Journal of Pediatric Surgery (2011) 46, 1412–1418 www.elsevier.com/locate/jpedsurg Ultrasound evaluation of prognosis in fetuses with posterior urethral valves Lisandra Stein Bernardes a,b,⁎, Remi Salomon c , Gunnar Aksnes d , Stephen Lortat-Jacob d , Alexandra Benachi a a Maternité. Hôpital Necker-Enfants Malades, APHP, Université Paris Descartes, France Hospital das Clinicas da faculdade de Medicina da Universidade de São Paulo, São Paulo, Brazil – Departamento de Obstetrícia e Ginecologia c Service de néphrologie pédiatrique. Hôpital Necker-Enfants Malades, APHP, Université Paris Descartes, France d Service de chirurgie pédiatrique. Hôpital Necker-Enfants Malades, APHP, Université Paris Descartes, France b Received 10 September 2010; revised 30 November 2010; accepted 10 December 2010 Key words: Posterior urethral valves; Prenatal; Ultrasound; Renal function Abstract Purpose: The aim of this study was to evaluate the ability of prenatal ultrasound markers to predict postnatal renal prognosis in fetuses with posterior urethral valves. Methods: Medical files on fetuses with prenatal diagnosis of posterior urethral valves from 2000 to 2006 were reviewed retrospectively. Data from prenatal follow-up included gestational age at diagnosis, ultrasound renal parenchyma evaluation, and presence and time of oligohydramnios onset. Prenatal parameters studied were correlated to postnatal renal function. Results: Thirty-one male fetuses were included. Six pregnancies were terminated. Of the remaining 25 pregnancies that were continued, 4 children had abnormal creatine and 21 normal creatinine levels at follow-up. Presence and time of oligohydramnios onset did not differ between groups (P = .43). Ultrasound detected bilateral renal abnormalities in 3 fetuses (75%) with altered renal function, and 10 fetuses (55%) with normal creatinine, at follow-up. Conclusions: None of the ultrasound parameters evaluated were able to reliably predict postnatal renal function. © 2011 Elsevier Inc. All rights reserved. 1. Background Posterior urethral valves (PUV) are the main cause of end-stage renal failure in male children [1-4]. In cases with ⁎ Corresponding author. Divisão de Obstetrícia e Ginecologia do Hospital das Clínicas da Universidade de São Paulo Av. Dr Enéas de Carvalho Aguiar, 255- Cerqueira César- 05403-000 / São Paulo- Brasil. Tel.: +55 11 30 69 62 09; fax: +55 11 30 69 62 09. E-mail address: lisbernardes@usp.br (L. Stein Bernardes). 0022-3468/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2010.12.010 prenatal suspicion of PUV, ultrasound assessment of prognosis is important to guide parents' counseling and indicate invasive evaluation of renal function [5-10]. Furthermore, if intrauterine procedures such as vesicoamniotic shunts are considered, the evaluation of renal function may help to adequately select patients eligible for the procedure [11]. Several ultrasound parameters have been proposed as predictors of long-term renal function, such as amniotic fluid index and renal parenchyma evaluation [6,8,12-22]. However, previous studies have Ultrasound evaluation with posterior urethral valves been inconclusive concerning the optimal ultrasound parameters for predicting long-term renal function. Furthermore, most previous studies have included multiple causes of low urinary tract obstruction (LUTO) in evaluation, confounding prenatal evaluations. Finally, ultrasound equipment used in these studies was heterogeneous, and most of these investigations were performed more than 10 years ago. The quality of renal parenchyma imaging has improved dramatically in recent years with the use of modern techniques such as high frequency probes and harmonic imaging. These improvements should provide a more accurate ultrasound evaluation of renal function. Complementary to ultrasound evaluation, some centers suggest the use of fetal biomarkers of renal impairment in selected cases. Biochemical evaluation of fetal renal function (urinary sodium and β2-microglobulin, serum β2-microglobulin) may be performed when ultrasound markers suggest renal impairment. However, controversies over biochemical analysis, as well as the optimal cutoff for each marker, are common in the literature [16,23,24]. Therefore, many centers choose not to perform invasive procedures to obtain blood and urine samples for biochemistry evaluation. Instead, centers rely exclusively on ultrasound markers for prognostic evaluation and counseling. For this reason, we aimed to evaluate the ability of ultrasound alone to predict renal failure in early childhood in children with prenatally detected PUV. A secondary objective was to describe survival and renal outcome for these fetuses. 2. Methods Records of all fetuses followed up in our prenatal care center between 2000 and 2006 for PUV were reviewed retrospectively. When PUV was suspected, careful follow-up was initiated, and ultrasound evaluation of renal function was performed. Our protocol was to perform biochemical urine or blood evaluation (sodium and β2-microglobulin in urine and β2-microglobulin alone in blood) whenever ultrasound markers suggested renal impairment. Termination of pregnancy was considered in those cases displaying abnormal results on ultrasound and biochemical evaluations. Ultrasound was performed using a Voluson 730 ultrasound machine (Kretztechnik, Zipf, Austria), with a 4- to 8-MHz transducer. Data used for prenatal evaluation were gestational age (GA) at diagnosis, renal parenchyma evaluation (echogenicity and differentiation), presence and GA at oligohydramnios and anamnios onset, and presence of urinary ascites or urinoma. Kidneys were considered hyperechogenic if cortical echogenicity was higher than the echogenicity of the adjacent liver (Fig. 1). Kidneys were considered to lack corticomedullary differentiation if there was no difference between the echogenicity of the cortex and of the medulla (Fig. 2). 1413 Fig. 1 Longitudinal view of a kidney showing renal hyperechogenicity with normal corticomedullary differentiation. Small arrows indicate kidney delimitation; big arrow, liver. Oligohydramnios was defined as an amniotic fluid index of less than 5 cm. Anamnios was defined as the absence of amniotic fluid on ultrasound examination. Prenatal vesicoamniotic shunts for management of PUV are not routinely performed at our center. The option of peritoneoamniotic shunt was discussed with parents if the fetus had urinary ascites and oligohydramnios. Termination of pregnancy was acceptable according to French law in the event of oligohydramnios or anamnios where both ultrasound and biochemical evaluations showed fetal renal failure [25,26]. When parents chose to terminate pregnancy, the corresponding autopsy data were recorded. Data considered for the postnatal evaluation were age at surgery, number of surgeries, first-month creatinine level, creatinine level between 1 and 2 years, creatinine level at last follow-up, and need for antihypertensive drugs. Children who had a 1-year creatinine level lower than 50 μmol/L (0.57 mg/dL) were considered to have normal Fig. 2 Longitudinal view of a kidney showing lack of differentiation with normal echogenicity. Arrows indicate kidney delimitation. 1414 renal function at 1 year. Renal function at follow-up was considered normal if creatinine levels lay within the reference range for age. This study was approved by our institutional review board. 