Congenital urinary tract obstruction

Best Practice & Research Clinical Obstetrics and Gynaecology Vol. 22, No. 1, pp. 97–122, 2008 doi:10.1016/j.bpobgyn.2007.08.007 available online at http://www.sciencedirect.com

7 Congenital urinary tract obstruction R.K. Morris *

MRCOG

Clinical Research Fellow Academic Department of Obstetrics and Gynaecology, Birmingham Women’s Hospital, University of Birmingham, Birmingham B15 2TG, UK

Mark D. Kilby

MD, MRCOG

Professor of Maternal and Fetal Medicine Fetal Medicine Centre, Birmingham Women’s Hospital, University of Birmingham, Birmingham B15 2TG, UK

Congenital urinary tract obstruction is a heterogenous condition with a varying natural history. Accurate diagnosis within the late-first and second trimesters allows for counselling of the parents and planning of multi-disciplinary care for the pregnancy and newborn. Antenatal investigations to predict postnatal renal function are of varying accuracy. However, some factors have been shown to be predictive of poor outcome in terms of renal function at birth and infancy. There is the possibility of in-utero intervention in these fetuses. Key words: congenital urinary tract obstruction; hydronephrosis; posterior urethral valves; vesico-amniotic shunting.

INTRODUCTION Congenital abnormalities of the genitourinary tract are one of the most common sonographically identified anomalies, with a rate of 1 in 250 to 1 in 1000.1 Obstructive uropathies account for the majority of cases: approximately 1 in 2000 pregnancies.2 Obstruction can be at the level of the ureteropelvic or uretovesical junction, or it might be urethral. It can be either unilateral or bilateral; although, by definition, obstruction at the level of the urethra must be bilateral and it might involve some or all of the urinary tract. The majority of genitourinary abnormalities are diagnosed during the detailed second-trimester scan (usually performed at 18–20 weeks). However, with the increasing use of first-trimester screening, more severe renal anomalies are being noted between 11 and 14 weeks using ultrasound. The findings on antenatal ultrasound will depend on * Corresponding author. Tel.: þ44 121 626 4535; Fax: þ44 121 415 4837. E-mail address: [email protected] (R.K. Morris). 1521-6934/$ - see front matter ª 2007 Elsevier Ltd. All rights reserved.

98 R. K. Morris and M. D. Kilby

the level and severity of the obstruction and, in the presence of a minor obstruction with normal liquor volume and in which dilatation of the renal tract might not occur until the third trimester, the diagnosis can be missed. Antenatal diagnosis allows for the planning of appropriate prenatal and postnatal care, consideration of prenatal intervention and psychological adjustment of the parents; however, minor anomalies can cause undue parental anxiety. For some anomalies there is also uncertainty regarding best management. The prognosis in upper urinary tract obstruction in the fetus is generally relatively good unless other co-existent congenital anomalies are present. However, the situation is very different for lower urinary tract obstruction (LUTO) because of its association with cystic renal dysplasia, oligohydramnios, pulmonary hypoplasia, mild dysmorphia and limb contractures. For this reason, prenatal in-utero therapy has been considered in selected cases of LUTO in an attempt to bypass the congenital urinary tract obstruction and attenuate the secondary developmental complications. Prenatal counselling in this situation is difficult because the prognostic value of the modalities used to assess fetal renal function are uncertain and the effectiveness of therapy remains to be established. In this chapter we discuss the epidemiology and aetiology of congenital urinary tract obstruction and the pathophysiology behind these conditions. Current options for investigation and management of the fetus are presented, with discussion of future research developments. EPIDEMIOLOGY Upper urinary tract obstruction can be detected consistently at the second-trimester ultrasound scan. The largest European database is from the EUROSCAN Group, which involves 20 registries from 12 European countries. Data have been acquired from 709,030 cases, with a mean prevalence of all types of renal malformation of 1.6 per 1000 births. The most frequent diagnosis reported was upper urinary tract dilatation in 309 (27%) of patients, with 259 (84%) being detected prenatally.3 Unilateral hydronephrosis is more common than bilateral4 and most cases – 88% of bilateral and 69% of unilateral – can be described as mild.5 Pelviureteric obstruction is the most common cause of hydronephrosis (reported incidences of 39–64%) with reflux accounting for a third (33%) and vesico-ureteric obstruction 9–14%.6–9 Only one study giving population-based information on LUTO and data from a regional congenital anomaly register (Northern Region) has been published in the medical literature. This study identified 113 registered cases in the 14-year period between 1984 and 1997 and noted that the registry had a high notification rate with an ascertainment level of 95%. The incidence of LUTO was calculated as 2.2 per 10,000 births (based on total birth denominator data) with posterior urethral valves (PUV; determined by postnatal investigations and autopsies) being the most common at 64% (1.4/10,000 births), followed by urethral atresia at 39% (0.7/10,000 births) and prune belly syndrome 4%.10 AETIOLOGY Upper urinary tract obstruction Ureteropelvic junction obstruction Ureteropelvic junction (UPJ) obstruction is the most common lesion of the fetal urinary tract. Its aetiology is unclear. Causes include abnormal recanalisation of the

Congenital urinary tract obstruction 99

ureter during development, atypical mucosal valves, ureteric duplication and urteral strictures. Extrinsic causes include ectopic renal vessels. In the majority of cases, the underlying cause is thought to be functional. Ureterovesical obstruction The aetiology of ureterovesical obstruction is diverse, with the most common cause being primary megaureter. Other causes include ureteric strictures, ureteric atresia, retrocaval ureter, vascular obstruction, diverticulum, valves, ureterocele and vesicoutereral reflux. These causes of upper urinary tract obstruction can be unilateral or bilateral. Lower urinary tract obstruction Urethral obstruction The most common cause of urethral obstruction is posterior urethral valves (PUV)11, which account for one-third of renal tract anomalies detected at autopsy following termination for ultrasound diagnosed fetal anomaly.12 The affected fetus is typically male. Other causes include urethral agenesis and urethral strictures. The earlier the diagnosis is made on ultrasound the greater the association with urethral atresia.13 Megacystis-microcolon-intestinal hypoperistalsis syndrome Megacystis-microcolon-intestinal hypoperistalsis syndrome (MMIHS) is a cloacal plate anomaly that is more often seen in females. It is characterised by a dilated bladder and hydronephrosis in the presence of normal or increased amniotic fluid. There is dysfunctional smooth muscle in the bladder and distal bowel. Examination of tissue from MMIHS patients has shown a lack of the a3-nicotinic acetylcholine receptor subunit, giving a possible explanation for the underlying pathogenesis of this condition in some cases.14 Occasionally, lack of diagnosis or intervention can result in ‘prune belly’ syndrome. This consists of a triad of features – deficient or absent anterior abdominal wall musculature, dilation of the proximal and distal urinary tract (hydronephrosis, megacystis) and bilateral cryptorchidism – present in the neonatal period. It is rare in females, in whom the uropathy and abdominal wall defect are associated with vaginal atresia, rectovaginal or rectovesical fistula, and bicornuate uterus. NATURAL HISTORY The hallmark of upper urinary tract obstruction is hydronephrosis; however, this is a common antenatal finding, affecting 1–5% of all pregnancies.5,15 It is important to realise that the hydronephrosis can be transient and functional and thus it is important to determine which fetuses have significant dilatation that will impact on renal function postnatally. This has resulted in many studies being directed at defining the natural history of hydronephrosis and the ultrasound criteria that define significant antenatal hydronephrosis, the most recent papers16–20 are shown in Table 1. A recent systematic review21 summarised the data for stabilisation of antenatal hydronephrosis (defined as decreasing pelvic diameter or not worsening). For mild hydronephrosis [anteroposterior renal pelvis diameter (APPD) < 12 mm] 98% (0.93–1.0) of patients stabilised; for severe hydronephrosis (APPD > 12 mm) 51%

Author and year (ref. no.)

Exclusions

Study design

Ahmad et al 200516

6 lost to follow up

Retrospective

Signorelli et al 200517

Multiple pregnancies, chromosomal or structural abnormalities Unselected

Retrospective

2 terminations of pregnancy Other fetal anomalies

Prospective

Sairam et al 200118

Jaswon et al 199919 Persutte et al 199720

Prospective

Prospective

Definition of hydronephrosis

Total number of patients

Number of cases

Prevalence

Resolution Third trimester

Postnatal

Surgical intervention (% hydronephrosis in third trimester) Mild ¼ 0 Moderate/ severe ¼ 3

Mid-trimester scan: Mild: AP > 5 mm and 7 mm Moderate: AP > 7 mm and 10 mm Severe: AP > 10 mm AP > 5 mm and <10 mm

e

67 (mild ¼ 38; moderate ¼ 20; severe ¼ 9)

e

74% of mild cases resolved spontaneously

e

280

e

51/280 (18.2%)

66/229 by 1 month (28.8%)

22/229 (9.6%)

AP  4 mm <23 weeks AP  10 mm > 28 weeks AP  5 mm at 20 weeks Mild AP  4 mm and <10 mm

11,465

268

2.3%

152/227 (67%)

48/75 (64%)

11/75 (14.7%)

7000

139

2%

e

1/104 (0.96%)

5529

306 (251 for follow up, 165 one ultrasound only)

5.5%

6/129 (4.7%)

47/104 (45.2%) 109/226 by 1 month (48.2%)

9/226 (4%)

AP, anterior–posterior renal pelvis diameter The first column details author and year of publication. Subsequent columns details exclusion criteria, whether study was retrospective or prospective, definition of hydronephrosis used in paper, total number of patients assessed, number of cases of hydronephrosis and prevalence, percentage of cases showing resolution in the third trimester and postnatally and finally percentage of patients requiring surgical intervention (as a percentage of those with hydronephrosis in the third trimester).

100 R. K. Morris and M. D. Kilby

Table 1. Recent papers studying the natural history of hydronephrosis, including details of study design, gestation, diagnostic criteria, spontaneous resolution and need for postnatal surgery.

