- Email: [email protected]
Society for Fetal Urology Classification vs Urinary Tract Dilation Grading System for Prognostication in Prenatal Hydronephrosis: A Time to Resolution Analysis
Society for Fetal Urology Classification vs Urinary Tract Dilation Grading System for Prognostication in Prenatal Hydronephrosis: A Time to Resolution Analysis Luis H. Braga,*,† Melissa McGrath,† Forough Farrokhyar, Kizanee Jegatheeswaran and Armando J. Lorenzo From the Division of Urology (LHB), McMaster Pediatric Surgery Research Collaborative (LHB, MM, KJ) and Office of Surgical Research Services (FF), Department of Surgery, and Department of Clinical Epidemiology and Biostatistics (LHB, FF), McMaster University, Hamilton and Division of Urology, Hospital for Sick Children (AJL) and Department of Surgery, University of Toronto (AJL), Toronto, Ontario, Canada
Purpose: The Urinary Tract Dilation grading system for prenatal hydronephrosis was introduced to address potential shortcomings of the Society for Fetal Urology classification. Hydronephrosis resolution is an important patient outcome and is frequently discussed during family counseling. We compared these 2 grading systems and their ability to predict time to hydronephrosis resolution. Materials and Methods: We prospectively screened 855 patients with prenatal hydronephrosis due to ureteropelvic junction obstruction-like hydronephrosis, nonrefluxing primary megaureter or vesicoureteral reflux between 2009 and 2015. Of the patients 454 were excluded due to surgery, late referral, absence of postnatal dilatation or presence of other anomalies, resulting in 401 eligible patients (of whom 81% were male) to be included for analyses. Hydronephrosis grades collected at baseline and last followup were compared to identify resolution trends through time. Hydronephrosis resolution was defined as renal pelvis anteroposterior diameter 10 mm or less at last followup. Time to resolution was analyzed using Cox proportion regression. Results: Of 401 patients 328 (82%) had resolution during a mean SD followup of 24 18 months (maximum 107). Cumulative resolution rate at 3 years was 98% for Society for Fetal Urology grade I hydronephrosis, 87% for grade II, 76% for grade III and 57% for grade IV. The 3-year hydronephrosis resolution rate was 90% for Urinary Tract Dilation postnatal grade 1 (low risk), 81% for grade 2 (intermediate risk) and 71% for grade 3 (high risk). Conclusions: Patients with distinctive baseline hydronephrosis grades (classified by Society for Fetal Urology or Urinary Tract Dilation system) had significantly different resolution times for hydronephrosis (p <0.001). Counseling families regarding time to resolution of prenatal hydronephrosis should remain the same whether using Society for Fetal Urology or Urinary Tract Dilation grading system. Key Words: child, hydronephrosis, prenatal diagnosis, urinary tract, prospective studies
0022-5347/18/1996-1615/0 THE JOURNAL OF UROLOGY® Ó 2018 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
https://doi.org/10.1016/j.juro.2017.11.077 Vol. 199, 1615-1621, June 2018 Printed in U.S.A.
Abbreviations and Acronyms APD ¼ anteroposterior diameter HN ¼ hydronephrosis P ¼ postnatal category PM ¼ nonrefluxing primary megaureter SFU ¼ Society for Fetal Urology UPJO ¼ ureteropelvic junction obstruction UTD ¼ Urinary Tract Dilation VUR ¼ vesicoureteral reflux Accepted for publication November 20, 2017. No direct or indirect commercial incentive associated with publishing this article. The corresponding author certifies that, when applicable, a statement(s) has been included in the manuscript documenting institutional review board, ethics committee or ethical review board study approval; principles of Helsinki Declaration were followed in lieu of formal ethics committee approval; institutional animal care and use committee approval; all human subjects provided written informed consent with guarantees of confidentiality; IRB approved protocol number; animal approved project number. Study received institutional ethics board approval. * Correspondence: Division of Urology, Department of Surgery, McMaster University, 1280 Main St. West, Hamilton, Ontario, Canada L8S 4K1 (telephone: 905-521-2100, ext. 73777; FAX: 905-570-8971; e-mail: [email protected]). † Equal study contribution.
www.jurology.com
j
1615
1616
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
PRENATAL hydronephrosis is relatively common, being reported in 1% to 5% of pregnancies.1 The most common etiologies of prenatal hydronephrosis are ureteropelvic junction obstruction-like or isolated hydronephrosis, vesicoureteral reflux and nonrefluxing primary megaureter.2 Prenatal hydronephrosis is seen in a wide spectrum of disorders, from so-called physiological dilatation of the urinary tract (which resolves spontaneously) to pathological conditions that may require surgical intervention. Time to resolution varies according to etiology and grade of hydronephrosis and is an important parameter to study. Grading prenatal HN severity using renal ultrasound parameters is common and is the basis for establishing logical and efficient treatment plans.3 Resolution of prenatal HN is a critical patient outcome measure and is frequently discussed with families during counseling. HN has been traditionally classified according to the Society for Fetal Urology system into low (I to II) and high grade (III to IV) disease.2 More recently the Urinary Tract Dilation classification system for HN was released to address potential shortcomings of the current SFU grading system, specifically to deal with inconsistencies between prenatally detected HN and postnatal management strategies within and across specialties.3 The goal of the multidisciplinary consensus was to create a single grading system that could be used during the prenatal and postnatal periods to describe urinary tract dilatation and facilitate communication between different fields. UTD is a 3-point system based on 6 different ultrasound observations (renal pelvis APD, calyceal dilatation, parenchymal thickness, parenchymal appearance, ureteral dilatation and bladder abnormalities) to stratify risk based on the most concerning observations.4 We compared the SFU and UTD grading systems and their ability to predict the spontaneous resolution of prenatal HN in a large prospective cohort of children who did not require corrective surgery. We hypothesized that both grading systems would be comparable in their ability to predict time to resolution according to HN grade.
METHODS Study Population and Inclusion and Exclusion Criteria This study is part of an ongoing larger prospective cohort study that involves all patients diagnosed with prenatal HN who were referred to our tertiary care pediatric hospital. After obtaining ethics board approval we screened 855 consecutive patients from 2008 to 2015. Of these patients 258 were excluded due to late presentation (age 24 months or older) or resolution of HN by first clinic visit. A total of 108 patients were excluded due to ectopic ureter,
ureterocele, posterior urethral valves, multicystic dysplastic kidney, neurogenic bladder or prune belly syndrome. We further excluded 83 children who underwent surgical procedures since HN resolution was arbitrarily impacted by the decision to intervene. Thus, 401 consecutive prenatal patients with etiologies of isolated HN (UPJO-like), hydroureteronephrosis (PM) and VUR were included in the study population.