3. Results Thirty-one fetuses were included. Median GA at diagnosis was 25 weeks (19-36 weeks). Six pregnancies (19.3%) were terminated. Median GA at TOP was 30 weeks (21-36 weeks). All these fetuses had bilateral renal hyperechogenicity, lack of differentiation, and oligohydramnios on prenatal evaluation. Median GA at oligohydramnios onset was 27.7 weeks (20-32 weeks). One fetus had urinary ascites at 35 weeks. No shunt procedures were performed. Blood and/or urine samples had been collected for all fetuses to evaluate glomerular renal function. Five fetuses (83.3%) had abnormal β2-microglobulin values. One had a normal blood β2-microglobulin value, but the urine sample showed abnormal β2-microglobulin. Because oligohydramnios was present at 20 weeks, TOP was performed at parents' request. Autopsy study revealed tubular cysts and a diminished glomerular row count in all cases. Twenty-five pregnancies (80.7%) were continued to birth and median follow-up after birth was 40 months (range, 12-79 months). Table 1 shows postnatal data and creatinine values at 1 year and at follow-up for these children. Four fetuses had urinary ascites at 25, 28, 33 and 34 weeks. Peritoneoamniotic shunts were performed in 3 of these fetuses. The fourth fetus was delivered at 33 weeks because of fetal distress. Creatinine at last follow-up was normal for 3 of the children (Table 2). One fetus had a unilateral urinoma and oligohydramnios at 34 weeks. Because there were no signs of compression, vesicoamniotic shunt was not proposed. Ultrasound showed bilateral hyperechogenicity but normal corticomedullary differentiation. Creatinine at last follow-up (38 months) was normal (Table 1). Of the 25 live-born children, 4 (16%) had renal impairment at follow-up. One child died on day 16 from renal failure. This child had normal renal echogenicity but bilateral lack of differentiation during prenatal assessment. Oligohydramnios appeared at 36 weeks. No prenatal samples were taken for biochemical evaluation. A second child had a creatinine level of 206 mmol/L (2.33 mg/dL) at last follow-up (79 months). Repeated ultrasound examination during fetal life in this patient showed normal renal parenchyma. A urine sample was taken at 23 WA because of a borderline amniotic fluid index. The B2MCG level was above the 95th percentile for GA, but sodium level was within the reference range for GA. Amniotic fluid became normal after 25 WA and throughout pregnancy, whereas oligohydramnios reappeared at 37 weeks. L. Stein Bernardes et al. Another 2 children had creatinine levels above the reference range at follow-up (30 and 75 months). Of the 6 children that had creatinine levels greater than 50 μmol/L (0.57 mg/dL) at 1 year, only 3 had creatinine levels higher than the reference range at follow-up. Twenty-one children (84%) had normal creatinine at follow-up. All children, except for the individual who died, were submitted to endoscopic resection of valves. Twenty-four children underwent the surgery within the first 5 days of life. Four children underwent the surgery at a later date. Three had late surgery because of preterm delivery, and one because prenatal and postnatal ultrasound findings were mild, and definitive diagnosis was reached at day 13 with surgery performed at day 15. 3.1. Prenatal ultrasound parameters and prognosis 3.1.1. Amniotic fluid volume Fifty-two percent of the completed pregnancies had oligohydramnios. Seventy-six percent of these children had normal creatinine at last follow-up. Median GA at first appearance of oligohydramnios was 33.2 weeks (±3.6 weeks). Three children (12%) had oligohydramnios before 32 weeks. Creatinine level at last follow-up was normal for all of these subjects. The 2 children who developed renal failure had oligohydramnios later in pregnancy (36 and 37 weeks). Four fetuses (16%) presented anamnios, and median GA at anamnios onset was 34 WA (33-35 weeks). Creatinine level at follow-up was normal for 3 of these cases. 3.2. Renal parenchyma evaluation Bilateral hyperechogenicity was present in 11 (44%) of the 25 live-born fetuses. Ten (90.9%) had normal creatinine at follow-up. Bilateral lack of differentiation was present in 12 fetuses (48%). Ten (83.3%) had normal creatinine at follow-up. Both bilateral lack of differentiation and hyperechogenicity were present in 9 fetuses (36%). Eight (88.8%) had normal creatinine levels at follow-up. Associated bilateral lack of differentiation, hyperechogenicity, and oligohydramnios occurred in 8 fetuses (32%), 7 (87.5%) of which had normal creatinine levels at follow-up. Table 3 summarizes ultrasound findings of the 31 fetuses evaluated. Specificity and sensitivity of ultrasound parameters for predicting postnatal renal failure are depicted in Table 4. 