Congenital urinary tract obstruction 101

(0.34–0.68) stabilised. This review did not analyse postnatal outcome, however; this has been addressed in a different paper, which performed a meta-analysis looking at antenatal hydronephrosis as a predictor of postnatal outcome.22 A total of 1645 citations was screened and 17 studies included (104,572 patients screened) with 1308 subjects with antenatal hydronephrosis. The risk of any postnatal pathology per degree of antenatal hydronephrosis was 11.9% (4.5–28.0) for mild hydronephrosis (APPD  7 mm second trimester/9 mm third trimester), 45.1% (25.3–66.6) for moderate hydronephrosis (APPD 7–10 mm second trimester/9–155 mm third trimester) and 88.3% (53.7– 98.0) for severe hydronephrosis (10 mm second trimester/15 mm third trimester). In LUTO, the effects are often more variable. Long-term urethral obstruction is potentially associated with cystic renal dysplasia, abnormal renal (glomerular and tubular) function leading to severe oligohydramnios, pulmonary hypoplasia and positional limb anomalies.1 If untreated, fetal LUTO carries a mortality rate of 45%, which is mainly due to the association of severe oligohydramnios in the mid-trimester with pulmonary hypoplasia.23 Of those who survive the neonatal period, 25–30% develop end-stage chronic renal impairment necessitating dialysis and/or transplantation.24 In fact, congenital obstructive uropathy accounts for up to 60% of all paediatric renal transplants.25 Several papers have reviewed the natural history of LUTO. Most of the series are small; their results are summarised in Table 2. These papers have also looked at the antenatal factors that correlate with postnatal outcome and include the timing of ultrasound detection, associated anomalies (structural or/and chromosomal) and degree of oligohydramnios. The latter appears to be the most important, with oligohydramnios in one case series26 having an overall mortality rate of 80%, predominantly due to the associated risks of pulmonary hypoplasia. Independent predictors of the occurrence of pulmonary hypoplasia include gestational age at which premature Table 2. Summary of papers showing natural history of lower urinary tract obstruction.124 Author and year (ref. No.)

Thomas et al 1985120 Mahoney et al 198526 Nakayama et al 1986121 Hayden et al 1988122 Reuss et al 1988123 Anumba et al 200510)

Total/mean values

Number of cases

Mortality

Cystic renal disease/chronic renal failure

Pulmonary hypoplasia

Associated structural or chromosomal anomalies

18 40 11 14 43 113

33% 63% 45% 64% 72% Prenatal detection 75% (includes TOP) Postnatal detection 53%

56% 45% 37% e 36% Detection before 24 weeks 67% Detection after 24 weeks 40%

30% 40% 48% 36% 10% Without shunting 26%

56% e e 43% 42% 23%

58%

47%

31%

239

With shunting 25% 41%

The first column includes the authors and date of publication. Subsequent columns detail the number of cases from each study with the mortality rates, percentage of cases developing cystic renal dysplasia or chronic renal failure, percentage with pulmonary hypoplasia and percentage with associated structural or chromosomal anomalies. The final row is a mean of all these results. TOP, termination of pregnancy. Reprinted from ‘Congenital lower urinary tract obstruction and the efficacy of vesicoamniotic Shunting’ by R Morris, K Khan & M Kilby In: Studd J, Tan S & Chervenak F (eds.). Progress in Obstetrics and Gynaecology 2006; vol. 17. pp. 78–97. ª 2006, with permission from Elsevier.

102 R. K. Morris and M. D. Kilby

membrane rupture occurs27; however, these data relate mainly to pregnancies that are complicated by amniorrhexis, and should therefore be extrapolated with caution. PATHOPHYSIOLOGY The consequences of congenital urinary tract obstruction are not the same of those of postnatal obstruction because of the differing physiology of the fetal kidney, the interaction of the fetal–placental unit, low renal blood flow and the immaturity of the fetal endocrine and vascular systems, in particular the renin–angiotensin system. In animal studies, the fetal sheep has been the most commonly utilised model, with the majority of the work being done by Michael Harrison and his group in San Francisco.28,29 Using this model, the group assessed the effect of ureteric and urethral obstruction, both partial and complete, and its subsequent correction on pulmonary and renal development (Figure 1). These experiments demonstrated that complete urethral obstruction produced severe hydronephrosis, hydroureter, megacystis and urinary ascites, as well as significant pulmonary hypoplasia (measured by weight and volume). However, Harrison and colleagues could not demonstrate that urethral obstruction caused cystic and dysplastic renal changes. It was also shown that some of the urinary tract dilatation resolved with decompression in utero.28,29 Further experiments producing complete unilateral ureteral obstruction in the lamb showed that kidneys were not

A

B

urachus

C

D

urethra amniotic cavity Pudenz catheter to amniotic cavity

Balloon cuff constrictor

E

Ameroid constrictor

Figure 1. Urethral ligation in sheep model. (after Harrison, Golbus and Filly 1990126). This diagram represents attempts made by Harrison and Filly to simulate urethral obstruction and congenital hydronephrosis in fetal lambs. (A) represents ligation of the urethra; (B) ligation of the urethra and urachus; (C) ligation of the urachus and intermittent occlusion of the urethra with a balloon constrictor; (D) ligation of urachus and pressure limited obstruction with a Pudenz catheter; (E) ligation of urachus and gradual occlusion of the urethra with an ameroid constrictor. Reprinted from ‘The unborn patient: prenatal diagnosis and treatment’ by Harrison et al. ª 1990, with permission from Elsevier.

Congenital urinary tract obstruction 103

grossly hydronephrotic but were dysplastic by histological criteria with a contralateral unobstructed normal control kidney. The changes were similar to those seen in the human fetus.30 Again, the team went on to look at the effect of in-utero decompression; their results showed that recovery of renal functional was directly proportional to the duration of in-utero decompression and inversely proportional to duration of obstruction. The decompression also either prevented or greatly ameliorated the development of renal dysplasia.31 Other animal models have reproduced the finding of renal dysplasia with ureteric obstruction including the primate.32 Harrison et al postulated that pulmonary hypoplasia was secondary to compression from the massively dilated urinary tract or urinary ascites, and not primary pulmonary malformation. They claim that the results of their studies showing that restoration of the amniotic fluid allows an increase in lung weight and a clinical improvement in respiratory function support this hypothesis. They have also shown the same results in rabbit models.33 In the fetal lamb, opossum and rabbit34–36, experiments have shown that complete unilateral ureteric obstruction results in fewer glomeruli and reduced glomerular surface area. This arrested renal development had also been observed in an experiment on fetal preparations from aborted fetuses, those with lower urinary tract obstruction showed an arrest of nephrogenesis.37 The applicability of animal models to human congenital disease remains controversial. Harrison et al could not reproduce the histological changes of renal dysplasia seen with urethral obstruction in the human fetus. Other authors, using in-vitro chick embryos, have shown that obstruction of the ureter of the developing kidney caused only hydronephrosis and could not demonstrate renal dysplasia. They postulated that obstruction might not be the sole cause of the renal dysplasia seen in bladder outlet obstruction.38 Further animal work using the ovine model has gone on to look at the specific mechanisms of congenital urinary tract obstruction, including the role of the renin–angiotensin system39,40 and the influence of certain growth factors such as insulin-like growth factor 1 (IGF-1).41 Other advances in the understanding of the pathogenesis of congenital urinary tract obstruction have occurred in the field of medical genetics. Over 30 specific genes have been identified in the development of the mammalian kidney and urinary tract. The fact that abnormalities of the genitourinary tract, if concurrently present, can sometimes take on a familial pattern and be associated with other structural anomalies or genetic syndromes gives weight to the theory of a genetic component to their development. Animal experiments in transgenic mice have shown that lack of the a3 subunit or both b2 and b4 subunits of the nicotinic acetylcholine receptor (nAChR) subunit results in a phenotype similar to that of MMIHS, suggesting a basis for that condition.42,43 In 2001, Richardson and his group examined tissue from patients with MMIHS phenotypes and controls. They found that the MMIHS tissue gave negative staining for the a3 subunit.14 Further work has looked at mutation analyses in the a3 and b4 genes in MMIHS families enabling the human gene encoding the b4subunit to be fully characterised. Analysis of disease families and controls identified numerous genetic variants.44 DIAGNOSIS AND ASSESSMENT Ultrasound Assessment of the fetal genitourinary tract forms a part of all routine ultrasound examinations. When abnormalities are detected there should be a detailed assessment

104 R. K. Morris and M. D. Kilby

focusing on amniotic fluid volume, renal size, parenchyma, collecting system and bladder size. Dilatation of the renal pelvis and fluid-filled areas as small as 1–2 mm can be visualised in utero using high-resolution real-time ultrasound.45 The types of abnormality that can be detected are shown in Table 3. As already stated, the hallmark of upper urinary tract obstruction is hydronephrosis, with the majority of cases being diagnosed at the detailed second-trimester ultrasound scan (Figure 2). The EUROSCAN study group reported data for the detection of renal abnormalities across 12 European countries. The highest prenatal detection rate was found in the UK (92%) and was due to the routinely performed ultrasound at 18–24 weeks.3 The accuracy of antenatal ultrasound to detect renal pathology for upper urinary tract obstruction depends on the definition of hydronephrosis antenatally and the diagnosis of obstruction postnatally. As shown in Table 1, these definitions vary. The majority of authors use the anteroposterior diameter of the renal pelvis to classify dilatation; the Society for Fetal Urology has a different classification system46, based on calyceal dilatation, but some authors feel that this system is less reproducible.47 Reported false-positive rates vary from 6% to 80% depending on the pelvic diameter cut-off used.45,48 Further studies have shown that accounting for gestational age in the definitions can improve accuracy.49 False-positive results are important because they lead to inappropriate investigation, testing and parental anxiety.50 In any assessment of accuracy of a diagnostic test it is important to consider false-negative results. However, this can be difficult with hydronephrosis because postnatal ultrasound is not a routine investigation in the newborn. A retrospective analysis of children presenting with urinary tract infection in infancy reported that 56% had had a normal prenatal ultrasound.51 In addition, some studies have indicated isolated cases in which prospective differentiation between ‘obstruction’ and ‘reflux’ using prenatal ultrasound can be difficult.52

Table 3. Urinary tract anomalies detectable by ultrasonography. Uropathy/renal failure

Bladder distension

Obstructive Ureteropelvic junction obstruction Ureterovesical junction obstruction Multicystic dysplastic kidney Ureterocele/ectopic ureter Posterior urethral valves Urethral atresia

Rarely if bilateral Rarely if bilateral Only if bilateral Rarely if bladder outlet obstructed Up to 50% of cases Always

No No No Possible Yes Yes

Non-obstructive Physiological dilatation (<1 cm) Vesicoureteral reflux Megacystis-megaureter-microcolon Prune belly syndrome Renal agenesis Infantile polycystic kidneys

No No No Yes with urethral atresia Stillbirth if bilateral Depends on degree of penetrance

No Possible Yes Yes No No

The first column shows the obstructive and non-obstructive types of urinary tract anomalies that can be detected at ultrasonography as described by Coplen et al.111 The second column describes the likelihood of the development of renal failure. The third column describes whether bladder distension would be present for each anomaly. Reprinted from Journal of Urology, vol. 157, Coplen DE. Prenatal intervention for hydronephrosis, 2270–2277. ª 1997 with permission from The American Urological Association.

