Fetal renal impairment

Fetal renal impairment

Tina Vanderheyden*, Sailesh Kumar, Nicholas M. FiskInstitute of Reproductive and Developmental Biology, Queen Charlotte's and Chelsea Hospital,Imperial College, Hammersmith Campus, Du Cane Road, London W12OSH, UK

Received 1 January 2003; accepted 31 January 2003

Summary Renal function in utero deals chiefly with urine production rather than theexcretion of metabolites, which are cleared by the placenta. Fetal renal impairment(FRI) in bilateral renal disease thus presents as oligohydramnios or anhydramnios; thiscan lead to lung hypoplasia and early neonatal death. As in the adult, FRI can bedivided into prerenal, renal and postrenal causes. Causes of prerenal FRI includeintrauterine growth restriction, unbalanced intertwin transfusion in monochorionictwins and maternal drug ingestion. Bilateral renal agenesis, multicystic dysplasia andboth the autosomal dominant and recessive forms of polycystic kidney disease areexamples of renal causes, whereas postrenal etiologies are usually caused by lowerurinary tract obstruction (LUTO). When both kidneys are affected and there is severemid-trimester oligohydramnios, the prognosis is poor. Although animal studies haveshown that prolonged LUTO leads to lung hypoplasia and renal damage, and thatdecompression of the fetal kidney in early pregnancy restores fetal pulmonary andrenal function, the value of fetal therapy such as vesico-amniotic shunting remainscontroversial, with a high procedure-related complication rate and a high incidence ofend-stage renal failure in childhood. Fetal cystoscopic treatment of posterior urethralvalves in utero may obviate some of these difficulties but remains an investigationalprocedure.© 2003 Elsevier Ltd. All rights reserved.

KEYWORDSLung hypoplasia;Fetal urine analysis;Fetal kidney function;Renin–angiotensinsystem;Renal agenesis;Maternal drug intake;Prenatal diagnosis;Posterior urethralvalves;Fetal therapy

Introduction

In utero, fetal fluid and electrolyte balance as wellas acid–base homeostasis is regulated and main-tained by the interaction of the placenta andmaternal blood. Thus, the placenta functions asan in vivo dialysis unit.1 Fetal glomeruli develop by8–9 weeks, tubular function commences after the14th week, and nephrogenesis continues until34–36 weeks gestation. Studies of fetal urine innormal pregnancies have shown that there is aninverse relationship between creatinine and sodiumlevels as gestation progresses. This is caused byincreasing maturity of the renal tubular system(Fig. 1).2,3 After 20 weeks, the kidneys provide over90% of the amniotic fluid. An adequate amount of

amniotic fluid is of vital importance for lung andskeletal development. Severe oligohydramnios oranhydramnios causes lung hypoplasia and softtissue deformities.4 The latter refers to the typicalphenotype of a fetus with low-set ears, wide-seteyes, micrognathia, limb contractures and talipes(Potter sequence).5

The adverse effect of anhydramnios on fetallung development is well documented in animalmodels6 and in neonates born with bilateral renalagenesis who die of lung hypoplasia, as severeoligohydramnios from 16 weeks onwards precludesmost further pulmonary development. In con-trast, oligohydramnios after the second trimester isunlikely to result in pulmonary hypoplasia as thecrucial canalicular phase of lung development(between 16 and 25 weeks) has been largely com-pleted by this stage.7,8 Severe oligohydramnios,defined as a deepest vertical pool of less than 1 cm,

* Corresponding author. Tel.: +44-20-8383-3998; fax:+44-20-8383-3507E-mail address: [email protected] (T. Vanderheyden).