4. Discussion Nineteen percent of the live-born children in our study developed renal failure during the observational period. This is a relatively low percentage compared with previous Prenatal and postnatal characteristics of the 25 fetuses whose pregnancies were pursued GA at GA at GA at GA of delivery Peritoneoaminiotic GA ascites diagnosis (WA) oligoamnios (WA) anamnios (WA) (WA) shunt (Y/N) or urinoma (WA) Creatinine at Renal function Age at Creatinine between Total at follow-up follow-up follow-up surgery (d) 12 and 24 mo (months) (μmol/L) (μmol/L) 19 36 No 37 3 Deceased 16 d 315 22 37 No 37 3 115 79 206 31 31 21 32 34 21 25 22 23 33 34 35 No 36 No No No 23 No 34 No No No No No No No 37 37 39 38 37 38 38 39 33 5 2 3 2 2 4 7 5 21 33 48 61 51 48 48 42 69 60 60 34 30 24 48 35 42 60 47 65 18 37 32 38 36 33 34 No No No No No No 37 39 40 5 7 5 30 26 18 48 30 27 47 75 23 28 No 28 No No 38 33 2 45 48 29 16 52 48 45 25 32 28 No No 34 No No 35 38 37 36 2 3 2 56 36 65 41 12 30 34 36 72 22 32 23 34 No 30 No No 33 36 38 33 7 3 27 43 50 38 42 43 39 53 Normal Normal Renal impairment Normal Normal Normal 22 32 28 33 33 No No 35 No 37 36 39 5 9 40 25 13 24 36 25 39 49 Normal Normal Normal Y Y Y N N 25 Urinary ascites 28 Urinary ascites 34 Urinary ascites 33 Urinoma 33 Urinary ascites 44 29 43 56 49 Renal impairment Renal impairment Normal Normal Normal Normal Normal Normal Normal Normal Normal Ultrasound evaluation with posterior urethral valves Table 1 Normal Normal Renal impairment Normal Normal 1415 1416 Table 2 L. Stein Bernardes et al. Follow-up of fetuses with urinary ascites GA at ascites Peritoneoamniotic shunt (Y/N) GA at oligohydramnios GA at anamnios Bilateral hyperechogenicity (Y/N) Bilateral lack of differentiation (Y/N) Total follow-up (mo) Renal function at follow-up 25 28 33 Y Y N 23 28 30 N N 33 N Y Y N Y N 48 52 30 34 Y 34 35 Y Y 42 Normal Normal Moderate renal impairment Normal reports. Miguelez et al [27] reported 45% mortality and 40% renal failure in prenatally diagnosed LUTO. Similarly, Anumba et al [28] reported 57% renal failure at 24 months after birth, in fetuses with LUTO. However, these studies included children with diverse obstructive uropathies, such as PUV, urethral atresia, and Prune-belly syndrome. In our series, only children with PUV were assessed. This could be the main reason explaining the prognostic differences. Recent studies involving long-term follow-up in children with prenatal diagnosis of PUV found a renal failure rate of around 30% up to adolescence [29,30]. This result is in-line with the present findings, especially when taking into account that the median follow-up in our series was only 3 years, and some children may progress to renal failure later in childhood. Different methods can be used to evaluate renal function in childhood [31]. Children with normal values of serum creatinine and no clinical complications do not routinely have glomerular filtration rate measurements taken or other renal function tests performed in our unit. Therefore, we used serum creatinine to evaluate renal function in this study. Table 3 Ultrasound findings for the 31 fetuses evaluated GA diagnosis Oligohydramnios, n (%) GA oligohydramnios TOP (6) Renal Good impairment prognosis (4) (21) 23.5 WA (20-32) 6 (100%) 25.7 WA (19-34) 3 (75%) 27.7 WA (21-36) 10 (47.6%) 27.7 WA (20-32) 2 (44.4%) 6 (100%) 35.6 WA (34-37) 1 (25%) 1 (25%) 32.0 WA (23-36) 3 (14.2%) 10 (47.6%) 2 (50%) 10 (47.6%) 1 (25%) 8 (38%) 1 (25%) 7 (33%) Anamnios Bilateral renal hyperechogenicity Bilateral lack of 6 (100%) differentiation 6 (100%) Bilateral lack of differentiation and hyperechogenicity 6 (100%) Bilateral lack of differentiation and hyperechogenicity and oligohydramnios However, we are well aware that many children presenting a serum creatinine level within the age-specific reference interval do not necessarily have 2 or even one normal kidney. Because the aim of this study was to evaluate the ability of ultrasound parameters alone to predict postnatal renal function, and given that many countries do not use biochemistry in their routine evaluation, biochemical evaluation was not included in our analysis. 