Congenital urinary tract obstruction 105

Figure 2. An ultrasound image showing hydronephrosis.

The detection of fetal LUTO using ultrasound has a good accuracy13,53, partly because at this level the anomalies of the renal tract and kidneys are also associated with secondary findings, such as oligohydramnios. The main sonographic findings associated with urethral (LUTO) obstruction include enlarged fetal bladder and proximal urethral dilatation and thickening of the bladder wall, which appears echogenic due to muscular hypertrophy (Figure 3).54 Ultrasonography can help in the differentiation of obstructive and non-obstructive causes of megacystis, with the association of increased echogenicity and oligohydramnios in the presence of bladder distension being predictive of an

Figure 3. An ultrasound image of a fetal bladder with lower urinary tract obstruction, the dilatation of the proximal urethra gives the ‘keyhole’ appearance.

106 R. K. Morris and M. D. Kilby

obstructive aetiology in approximately 87% of cases53; it is, however, of limited value in differentiating PUV from other causes of LUTO.53,55 The correlation between ultrasound and anatomical findings in LUTO has been assessed by Robyr et al28 in a small cohort study. Detailed post-mortem examinations were carried out on 24 male fetuses after termination of pregnancy. Presentation was prior to 25 weeks gestation with ultrasound evidence of isolated severe LUTO. In 20 cases, PUV were suspected antenatally and urethral atresia in none. At post mortem, urethral atresia was demonstrated in six cases and PUV in nine. Hydronephrosis was more frequent in cases with PUV (8/9) and urethral stenosis (6/8) than with urethral atresia (0/6). In LUTO presenting in the first and second trimester, hyperechogenic kidneys were predictive of renal dysplasia in 95% of cases. The association of a sagittal diameter of the bladder of at least 40 mm with hydronephrosis before 28 weeks was predictive of PUV with a positive (PPV) and negative (NPV) predictive value of 44.4% and 66.6%, respectively. Absence of hydronephrosis and a sagittal diameter of the bladder of less than 40 mm were predictive of urethral atresia or stenosis with a PPV and NPV of 100% and 47.6%, respectively. The absence of hydronephrosis was predictive of urethral atresia with a PPV and NPV of 66.6% and 100%, respectively.13 This measure of fetal bladder sagittal diameter has been assessed in other studies and a measurement of greater than 95% of the upper limit of normal shown to be associated with a postnatal diagnosis of uropathy.56 Detection of PUV varies with gestation, with less than 50% being detected on routine second-trimester scans performed before 24 weeks. Scanning after 28 weeks can increase detection to 80%.57,58 Oliveira et al performed a review of 148 cases of children with fetal hydronephrosis. A number of variables were assessed but only two were identified as independent predictors of fetal urethral obstruction: oligohydramnios [odds ratio (OR) ¼ 5, 95% confidence interval (CI) ¼ 1.3–15, P ¼ 0.01] and megacystis (OR ¼ 9, 95% CI ¼ 2.0–40, P ¼ 0.004). The sensitivity and specificity of the combination of both variables were 60 and 98.5%, respectively.59 It is important to differentiate true urethral obstruction from MMIHS, which is more common in females and has a very poor prognosis. As it can be difficult to differentiate on ultrasound, it is important to consider this diagnosis carefully in a female with LUTO but normal liquor volume. The pathogenesis of this condition might be explained by lack of the a3 subunit or both b2 and b4 subunits of the nicotinic acetylcholine receptor subunit, as seen in transgenic mice with a similar phenotype to MMIHS.42,43 Examination of tissue from MMIHS patients showed negative staining for the a3 subunit.14 Several case reports of MMIHS in sibling pairs provide evidence for an autosomal recessive inheritance60–62; genetic analysis of the human gene encoding the b4 subunit of disease families and controls has identified numerous genetic variants.44 Magnetic resonance imaging MRI is increasingly being used in prenatal diagnosis when ultrasound has proved unsatisfactory. Developments in MRI techniques, particularly the advent of single-shot fastspin echo (SSFSE) techniques, have allowed MRI of the fetus to be performed without sedation.63 A lack of data means that MRI is still best avoided in the first trimester; contrast media should also be avoided. The area in which MRI has been shown to be most beneficial is in the imaging of the fetal CNS, and the majority of the literature relates to its use in this area. The use of MRI in the imaging of the fetal pelvis is less well documented, although small studies have been done to assess its use in fetal

Congenital urinary tract obstruction 107

urinary tract anomalies and its benefit might lie in the imaging of those patients with oligohydramnios in whom ultrasound imaging is suboptimal. Poutamo et al have documented that additional information by MRI was gained from five fetuses out of 12 pregnancies with oligohydramnios.64 In 2004, Cassart et al looked at 16 third-trimester fetuses with suspected bilateral urinary tract anomalies following ultrasonography. The addition of MRI to sonography modified the diagnosis in five cases, in four of which it actually resulted in a diagnosis that changed the decision to continue or terminate the pregnancy.65 In the future, this imaging modality might improve the accuracy of diagnosis. In-utero percutaneous cystoscopy For LUTO, the option of cystoscopy exists to aid diagnosis. Quintero and colleagues investigated 11 patients with sonographic findings of lower obstructive uropathy and performed cystoscopy with a 0.7-mm fibreoptic endoscope.66 The endoscopic appearance of the proximal urethra was in agreement with the sonographic image in 10/11 fetuses. This combined sonographic/endoscopic technique allowed a diagnosis of prune belly syndrome in one fetus, megacystis-microcolon in another and bladder-outlet obstruction in seven others. Although there was direct visualisation of the proximal urethra, the investigators could not distinguish between PUV and urethral atresia due to difficulty in negotiating the posterior urethra. One advantage of this technique was that in two cases the procedure was extended to allow introduction of a urethral vesicoamniotic shunt.66 In-utero percutaneous cystoscopy has also been performed in European centres. Fisk’s group reported cystoscopy in 13 fetuses, with successful visualisation of the bladder wall in 12 and of the bladder neck in 11.67 The cause of the obstruction was identified in five cases (four PUV and one urethral atresia). Therapies utilised included saline hydro-ablation and/or guide-wire passage, and were successful in 60% cases.67 This assessment is currently undergoing further audit as recommended by the National Institute of Health and Clinical Excellence (NICE) in the UK (see: http://www.guidance.nice.org.uk/IPG205/?c¼91520). Antenatal assessment When a diagnosis of urinary tract obstruction is made, further investigations must be performed. In the first instance, a detailed ultrasound, by an experienced operator, should be performed to determine fetal sex, assess liquor volume, assess renal architecture and examine for other structural anomalies. Fetal karyotyping should be offered; as shown in Table 1, there is a high incidence of chromosomal abnormalities with LUTO. If there is severe oligohydramnios then amnioinfusion might be needed to allow accurate evaluation of fetal structures. Table 4 shows the other structural anomalies associated with LUTO. Mild unilateral hydronephrosis (in the absence of other soft markers) is usually an isolated condition; with bilateral hydronephrosis there is also an association with other structural malformations and chromosomal defects. Nicolaides et al report data from a 6-year period during which blood karyotyping was performed in 682 fetuses with renal defects. The overall incidence of chromosomal abnormalities was 12%, with an incidence in females (18%) almost double that in males (10%). The risk for fetal chromosomal anomalies was three times higher when there was an isolated renal defect and 30 times higher when there were additional malformations.68

108 R. K. Morris and M. D. Kilby

Table 4. Non-urinary tract anomalies associated with lower urinary tract obstruction (LUTO). Associated anomalies

Prenatal LUTO (n ¼ 56)

Other renal Chromosomal Cardiac Rectal atresia Central nervous system Multiple Cloacal dysgenesis

12 3 2 2 1 2 1

(21.4%) (5.3%) (3.6%) (3.6%) (1.8%) (3.6%) (1.8%)

This table shows data from a paper by Anumba et al10 and shows in column one the types of anomalies associated with lower urinary tract obstruction, column two shows the percentage of each of these anomalies seen in prenatally diagnosed cases of LUTO. Reprinted from ‘Diagnosis and outcome of fetal lower urinary tract obstruction in the northern region of England’ By D O Anumba et al. ª 2005 John Wiley & Sons Ltd. Reproduced with permission.

Predictors of outcome that can be identified on ultrasound include: gestational age at diagnosis (detection at or before 24 weeks of gestation has been shown to predict poor outcome)26,57,69; liquor volume; oligohydramnios predicting poor outcome (RR 10.6 (2.7–77)70 and kidney appearance; evidence of renal dysplasia, such as cortical cysts; and echogenic parenchyma, although the last might be of limited value.10,71,72 This has resulted in investigation into other methods of determining renal function in utero, including analysis of fetal urine, serum or amniotic fluid and finally renal biopsy. Fetal urine Assessing fetal renal function in early pregnancy has been utilised in fetal triage before performing in-utero therapy. This is the most commonly used investigation to inform prognosis. Tubular dysfunction can be indicated by raised urinary sodium levels, with values <90 mmol/L being normal at 20–30 weeks of gestation. Fetal urinary sodium or chloride values in excess of 100 mmol/L have also been shown to be highly predictive of fetal or perinatal death from terminal renal or pulmonary failure.72,73 Other electrolytes (calcium, potassium and phosphate) mainly reflect tubular function. Glomerular function can be assessed by a low-molecular-weight protein, b2microglobulin, which is filtered by the glomeruli. In normal kidneys, >99.9% is reabsorbed and metabolised in the proximal tubules; in renal disease with damage to this area, b2-microglobulin is excreted in the urine, although this pathophysiology has not been proven in the fetal kidney. Lipitz et al reported that a b2-microglobulin >13 mg/L was invariably associated with a fatal outcome.74 Fetal serum b2microglobulin has been used as an index of fetal glomerular filtration rate (GFR) and in the prediction of postnatal GFR.75,76 It has been suggested that, in the presence of urinary tract obstruction, the first urine sampled does not accurately reflect kidney function, as this urine might have been present in the bladder for some time. With sequential sampling there might be an improvement in renal function secondary to the decompression. However, values that remain stable or increase are more indicative of renal dysplasia. This has led two groups to look at the changes in electrolyte values in urine with sequential sampling, demonstrating that electrolyte values drop when fetal urine is re-sampled

