the third trimester (n = 1, 0.9%), and postnatally within the 1st month (n = 4, 3.96%), within 6 months (n = 9, 8.91%), and between 6 and 24 months age (n = 21, 20.79%). There was partial resolution in another 81 (69.82%) units who were also managed conservatively.

Figure 1

Algorithm showing final diagnosis and management of study patients

All 71 patients with UPJO (76 units: Left 38, right 28, bilateral 5) underwent Anderson Hynes dismembered pyeloplasty at a mean age of 5.6 months (range 15 days–48 months). Initial bilateral percutaneous nephrostomy was done in four patients, two for bilateral pyonephrosis at the age of 15 days and 1 month and two for abdominal mass and gross HDN at the age of 1 month and 2 months.

MCU was done in 95 (57.57%) of 165 patients. VUR was seen in 25 patients (27 units).

In concomitant ipsilateral VUR and UPJO, the antenatal scan had shown unilateral HDN in 13, and unilateral HDUN in 2 cases. Grade of VUR seen was grade 2 in 2, grade 3 in 5, grade 4 in 2, and grade 5 in 6. In isolated VUR (Left 3, Right 4, Bilateral 2), antenatal scans had shown unilateral HDN (n = 6), bilateral HDN (n = 2) and unilateral HDUN (n = 1). Grade of VUR seen were grade 2 in 2, grade 3 in 6, grade 4 in 2 and grade 5 in 1. Cystoscopic injection of dextranomer hyaluronic acid polymer was successful in the 7 units where it was given while the rest were managed conservatively for the VUR.

Antenatal HDUN was noted in nine patients (R 6, L 3), with postnatal final diagnosis of partial VUJO (n = 4), VUR (n = 3), Megaureter (n = 1) and MCDK with VUR (n = 1). In partial VUJO patients (n = 7), antenatal scans had shown 4 with HDUN, and 3 with HDN. Three were diagnosed in the 2nd trimester and 4 in the 3rd trimester. Mean APD in second and third trimester was 8.96 ± 1.8056 and 11.46 ± 4.150 mm. Postnatal follow-up EC scans and USG showed resolution of HDUN with time.

All five patients with MCDK had cystic dilatation of kidney on antenatal scan with 1 having HDUN. Four were diagnosed in the 2nd trimester and one in the 3rd trimester. All showed reduction in size of the cysts with time on serial USGs. In MCDK, the importance of APD measurement reduces as cysts do not communicate with each other or the pelvis and therefore these cases were not considered during final analysis.

Of the two patients with megaureter, antenatal scan had shown HDUN in one in the 2nd trimester and HDN in the other in the 3rd trimester. Follow-up EC scans showed no deterioration in function with slow improvement in drainage. They were kept on uroprophylaxis and did not have any breakthrough UTIs.

Postnatal management based on the degree of HDN correlated similarly in 2nd and 3rd trimester [Table 1]. Mean APD in the second trimester, third trimester and postnatal 6 weeks with all available data in conservatively managed patients (n = 138 units) was 6.42 ± 2.19 mm, 9.75 ± 4.58 mm and 8.114 ± 3.1992 and in surgically managed patients (n = 81 units) was 12.15 ± 5.65 mm, 20.86 ± 8.46 mm and 22.448 ± 8.7068 respectively [P = 0.000].

Table 1

Correlation of antenatal degree of hydronephrosis with postnatal management

Postnatal management	Degree of hydronephrosis based on antenatal anteroposterior diameter of pelvis in mm
	Second trimester				Third trimester
	Mild (4-<7), n (%)	Moderate (7-<10), n (%)	Severe (>10), n (%)	Total	Mild (7-<10), n (%)	Moderate (10-<15), n (%)	Severe (>15), n (%)	Total
Conservative	47 (88.68)	25 (65.79)	4 (14.29)	76	19 (90.48)	34 (62.97)	7 (13.73)	60
Surgical	6 (11.32)	13 (34.21)	24 (85.71)	43	2 (9.52)	20 (37.03)	44 (86.27)	66
Total	53	38	28	119	21	54	51	126

In the second trimester, the area under ROC was 0.848, whereas in third trimester, it was 0.897 [Figure 2]. An APD of 8.2 mm in the second trimester was associated with a sensitivity of 74% and specificity of 82% for patients at risk for postnatal surgery. Similarly, an APD of 12.85 mm in the third trimester was associated with a sensitivity of 83.3% and a specificity of 78.8% for patients at risk for postnatal surgery.