Independent Variables, Outcomes of Interest and Followup Predictors of resolution of prenatal hydronephrosis were decided a priori and included 4 SFU grades, 3 UTD categories and APD measured at the initial ultrasound visit. Severity of HN was categorized as low (grades I and II) or high (III and IV) according to the SFU grading system using baseline sagittal ultrasound image series.2 SFU grades were prospectively collected at baseline and last followup visits by our fellowship trained pediatric radiologists (with extensive training in SFU and UTD classification systems) following institutional protocol (ie no pretest patient hydration to minimize measurement bias). Using the same sagittal ultrasound image series as for grading based on the SFU system, severity of UTD was categorized as P1, P2 or P3 as described previously.3 The baseline renal pelvis APD variable was coded as larger than 16 mm or 16 mm or smaller,5 and 20 mm or smaller or larger than 20 mm,6 according to previous studies. Outcome of interest was resolution of prenatal hydronephrosis, which was defined as 2 consecutive ultrasounds showing either SFU grade I or less or a renal pelvis APD of 10 mm or less for patients with isolated hydronephrosis, or ureteral dilatation less than 7 mm for those with PM at the last followup clinic visit. These resolution criteria were applicable for patients with SFU grade II to IV or UTD P1 to P3 HN. Children born with SFU grade I HN or APD less than 10 mm had to demonstrate a reduction (improvement) to SFU grade 0 to be considered to have resolution. Two consecutive ultrasounds indicating resolution were required to account for individual fluctuations that could be influenced by patient hydration status, bladder urine volume or technical variability. This prospective study was designed to follow children for a minimum of 5 years or until the HN resolved, whichever occurred first. Patients were assessed by ultrasound at each clinic visit and assigned a SFU grade, a UTD category and an APD value. All infants were followed for a minimum of 6 months. The figure outlines the standardized followup protocol for prenatal hydronephrosis. Those with bilateral prenatal hydronephrosis were categorized based on the kidney with the higher SFU grade, UTD category or APD value.
Statistical Analyses and Sample Size Calculation Continuous data with normal distribution were reported as mean SD. Univariate analyses comparing HN resolution rates between SFU and UTD systems were conducted using chi-square testing. Tables were created to illustrate the cumulative rate of unresolved prenatal hydronephrosis by SFU grades, UTD category, APD cutoff
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
1617
Followup ultrasound (u/s) protocol time line in 401 patients. Asterisk indicates testing at clinician discretion. Voiding cystourethrogram interval was 6 to 12 weeks, and to allow for proper kidney maturation renal scan interval was 8 to 12 weeks.
values and hydronephrosis etiologies through time. Differences between overall hydronephrosis resolution times for each SFU grade and UTD category, APD cutoffs and hydronephrosis etiologies were assessed using log-rank testing. A priori defined subgroup analyses explored resolution patterns according to 3 different HN etiologies, that is isolated hydronephrosis (UPJO-like), hydroureteronephrosis (PM) and VUR (supplementary tables 1 to 3, http://jurology.com/). Statistical analyses were conducted using SPSSÒ, version 22, where p <0.05 was considered statistically significant. Based on expected resolution rates of 64% and 29% for patients with low and high grade prenatal HN, respectively, at 18 months,7 and assuming 90% power and an a error of 5%, the sample size required to properly answer the research question was 400 patients, accounting for 10% loss to followup. We assumed the accrual time for study period to be 84 months.
RESULTS The study population consisted of 401 prenatal HN patients, including 76 females (19%) and 325 males (81%). Mean SD patient age at first clinic visit was 4.5 5.2 months. Mean renal pelvis APD at first clinic visit was comparable for patients with SFU grade I/II and those with UTD P1 hydronephrosis (table 1). Comparable renal pelvis APDs were also seen in children with SFU grade III and those with UTD P2, and in patients with SFU grade IV and those with UTD P3 hydronephrosis. Overall, 328 children (82%) had resolution of HN during a mean SD followup of 24 18 months (range 6 to 107).
The cumulative rate of resolved HN in 401 patients was 36% at 1 year, 66% at 2 years, 83% at 3 years, 92% at 4 years, 95% at 5 years and 97% after 6 years. Cumulative rates of HN resolution according to SFU grade and UTD category are outlined in table 2. When the same patients were stratified by initial renal pelvis APD cutoffs of 16 mm or less and greater than 16 mm, the overall cumulative HN resolution rates for these 2 subgroups were significantly different (p <0.001). Similar findings were observed when APD cutoffs of 20 mm or less and greater than 20 mm were compared (p ¼ 0.003, table 3). Patients with nonrefluxing primary megaureter had significantly lower overall cumulative spontaneous hydronephrosis resolution rates compared to those with isolated hydronephrosis and VUR when classified either by SFU grade (p ¼ 0.009) or UTD category (p ¼ 0.048, table 4). Table 1. Renal pelvis APD at first clinic visit by HN grade/ category
SFU grade: I II III IV UTD category: P1 P2 P3
No. Pts
Mean APD (mm)
60 151 157 33
5 7 12 18
169 194 38
7 11 17
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
1618
Table 2. Cumulative HN resolution rates by grade/category No. SFU Grade (%)* Followup 6 Mos 1 Yr 1.5 Yrs 2 Yrs 2.5 Yrs 3 Yrs 3.5 Yrs 4 Yrs 4.5 Yrs 5 Yrs
I 60 44 44 12 6 3
II
Table 4. Cumulative rates of HN resolution by etiology according to SFU and UTD classifications
No. UTD Category (%)†
III
IV
P1
P2
P3
(27) 151 (18) 157 (11) 33 (6) 169 (22) 194 (20) 38 (11) (57) 120 (37) 132 (32) 21 (11) 130 (48) 145 (42) 25 (36) (78) 90 (59) 90 (45) 19 (25) 85 (70) 94 (56) 17 (51) (88) 55 (73) 69 (56) 15 (35) 46 (80) 68 (69) 12 (59) (94) 37 (82) 52 (66) 9 (50) 30 (87) 43 (76) 12 (59) (98) 23 (87) 39 (76) 7 (57) 17 (90) 32 (81) 7 (71) 17 (92) 27 (82) 5 (74) 12 (92) 25 (84) 7 (71) 11 (95) 18 (86) 3 (83) 9 (94) 16 (90) 3 (81) 7 (96) 12 (88) 2 (91) 7 (96) 16 (90) 3 (81) 5 (98) 10 (91) 5 (99) 10 (92) 2 (100)
* Log-rank p <0.001 for overall HN resolution rate. † Log-rank p ¼ 0.011 for overall HN resolution rate.
Separate subgroup analyses involving cumulative hydronephrosis resolution rates in patients with isolated hydronephrosis (UPJO-like), PM and VUR are displayed in supplementary tables 1 to 3 (http:// jurology.com/).