4.1. Prenatal ultrasound evaluation Many studies have evaluated the extent to which prenatal ultrasound findings predict postnatal renal function in fetuses with PUV [7,8,14,22,32-34]. The prenatal diagnosis was one of the first variables evaluated. Children with prenatal diagnosis of PUV tend to have poorer renal function than those with postnatal diagnosis [7]. This group is also more likely to have pulmonary hypoplasia because of early onset of oligohydramnios [7,9]. Studies evaluating GA at diagnosis have shown a higher rate of renal failure in children who had early diagnosis, especially if diagnosis was made before 24 WA [9,14]. In our study, children who developed renal failure had an earlier diagnosis, with a tendency toward significance (P = .055). The small number of children with renal failure in our series (19%) could explain the absence of significance. Amniotic fluid volume has also been considered a predictor of postnatal renal function. Oliveira et al [33] Table 4 Sensitivity and specificity for the ultrasound parameters evaluated Oligohydramnios, n (%) Anamnios Bilateral renal hyperechogenicity Bilateral lack of differentiation Bilateral lack of differentiation and hyperechogenicity Bilateral lack of differentiation and hyperechogenicity and oligohydramanios Sensitivity Specificity 90% 30% 70% 80% 70% 52% 86% 52% 52% 62% 25% 67% Ultrasound evaluation with posterior urethral valves described a high relative risk (odds ratio, 10.6) for renal impairment when oligohydramnios was present before birth. Several other studies have shown a higher prevalence of renal impairment in children who had oligohydramnios [6-8,16,27]. However, akin to the present study, these previous studies also showed that normal amniotic index did not preclude renal impairment because approximately 25% of fetuses with normal amniotic index may develop renal failure [22,34]. In a recent systematic review, Morris et al [35] concluded that the presence of oligoamnios had a positive likelihood ratio for predicting renal failure ranging from 0.8 to 19 across the studies. In our series, oligohydramnios was present in 75% of the fetuses with renal failure. However, it was also present in 47.6% of fetuses with normal creatinine at follow-up, and no significant difference in oligohydramnios presence or GA was found when they appeared in both groups. A high predictive value for postnatal severe renal failure of the prenatal variables used to decide whether TOP can be carried out is imperative. The present study clearly showed that amniotic fluid volume alone was insufficient to predict postnatal renal function. Even if oligohydramnios is associated with poorer renal function, it did not confirm the condition because 47.6% of children with normal creatinine had oligohydramnios at a median GA of 32 WA (23-36 WA). In addition to amniotic fluid index and GA at diagnosis, ultrasound characteristics of the renal parenchyma have also been studied as a predictor of postnatal renal function [6,8,27,35-37]. Anumba et al [28] not only found renal impairment in all children who had bilateral renal hyperechogenicity or cysts on prenatal ultrasound but also found that 38% of the fetuses with a normal renal parenchyma developed renal failure after birth. Similarly, Miguelez et al [27] found renal impairment not only in 60% of fetuses with bilateral hyperechogenicity but also in 20% of those fetuses with normal renal echogenicity. In our series, 36% of the live-born children had bilateral