Figure 2

Receiver operating characteristic curve for renal pelvis dimensions in the (a) second trimester and (b) third trimester

All the patients did not have antenatal scans of both 2nd and 3rd trimester. In some cases, which were first referred to us in the 3rd trimester, the scans done in the 2nd trimester elsewhere only mentioned the HDN as mild, moderate, or severe without giving the exact dimensions. Their data were not included for comparison of 2nd and 3rd trimester scans.

In a subset of 50 units where both APD of 2nd and 3rd trimester was available, the mean increase in APD of pelvis was seen to be higher in those requiring surgery compared to conservatively managed units [Table 2].

Table 2

Mean renal pelvis dimensions in second and third trimester and correlation with type of management

Type of management	n (units)	APD of pelvis, mean±SD (range) in mm		Co-relation	P	Paired differences	P
		Second trimester	Third trimester
Conservative	29	6.77±2.18 (3-13)	10.31±5.34 (3.9-28)	0.666	0.000	3.548±4.219	0.000
Surgical	21	13.43±6.090 (4–28)	21.69±7.538 (8-40)	0.649	0.001	8.261±5.857	0.002

APD: Anteroposterior diameter, SD: Standard deviation

In a subset of 37 units, APD of pelvis of 2nd and 3rd trimester and 6-week postnatal period were available [Table 3]. Mean APD in conservatively managed group in second trimester, third trimester and 6 weeks after birth was 7.20 ± 2.018 mm, 10.46 ± 5.416 mm and 8.62 ± 2.756 mm. In the surgically managed group, this was found to be 14.80 ± 6.192 mm, 22.37 ± 8.265 mm and 27.71 ± 13.067 mm. The change in dimensions was found to be statistically significant between the 2nd and 3rd trimester in both groups but the difference was higher in the surgical group. The difference between the 2nd trimester and postnatal USG APD was significant only in the surgically managed group.

Table 3

Change in renal pelvis dimensions between antenatal and postnatal period and correlation with type of management

Change in APD between different time periods	Conservatively managed (n=22)		Surgically managed (n=15)
	Mean difference	P	Mean difference	P
Between second and third trimester	3.265±0.668	0.006	7.57±1.703	0.002
Between second trimester and postnatal period	1.423±0.612	0.088	12.91±3.247	0.004
Between third trimester and postnatal period	1.841±1.171	0.301	5.340±2.614	0.181

APD: Anteroposterior diameter, SD: Standard deviation

DISCUSSION

With the routine use of antenatal USG, the diagnosis of ANH has increased. Detection in early pregnancy helps in proper counselling of parents by pediatric surgeons and allay their fears regarding requirement of surgery. The severity of antenatal HDN appears to be directly related to urological anomalies like UPJO among others. However, many infants with mild ANH, which are likely to resolve spontaneously tend to undergo a lot of unnecessary postnatal investigations, and there is a need for further data on this subject.

In this study, 165 patients diagnosed with ANH were followed up postnatally. The M: F ratio in our study was 3.7:1 and was very similar to the study of Aksu et al. who had a M: F ratio of 3.58:1 among 156 patients.[5] In our study, the chances of requiring surgical intervention based on the 2nd and 3rd trimester pelvis APD were similar being 11.32% and 9.52% in mild HDN, 34.21% and 37.03% in moderate HDN, and 85.71% and 86.27% in severe HDN. Lee et al. also had similar results in 1308 patients where the risk of postnatal pathology in mild, moderate, and severe HDN was 11.9%, 45.1%, and 88.3%.[6] Passerotti et al. also showed that the risk of postnatal pathology increases progressively from 29.6% with mild HDN to 96.3% in severe HDN.[7] Wollenberg et al. in their study have shown that none of the patients with mild HDN, 23% with moderate HDN and 64% with severe HDN had associated UTI or required surgery.[8] Grignon et al. in their study showed that an APD of > 20 mm, was associated with significant abnormality requiring surgery or long-term follow-up in about 94% of patients whereas, if the APD was between 10 and 15 mm, then half had an abnormality, and if the APD was <10 mm, then only 3% had an abnormality.[9] In the studies done by Sairam et al. and Broadley et al., 31% and 36% of the fetuses with APD ≥10 mm on third trimester USG required surgery.[1011]