Seminars in Neonatology (2003) 8, 279–289

Seminars inNEONATOLOGY

www.elsevierhealth.com/journals/siny

1084-2756/03/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.doi:10.1016/S1084-2756(03)00022-8

for more than 14 days before 25 weeks gestation isassociated with neonatal mortality of more than90%.9–11

Causes of fetal renal impairment may be con-sidered as prerenal, renal and postrenal. Sinceoligo- or anhydramnios is usually the result of dam-age to both kidneys, only causes of bilateral renalimpairment will be considered here. Isolated uni-lateral renal impairment has an excellent prognosisfor fetal, neonatal and adult kidney function whencontralateral kidney function is normal.12 In animalstudies of unilateral impairment, the contralateralkidney usually shows compensatory renal growth.This phenomenon is also seen in human fetuseswith unilateral renal agenesis or multicystic kidneydisease (MCDK).13–15

Pre-renal fetal renal impairment

Intrauterine growth restriction

Intrauterine growth restriction (IUGR) caused byplacental insufficiency is often associated with oli-gohydramnios. Chronic hypoxemia is characterizedby a redistribution of oxygenated blood to thebrain, heart and adrenal glands and away fromnon-vital peripheral organs such as the kidneys.Animal models and observational studies on humanfetuses have shown that the urine production ratein fetuses with IUGR is reduced.16–18

An experimental IUGR rabbit model,18 achievedby ligation of 25–30% of the uteroplacental vesselsin the last-third of gestation, showed that thenumber of glomeruli was significantly reducedat birth compared with non-IUGR controls. The

reduction in the number of glomeruli is believed tocause a decrease in glomerular filtration capacityand adversely affect renal development. This leadsto impaired renal function in utero, as well as inthe neonatal period, and has been implicated inlate cardiovascular sequelae in adults, such ashypertension.19

The fetal renin–angiotensin system (RAS) iscrucially important in the maintenance of glomeru-lar filtration and urine production, as well as renalperfusion. Abnormal maturation of the RAS caninfluence renal function and blood pressure controlin survivors of severe IUGR. One study comparedrenin expression in the kidneys of 17 stillbornfetuses with IUGR and 26 unexplained stillborns ofappropriate weight.20 In the control group, therewas a shift in renin-containing cells from the innerto the outer zone of the renal cortex with advanc-ing gestational age, consistent with these cellsbeing located in the outer renal cortex in healthyadult kidneys. In the IUGR group, this shift of renin-containing cells did not occur. In another study, 25of 90 IUGR fetuses showed increased medullaryechogenicity on ultrasound, which increased therisk of abnormal outcome 1.5-fold compared withthose with isoechoic kidneys. The medullae arevery sensitive to hypoxemia, and these authorsargue that fetal medullar hyperechogenicity is apathologic sign of in utero hypoxia.21,22

Twin–twin transfusion syndrome

Twin–twin transfusion syndrome (TTTS) compli-cates 10–15% of monochorionic twin pregnancies. Itis caused by an imbalance of intertwin transfusionthrough connecting vascular anastomoses andpresents on ultrasound with an oligohydramnios–polyhydramnios sequence. There also is discord-ance in bladder dynamics, the recipient having alarge polyuric bladder and the donor a small, ofteninvisible, bladder.

A recent series23 found signs of renal tubulardysgenesis in 11 of 21 donor twins but in none of the17 recipient co-twins, nor in any twins or the twinsof unknown chorionicity without TTTS.23 Owing tothe perfusion discordance in TTTS, renal perfusionand glomerular filtration are likely to be reduced inthe donor twin. Upregulation of renin synthesis isknown to be induced by reduced renal perfusion,leading to vasoconstriction and a further reductionin glomerular perfusion and filtration. Recent evi-dence shows increased renin protein and messengerRNA in donor kidneys. In addition, the histologyof these kidneys shows tubular dysgenesis resem-bling that seen in fetuses exposed in utero toangiotensin-converting enzyme inhibitors (ACEIs).24

Fig. 1 Mean and 95% data intervals for fetal urinary sodiumthroughout gestation in 26 control fetuses (B) (reprinted fromreference 2, with permission from Elsevier Science).