Congenital urinary tract obstruction 109

after 24 hours.77,78 Johnson et al went on to show that sequential fetal urine analysis improves the discriminatory accuracy of identifying fetuses with severe underlying renal damage from those that might benefit from vesico-amniotic shunting. A systematic review to evaluate the clinical usefulness of analysis of fetal urine in the prediction of poor postnatal renal function in cases of congenital urinary tract obstruction concluded that the current evidence is insufficient.79 Twenty-three articles met the selection criteria, including a total of 572 women. The two most accurate tests were calcium > 95th centile for gestation (positive likelihood ratio 6.65, 0.23–1.96; negative likelihood ratio 0.19, 0.05–0.74) and sodium > 95th centile for gestation (likelihood ratio positive 4.46, 1.71–11.6; likelihood ratio negative 0.39, 0.17–0.88). Many of the studies were small-case cohort series with different control groups. Normal ranges were also poorly defined in this group, measurement variation with gestational age is not often taken into account and the outcome parameters also vary considerably. The results from this review are summarised in a Forest plot (Figure 4). Fetal urine N-acetyl-B-glucosaminidase was examined by Lipitz et al and was found to have no correlation with outcome. Other groups have found raised levels of cystatin C80, urinary insulin-like growth factor 1 (UIGF-1) and urinary insulin-like binding protein 3 (UIGFBP-3)41 in cases of bilateral obstructive uropathy. UIGF-1 had a sensitivity of 90% and specificity of 88% for predicting serum creatinine > 50 mmol/L; UIGFBP-3 has a sensitivity of 80% and specificity of 88%. Fetal urinalysis has also been used to select suitable candidates for intrauterine shunting and numerous studies have been published on its accuracy.41,75,81–83 In this case scenario, urinary sodium and calcium have the best accuracy (sensitivity 70–100%, specificity 60–80%) but for a single measure urine b2-microglobulin has the best accuracy. More recently, a newer technique – proton nuclear magnetic resonance spectroscopy – has been used to assess amino acid levels in fetal urine, particularly levels of alanine/valine and valine/threonine. In small case series, amino acid measurement has been shown to be more accurate than any of the other investigated electrolytes in predicting postnatal renal function.81 With further research this might prove to be a useful diagnostic test both antenatally and postnatally. Fetal serum Serum creatinine cannot be used to assess renal function in the fetus because it crosses the placenta and is cleared by the mother.84 This, however, is not the case for serum microglobulins such as a1-microglobulin, retinol binding protein and b2microglobulin due to their molecular weight (<40,000 Da), which means they cannot cross the placenta but are still filtered by the glomerulus. These serum measures can thus provide a better quantification of glomerular function.85 Another advantage of serum measurement of b2-microglobulin is that concentrations do not seem to vary with gestational age86–88, although this has been disputed in one paper demonstrating a decrease in levels from 31 weeks.89 a1-microglobulin has been assessed in serum, urine and amniotic fluid. Microproteins in urine have been demonstrated to be of fetal origin.90 In 1996, Cobet et al looked at blood from nine fetuses with severe bilateral renal dysplasia or agenesis and found elevated levels of a1-microglobulin.91 Thus serum monitoring, while technically more difficult and with increased risks for the fetuses conveys advantages over urine sampling that include: better reflection of

110 R. K. Morris and M. D. Kilby Test and author Sodium

Threshold

Bunduki et al 1998

Sodium > 100mEq/l

Bussieres et al 1995

Sodium > 50mmol/l

Negative LR (95% CI)

Positive LR (95% CI)

0.54

(0.22 - 1.32)

7.00

(0.42 - 116.40)

0.19

(0.04 - 0.84)

3.26

(1.43 - 7.45)

0.06

(0.00 - 0.93)

9.44

(0.68 - 131.60)

0.06

(0.00 - 0.89)

4.24

(1.86 - 9.67)

0.03

(0.00 - 0.40)

34.45

(0.35 - 0.81) (0.18 - 0.50) (0.06 - 0.66)

32.30 (1.98 - 525.70) 10.07 (2.13 - 47.64) 3.71 (1.22 - 11.26)

Sodium > 95th centile for gestation Sodium > 2sd

0.53 0.30 0.20 0.58

(0.32 - 1.02)

Bussieres et al 1995

Beta 2 > 2.5mg/l

0.23

(0.08 - 0.70)

26.27

Daikha-Dahmane et al 1997

Beta 2 > 10mg/l

0.36

(0.18 - 0.70)

5.48

Freedman et al 1996 Freedman et al 1996

Beta 2 > 2mg/l Beta 2 > 6mg/l

0.08 0.63

(0.00 - 1.50) (0.45 - 0.88)

1.31 (0.96 - 1.78) 11.08 (0.70 - 176.24)

Johnson et al 1995 Johnson et al 1995

Beta 2 > 4mg/l (first urine) Beta 2 > 4mg/l (second urine)

0.67 0.77

(0.42 - 1.07) (0.54 - 1.12)

11.20 (0.64 - 194.77) 8.00 (0.43 - 150.09)

Mandelbrot et al 1993

Beta 2 > 2mg/l

0.48

(0.27 - 0.85)

15.00

Muller et al 1999 Muller et al 1999

Beta 2 > 5mg/lb Beta 2 > 5mg/lc

0.03 0.51

(0.00 - 0.47) (0.32 - 0.82)

5.44 7.14

(3.04 - 9.72) (2.02 - 25.27)

Johnson et al 1994

Calcium > 8mg/dl

0.14

(0.01 - 2.23)

1.34

(1.01 - 1.77)

Nicolaides et al 1992

Calcium > 95th centile for gestation

0.26

(0.14 - 0.46)

27.00

Nicolini et al 1992

Calcium > 95th centile for gestation

0.05

(0.00 - 0.85)

2.37

(1.16 - 4.84)

0.54

(0.22 - 1.32)

7.00

(0.42 - 116.40)

0.06 0.04

(0.00 - 0.93) (0.00 - 0.68)

9.44 3.35

(0.68 - 131.60) (1.65 - 6.81)

Grannum et al 1989 Johnson et al 1995 Muller et al 1999 Muller et al 1999 Nicolaides et al 1992 Nicolini et al 1992 Miguelez et al 2006

Sodium > 100mmol/l Sodium > 100mg/dla Sodium > 75mmol/lb Sodium > 75mmol/lc Sodium > 95th centile for gestation

9.90

(7.10 - 167.10)

(0.61 - 161.74)

Beta 2 microglobulin (1.68 - 410.77) (0.40 - 74.20)

(0.94 - 238.24)

Calcium

Chloride Bunduki et al 1998

Chloride > 90mEq/l

Grannum et al 1989 Johnson et al 1995

Chloride > 90mmol/l Chloride > 90mg/dl (second urine)

(1.74 - 418.12)

Osmolality Grannum et al 1989

Osmolality > 210mOsm/l

0.07

(0.00 - 1.07)

7.50

(0.56 - 100.87)

Johnson et al 1995

Osmolality > 200mOsm/l (last urine)

0.06

(0.00 - 0.83)

5.45

(2.09 - 14.25)

Reuss et al 1987

Osmolality > 210mOsm/kg

0.17

(0.01 - 2.98)

1.83

(0.58 - 5.83)

Total protein > 20mg/dl

0.39

(0.18 - 0.83)

6.02

(1.75 - 20.69)

Bussieres et al 1995

IGF-1 > 50pg/l

0.16

(0.04 - 0.72)

5.87

(1.83 - 18.90)

Bussieres et al 1995

IGFBP-3 > 30ng/l

0.27

(0.09 - 0.81)

5.25

(1.60 - 17.25)

Eugene et al 1994

Valine/threonine ratio NMR

0.17

(0.05 - 0.53)

5.41

(2.03 - 14.40)

Nicolaides et al 1992

0.20

(0.10 - 0.41)

9.77

(2.09 - 45.59)

Muller et al 1999

Creatinine < 5th centile for gestation) CystatinC > 1mg/lb

0.21

(0.08 - 0.54)

(2.69 - 9.84)

Muller et al 1999

CystatinC > 1mg/lc

5.14

0.65

(0.46 - 0.93)

8.50

(1.61 - 44.87)

0.14

(0.01 - 1.84)

10.50

Total protein Johnson et al 1994

Other urinary analytes

Combined urinary analytes Qureshi et al 1996

Sodium, chloride, osmolality, Calcium, β2 microglobulin, total protein

(0.72 - 153.07)

a second urine b severe renal disease c mild renal disease

0.01

1

1

Negative LR

Positive LR

100.0

Figure 4. Individual results for index tests likely to be at least moderately useful (either positive likelihood ratio  5 or negative likelihood ratio  0.2) in prediction of postnatal renal function in fetuses with obstructive uropathy.79 The diagram shows results for all studies in the systematic review with either a positive likelihood ratio  5 or a negative likelihood ratio  0.2. The first column details the urinary analyte under investigation and the author and year for the paper. The second column details the threshold used for the urinary analyte in the paper. The figure shows the respective positive and negative likelihood ratio for the test with 95% confidence intervals as a Forest plot. Reprinted form ‘Systematic review of accuracy of fetal urine analysis of predict poor postnatal renal function in cases of congenital urinary tract obstruction’ by R K Morris et al. ª 2007 John Wiley & Sons Ltd. Reproduced with permission.

glomerular function, possibility of serial monitoring, option of testing after placement of vesico-amniotic shunt and testing when there is minimal fetal urine (Table 5). Fetal renal biopsy Small case series in the literature report fetal renal biopsy and compare this with histological findings, ultrasound and fetal urinalysis in fetuses with bilateral obstructive uropathy.92,93 The biopsies are performed under ultrasound guidance using fine-needle aspiration. However, although no maternal complications were reported, the success

Congenital urinary tract obstruction 111

Table 5. Summary of papers looking at accuracy of fetal serum measurements in predicting renal dysplasia/chronic renal failure (from ref. 124). Reprinted from ‘Congenital lower urinary tract obstruction and the efficacy of vesicoamniotic Shunting’ by R Morris, K Khan & M Kilby In: Studd J, Tan S & Chervenak F (eds.). Progress in Obstetrics and Gynaecology 2006; vol. 17. pp. 78–97. ª 2006, with permission from Elsevier. Parameter

b2-microglobulin (mg/L)

b2-microglobulin (mg/L) Cystatin C (mg/L)

b2-microglobulin (mg/L)

Threshold After data from Berry et al 1995 (86) n ¼ 15 (post-mortem and serum creatinine at birth) >5.6 mg/L (control group) After data from Bo¨kenkamp et al 2001125 n ¼ 84 (serum creatinine at birth) Dynamic upper limit from control group Dynamic upper limit from control group After data from Dommergues et al 200076 n ¼ 61 (serum creatinine >50 mmol/L postnatal) >5 mg/L