Our study focused to get a more accurate cut off value of antenatal APD suggestive of postnatal surgery so that prognostication could be given to parents during the pregnancy. Based on the area under the ROC curves with cut off value of APD of 8.2 mm in the second trimester, risk for postnatal surgery was associated with a sensitivity of 74% and specificity of 82%. Similarly, risk of surgery in the third trimester with an APD of 12.85 mm was associated with a sensitivity of 83.3% and a specificity of 78.8%. Based on the area under the ROC curves, the third-trimester APD had higher diagnostic value than second-trimester APD in predicting postnatal surgery. Kim et al. in their study with 187 patients, stated that by using a cut off APD of 11 mm at second trimester, the sensitivity and specificity of predicting postnatal surgery was 70% and 69.2% whereas with a cut off of 15 mm in the late third trimester, predicting postnatal surgery was associated with a sensitivity and specificity of 74.5% and 70.4%.[12] These figures show that a single 2nd or 3rd trimester APD may be sufficient to predict the possibility of surgical intervention. This information could be utilized for counseling of the parents who are more concerned about the requirement of any surgery in the postnatal period.

In our study, the mean APD in third trimester in conservatively managed patients was 9.75 ± 4.58 mm whereas in those managed surgically, it was 20.86 ± 8.46 mm which was statistically significant (P ≤ 0.001). These are similar to those of Coplen et al. who had a mean APD of 11.8 mm in nonobstructive dilatation with nearly two-fold increase in those with obstruction who had an APD of 22.3 mm.[13] In their study, based on ROC, 15 mm APD was used as a threshold to discriminate obstruction in 80% of fetuses with a sensitivity of 73% and specificity of 82%. In our study, an APD of 12.85 mm and area under ROC curve of 0.897 in the third trimester was associated with a sensitivity of 83.3% and specificity of 78.8% of risk of postnatal surgery.

Transient or nonobstructive dilatation of the pelvi-calyceal system is the most important cause of ANH and has to be differentiated from others needing surgical intervention. Sairam et al. noted the incidence of transient HDN to be between 41% and 88%.[10] In our study also transient HDN was seen in 116 of 219 units (52.96%).

MCUG forms an important part of the postnatal evaluation.[14] Zerin et al. found VUR in 49 (38%) of 130 neonates with ANH.[15] Aksu et al. in their study found VUR in 24 patients (16.6%).[5] In our study MCUG could be done only in 95 of 165 (57.57%) patients. Among these, 24 (25.26%) patients had VUR on one or either side. VUR may be associated with ipsilateral UPJO. Hollowell et al. reported 14% coexistence of UPJO and VUR.[16] In our study, 15 of 71 patients with UPJO had coexisting VUR giving a 21.1% coexistence of the 2 conditions. An earlier study also showed that 35 of 135 patients (25.9%) with UPJO had ipsilateral VUR.[17]

Our data shows that VUJO cannot be confidently diagnosed antenatally. HDUN can be due to several causes, i.e. VUJO, VUR, and megaureter. In our series in some cases only HDN was detected antenatally. The mean APD in second and third trimester in patients who were finally diagnosed with partial VUJO was 8.96 ± 1.8056 and 11.46 ± 4.150 mm, respectively. Although the APD was significant as per our study and most likely to need surgery, these findings suggest that the cutoff suggested is for patients with HDN and not HDUN. Postnatal follow up EC scans and USG showed resolution of HDUN in all these cases with time.

As part of the study was retrospective, in many cases only the second or the third trimester USG was available. In several USGs done prior to referral, exact pelvis dimensions were not available. Hence it was difficult to compare the change in APD between second and third trimester scans of the same patient in all the cases. MCU could be done only in 95 (57.57%) of 165 patients, and therefore, some cases of VUR may have been missed. All the USGs were not done by the same clinician or at the same center, and therefore, observer bias was possible in the calculation of APD. Due to the nature of referral, there was a higher rate of UPJO in our series. However, we feel that we were able to derive APD values with more accuracy for patients in need of surgery.

CONCLUSION

From our study, we derived a cut off value of APD of 8.2 mm and 12.85 mm in the second and third trimester, above which there were higher chances of having urological anomalies requiring postnatal surgery. This had a sensitivity and specificity of 74% and 82% in the second trimester and 83.3% and 78.8% in the third trimester, respectively. The mean increase in APD between the second and third trimesters was also found to be much more in surgically managed (8.261 ± 5.857 mm) than in conservatively managed patients (3.548 ± 4.219 mm). These 2 parameters will be useful during antenatal counseling.

Ethical approval

The study was approved by the Institute Ethics Committee and all subjects gave informed consent.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.