DISCUSSION There is a large body of evidence exploring outcomes in prenatal HN cases. The majority of these series have focused on somewhat subjective end points such as need for surgical intervention and development of urinary tract infections. However, there remains a paucity of data regarding information frequently asked about by parents and caretakers regarding time to HN resolution. Although previous studies have provided prognostic data based on different HN grades, most have solely focused on the SFU grading system8e11 or renal pelvis APD measurements.12e14 To our knowledge, none has compared how the SFU and UTD grading systems perform regarding ability to predict time to HN resolution. This is a critical question as the UTD system is expected to be increasingly used by health care providers from different specialties dealing with this patient population. Although this system has been designed through a multidisciplinary Table 3. Cumulative HN resolution rates by APD cutoffs of 16 and 20 mm No. 16 mm Cutoff (%)*
Followup (yrs) 1 2 3 4 5 6
Baseline APD 16 mm or Less 288 146 93 42 15 6
(10) (62) (80) (91) (96) (98)
Baseline APD Greater than 16 mm 33 24 13 10 8 3
No. 20 mm Cutoff (%)† Baseline APD 20 mm or Less
(36) (50) (61) (65) (74) (90)
* Log-rank p <0.001 for overall HN resolution rate. † Log-rank p ¼ 0.003 for overall HN resolution rate.
309 161 101 48 20 8
(35) (62) (78) (89) (95) (97)
Baseline APD Greater than 20 mm 12 9 5 5 3 1
(9) (29) (43) (43) (81) (99)
Followup (yrs)
No. VUR (%)
No. Nonrefluxing Primary Megaureter (%)
No. UPJO-Like HN (%)
SFU* 1 2 3 4 5
60 25 10 3 2
(36) (64) (81) (93) (96)
1 2 3 4 5
60 25 10 3 2
(44) (70) (83) (94) (97)
48 29 19 9 6
(29) (47) (64) (80) (90)
209 94 39 19 10
(38) (72) (87) (93) (96)
45 23 13 7 5
(40) (61) (71) (80) (97)
195 78 33 18 12
(45) (77) (84) (93) (92)
UTD†
* Log-rank p ¼ 0.009 for overall HN resolution rate by etiology. † Log-rank p ¼ 0.048 for overall HN resolution rate by etiology.
consensus and has gained increased popularity across different specialties and countries, its value stills needs to be confirmed. Our analyses indicate that the SFU and UTD classifications similarly discriminate among patients in terms of HN resolution rates. In addition, regardless of the classification system used to grade HN, higher grades of prenatal HN are associated with longer spontaneous resolution times. The inverse relationship between higher HN grades and lower resolution rates, regardless of etiology, has been observed in other studies.15e18 Yang et al reported spontaneous UPJO-like HN resolution rates of 92% for SFU grade I, 87% for grade II, 30% for grade III and 0% for grade IV at a mean interval of 13 months.12 Recently Bader et al revealed that UTD P3 patients were significantly more likely not to have spontaneous resolution and to ultimately undergo pyeloplasty, according to the United Kingdom criteria.13 This same pattern of HN resolution was also observed in children with PM (hydroureteronephrosis). Patients with SFU grade I to III HN had a 71% resolution rate, compared to 55% for those with grade IV HN, at a mean followup of 36 months.10 Similarly McLellan et al reported that PM patients with SFU grade I to III HN had spontaneous resolution at a median age of 26 months, while those with SFU grade IV HN had slow resolution (median 49 months) and were more likely to undergo surgery.8 Even when HN severity was graded by renal pelvis APD measurement, lower HN grade cases resolved faster than the more severe ones (table 3). Similar to our findings, Aksu et al observed a negative correlation between renal pelvis APD and HN resolution rate (p <0.01).14 Patients with APDs of 5 to 9 mm, 10 to 14 mm and 15 mm or greater showed HN improvement of 22%, 6% and 0%, respectively. Using a different definition to describe mild HN, a systematic review by Sidhu et al
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
demonstrated that all UPJO-like cases with APD smaller than 12 mm resolved without specifying time to resolution.10 In contrast, some authors have reported resolution rates of 52% to 65% in patients with APD greater than 12 mm,6,19 while others have found a 75% resolution rate for those with APD ranging from 13 to 19 mm, and 43% for those with APD larger than 19 mm.20 To add to the variability on this issue, evidence has revealed that 100% of patients with APD 10 to 40 mm have resolution of HN, compared to 0% when the APD is larger than 40 mm.21 This variation probably occurs because the analyses have included surgical cases, a strategy that introduces challenges due to the subjective nature of needing surgical intervention, as well as the abrupt improvement in HN expected in most cases postoperatively. In the present analysis we strictly defined our patient population and went a step further, calculating rates of HN resolution according to the 3 different etiologies in an attempt to adjust for potential confounding factors (supplementary tables 1 to 3, http://jurology.com/). Finally, our study has to be contrasted with the findings reported by Hodhod et al.22 In stark distinction to their results, we found that the SFU and UTD grading systems were equally accurate at predicting time to HN resolution and discriminating resolution time according to HN severity. Hodhod et al reported that multivariate analyses favored UTD category as a predictor of surgical intervention, while SFU grade predicted resolution.22 Due to the aforementioned methodological differences and the prospective nature of our analysis, as well as the attention to separate assessment of predictive variables (due to risk of multicollinearity), our results challenge those findings and provide reassurance that both systems have similar predictive performance. As with most research studies, this prospective analysis has limitations that deserve acknowledgment. Due to the absence of randomization, we could not adjust for any unknown prognostic factors
1619
that may have biased resolution rates. It is also noteworthy that the results of this study depend extensively on measurements of SFU and UTD grades to track prenatal HN progression. Measurement bias could have been introduced if parents hydrated infants 30 to 60 minutes before ultrasound, indicating a protocol deviation. A prior study found APD measurements increased by a mean of 35% following hydration (p <0.01), with 31% of the kidneys showing an increase of greater than 50%.23 Such significant changes in measurement could have either decreased resolution rates (due to inflated measurements) or prolonged time to resolution. Furthermore, assessments of prenatal HN using both grading systems can be limited by their low interrater reliability for moderate cases (SFU grade III, UTD P2).24,25 Despite these limitations, we propose there is value in our data, as they are powered by a considerable sample size combined with a strict requirement of 2 consecutive ultrasounds to document HN resolution. We present novel information on resolution rates for different HN etiologies by comparing 2 grading systems in the hope of adding knowledge to a growing body of literature exploring the new UTD classification system.
CONCLUSIONS Based on our time to event analysis, children with higher HN grades (SFU III to IV and UTD P2 to P3) had longer spontaneous resolution times compared to those with lower grades (SFU I to II and UTD P1). Approximately 90% of patients with SFU grade I to II/UTD P1 had resolution of HN during a mean followup of 2 years, compared to 75% of those with SFU III/UTD P2 and 50% of those with SFU IV/ UTD P3. When counseling families, time to resolution of prenatal HN should remain the same whether using the SFU or UTD classification system.
REFERENCES 1. Woodward M and Frank D: Postnatal management of antenatal hydronephrosis. BJU Int 2002; 89: 149. 2. Nguyen HT, Herndon CD, Cooper C et al: The Society for Fetal Urology consensus statement on the evaluation and management of antenatal hydronephrosis. J Pediatr Urol 2010; 6: 212. 3. Nguyen HT, Benson CB, Bromley B et al: Multidisciplinary consensus on the classification of prenatal and postnatal urinary tract dilation (UTD classification system). J Pediatr Urol 2014; 10: 982.