Sensitivity

Specificity

80%

98.6%

90% 63.6%

85.5% 91.8%

67%

100%

rate in obtaining fetal material was poor (50%) thus limiting the use of the procedure. In a third of cases, renal histology added to the diagnosis and in one changed the prognosis so that termination was not performed.93 A review of biomarkers for congenital obstructive uropathy reported some exciting future developments. New techniques include immunohistochemical analysis and laser capture microscopy to look at the cellular response of the developing kidney to urinary tract obstruction. These included components of the renin–angiotensin system, transforming growth factor-b, monocyte chemoattractant protein-1 and epidermal growth factor.94 If a normal karyotype is confirmed in a fetus with LUTO and no other congenital anomalies, then consideration can be given to in-utero treatment. THERAPEUTIC OPTIONS A full assessment of the fetus allows the clinician to determine those in whom antenatal intervention might be beneficial, allows parents to be counselled and make informed decisions and enables the appropriate antenatal, intrapartum and postnatal care to be planned. In the presence of bilateral obstruction, no other structural or chromosomal anomalies, then intervention might be warranted, with the aim of relieving obstruction and thus improving kidney function, urine production and hence liquor volume with its effects on pulmonary development. The theoretical basis for intervention comes from Harrison and his group’s work on the fetal lamb (discussed above). For other fetuses, conservative management might be more appropriate. Termination of pregnancy In the presence of adverse prognostic criteria (severe oligohydramnios), early gestation of diagnosis and echogenic/cystic kidneys, termination of pregnancy must be discussed. Although therapeutic options for antenatal intervention exist, their effectiveness is far from proven and LUTO remains a condition with high mortality and morbidity. However, when parents make the difficult decision to terminate a pregnancy

112 R. K. Morris and M. D. Kilby

for fetal abnormality it must be recognised that persistent adverse psychological and social reactions are much more common than in those having termination for social reasons.95 It is therefore important to support parents both before and after this difficult decision. There are at present no data on the effects of termination for urinary tract abnormalities. Conservative management Isolated antenatal hydronephrosis can be managed conservatively. For those fetuses detected in the second trimester, a follow-up scan should be arranged in the third trimester to assess the level of dilatation due to the natural history of this condition. Those fetuses with persistent significant dilatation should be referred to the paediatricians postnatally for follow up. In the case of LUTO, conservative management includes regular antenatal monitoring to assess bladder dilatation, renal architecture and hydronephrosis. There is also a risk of intrauterine death. Parents should be adequately counselled by a paediatrician and booked for delivery in a unit with appropriate neonatal facilities. There are no contraindications to vaginal delivery; very occasionally, cystic kidneys might become sufficiently enlarged to obstruct labour necessitating operative delivery. Open fetal surgery The advances in technology for prenatal diagnosis, technical aspects of fetal intervention, maternal anaesthesia and drugs for tocolysis have all made fetal surgery possible over the last 20 years. The case for fetal surgery has been proven for some conditions where there is well-defined pathophysiology and natural history and high levels of mortality without intervention, for example congenital cystic adenomatoid malformation and fetal sacrococcygeal teratoma.96,97 The case is less clear for obstructive uropathy. Open hysterotomy with direct fetal approach served as the initial model for access to the fetus. Although these techniques were feasible, they were limited by the high complication rate associated with open surgery. The mother requires a laparotomy for the surgery and then one for delivery. There is also almost 100% complication by preterm labour immediately postoperatively, requiring extensive tocolytic use.98 Harrison et al reported the first successful in-utero decompression for hydronephrosis with open fetal surgery in 198199; they followed this with eight fetal surgeries (vesicostomies) performed at 17–24 weeks gestation for obstructive uropathy.99–101 There have been no new reports of open fetal surgery for obstructive uropathy since 1988102, due to the complications associated with open fetal surgery; adverse effects on future reproductive outcomes [uterine uterine dehiscence/rupture (12%/ 6%), caesarean hysterectomy (3%) and antepartum haemorrhage requiring transfusion (9%)]103, the risk of neurological injury to the fetus104 and the development of new fetoscopic techniques. Fetal surgery is also now limited to intervention for those fetuses with LUTO. Minimal invasive therapy Fetal cystoscopy Following on from its use as a diagnostic tool, cystoscopy has been used to administer therapy in LUTO including ablation of PUV105 and positioning of urethral

Congenital urinary tract obstruction 113

vesico-amniotic shunts.66 In 2003, the same treatment was reported from a European centre by Wesh et al, who successfully treated 6/10 fetuses by hydro-ablation or guidewire passage; five of these fetuses survived.67 Other techniques include fetal hydrolaparoscopy and endoscopic cystotomy. However, the authors advise that these techniques should be used only in complicated cases in which conventional shunting is not possible.102 Complications of minimalaccess fetal surgery include bleeding, preterm labour, chorioamniotic membrane separation and preterm prelabour rupture of membranes. The latter is the most common complication reported with fetoscopic techniques: 6–10% in single-port procedures106 and up to 60% in multiple-port procedures.107 Although technically more complicated, it is felt that cystoscopy might have advantages over shunting in that it can allow restoration of normal fetal bladder dynamics as opposed to the situation of chronic bladder decompression seen with shunting.108

Vesico-amniotic shunting and vesicocentesis This was first reported in 1982 by Golbus et al109 and is currently the most commonly used method to relieve urinary tract obstruction. Its efficacy, however, remains unproven. The procedure involves placement of a double pig-tailed catheter into the fetal bladder percutaneously. The first large case series of shunting reported in the literature was from the International Fetal Surgery Register in 1986, which audited 73 fetuses with ultrasound features of lower urinary tract obstruction that had been treated with indwelling vesico-amniotic shunts.110 Overall survival rates were 41%, although the authors felt that with appropriate patient selection perinatal mortality could be reduced to as low as 30%. In 1997 Coplen et al reviewed the five largest series of prenatal intervention comprising 169 cases of successful percutaneous shunt placements over 14 years (three from a single institution, one fetal registry and one review).111 The reported survival rate was 47%, most fetuses had oligohydramnios and failure to restore amniotic fluid volume was associated with 100% mortality. The survivors had a high proportion of end-stage renal disease (40%). The authors noticed that limiting intervention to fetuses with good prognosis improved survival and resulted in a lower incidence of renal failure in survivors. There was a high incidence of shunt-related complications: 45% of cases included shunt blockage, 25% shunt migration, 20% preterm labour and miscarriage, 5–15% amniorrhexis and 2–5% chorioamniotis and iatrogenic gastroschisis.112 In 2003, our group (Clark et al) published a systematic review and meta-analysis of the literature on prenatal therapy for bladder outflow obstruction.113 The review identified 16 observational studies, of which nine were uncontrolled case series (n ¼ 146 fetuses) and seven controlled case series (n ¼ 195 fetuses). The review demonstrated a lack of high-quality evidence to reliably inform clinical practice relating to prenatal bladder drainage in fetuses with ultrasound evidence of lower urinary tract obstruction. The literature consisted of many small studies, which used different criteria for patient selection, different surgical techniques and different outcome measures. The inclusion of a selective population with a heterogeneous mix of underlying aetiologies can result in reporting bias and overestimation of effectiveness if patients are selected for their favourable outcome. In the meta-analysis, confidence intervals were wide due to the small numbers of patients and uncertainty around the estimates of effect. There have also been comments that vesico-amniotic shunting

114 R. K. Morris and M. D. Kilby

uses ‘therapy’ in the form of shunts that are prone to high rates of bloackage/dislodgement and are therefore unreliable (R. Quintero, personal communication, 2007). The meta-anlaysis demonstrated an improvement in overall perinatal survival with vesico-amniotic drainage in utero (open, fetoscopic or percutaneous) compared to the non-drainage group (OR 2.5, 95% CI 1.0–5.9, P < 0.03). Subgroup analysis, however, indicated that this improved survival was predominantly noted in fetuses with a defined ‘poor prognosis’ (based on combination of ultrasound appearance and fetal urinary analytes) where there appeared to be marked improvement (OR 8.0, 95% CI 1.2–52.9, P < 0.03) (Figure 5). The review was unable to comment on the indication or timing of the vesicoamniotic shunting because the studies included were relatively small. In addition, heterogeneity of presentation and severity make evaluation of the literature more complex. It might be that the severest phenotypes, presenting with severe oligohydramnios before 20 weeks, are not ‘rescued’ by fetal therapy but that the milder phenotypes would benefit from vesico-amniotic shunting later in gestation to prevent deterioration in renal reserve. The answers to these questions are unknown. POSTNATAL MANAGEMENT The initial management of fetuses with LUTO can require intensive care, in particular respiratory support; once stabilised, postnatal investigations can be commenced. The initial assessment should include a thorough examination looking for abdominal masses (kidney, bladder) and for signs related to congenital syndromes, such as deficient abdominal wall musculature or genital abnormalities (e.g. undescended testes). A renal Effect of in-utero bladder drainage on perinatal and postnatal survival Outcome improved Outcome improved with with no drainage drainage Perinatal Survival All studies (n=4)

2.50 (1.00, 5.90)

Good prognosis (n=2)

2.80 (0.70, 10.80)

Poor prognosis (n=2)

8.00 (1.20, 52.90)

Postnatal survival All studies (n=4)

2.30 (0.90, 5.60)

Good prognosis (n=2)

1.90 (0.40, 8.90)

Poor prognosis (n=2)

9.30 (1.40, 62.00)

0.2

0.5

1

2

5

10

100

odds ratio (95% confidence interval) Figure 5. Forrest plot showing summary of effects of prenatal bladder drainage, using summary odds ratio, on perinatal and postnatal survival in fetuses with ultrasound evidence of lower urinary tract obstruction (analysis corrected for voluntary pregnancy terminations).