hyperechogenicity and lack of differentiation on prenatal ultrasound. Eighty-eight percent of the sample had a normal creatinine level at follow-up. These findings are consistent with those of previous studies. Thus, although ultrasound evaluation of renal parenchyma in the fetus serves to guide further investigation, used alone, it is unable to predict renal function after birth or during childhood. Finally, when bilateral hyperechogenicity, lack of differentiation, and oligohydramnios were all observed in a fetus, renal function after birth was frequently normal. Seven of the 8 fetuses who met these criteria did not have renal failure at follow-up. In conclusion, this study represented an opportunity to evaluate the follow-up of children with prenatal diagnosis of PUV seen at a single center in which vesicoamniotic shunt was not performed routinely. The overall outcome in our 1417 series was good, but 19% of pregnancies were terminated because of the renal pathology observed in the fetus. Despite improved ultrasound technology, none of the prenatal ultrasound parameters accurately predicted postnatal renal function. Acknowledgments We would like to thank MC Aubry, JP Aubry, B Barthe, S Chemouny, C Morere, and J Martinovic for their contribution to this study. We would also like to thank CAPES and COFECUB for supporting the international cooperation. References [1] Shroff R, Wright E, Ledermann S, et al. Chronic hemodialysis in infants and children under 2 years of age. Pediatr Nephrol 2003;18(4): 378-83. [2] Anochie I, Eke F. Obstructive uropathy in childhood, as seen in University of Port Harcourt Teaching Hospital, Nigeria. Niger J Med 2004;13(2):136-9. [3] Rahman MH, Karim MA, Hoque E, et al. Chronic renal failure in children. Mymensingh Med J 2005;14(2):156-9. [4] Zurowska A, Zagozdzon I, Bałasz I, et al. Congenital and genetic related causes of end-stage renal disease—data from Polish Registry of Renal Replacement Therapy in Children 2000-2004. Przegl Lek 2006; 63(Suppl 3):57-9 Polish. [5] Reuss A, Wladimiroff JW, Scholtmeijer RJ, et al. Prenatal evaluation and outcome of fetal obstructive uropathies. Prenat Diagn 1988;8(2): 93-102. [6] Mandelbrot L, Dumez Y, Muller F, et al. Prenatal prediction of renal function in fetal obstructive uropathies. J Perinat Med 1991;19 (Suppl 1):283-7. [7] Reinberg Y, de Castano I, Gonzalez R. Prognosis for patients with prenatally diagnosed posterior urethral valves. J Urol 1992;148(1): 125-6. [8] Hutton KA, Thomas DF, Davies BW. Prenatally detected posterior urethral valves: qualitative assessment of second trimester scans and prediction of outcome. J Urol 1997;158(3 Pt 2):1022-5. [9] El-Ghoneimi A, Desgrippes A, Luton D, et al. Outcome of posterior urethral valves: to what extent is it improved by prenatal diagnosis? J Urol 1999;162(3 Pt 1):849-53. [10] Herndon CD, Ferrer FA, Freedman A, et al. Consensus on the prenatal management of antenatally detected urological abnormalities. J Urol 2000;164(3 Pt 2):1052-6. [11] Pluto Collaborative Study group, Kilby M, Khan K, Morris K, Daniels J, Gray R, Magill L, Martin B, Thompson P, Alfirevic Z, Kenny S, Bower S, Sturgiss S, Anumba D, Mason G, Tydeman G, Soothill P, Brackley K, Loughna P, Cameron A, Kumar S, Bullen P. PLUTO trial protocol: percutaneous shunting for lower urinary tract obstruction randomised controlled trial. BJOG. 2007;114(7):904-5, e1-4. [12] Glazer GM, Filly RA, Callen PW. The varied sonographic appearance of the urinary tract in the fetus and newborn with urethral obstruction. Radiology 1982;144(3):563-8. [13] Nicolaides KH, Cheng HH, Snijders RJ, et al. Fetal urine biochemistry in the assessment of obstructive uropathy. Am J Obstet Gynecol 1992; 166(3):932-7. [14] Hutton KA, Thomas DF, Arthur RJ, et al. Prenatally detected posterior urethral valves: is gestational age at detection a predictor of outcome? J Urol 1994;152(2 Pt 2):698-701. 