4. Braga LH, McGrath M, Farrokhyar F et al: Associations of initial Society for Fetal Urology grades and urinary tract dilatation risk groups with clinical outcomes in patients with isolated prenatal hydronephrosis. J Urol 2017; 197: 831. 5. Dias CS, Silva JM, Pereira AK et al: Diagnostic accuracy of renal pelvic dilatation for detecting surgically managed ureteropelvic junction obstruction. J Urol 2013; 190: 661. 6. Dhillon HK: Prenatally diagnosed hydronephrosis: the Great Ormond Street experience. Br J Urol 1998; 81: 39.
7. G€okaslan F, Yalc¸ınkaya F, Fit€oz S et al: Evaluation and outcome of antenatal hydronephrosis: a prospective study. Ren Fail 2012; 34: 718. 8. McLellan DL, Retik AB, Bauer SB et al: Rate and predictors of spontaneous resolution of prenatally diagnosed primary nonrefluxing megaureter. J Urol 2002; 168: 2177. 9. Jaswon MS, Dibble L, Puri S et al: Prospective study of outcome in antenatally diagnosed renal pelvis dilatation. Arch Dis Child Fetal Neonatal Ed 1999; 80: 135.
1620
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
10. Sidhu G, Beyene J and Rosenblum ND: Outcome of isolated antenatal hydronephrosis: a systematic review and meta-analysis. Pediatr Nephrol 2006; 21: 218. 11. Zee RS, Herndon CD, Cooper CS et al: Time to resolution: a prospective evaluation from the Society for Fetal Urology hydronephrosis registry. J Pediatr Urol 2017; 13: 316.e1. 12. Yang Y, Hou Y, Niu ZB et al: Long-term follow-up and management of prenatally detected isolated hydronephrosis. J Pediatr Surg 2010; 45: 1701. 13. Bader M, Sur H, Robb A et al: Efficacy of Upper Tract Dilation classification in predicting the need for pyeloplasty compared to the UK approach. Presented at annual meeting of European Society for Pediatric Urology, Barcelona, Spain, April 19-22, 2017. 14. Aksu N, Yavascan O, Kangun M et al: Postnatal management of infants with antenatally detected hydronephrosis. Pediatr Nephrol 2005; 20: 1253. 15. Alconcher LF and Tombesi MM: Natural history of bilateral mild isolated antenatal hydronephrosis
conservatively managed. Pediatr Nephrol 2012; 27: 1119. 16. Coelho GM, Bouzada MC, Pereira AK et al: Outcome of isolated antenatal hydronephrosis: a prospective cohort study. Pediatr Nephrol 2007; 22: 1727. 17. Takla NV, Hamilton BD, Cartwright PC et al: Apparent unilateral ureteropelvic junction obstruction in the newborn: expectations for resolution. J Urol 1998; 160: 2175. 18. Rodriguez LV, Lock J, Kennedy WA et al: Evaluation of sonographic renal parenchymal area in the management of hydronephrosis. J Urol 2001; 165: 548. 19. Kletscher B, De Badiola F and Gonzalez R: Outcome of hydronephrosis diagnosed antenatally. J Pediatr Surg 1991; 26: 455. 20. Ransley PG, Dhillon HK, Gordon I et al: The postnatal management of hydronephrosis diagnosed by prenatal ultrasound. J Urol 1990; 144: 584.
21. Josephson S, Dhillon HK and Ransley PG: Postnatal management of antenatally detected, bilateral hydronephrosis. Urol Int 1993; 51: 79. 22. Hodhod A, Capolicchio JP, Jednak R et al: Evaluation of Urinary Tract Dilation classification system for grading postnatal hydronephrosis. J Urol 2016; 195: 725. 23. Gleason J, Bator E, Bowlin P et al: The impact of the state of hydration on the degree of hydronephrosis in children: towards a uniform standard of ultrasonographic assessment. Presented at annual pediatric urology fall congress of Societies for Pediatric Urology, Miami, Florida, October 24-26, 2014. 24. Keays MA, Guerra LA, Mihill J et al: Reliability assessment of Society for Fetal Urology ultrasound grading system for hydronephrosis. J Urol, suppl., 2008; 180: 1680. 25. Rickard M, Lorenzo AJ, Farrokhyar F et al: Six of one, half a dozen of the other: a measure of multidisciplinary inter/intra-rater reliability of the Society for Fetal Urology and Urinary Tract Dilation grading systems for hydronephrosis. J Pediatr Urol 2017; 13: 80.e1.
EDITORIAL COMMENT Fetal hydronephrosis remains a frequent clinical conundrum. Effective communication between maternal-fetal medicine, radiology, urology, nephrology and pediatrics is paramount. Toward this end, the Society for Fetal Urology recently convened experts to develop the UTD grading system to address deficiencies of the SFU grading system (reference 3 in article). Using a prospective database, Braga et al found that the 2 grading systems performed similarly in predicting postnatal resolution of hydronephrosis and confirmed what we see in practice, ie mild to moderate hydronephrosis typically resolves. SFU grade I to II is similar to UTD P1 hydronephrosis, and resolution was seen in 87% to 98% and 90% of patients, respectively. SFU grade III and UTD P2 hydronephrosis also had similar resolution rates (76% to 81%). Resolution in
SFU grade IV and UTD P3 hydronephrosis was intermediate (57% to 71%) and not as equivalent. This study was not designed to answer the most critical questiondhow do the 2 grading systems compare in terms of identifying which children need surgical intervention? Future research is needed to better clarify who requires greater attention. In the meantime this article is reassuring in that the 2 grading systems perform similarly in predicting resolution of hydronephrosis, and adoption of one of the systems over the other is not required if a clinical team has a preference. John S. Wiener Section of Pediatric Urology Duke University Medical Center Durham, North Carolina
REPLY BY AUTHORS Predicting the need for surgical intervention is an important outcome. However, to date, a conclusive test or imaging modality to definitively identify cases that will progress to this end point is lacking. Our group has been motivated to find new ways to
forecast which patients will ultimately undergo surgery, having applied the renal parenchyma-tohydronephrosis area ratio to determine hydronephrosis trends toward resolution or surgical intervention.1 More recently we have been using
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
machine learning predictive capability to identify those more likely to undergo surgery in our large prospective prenatal hydronephrosis database.2 Equally important to determining who is at risk for requiring surgery is to test the validity of any proposed grading system to accurately and reliably stratify those who will eventually reach clinically relevant outcomes (such as febrile urinary tract infection or surgery). With that in mind our group previously reported the predictive value of SFU vs UTD classification regarding these critical outcomes (reference 4 in article). We found similar results to the current study, showing that
1621
regardless of the grading system used, patients who underwent surgery or had febrile urinary tract infection could be properly segregated from those who did not. By reviewing the first postnatal ultrasound we may not yet be able to accurately predict which prenatal hydronephrosis cases will have deterioration of the condition and ultimately need surgical intervention based on current imaging modalities and statistical equations. As technology continues to advance, new high tech refined machine learning models may allow us to do exactly that in the near future.