Congenital urinary tract obstruction 115

ultrasound should be arranged within the first week of life and repeated at 4–6 week of age. It is this investigation that determines the subsequent tests. The first postnatal ultrasound should not be performed within the first few days of life, when the kidneys produce relatively little urine, as this can give rise to false-negative results. The ultrasound should assess for renal pelvic dilatation, calyceal dilatation, pelvic or ureteral wall thickening and signs of renal dysplasia. When such abnormalities are found, significant pathology is found in 70% and further investigation is required however, if the postnatal renal ultrasound is normal only 3% will have pathology, usually low-grade vesico-ureteric reflux.114 Current recommendations are for antibiotic prophylaxis to be commenced routinely, although this has not been examined in the context of a randomised controlled trial. Further investigations can include micturating cystourethrography, diuretic renography, MRI and nuclear medicine scans to determine degree of reflux and obstruction, image the urinary tract and determine renal function. Biochemical investigations include serum creatinine and in some centres urinary sodium, calcium and b2-microglobulin. Surgical intervention should be considered in hydronephrosis only after a period of observation, because of the possibility of spontaneous resolution. It is important to realise, however, that with true obstruction early surgical intervention can help preserve renal function.115 OUTCOME The systematic review by Clark et al looked at perinatal survival as an outcome to assess the effectiveness of bladder drainage in utero. However, there is concern about the effectiveness of in-utero intervention in the long term. Cohort and case series with long-term follow-up have been reported in the literature. Holmes et al looked at the medium-term outcome of 36 patients, with favourable fetal electrolytes and oligohydramnios, treated at the University of California over an 18-year period.116 Fifty-seven per cent of the patients with PUV survived, of whom 63% had chronic renal impairment and underwent urinary diversion/reconstruction procedures. These longterm data indicate that whereas antenatal intervention might prevent intrauterine death or perinatal death due to pulmonary hypoplasia, serious morbidity is associated with this condition, which must be taken into account when counselling the parents. There have been other reports on the medium-term sequelae of this condition. In a case series reported by Freedman et al in 1999, 34 patients were followed to a mean age of 54.3 months.117 Five (36%) had renal failure and successful transplantation, six (43%) had renal insufficiency and six (43%) had normal renal function. The authors noted that height was below the 25th percentile in 12 (86%), with seven (50%) below the 5th percentile. Despite normal renal function or successful transplantation in those with abnormal function, it must be noted that there might be some bladder dysfunction. Fourteen per cent of children were incontinent, of whom 50% were acceptably continent; the remaining 36% had not yet commenced toilet training.117 Reassuringly, however, data have been reported on male children with LUTO followed until a mean of 5.83 years of age showing that overall 1-year survival was 91%. Despite the need for dialysis and renal transplantation in a third of survivors, the children were neurodevelopmentally normal, had acceptable renal and bladder function, and reported a satisfactory quality of life.118 There are not many long-term data for these children because the majority of those who had intervention antentally are only just reaching adulthood. One study by

116 R. K. Morris and M. D. Kilby

Holmdahl et al has looked at long-term follow up of 19 boys with PUV to 44 years of age and found that 32% were uraemic, 21% had moderate renal failure and that 48% had not been checked since adolescence. There were signs of bladder dysfunction in 40%, all with symptoms of detrusor weakness; however, all were continent. The ability to father children depended on whether or not the man was uraemic.119 However, these individuals were born before the advent of antenatal intervention and thus received only postnatal treatment. Holmdahl et al point out the importance of extended long-term follow-up for these patients. THE FUTURE In the future we can expect to see advances in the antenatal diagnosis of congenital urinary tract anomalies with the use of techniques such as MRI. Research in the areas of biochemical markers and proton nuclear magnetic resonance spectroscopy may help with the identification of patients most likely to benefit from intervention antenatally. With development in the area of minimally invasive fetal surgery, new treatment modalities such as fetal cystoscopy might allow early and more successful prenatal treatment. PLUTO PLUTO is a randomised, controlled trial designed to investigate the role of vesico-amniotic shunting in babies with moderate or severe lower urinary tract obstruction. It is funded by the charity Wellbeing for Women and sponsored by the University of Birmingham. Recruitment commenced in September 2005 and, currently, 13 centres in the UK have agreed to take part in the study and to receive ethics committee and Trust approval to act as collaborating centres. A further 10 centres are going through the administrative process to take part in the study. Power calculations of this study indicate that the aim over a 5-year period is to recruit 200 pregnancies, which will be randomised to either fetal vesico-amniotic shunting or conservative observation. In tandem with this, a prospective registry will run alongside the trial for women who are ineligible when either clinicians or patients feel that they cannot consent to randomisation. It is hoped that fetal urine will be sampled and analysed for urinary calcium, sodium and urinary b2-microglobulin. There will be no gestational limit for entry to the trial. However, subgroup analysis will be performed based on gestational age at diagnosis. It is hoped that this will give evidence for the most effective timing of vesico-amniotic shunting and, indeed, whether vesico-amniotic shunting of fetuses with lower urinary tract obstruction into the third trimester is of benefit. Initially, follow-up will be over a 12-month period, with primary outcome measures being perinatal mortality and markers of acute renal failure within this time period. Secondary measures of outcome will be noted in terms of the degree of reflux on a micturating cystourethrogram, bladder-wall thickness and fetal pelvi-ureteric dilatation on postnatal ultrasound. Serious adverse events, including procedure-related loss, prelabour ruptured membranes, preterm labour and shunt complications will also be collected. It is hoped that further funding will be obtained so that long-term follow-up of up to 5 years can assess both developmental outcomes and long-term complications such as the presence of urinary incontinence. Further information on this study can be obtained at http://www.pluto.bham.ac.uk

Congenital urinary tract obstruction 117

SUMMARY Congenital urinary tract obstruction is a heterogeneous condition. The natural history of the condition is variable, with upper urinary tract obstruction being associated with a high level of spontaneous resolution. By contrast, lower urinary tract obstruction is associated with high morbidity and mortality. Technical advances have allowed the introduction of tests to determine renal function and options for in-utero treatment. Despite this, the evidence can still not identify an acceptable method to predict long term renal function nor can we be certain that prenatal intervention is successful. Further research in to these important areas is required.

Practice points  Upper urinary tract obstruction, particularly if unilateral, is associated with a good outcome. LUTO is a disease of high mortality and morbidity.  Urinary tract obstruction can be diagnosed at second-trimester ultrasound.  Prognostic antenatal factors include level of obstruction, presence of other structural or chromosomal anomalies, amniotic fluid volume, kidney appearance, gestation of diagnosis and fetal sex.  Antenatal intervention is available but efficacy is not proven.  Management should be multi-disciplinary, with early involvement of neonatologists and paediatric urologists.

Research agenda  Further animal work to define mechanisms of kidney damage.  Identification of accurate urine or serum marker in the fetus to predict renal damage.  Recruitment to PLUTO to determine efficacy of vesico-amniotic shunting.

ACKNOWLEDGEMENTS Dr Morris is a clinical research fellow supported by a project grant from Wellbeing of Women. The PLUTO trial is funded by Wellbeing of Women and sponsored by the University of Birmingham. Professor Mark Kilby and Professor Khalid Khan are principal investigators for the PLUTO trial. REFERENCES 1. Merrill DC & Weiner CP. Urinary tract obstruction. In: Fisk NM & Moise Jr. KJ (eds.). Fetal therapy: invasive and transplacental. Cambridge, UK: Cambridge University Press, 1997, pp. 273–286.

118 R. K. Morris and M. D. Kilby 2. Kaplan GW & Scherz HL. Infravesical obstruction. In: Kelalis PP, King LR & Belman AB (eds.). Clinial pediatric urology. 3rd edn. Philadelphia, PA, USA: W.B.Saunders, 1992, pp. 821–864. *3. Wiesel A, Queisser-Luft A, Clementi M et al. Prenatal detection of congenital renal malformations by fetal ultrasonographic examination: an analysis of 709,030 births in 12 European countries. Eur J Med Genet 2005; 48(2): 131–144. 4. Mandell J, Peters C & Retik A. Current concepts in the perinatal diagnosis and mangement of hydronephrosis. Urol Clin North Am 1990; 17: 247–262. 5. Blyth B, Snyder HM & Duckett JW. Antenatal diagnosis and subsequent management of hydronephrosis [Review] [55 refs]. J Urol 1993 Apr; 149(4): 693–698. 6. Gloor JM. Management of prenatally detected fetal hydronephrosis [see comment]. [Review] [47 refs]. Mayo Clin Proc 1995 Feb; 70(2): 145–152. 7. Scott JE & Renwick M. Northern Region Fetal Abnormality Survey results 1987. J Pediatr Surg 1990 Apr; 25(4): 394–397. 8. Blachar A, Blachar Y, Livne PM et al. Clinical outcome and follow-up of prenatal hydronephrosis. Pediatr Nephrol 1994 Feb; 8(1): 30–35. 9. Ismaili K, Hall M, Donner C et al. Results of systematic screening for minor degrees of fetal renal pelvis dilatation in an unselected population. Am J Obstet Gynecol 2003 Jan; 188(1): 242–246. *10. 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 Jan; 25(1): 7–13. 11. Steinhardt G, Hogan W, Wood E et al. Long-term survival in an infant with urethral atresia. J Urol 1990; 143: 336–337. 12. Brand JR, Kaminopetros P, Cave C et al. Specificity of antenatal ultrasound in the Yorkshire Region: a prospective study of 2261 ultrasound detected anomalies. Br J Obstet Gynaecol 1994; 101: 392–397. *13. Robyr R, Benachi A, ikha-Dahmane F et al. Correlation between ultrasound and anatomical findings in fetuses with lower urinary tract obstruction in the first half of pregnancy. Ultrasound Obstet Gynecol 2005 May; 25(5): 478–482. 14. Richardson C, Morgan J, Jasani B et al. Megacystis-Microcolon-Intestinal Hypoperistalsis Syndrome and the absence of the alpha3 nicotinic acetylcholine receptor subunit. Gastroenterology 2001; 121(2): 350–357. 15. Gunn TR, Mora JD & Pease P. Antenatal diagnosis of urinary tract abnormalities by ultrasonography after 28 weeks’ gestation: incidence and outcome. Am J Obstet Gynecol 1995 Feb; 172(2 Pt 1): 479–486. 16. Ahmad G & Green P. Outcome of fetal pyelectasis diagnosed antenatally. J Obstet Gynaecol 2005 Feb; 25(2): 119–122. 17. Signorelli M, Cerri V, Taddei F et al. Prenatal diagnosis and management of mild fetal pyelectasis: implications for neonatal outcome and follow-up. Eur J Obstet Gynecol Reprod Biol 2005 Feb 1; 118(2): 154–159. 18. Sairam S, Al-Habib A, Sasson S et al. Natural history of fetal hydronephrosis diagnosed on mid-trimester ultrasound. Ultrasound Obstet Gynecol 2001 Mar; 17(3): 191–196. 19. Jaswon MS, Dibble L, Puri S et al. Prospective study of outcome in antenatally diagnosed renal pelvis dilatation [see comment]. Arch Dis Child Fetal Neonatal Ed 1999 Mar; 80(2): F135–F138. 20. Persutte WH, Koyle M, Lenke RR et al. Mild pyelectasis ascertained with prenatal ultrasonography is pediatrically significant [see comment]. Ultrasound Obstet Gynecol 1997 Jul; 10(1): 12–18. 21. Sidhu G, Beyene J & Rosenblum N. Outcome of isolated antenatal hydronephrosis: a systematic review and meta-analysis. Pediatr Nephrol 2006; 21(2): 218–224. 22. Lee RS, Cendron M, Kinnamon DD et al. Antenatal hydronephrosis as a predictor of postnatal outcome: a meta-analysis. Pediatrics 2006 Aug; 118(2): 586–593. 23. Freedman AL, Johnson MP & Gonzalez R. Fetal therapy for obstructive uropathy: past, present, future? [Review] [57 refs]. Pediatr Nephrol 2000 Feb; 14(2): 167–176. 24. Parkhouse HF, Barratt TM, Dillon MJ et al. Long term outcome of boys with posterior urethral valves. Br J Urol 1988; 62: 59–62. 25. Parkhouse HF & Woodhouse CR. Long-term status of patients with posterior urethral valves. Urol Clin North Am 1990; 17: 373–378. 26. Mahoney BS, Callen PW & Filly RA. Fetal urethral obstruction: US evaluation. Radiology 1985; 157: 221–224.