280 T. Vanderheyden et al.

It is speculated that high angiotensin II levels aretransferred from the donor to the recipient downvascular anastomoses in which blood flows pre-dominantly towards the recipient. Activation of theRAS aggravates the hypervolemic recipient, con-tributing to systemic hypertension and myocardialdysfunction. The same studies have shown a down-regulation of the RAS in recipient kidneys. Histologyshows large and congested recipient kidneys withmultiple hemorrhagic infarcts, as well as hyper-trophy of mesangial cells as seen in hypertensivenephropathy. Discordant activation of the RAS isthus implicated in the discordant renal functionand pathology seen in these twins in utero.24–26

Because postnatal renal impairment and/or tubulardysgenesis have been observed in 12–30% of donortwins,27–29 postnatal follow-up of renal function isrecommended, as is blood pressure monitoring inchildhood and later life.30

Maternal drug intake

Some drugs taken by the mother can impair fetalrenal function and cause oligohydramnios. Mice, inwhich the angiotensinogen and angiotensin-converting enzyme genes are inactivated, showstructural and functional renal anomalies similar tothose seen in fetuses exposed to ACEIs.31 ACEIs canseverely affect renal development and functionwhen used in the second and third trimesters ofpregnancy, leading to tubular dysgenesis, oligo-hydramnios, growth restriction, hypocalvaria,pulmonary hypoplasia, neonatal anuria and still-birth.32 These anomalies are thought to be causedby a direct action of ACEIs on the fetal RAS byreducing the concentration of angiotensin II33 andsecondary fetoplacental ischemia resulting frommaternal hypotension and a decrease in fetal–placental blood flow.34 No adverse fetal effectshave been linked to the first trimester use of ACEIs.The ACEI Registry found no evidence of renaltubular dysplasia in the infants of 66 women whoself-reported exposure only in the first trimester.35

Non-steroidal anti-inflammatory drugs (NSAIDs)can also influence fetal renal function and develop-ment. Cyclo-oxygenase type 1 (COX-1) inhibitorssuch as indomethacin definitely reduce urine out-put and may lead to oligohydramnios and renaldysfunction. A study comparing preterm neonatesexposed in utero to indomethacin used as a toco-lytic with preterm controls showed that the formerhad a lower urine output and a higher serum creati-nine concentration during the first 3 days afterbirth and that this effect in the neonate was moreprofound when there was a short interval betweenthe last dose and delivery.36

It was originally hoped that COX-2 inhibitors,being developed as tocolytics, would avoidthis effect. Because prostaglandin synthesis isincreased in the fetal membranes in preterm labor,this reaction being catalyzed by COX-2, suppressionof COX-2 is a logical therapeutic target for tocoly-sis. However, in COX-2 knockout mice, although noanomalies in renal function or structure were notedat birth, morphological alterations occurred earlyin the postnatal period (nephrogenesis lasting until3 weeks postpartum in mice), the mice eventuallydying of renal failure by 8 weeks of age. The histo-logic changes seen in the COX-2 knockout micewere similar to those observed in the kidneys ofhuman fetuses exposed to a prolonged use ofNSAIDs (both non-specific COX-1 and COX-2 inhibi-tors) in the third trimester of pregnancy.37 Further-more, human fetal kidneys have recently beenshown to have an increased expression of COX-2compared with adult kidneys, suggesting thatCOX-2 is constitutively expressed in the humanfetal kidney and that inhibition by prolongedNSAID use can lead to morphological and functionaldeficits.37

In obstetrics, selective NSAIDs have recentlybeen used to treat preterm labor and poly-hydramnios in order to reduce the amount ofamniotic fluid in monoamniotic twins and thus pre-vent cord entanglement.38 A double-blind ran-domized study in our group compared the fetalside-effects of maternally administered indo-methacin (n�10), sulindac (n�10) and nimesulide(n�10) for the short-term treatment of pretermlabor. Indomethacin is a non-selective inhibitor ofCOX-1 and COX-2.39 Sulindac is also a non-selectiveCOX inhibitor but has to be metabolized in the liverto become active, and this metabolic process islimited in the fetus. Nimesulide is 25-fold moreselective for COX-2 than indomethacin and alsoblocks calcium channels. Notwithstanding thesedifferences, each drug caused a similar reduction infetal urine production and amniotic fluid index.Thus, nimesulide, although a selective COX-2inhibitor, caused short-term side-effects similar tothose of indomethacin and sulindac. In a casereport from the same group,40 nimesulide wasadministered for 17 weeks during pregnancy with-out fetal side-effects. Nevertheless, prolongedexposure without titration may lead to oligo-hydramnios.