REFERENCES 1. Rickard M, Lorenzo AJ, Braga LH et al: Parenchyma-to-hydronephrosis area ratio is a promising outcome measure to quantify upper tract changes in infants with highgrade prenatal hydronephrosis. Urology 2017; 104: 166. 2. Lorenzo AJ, Rickard M and Braga LH: Predictive analytics and modeling employing machine learning technology: the next step in data sharing, analysis and individualized counseling explored with a large, prospective prenatal hydronephrosis database. Presented at annual meeting of Societies for Pediatric Urology, Boston, Massachusetts, May 12-14, 2017.
Purpose: The Urinary Tract Dilation grading system for prenatal hydronephrosis was introduced to address potential shortcomings of the Society for Fetal Urology classification. Hydronephrosis resolution is an important patient outcome and is frequently discussed during family counseling. We compared these 2 grading systems and their ability to predict time to hydronephrosis resolution. Materials and Methods: We prospectively screened 855 patients with prenatal hydronephrosis due to ureteropelvic junction obstruction-like hydronephrosis, nonrefluxing primary megaureter or vesicoureteral reflux between 2009 and 2015. Of the patients 454 were excluded due to surgery, late referral, absence of postnatal dilatation or presence of other anomalies, resulting in 401 eligible patients (of whom 81% were male) to be included for analyses. Hydronephrosis grades collected at baseline and last followup were compared to identify resolution trends through time. Hydronephrosis resolution was defined as renal pelvis anteroposterior diameter 10 mm or less at last followup. Time to resolution was analyzed using Cox proportion regression. Results: Of 401 patients 328 (82%) had resolution during a mean SD followup of 24 18 months (maximum 107). Cumulative resolution rate at 3 years was 98% for Society for Fetal Urology grade I hydronephrosis, 87% for grade II, 76% for grade III and 57% for grade IV. The 3-year hydronephrosis resolution rate was 90% for Urinary Tract Dilation postnatal grade 1 (low risk), 81% for grade 2 (intermediate risk) and 71% for grade 3 (high risk). Conclusions: Patients with distinctive baseline hydronephrosis grades (classified by Society for Fetal Urology or Urinary Tract Dilation system) had significantly different resolution times for hydronephrosis (p <0.001). Counseling families regarding time to resolution of prenatal hydronephrosis should remain the same whether using Society for Fetal Urology or Urinary Tract Dilation grading system. Key Words: child, hydronephrosis, prenatal diagnosis, urinary tract, prospective studies
0022-5347/18/1996-1615/0 THE JOURNAL OF UROLOGY® Ó 2018 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
https://doi.org/10.1016/j.juro.2017.11.077 Vol. 199, 1615-1621, June 2018 Printed in U.S.A.
Abbreviations and Acronyms APD ¼ anteroposterior diameter HN ¼ hydronephrosis P ¼ postnatal category PM ¼ nonrefluxing primary megaureter SFU ¼ Society for Fetal Urology UPJO ¼ ureteropelvic junction obstruction UTD ¼ Urinary Tract Dilation VUR ¼ vesicoureteral reflux Accepted for publication November 20, 2017. No direct or indirect commercial incentive associated with publishing this article. The corresponding author certifies that, when applicable, a statement(s) has been included in the manuscript documenting institutional review board, ethics committee or ethical review board study approval; principles of Helsinki Declaration were followed in lieu of formal ethics committee approval; institutional animal care and use committee approval; all human subjects provided written informed consent with guarantees of confidentiality; IRB approved protocol number; animal approved project number. Study received institutional ethics board approval. * Correspondence: Division of Urology, Department of Surgery, McMaster University, 1280 Main St. West, Hamilton, Ontario, Canada L8S 4K1 (telephone: 905-521-2100, ext. 73777; FAX: 905-570-8971; e-mail: [email protected]). † Equal study contribution.
www.jurology.com
j
1615
1616
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
PRENATAL hydronephrosis is relatively common, being reported in 1% to 5% of pregnancies.1 The most common etiologies of prenatal hydronephrosis are ureteropelvic junction obstruction-like or isolated hydronephrosis, vesicoureteral reflux and nonrefluxing primary megaureter.2 Prenatal hydronephrosis is seen in a wide spectrum of disorders, from so-called physiological dilatation of the urinary tract (which resolves spontaneously) to pathological conditions that may require surgical intervention. Time to resolution varies according to etiology and grade of hydronephrosis and is an important parameter to study. Grading prenatal HN severity using renal ultrasound parameters is common and is the basis for establishing logical and efficient treatment plans.3 Resolution of prenatal HN is a critical patient outcome measure and is frequently discussed with families during counseling. HN has been traditionally classified according to the Society for Fetal Urology system into low (I to II) and high grade (III to IV) disease.2 More recently the Urinary Tract Dilation classification system for HN was released to address potential shortcomings of the current SFU grading system, specifically to deal with inconsistencies between prenatally detected HN and postnatal management strategies within and across specialties.3 The goal of the multidisciplinary consensus was to create a single grading system that could be used during the prenatal and postnatal periods to describe urinary tract dilatation and facilitate communication between different fields. UTD is a 3-point system based on 6 different ultrasound observations (renal pelvis APD, calyceal dilatation, parenchymal thickness, parenchymal appearance, ureteral dilatation and bladder abnormalities) to stratify risk based on the most concerning observations.4 We compared the SFU and UTD grading systems and their ability to predict the spontaneous resolution of prenatal HN in a large prospective cohort of children who did not require corrective surgery. We hypothesized that both grading systems would be comparable in their ability to predict time to resolution according to HN grade.
METHODS Study Population and Inclusion and Exclusion Criteria This study is part of an ongoing larger prospective cohort study that involves all patients diagnosed with prenatal HN who were referred to our tertiary care pediatric hospital. After obtaining ethics board approval we screened 855 consecutive patients from 2008 to 2015. Of these patients 258 were excluded due to late presentation (age 24 months or older) or resolution of HN by first clinic visit. A total of 108 patients were excluded due to ectopic ureter,
ureterocele, posterior urethral valves, multicystic dysplastic kidney, neurogenic bladder or prune belly syndrome. We further excluded 83 children who underwent surgical procedures since HN resolution was arbitrarily impacted by the decision to intervene. Thus, 401 consecutive prenatal patients with etiologies of isolated HN (UPJO-like), hydroureteronephrosis (PM) and VUR were included in the study population.