Congenital urinary tract obstruction 119 27. Vergani P, Ghidini A, Locatelli A et al. Risk factors for pulmonary hypoplasia in second-trimester premature rupture of membranes. Am J Obstet Gynecol 1994; 170(5 Pt 1): 1359–1364. *28. Harrison MR, Ross NA & Noall RA. Correction of congenital hydronephrosis in utero I: the model: fetal urethral obstruction produces hydronephrosis and pulmonary hypoplasia in fetal lambs. J Pediatr Surg 1983; 18: 247. 29. Harrison MR, Nakayama DK, Noall RA et al. Correction of congenital hydronephrosis in utero. II: decompression reverses the effects of obstruction on the fetal lung and urinary tract. J Pediatr Surg 1982; 17: 965. 30. Glick PL, Harrison MR, Noall RA et al. Correction of congenital hydronephrosis in utero III. Early mid-trimester ureteral obstruction produces renal dysplasia. J Pediatr Surg 1983; 18(6). 31. Glick PL, Harrison MR & Adzick NS. Correction of congenital hydronephrosis in utero IV: in utero decompression prevents renal dysplasia. J Pediatr Surg 1984; 19(6). 32. Tarantal AF, Han V, Cochrum K et al. Fetal rhesus monkey model of obstructive renal dysplasia. Kidney Int 2001; 59: 446–456. 33. Nakayama DK, Glick PL, Villa RL et al. Experimental pulmonary hypoplasia due to oligohydramnios and its reversal by relieving thoracic compression. J Pediatr Surg 1983; 18: 347. 34. Peters C, Carr MC, Lais A et al. The response of the fetal kidney to obstruction. J Urol 1992; 148: 503–509. 35. Steinhardt GF, Salinas-Madrigal L & deMello D. Experimental ureteral obstruction in the fetal opossum: histologic assessment. J Urol 1994; 152: 2133–2138. 36. McVary K & Maizels M. Urinary obstruction reduces glomerulogenesis in the developing kidney: a model in the rabbit. J Urol 1989; 142: 646–651. 37. Gasser B, Mauss Y & Ghnassia J. A quanatitative study of normal nephrogenesis in the human fetus: its implication in the natural history of kidney changes due to low obstructive uropathies. Fetal Diagn Ther 1993; 8: 371–384. 38. Berman DJ & Maizels M. The role of urinary obstruction in the genesis of renal dysplasia. A model in the chick embryo. J Urol 1982; 128(5): 1091–1096. 39. Ayan S. Partial ureteral obstruction dysregulates the renal renin-angiotensin system in the fetal sheep kidney. Urology 2001; 58(2): 301–306. 40. Gobet R. Renal renin-angiotensin system dysregulation caused by partial bladder outlet obstruction in the fetal sheep. Kidney Int 1999; 56(5): 1654–1661. 41. Bussieres L, Laborde K, Souberbielle JC et al. Fetal urinary insulin-like growth factor I and binding protein 3 in bilateral obstructive uropathies. Prenat Diagn 1995 Nov; 15(11): 1047–1055. 42. Xu W, Gelber S, Orr-Urteger A et al. Megacystis, mydriasis and ion channel defect in mice lacking the alpha3 neuronal nicotinic acetylcholine receptor. Proc Natl Acad Sci U S A 1999; 96: 5746–5751. 43. Xu W, Orr-Urteger A, Nigro F et al. Multiple autonomic dysfunction in mice lacking the beta2 and beta4 subunits of neuronal nicotinic acetylcholine receptors. J Neurosci 1999; 19: 9298–9905. 44. Lev-Lehman E, Bercovich D, Xu W et al. Characterization of the human beta4 nAChR gene and polymorphisms in CHRNA3 and CHRNB4. J Hum Genet 2001; 46(7): 362–366. 45. Arger PH, Coleman BG, Mintz MC et al. Routine fetal genitourinary tract screening. Radiology 1985; 156: 485–489. 46. Fernbach SK, Maizels M & Conway JJ. Ultrasound grading of hydronephrosis: introduction to the system used by the Society for Fetal Urology. Pediatr Radiol 1993; 23(6): 478–480. 47. Phan V, Traubici J, Hershenfield B et al. Vesicoureteral reflux in infants with isolated hydronephrosis. Pediatr Nephrol 2003; 18: 1224–1228. 48. Grignon A, Filiatrault D & Homsy Y. Ureteropelvic junction stenosis: antenatal ultrasonographic diagnosis, postnatal investigation and follow-up. Radiology 1986; 160: 649–651. 49. Podevin G, Mandelbrot L, Vuillard E et al. Outcome of urological abnormalities prenatally diagnosed by ultrasound. Fetal Diagn Ther 1996 May; 11(3): 181–190. 50. Harding LJ, Malone PS & Wellesley DG. Antenatal minimal hydronephrosis: is its follow-up an unnecessary cause of concern? Prenat Diagn 1999 Aug; 19(8): 701–705. 51. Lakhoo K, Thomas DFM, Fuenfer M et al. Failure of pre-natal ultrasonography to prevent urinary infection associated with underlying urological abnormalities. Br J Urol 1996; 77(6): 905–908. 52. Broadley P, McHugo J, Morgan I et al. The 4 year outcome following the demonstration of bilateral renal pelvic dilatation on pre-natal renal ultrasound. Br J Radiol 1999 Mar; 72(855): 265–270.

120 R. K. Morris and M. D. Kilby 53. Kaefer M, Peters CA, Retik AB et al. Increased renal echogenicity: a sonographic sign for differentiating between obstructive and nonobstructive etiologies of in utero bladder distension. J Urol 1997 Sep; 158(3 Pt 2): 1026–1029. 54. McHugo J & Whittle M. Enlarged fetal bladders: aetiology, management and outcome [Review] [9 refs]. Prenat Diagn 2001 Nov; 21(11): 958–963. 55. Abbott JF, Levine D & Wapner R. Posterior urethral valves: inaccuracy of prenatal diagnosis. Fetal Diagn Ther 1998 May; 13(3): 179–183. 56. Maizels M, Alpert SA, Houston JT et al. Fetal bladder sagittal length: a simple monitor to assess normal and enlarged fetal bladder size, and forecast clinical outcome. J Urol 2004 Nov; 172(5 Pt 1): 1995–1999. 57. 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 Aug; 152(2 Pt 2): 698–701. 58. Dinneen MD, Dhillon HK, Ward HC et al. Antenatal diagnosis of posterior urethral valves [see comment]. Br J Urol 1993 Sep; 72(3): 364–369. *59. Oliveira EA, Diniz JS, Cabral AC et al. Predictive factors of fetal urethral obstruction: a multivariate analysis. Fetal Diagn Ther 2000 May; 15(3): 180–186. 60. Winter R & Knowles S. Megacystis-microcolon-intestinal-hypoperistalsis syndrome: confirmation of autosomal recessive inheritance. J Med Genet 1986; 23: 360–362. 61. Anneren G, Meurling S & Olsen L. Megacystis-microcolon-intestinal hypoperistalsis syndrome (MMIHS), an autosomal recessive disorder: clinical reports and review of the literature. Am J Med Genet 1991; 41: 251–254. 62. Stamm E, King G & Thickman D. Megacystis-microcolon-intestinal hypoperistalsis syndrome: prenatal identification in siblings and review of the literature [Review] [29 refs]. J Ultrasound Med 1991 Oct; 10(10): 599–602. 63. Levine D, Hatabu H, Gaa J et al. Fetal anatomy revealed with fast MR sequences. AJR Am J Roentgenol 1996; 167: 905–908. 64. Poutamo J, Vanninen R, Partanen K et al. Diagnosing fetal urinary tract abnormalities: benefits of MRI compared to ultrasonography. Acta Obstet Gynecol Scand 2000 Jan; 79(1): 65–71. 65. Cassart M, Massez A, Metens T et al. Complementary role of MRI after sonography in assessing bilateral urinary tract anomalies in the fetus. AJR Am J Roentgenol 2004 Mar; 182(3): 689–695. 66. Quintero RA, Johnson MP, Romero R et al. In-utero percutaneous cystoscopy in the management of fetal lower obstructive uropathy. Lancet 1995 Aug 26; 346(8974): 537–540. 67. Welsh A, Agarwal S, Kumar S et al. Fetal cystoscopy in the management of fetal obstructive uropathy: experience in a single European center [see comment]. Prenat Diagn 2003 Dec 30; 23(13): 1033–1041. 68. Nicolaides KH, Cheng HH, Abbas A et al. Fetal renal defects: associated malformations and chromosomal defects. Fetal Diagn Ther 1992; 7(1): 1–11. 69. Barker AP, Cave MM, Thomas DF et al. Fetal pelvi-ureteric junction obstruction: predictors of outcome. Br J Urol 1995 Nov; 76(5): 649–652. 70. Oliveira EA, Rabelo EA, Pereira AK et al. Prognostic factors in prenatally-detected posterior urethral valves: a multivariate analysis. Pediatr Surg Int 2002 Dec; 18(8): 662–667. 71. Estroff JA, Mandell J & Benacerraf BR. Increased renal parenchymal echogenicity in the fetus: importance and clinical outcome. Radiology 1991 Oct; 181(1): 135–139. 72. Crombleholme TM, Harrison MR, Golbus MS et al. Fetal intervention in obstructive uropathy: prognostic indicators and efficacy of intervention. Am J Obstet Gynecol 1990 May; 162(5): 1239–1244. 73. 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 Aug; 20(4): 376–387. 74. Lipitz S, Ryan G, Samuell C et al. Fetal urine analysis for the assessment of renal function in obstructive uropathy. Am J Obstet Gynecol 1993 Jan; 168(1 Pt 1): 174–179. 75. Johnson MP, Bukowski TP, Reitleman C et al. In utero surgical treatment of fetal obstructive uropathy: a new comprehensive approach to identify appropriate candidates for vesicoamniotic shunt therapy. Am J Obstet Gynecol 1994 Jun; 170(6): 1770–1776. 76. Dommergues M, Muller F, Ngo S et al. Fetal serum beta2-microglobulin predicts postnatal renal function in bilateral uropathies. Kidney Int 2000 Jul; 58(1): 312–316.