1418 [15] Johnson MP, Corsi P, Bradfield W, et al. Sequential urinalysis improves evaluation of fetal renal function in obstructive uropathy. Am J Obstet Gynecol 1995;173(1):59-65. [16] Bensman A. Prognosis of prenatal uropathies: should fetal urine tests be continued? J Gynecol Obstet Biol Reprod (Paris) 1996;25(5):510-4. [17] Guez S, Assael BM, Melzi ML, et al. Shortcomings in predicting postnatal renal function using prenatal urine biochemistry in fetuses with congenital hydronephrosis. J Pediatr Surg 1996;31(10):1401-4. [18] Qureshi F, Jacques SM, Seifman B, et al. In utero fetal urine analysis and renal histology correlate with the outcome in fetal obstructive uropathies. Fetal Diagn Ther 1996;11(5):306-12. [19] Daïkha-Dahmane F, Dommergues M, Muller F, et al. Development of human fetal kidney in obstructive uropathy: correlations with ultrasonography and urine biochemistry. Kidney Int 1997;52(1): 21-32. [20] Nicolini U, Spelzini F. Invasive assessment of fetal renal abnormalities: urinalysis, fetal blood sampling and biopsy. Prenat Diagn 2001; 21(11):964-9. [21] Mussap M, Fanos V, Pizzini C, et al. Predictive value of amniotic fluid cystatin C levels for the early identification of fetuses with obstructive uropathies. BJOG 2002;109(7):778-83. [22] Zaccara A, Giorlandino C, Mobili L, et al. Amniotic fluid index and fetal bladder outlet obstruction. Do we really need more. J Urol 2005; 174(4 Pt 2):1657-60. [23] Elder JS, O'Grady JP, Ashmead G, et al. Evaluation of fetal renal function: unreliability of fetal urinary electrolytes. J Urol 1990;144 (2 Pt 2):574-8. [24] Morris RK, Quinlan-Jones E, Kilby MD, et al. Systematic review of accuracy of fetal urine analysis to predict poor postnatal renal function in cases of congenital urinary tract obstruction. Prenat Diagn 2007;27 (10):900-11. [25] Muller F, Dommergues M, Mandelbrot L, et al. Fetal urinary biochemistry predicts postnatal renal function in children with bilateral obstructive uropathies. Obstet Gynecol 1993;82(5):813-20. L. Stein Bernardes et al. [26] Dommergues M, Muller F, Ngo S, et al. Fetal serum beta2microglobulin predicts postnatal renal function in bilateral uropathies. Kidney Int 2000;58(1):312-6. [27] Miguelez J, Bunduki V, Yoshizaki CT, et al. Fetal obstructive uropathy: is urine sampling useful for prenatal counselling? Prenat Diagn 2006;26(1):81-4. [28] Anumba DO, Scott JE, Plant ND, et al. Diagnosis and outcome of fetal lower urinary tract obstruction in the northern region of England. Prenat Diagn 2005;25(1):7-13. [29] Kousidis G, Thomas DF, Morgan H, et al. The long-term outcome of prenatally detected posterior urethral valves: a 10 to 23-year follow-up study. BJU Int 2008. [30] Ylinen E, Ala-Houhala M, Wikström S. Prognostic factors of posterior urethral valves and the role of antenatal detection. Pediatr Nephrol 2004;19(8):874-9. [31] Atiyeh BA, Dabbagh SS, Gruskin AB. Evaluation of renal function during childhood. Pediatr Rev 1996;17(5):175-80. [32] Wilson RD, Morrison MG, Wittmann BK, et al. Clinical follow-up of fetal urinary tract anomalies diagnosed prenatally by ultrasound. Fetal Ther 1988;3(3):141-51. [33] Oliveira EA, Rabelo EA, Pereira AK, et al. Prognostic factors in prenatally-detected posterior urethral valves: a multivariate analysis. Pediatr Surg Int 2002;18(8):662-7. [34] Eckoldt F, Heling KS, Woderich R, et al. Posterior urethral valves: prenatal diagnostic signs and outcome. Urol Int 2004;73(4):296-301. [35] Morris RK, Malin GL, Khan KS, et al. Antenatal ultrasound to predict postnatal renal function in congenital lower urinary tract obstruction: systematic review of test accuracy. BJOG 2009;116(10):1290-9. [36] Glick PL, Harrison MR, Golbus MS, et al. Management of the fetus with congenital hydronephrosis II: prognostic criteria and selection for treatment. J Pediatr Surg 1985;20(4):376-87. [37] Sarhan O, Zaccaria I, Macher MA, et al. Long-term outcome of prenatally detected posterior urethral valves: single center study of 65 cases managed by primary valve ablation. J Urol 2008;179(1):307-12.