Independent Variables, Outcomes of Interest and Followup Predictors of resolution of prenatal hydronephrosis were decided a priori and included 4 SFU grades, 3 UTD categories and APD measured at the initial ultrasound visit. Severity of HN was categorized as low (grades I and II) or high (III and IV) according to the SFU grading system using baseline sagittal ultrasound image series.2 SFU grades were prospectively collected at baseline and last followup visits by our fellowship trained pediatric radiologists (with extensive training in SFU and UTD classification systems) following institutional protocol (ie no pretest patient hydration to minimize measurement bias). Using the same sagittal ultrasound image series as for grading based on the SFU system, severity of UTD was categorized as P1, P2 or P3 as described previously.3 The baseline renal pelvis APD variable was coded as larger than 16 mm or 16 mm or smaller,5 and 20 mm or smaller or larger than 20 mm,6 according to previous studies. Outcome of interest was resolution of prenatal hydronephrosis, which was defined as 2 consecutive ultrasounds showing either SFU grade I or less or a renal pelvis APD of 10 mm or less for patients with isolated hydronephrosis, or ureteral dilatation less than 7 mm for those with PM at the last followup clinic visit. These resolution criteria were applicable for patients with SFU grade II to IV or UTD P1 to P3 HN. Children born with SFU grade I HN or APD less than 10 mm had to demonstrate a reduction (improvement) to SFU grade 0 to be considered to have resolution. Two consecutive ultrasounds indicating resolution were required to account for individual fluctuations that could be influenced by patient hydration status, bladder urine volume or technical variability. This prospective study was designed to follow children for a minimum of 5 years or until the HN resolved, whichever occurred first. Patients were assessed by ultrasound at each clinic visit and assigned a SFU grade, a UTD category and an APD value. All infants were followed for a minimum of 6 months. The figure outlines the standardized followup protocol for prenatal hydronephrosis. Those with bilateral prenatal hydronephrosis were categorized based on the kidney with the higher SFU grade, UTD category or APD value.
Statistical Analyses and Sample Size Calculation Continuous data with normal distribution were reported as mean SD. Univariate analyses comparing HN resolution rates between SFU and UTD systems were conducted using chi-square testing. Tables were created to illustrate the cumulative rate of unresolved prenatal hydronephrosis by SFU grades, UTD category, APD cutoff
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
1617
Followup ultrasound (u/s) protocol time line in 401 patients. Asterisk indicates testing at clinician discretion. Voiding cystourethrogram interval was 6 to 12 weeks, and to allow for proper kidney maturation renal scan interval was 8 to 12 weeks.
values and hydronephrosis etiologies through time. Differences between overall hydronephrosis resolution times for each SFU grade and UTD category, APD cutoffs and hydronephrosis etiologies were assessed using log-rank testing. A priori defined subgroup analyses explored resolution patterns according to 3 different HN etiologies, that is isolated hydronephrosis (UPJO-like), hydroureteronephrosis (PM) and VUR (supplementary tables 1 to 3, http://jurology.com/). Statistical analyses were conducted using SPSSÒ, version 22, where p <0.05 was considered statistically significant. Based on expected resolution rates of 64% and 29% for patients with low and high grade prenatal HN, respectively, at 18 months,7 and assuming 90% power and an a error of 5%, the sample size required to properly answer the research question was 400 patients, accounting for 10% loss to followup. We assumed the accrual time for study period to be 84 months.
RESULTS The study population consisted of 401 prenatal HN patients, including 76 females (19%) and 325 males (81%). Mean SD patient age at first clinic visit was 4.5 5.2 months. Mean renal pelvis APD at first clinic visit was comparable for patients with SFU grade I/II and those with UTD P1 hydronephrosis (table 1). Comparable renal pelvis APDs were also seen in children with SFU grade III and those with UTD P2, and in patients with SFU grade IV and those with UTD P3 hydronephrosis. Overall, 328 children (82%) had resolution of HN during a mean SD followup of 24 18 months (range 6 to 107).
The cumulative rate of resolved HN in 401 patients was 36% at 1 year, 66% at 2 years, 83% at 3 years, 92% at 4 years, 95% at 5 years and 97% after 6 years. Cumulative rates of HN resolution according to SFU grade and UTD category are outlined in table 2. When the same patients were stratified by initial renal pelvis APD cutoffs of 16 mm or less and greater than 16 mm, the overall cumulative HN resolution rates for these 2 subgroups were significantly different (p <0.001). Similar findings were observed when APD cutoffs of 20 mm or less and greater than 20 mm were compared (p ¼ 0.003, table 3). Patients with nonrefluxing primary megaureter had significantly lower overall cumulative spontaneous hydronephrosis resolution rates compared to those with isolated hydronephrosis and VUR when classified either by SFU grade (p ¼ 0.009) or UTD category (p ¼ 0.048, table 4). Table 1. Renal pelvis APD at first clinic visit by HN grade/ category
SFU grade: I II III IV UTD category: P1 P2 P3
No. Pts
Mean APD (mm)
60 151 157 33
5 7 12 18
169 194 38
7 11 17
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
1618
Table 2. Cumulative HN resolution rates by grade/category No. SFU Grade (%)* Followup 6 Mos 1 Yr 1.5 Yrs 2 Yrs 2.5 Yrs 3 Yrs 3.5 Yrs 4 Yrs 4.5 Yrs 5 Yrs
I 60 44 44 12 6 3
II
Table 4. Cumulative rates of HN resolution by etiology according to SFU and UTD classifications
No. UTD Category (%)†
III
IV
P1
P2
P3
(27) 151 (18) 157 (11) 33 (6) 169 (22) 194 (20) 38 (11) (57) 120 (37) 132 (32) 21 (11) 130 (48) 145 (42) 25 (36) (78) 90 (59) 90 (45) 19 (25) 85 (70) 94 (56) 17 (51) (88) 55 (73) 69 (56) 15 (35) 46 (80) 68 (69) 12 (59) (94) 37 (82) 52 (66) 9 (50) 30 (87) 43 (76) 12 (59) (98) 23 (87) 39 (76) 7 (57) 17 (90) 32 (81) 7 (71) 17 (92) 27 (82) 5 (74) 12 (92) 25 (84) 7 (71) 11 (95) 18 (86) 3 (83) 9 (94) 16 (90) 3 (81) 7 (96) 12 (88) 2 (91) 7 (96) 16 (90) 3 (81) 5 (98) 10 (91) 5 (99) 10 (92) 2 (100)
* Log-rank p <0.001 for overall HN resolution rate. † Log-rank p ¼ 0.011 for overall HN resolution rate.
Separate subgroup analyses involving cumulative hydronephrosis resolution rates in patients with isolated hydronephrosis (UPJO-like), PM and VUR are displayed in supplementary tables 1 to 3 (http:// jurology.com/).
DISCUSSION There is a large body of evidence exploring outcomes in prenatal HN cases. The majority of these series have focused on somewhat subjective end points such as need for surgical intervention and development of urinary tract infections. However, there remains a paucity of data regarding information frequently asked about by parents and caretakers regarding time to HN resolution. Although previous studies have provided prognostic data based on different HN grades, most have solely focused on the SFU grading system8e11 or renal pelvis APD measurements.12e14 To our knowledge, none has compared how the SFU and UTD grading systems perform regarding ability to predict time to HN resolution. This is a critical question as the UTD system is expected to be increasingly used by health care providers from different specialties dealing with this patient population. Although this system has been designed through a multidisciplinary Table 3. Cumulative HN resolution rates by APD cutoffs of 16 and 20 mm No. 16 mm Cutoff (%)*
Followup (yrs) 1 2 3 4 5 6
Baseline APD 16 mm or Less 288 146 93 42 15 6
(10) (62) (80) (91) (96) (98)
Baseline APD Greater than 16 mm 33 24 13 10 8 3
No. 20 mm Cutoff (%)† Baseline APD 20 mm or Less
(36) (50) (61) (65) (74) (90)
* Log-rank p <0.001 for overall HN resolution rate. † Log-rank p ¼ 0.003 for overall HN resolution rate.