Congenital urinary tract obstruction 121 77. Nicolini U, Tannirandorn Y, Vaughan J et al. Further predictors of renal dysplasia in fetal obstructive uropathy: bladder pressure and biochemistry of ‘fresh’ urine. Prenat Diagn 1991 Mar; 11(3): 159–166. 78. Johnson MP, Corsi P, Bradfield W et al. Sequential urinalysis improves evaluation of fetal renal function in obstructive uropathy [see comment]. Am J Obstet Gynecol 1995 Jul; 173(1): 59–65. *79. Morris RK, Quinlan-Jones E, Kilby M 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 Jul 4; [Epub ahead of print]. 80. Muller F, Bernard M-A, Benkirane A et al. Fetal urine cystatin C as a predictor of postnatal renal function in bilateral uropathies. Clin Chem 1999; 45(12): 2292–2293. 81. Eugene M, Muller F, Dommergues M et al. Evaluation of postnatal renal function in fetuses with bilateral obstructive uropathies by proton nuclear magnetic resonance spectroscopy. Am J Obstet Gynecol 1994 Feb; 170(2): 595–602. 82. Foxall PJ, Bewley S, Neild GH et al. Analysis of fetal and neonatal urine using proton nuclear magnetic resonance spectroscopy. Arch Dis Child Fetal Neonatal Ed 1995 Nov; 73(3): F153–F157. 83. Daikha-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 Jul; 52(1): 21–32. 84. Nolte S, Mueller B & Pringsheim W. Serum alpha1-microglobulin and beta2-microglobulin for the estimation of fetal glomerular renal function. Pediatr Nephrol 1991; 5(5): 573–577. *85. Nicolini U & Spelzini F. Invasive assessment of fetal renal abnormalities: urinalysis, fetal blood sampling and biopsy [Review] [46 refs]. Prenatal Diagn 2001 Nov; 21(11): 964–969. 86. Berry SM, Lecolier B, Smith RS et al. Predictive value of fetal serum beta2-microglobulin for neonatal renal function. Lancet 1995; 345(8960): 1277–1278. 87. Ciardelli V, Rizzo N, Farina A et al. Prenatal evaluation of fetal renal function based on serum beta(2)microglobulin assessment. Prenatal Diagn 2001 Jul; 21(7): 586–588. 88. Tassis B, Trespidi L, Tirelli AS et al. Serum B2-microglobulin in fetuses with urinary tract anomlaies. Am J Obstet Gynecol 1997; 176: 54–57. 89. Nolte S. Estimation of fetal renal function by microprotein determination (alpha1- and beta2-microglobulin) in fetal blood. Zeitschrift fur Geburtshilfe und Perinatologie 1991; 195(4): 153–158. 90. Cagdas A, Aydinli K, Irez T et al. Evaluation of the fetal kidney maturation by assessment of amniotic fluid alpha-1 microglobulin levels. Eur J Obstet Gynecol Reprod Biol 2000; 90(1): 55–61. 91. Cobet G, Gummelt T, Bollmann R et al. Assessment of serum levels of alpha-1-microglobulin, beta2-microglobulin, and retinol binding protein in the fetal blood. A method for prenatal evaluation of renal function. Prenatal Diagn 1996 Apr; 16(4): 299–305. 92. Greco P, Loverro G, Caruso G et al. The diagnostic potential of fetal renal biopsy. Prenatal Diagn 1993; 13: 551–556. 93. Bunduki V, Saldanha LB, Sadek L et al. Fetal renal biopsies in obstructive uropathy: feasibility and clinical correlations – preliminary results. Prenat Diagn 1998 Feb; 18(2): 101–109. 94. Chevalier RL. Biomarkers of congenital obstructive nephropathy: past, present and future. J Urol 2004; 172(3): 852–857. 95. Donnai P. Attitudes of patients after ‘‘genetic’’ termination of pregnancy. Br Med J 1981; 282: 621–622. 96. Flake AW, Harrison MR & Adzick NS. Fetal sacrococcygeal teratoma. J Pediatr Surg 1986; 21(7): 563–566. 97. Adzick NS, Harrison MR, Flake AW et al. Fetal surgery for cystic adenomatoid malformation of the lung. J Pediatr Surg 1993 Jun; 28(6): 806–812. 98. Farmer DL. Fetal surgery: a brief review [Review] [17 refs]. Pediatr Radiol 1998 Jun; 28(6): 409–413. 99. Harrison MR, Golbus MS, Filly RA et al. Fetal surgery for congenital hydronephrosis. N Engl J Med 1982 Mar 11; 306(10): 591–593. 100. Harrison MR, Golbus MS, Filly RA et al. Fetal hydronephrosis: selection and surgical repair. J Pediatr Surg 1987 Jun; 22(6): 556–558. 101. Crombleholme TM, Harrison MR, Langer JC et al. Early experience with open fetal surgery for congenital hydronephrosis. J Pediatr Surg 1988 Dec; 23(12): 1114–1121.

122 R. K. Morris and M. D. Kilby 102. Quintero RA, Morales WJ, Allen MH et al. Fetal hydrolaparoscopy and endoscopic cystotomy in complicated cases of lower urinary tract obstruction. Am J Obstet Gynecol 2000 Aug; 183(2): 324–330. 103. Wilson RD, Johnson MP, Flake AW et al. Reproductive outcomes after pregnancy complicated by maternal-fetal surgery. Am J Obstet Gynecol 2004 Oct; 191(4): 1430–1436. 104. Bealer JF, Raisanen J, Skarsgard ED et al. The incidence and spectrum of neurological injury after open fetal surgery. J Pediatr Surg 1995 Aug; 30(8): 1150–1154. 105. Quintero RA, Hume R, Smith C et al. Percutaneous fetal cystoscopy and endoscopic fulguration of posterior urethral valves [see comment]. Am J Obstet Gynecol 1995 Jan; 172(1 Pt 1): 206–209. 106. Ville Y, Hecher K, Gagnon A et al. Endoscopic laser coagulation in the mangament of severe twinto-twin transfusion syndrome. Br J Obstet Gynaecol 1998; 105: 446–453. 107. Deprest J, Van Ballaer P & Evrard V. Experience with fetoscopic cord ligation. Eur J Obstet Gynecol Reprod Biol 1998; 81: 157–164. 108. Agarwal SK & Fisk NM. In utero therapy for lower urinary tract obstruction [Review] [29 refs]. Prenat Diagn 2001 Nov; 21(11): 970–976. 109. Golbus MS, Harrison MR & Filly RA. In utero treatment of urinary tract obstruction. Am J Obstet Gynecol 1982; 142(4): 383–388. 110. Manning FA, Harrison MR & Rodeck C. Catheter shunts for fetal hydronephrosis and hydrocephalus. Report of the International Fetal Surgery Registry. N Engl J Med 1986 Jul 31; 315(5): 336–340. 111. Coplen DE. Prenatal intervention for hydronephrosis [Review] [63 refs]. J Urol 1997 Jun; 157(6): 2270–2277. 112. Rodeck CH & Nicolaides KH. Ultrasound guided invasive procedures in obstetrics [Review] [85 refs]. Clin Obstet Gynaecol 1983 Dec; 10(3): 515–539. *113. Clark TJ, Martin WL, Divakaran TG et al. Prenatal bladder drainage in the management of fetal lower urinary tract obstruction: a systematic review and meta-analysis [Review] [36 refs]. Obstet Gynecol 2003 Aug; 102(2): 367–382. 114. Ismaili K, Avni FE, Wissing KM et al. Long-term clinical outcome of infants with mild and moderate fetal pyelectasis: validation of neonatal ultrasound as a screening tool to detect significant nephrouropathies. J Pediatr 2004 Jun; 144(6): 759–765. 115. King LR, Coughlin P & Bloch E. The case for immediate pyeloplasty in the neonate with ureteropelvic junction obstruction. J Urol 1984; 132: 725–728. *116. Holmes N, Harrison MR & Baskin LS. Fetal surgery for posterior urethral valves: long-term postnatal outcomes. Pediatrics 2001 Jul; 108(1): E7. 117. Freedman AL, Johnson MP, Smith CA et al. Long-term outcome in children after antenatal intervention for obstructive uropathies [see comment]. Lancet 1999 Jul 31; 354(9176): 374–377. *118. Biard JM, Johnson MP, Carr MC et al. Long-term outcomes in children treated by prenatal vesicoamniotic shunting for lower urinary tract obstruction. Obstet Gynecol 2005 Sep; 106(3): 503–508. 119. Holmdahl, Gundela, Sillen et al. Boys with posterior urethral valves: outcome concerning renal function, bladder function and paternity at ages 31 to 44 years. J Urol 2005; 174(3): 1031–1034. 120. Thomas DF, Irving HC & Arthur RJ. Pre-natal diagnosis: how useful is it? Br J Urol 1985; 57: 784–787. 121. Nakayama DK, Harrison MR & de Lomimier AA. Prognosis of posterior urethral valaves presenting at birth. J Pediatr Surg 1986; 21: 43–45. 122. Hayden SA, Russ PD, Pretorius DH et al. Posterior urethral obstruction: prenatal sonographic findings and clinical outcome in 14 cases. J Ultrasound Med 1988; 7: 371. 123. Reuss A, Wladimiroff JW, Stewart PA et al. Non-invasive management of fetal obstructive uropathy. Lancet 1988; ii: 949–951. 124. Morris R, Khan K & Kilby M. Congenital lower urinary tract obstruction and the efficacy of vesicoamniotic shunting. In: Studd J, Tan S & Chervenak F (eds.). Progress in Obstetrics and Gynaecology 2006; vol. 17. London: Elsevier, 2006, pp. 78–97. 125. Bokenkamp A, Dieterich C, Dressler F et al. Fetal serum concentrations of cystatin C and beta2microglobulin as predictors of postnatal kidney function. Am J Obstet Gynecol 2001; 185(2): 468–475. 126. The unborn patient: prenatal diagnosis and treatment. 2nd edn. Phiadelphia: W.B.Saunders Company, 1990.