309 161 101 48 20 8
(35) (62) (78) (89) (95) (97)
Baseline APD Greater than 20 mm 12 9 5 5 3 1
(9) (29) (43) (43) (81) (99)
Followup (yrs)
No. VUR (%)
No. Nonrefluxing Primary Megaureter (%)
No. UPJO-Like HN (%)
SFU* 1 2 3 4 5
60 25 10 3 2
(36) (64) (81) (93) (96)
1 2 3 4 5
60 25 10 3 2
(44) (70) (83) (94) (97)
48 29 19 9 6
(29) (47) (64) (80) (90)
209 94 39 19 10
(38) (72) (87) (93) (96)
45 23 13 7 5
(40) (61) (71) (80) (97)
195 78 33 18 12
(45) (77) (84) (93) (92)
UTD†
* Log-rank p ¼ 0.009 for overall HN resolution rate by etiology. † Log-rank p ¼ 0.048 for overall HN resolution rate by etiology.
consensus and has gained increased popularity across different specialties and countries, its value stills needs to be confirmed. Our analyses indicate that the SFU and UTD classifications similarly discriminate among patients in terms of HN resolution rates. In addition, regardless of the classification system used to grade HN, higher grades of prenatal HN are associated with longer spontaneous resolution times. The inverse relationship between higher HN grades and lower resolution rates, regardless of etiology, has been observed in other studies.15e18 Yang et al reported spontaneous UPJO-like HN resolution rates of 92% for SFU grade I, 87% for grade II, 30% for grade III and 0% for grade IV at a mean interval of 13 months.12 Recently Bader et al revealed that UTD P3 patients were significantly more likely not to have spontaneous resolution and to ultimately undergo pyeloplasty, according to the United Kingdom criteria.13 This same pattern of HN resolution was also observed in children with PM (hydroureteronephrosis). Patients with SFU grade I to III HN had a 71% resolution rate, compared to 55% for those with grade IV HN, at a mean followup of 36 months.10 Similarly McLellan et al reported that PM patients with SFU grade I to III HN had spontaneous resolution at a median age of 26 months, while those with SFU grade IV HN had slow resolution (median 49 months) and were more likely to undergo surgery.8 Even when HN severity was graded by renal pelvis APD measurement, lower HN grade cases resolved faster than the more severe ones (table 3). Similar to our findings, Aksu et al observed a negative correlation between renal pelvis APD and HN resolution rate (p <0.01).14 Patients with APDs of 5 to 9 mm, 10 to 14 mm and 15 mm or greater showed HN improvement of 22%, 6% and 0%, respectively. Using a different definition to describe mild HN, a systematic review by Sidhu et al
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
demonstrated that all UPJO-like cases with APD smaller than 12 mm resolved without specifying time to resolution.10 In contrast, some authors have reported resolution rates of 52% to 65% in patients with APD greater than 12 mm,6,19 while others have found a 75% resolution rate for those with APD ranging from 13 to 19 mm, and 43% for those with APD larger than 19 mm.20 To add to the variability on this issue, evidence has revealed that 100% of patients with APD 10 to 40 mm have resolution of HN, compared to 0% when the APD is larger than 40 mm.21 This variation probably occurs because the analyses have included surgical cases, a strategy that introduces challenges due to the subjective nature of needing surgical intervention, as well as the abrupt improvement in HN expected in most cases postoperatively. In the present analysis we strictly defined our patient population and went a step further, calculating rates of HN resolution according to the 3 different etiologies in an attempt to adjust for potential confounding factors (supplementary tables 1 to 3, http://jurology.com/). Finally, our study has to be contrasted with the findings reported by Hodhod et al.22 In stark distinction to their results, we found that the SFU and UTD grading systems were equally accurate at predicting time to HN resolution and discriminating resolution time according to HN severity. Hodhod et al reported that multivariate analyses favored UTD category as a predictor of surgical intervention, while SFU grade predicted resolution.22 Due to the aforementioned methodological differences and the prospective nature of our analysis, as well as the attention to separate assessment of predictive variables (due to risk of multicollinearity), our results challenge those findings and provide reassurance that both systems have similar predictive performance. As with most research studies, this prospective analysis has limitations that deserve acknowledgment. Due to the absence of randomization, we could not adjust for any unknown prognostic factors
1619
that may have biased resolution rates. It is also noteworthy that the results of this study depend extensively on measurements of SFU and UTD grades to track prenatal HN progression. Measurement bias could have been introduced if parents hydrated infants 30 to 60 minutes before ultrasound, indicating a protocol deviation. A prior study found APD measurements increased by a mean of 35% following hydration (p <0.01), with 31% of the kidneys showing an increase of greater than 50%.23 Such significant changes in measurement could have either decreased resolution rates (due to inflated measurements) or prolonged time to resolution. Furthermore, assessments of prenatal HN using both grading systems can be limited by their low interrater reliability for moderate cases (SFU grade III, UTD P2).24,25 Despite these limitations, we propose there is value in our data, as they are powered by a considerable sample size combined with a strict requirement of 2 consecutive ultrasounds to document HN resolution. We present novel information on resolution rates for different HN etiologies by comparing 2 grading systems in the hope of adding knowledge to a growing body of literature exploring the new UTD classification system.
CONCLUSIONS Based on our time to event analysis, children with higher HN grades (SFU III to IV and UTD P2 to P3) had longer spontaneous resolution times compared to those with lower grades (SFU I to II and UTD P1). Approximately 90% of patients with SFU grade I to II/UTD P1 had resolution of HN during a mean followup of 2 years, compared to 75% of those with SFU III/UTD P2 and 50% of those with SFU IV/ UTD P3. When counseling families, time to resolution of prenatal HN should remain the same whether using the SFU or UTD classification system.
REFERENCES 1. Woodward M and Frank D: Postnatal management of antenatal hydronephrosis. BJU Int 2002; 89: 149. 2. Nguyen HT, Herndon CD, Cooper C et al: The Society for Fetal Urology consensus statement on the evaluation and management of antenatal hydronephrosis. J Pediatr Urol 2010; 6: 212. 3. Nguyen HT, Benson CB, Bromley B et al: Multidisciplinary consensus on the classification of prenatal and postnatal urinary tract dilation (UTD classification system). J Pediatr Urol 2014; 10: 982.
4. Braga LH, McGrath M, Farrokhyar F et al: Associations of initial Society for Fetal Urology grades and urinary tract dilatation risk groups with clinical outcomes in patients with isolated prenatal hydronephrosis. J Urol 2017; 197: 831. 5. Dias CS, Silva JM, Pereira AK et al: Diagnostic accuracy of renal pelvic dilatation for detecting surgically managed ureteropelvic junction obstruction. J Urol 2013; 190: 661. 6. Dhillon HK: Prenatally diagnosed hydronephrosis: the Great Ormond Street experience. Br J Urol 1998; 81: 39.
7. G€okaslan F, Yalc¸ınkaya F, Fit€oz S et al: Evaluation and outcome of antenatal hydronephrosis: a prospective study. Ren Fail 2012; 34: 718. 8. McLellan DL, Retik AB, Bauer SB et al: Rate and predictors of spontaneous resolution of prenatally diagnosed primary nonrefluxing megaureter. J Urol 2002; 168: 2177. 9. Jaswon MS, Dibble L, Puri S et al: Prospective study of outcome in antenatally diagnosed renal pelvis dilatation. Arch Dis Child Fetal Neonatal Ed 1999; 80: 135.
1620
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
10. Sidhu G, Beyene J and Rosenblum ND: Outcome of isolated antenatal hydronephrosis: a systematic review and meta-analysis. Pediatr Nephrol 2006; 21: 218. 11. Zee RS, Herndon CD, Cooper CS et al: Time to resolution: a prospective evaluation from the Society for Fetal Urology hydronephrosis registry. J Pediatr Urol 2017; 13: 316.e1. 12. Yang Y, Hou Y, Niu ZB et al: Long-term follow-up and management of prenatally detected isolated hydronephrosis. J Pediatr Surg 2010; 45: 1701. 13. Bader M, Sur H, Robb A et al: Efficacy of Upper Tract Dilation classification in predicting the need for pyeloplasty compared to the UK approach. Presented at annual meeting of European Society for Pediatric Urology, Barcelona, Spain, April 19-22, 2017. 14. Aksu N, Yavascan O, Kangun M et al: Postnatal management of infants with antenatally detected hydronephrosis. Pediatr Nephrol 2005; 20: 1253. 15. Alconcher LF and Tombesi MM: Natural history of bilateral mild isolated antenatal hydronephrosis
conservatively managed. Pediatr Nephrol 2012; 27: 1119. 16. Coelho GM, Bouzada MC, Pereira AK et al: Outcome of isolated antenatal hydronephrosis: a prospective cohort study. Pediatr Nephrol 2007; 22: 1727. 17. Takla NV, Hamilton BD, Cartwright PC et al: Apparent unilateral ureteropelvic junction obstruction in the newborn: expectations for resolution. J Urol 1998; 160: 2175. 18. Rodriguez LV, Lock J, Kennedy WA et al: Evaluation of sonographic renal parenchymal area in the management of hydronephrosis. J Urol 2001; 165: 548. 19. Kletscher B, De Badiola F and Gonzalez R: Outcome of hydronephrosis diagnosed antenatally. J Pediatr Surg 1991; 26: 455. 20. Ransley PG, Dhillon HK, Gordon I et al: The postnatal management of hydronephrosis diagnosed by prenatal ultrasound. J Urol 1990; 144: 584.
21. Josephson S, Dhillon HK and Ransley PG: Postnatal management of antenatally detected, bilateral hydronephrosis. Urol Int 1993; 51: 79. 22. Hodhod A, Capolicchio JP, Jednak R et al: Evaluation of Urinary Tract Dilation classification system for grading postnatal hydronephrosis. J Urol 2016; 195: 725. 23. Gleason J, Bator E, Bowlin P et al: The impact of the state of hydration on the degree of hydronephrosis in children: towards a uniform standard of ultrasonographic assessment. Presented at annual pediatric urology fall congress of Societies for Pediatric Urology, Miami, Florida, October 24-26, 2014. 24. Keays MA, Guerra LA, Mihill J et al: Reliability assessment of Society for Fetal Urology ultrasound grading system for hydronephrosis. J Urol, suppl., 2008; 180: 1680. 25. Rickard M, Lorenzo AJ, Farrokhyar F et al: Six of one, half a dozen of the other: a measure of multidisciplinary inter/intra-rater reliability of the Society for Fetal Urology and Urinary Tract Dilation grading systems for hydronephrosis. J Pediatr Urol 2017; 13: 80.e1.
EDITORIAL COMMENT Fetal hydronephrosis remains a frequent clinical conundrum. Effective communication between maternal-fetal medicine, radiology, urology, nephrology and pediatrics is paramount. Toward this end, the Society for Fetal Urology recently convened experts to develop the UTD grading system to address deficiencies of the SFU grading system (reference 3 in article). Using a prospective database, Braga et al found that the 2 grading systems performed similarly in predicting postnatal resolution of hydronephrosis and confirmed what we see in practice, ie mild to moderate hydronephrosis typically resolves. SFU grade I to II is similar to UTD P1 hydronephrosis, and resolution was seen in 87% to 98% and 90% of patients, respectively. SFU grade III and UTD P2 hydronephrosis also had similar resolution rates (76% to 81%). Resolution in
SFU grade IV and UTD P3 hydronephrosis was intermediate (57% to 71%) and not as equivalent. This study was not designed to answer the most critical questiondhow do the 2 grading systems compare in terms of identifying which children need surgical intervention? Future research is needed to better clarify who requires greater attention. In the meantime this article is reassuring in that the 2 grading systems perform similarly in predicting resolution of hydronephrosis, and adoption of one of the systems over the other is not required if a clinical team has a preference. John S. Wiener Section of Pediatric Urology Duke University Medical Center Durham, North Carolina
REPLY BY AUTHORS Predicting the need for surgical intervention is an important outcome. However, to date, a conclusive test or imaging modality to definitively identify cases that will progress to this end point is lacking. Our group has been motivated to find new ways to
forecast which patients will ultimately undergo surgery, having applied the renal parenchyma-tohydronephrosis area ratio to determine hydronephrosis trends toward resolution or surgical intervention.1 More recently we have been using
GRADING SYSTEMS FOR PRENATAL HYDRONEPHROSIS
machine learning predictive capability to identify those more likely to undergo surgery in our large prospective prenatal hydronephrosis database.2 Equally important to determining who is at risk for requiring surgery is to test the validity of any proposed grading system to accurately and reliably stratify those who will eventually reach clinically relevant outcomes (such as febrile urinary tract infection or surgery). With that in mind our group previously reported the predictive value of SFU vs UTD classification regarding these critical outcomes (reference 4 in article). We found similar results to the current study, showing that
1621
regardless of the grading system used, patients who underwent surgery or had febrile urinary tract infection could be properly segregated from those who did not. By reviewing the first postnatal ultrasound we may not yet be able to accurately predict which prenatal hydronephrosis cases will have deterioration of the condition and ultimately need surgical intervention based on current imaging modalities and statistical equations. As technology continues to advance, new high tech refined machine learning models may allow us to do exactly that in the near future.
REFERENCES 1. Rickard M, Lorenzo AJ, Braga LH et al: Parenchyma-to-hydronephrosis area ratio is a promising outcome measure to quantify upper tract changes in infants with highgrade prenatal hydronephrosis. Urology 2017; 104: 166. 2. Lorenzo AJ, Rickard M and Braga LH: Predictive analytics and modeling employing machine learning technology: the next step in data sharing, analysis and individualized counseling explored with a large, prospective prenatal hydronephrosis database. Presented at annual meeting of Societies for Pediatric Urology, Boston, Massachusetts, May 12-14, 2017.