Maternal cocaine use adversely influences fetalrenal function by hypoperfusion and thus influencesthe fetal RAS. It is also associated with oligo-hydramnios as well as other fetal vascular compli-cations. One study compared the histology of renal

Fetal renal impairment 281

arteries from 29 stillborn fetuses exposed tococaine and 22 stillborn fetuses with an unknownetiology, reporting a significant thickening of theinterlobular arterial wall of the fetal kidney andnarrowing of the lumen of renal arteries in exposedfetuses.41 This correlates with the clinical findingthat fetuses exposed to cocaine have higher renalartery resistance indices and a significant decreasein urine output, compared with control fetuses.42,43

Renal cause of fetal renal impairment

Renal agenesis

Bilateral renal agenesis is a lethal conditiondiagnosed antenatally when anhydramnios, absentkidneys and a non-visualized bladder are noted onultrasound scanning. Death usually occurs soonafter birth from respiratory failure secondary topulmonary hypoplasia. Fetal kidneys can usually beidentified on ultrasound as early as the first trimes-ter. They appear as oval echogenic structures oneither side of the spine. The bladder should bevisible by 13 weeks onwards. In the second tri-mester, from around 18 to 19 weeks, severe oligo-hydramnios can hamper demonstration of emptyrenal fosse, and there may be confusion withthe adrenal glands. Severe IUGR with oligo- oranhydramnios and premature rupture of the mem-branes are important differential diagnoses.44

Some investigators have used transvaginal scanningwhen the fetus is in breech presentation, or amnio-infusion,45 to improve ultrasound visualization,thus allowing a definitive diagnosis to be made.

In 33 pregnancies referred with severe oligo-hydramnios to our center, high-resolution colorDoppler ultrasonography was used to establish thepresence or absence of renal arteries. Non-visualization of the renal arteries indicated renalagenesis or a non-functioning multicystic kidney.All fetuses that prenatally had two renal arteriesvisualized with Doppler scanning had the presenceof two normal kidneys confirmed on postnatal orpostmortem examination. Doppler flow signals inrenal arteries can be an accurate, non-invasive wayto predict absent renal function such as in renalagenesis or MCDK, and the technique has, in ourexperience, largely replaced amnio-infusion for thediagnosis of renal agenesis (Fig. 2).46

It is well established that it is the secondaryoligohydramnios, rather than the renal agenesis perse, that causes the respiratory difficulties. First,animal studies have shown that the correction ofoligohydramnios prevents pulmonary hypoplasia. Ina fetal lamb model, lower urinary tract obstruction

(LUTO) was induced by ligating the urethra andurachus. In half of the fetuses, a secondprocedure—marsupialization of the bladder—wasperformed to decompress the fetal kidneys. Thehistology of these fetal lungs was similar to thatseen in lambs without LUTO. The lung had areduced number of airway branches but adequatealveolar development, whereas the kidneys showeda lesser degree of tissue damage.47 Second,Cameron et al.48 showed, in a single case, thatserial amnio-infusion in bilateral renal agenesisprevented lung hypoplasia and signs of fetal com-pression. As renal transplantation is not feasible inthe neonatal period, this is not a managementoption for pregnancies with renal agenesis, butexperimentally it validates the importance ofthe presence of amniotic fluid in preventing lunghypoplasia.

Another example of normal lung development infetuses with renal agenesis is seen in Fraser syn-drome (cryptophthalmos–syndactyly syndrome), anautosomal recessive condition in which laryngealatresia and bilateral renal agenesis can co-exist.Despite oligohydramnios, the lungs appear enlargedand hyperechogenic on ultrasound as the lungsecretions that normally exit the trachea with fetalbreathing movements accumulate because of thelaryngeal atresia.49 Indeed, the prevention ofoligohydramnios-related lung hypoplasia by tra-cheal occlusion in animal studies has led to thedevelopment of clinical techniques to occlude thetrachea of fetuses with diaphragmatic hernia, thuspreventing lung hypoplasia.

Fig. 2 Color Doppler scan of the renal arteries. The arrowindicates a near-field kidney in coronal view.


Multicystic dysplastic kidney disease

MCDK is an anomaly that arises from a failedco-ordination of development of the metanephrosand the branching ureteric bud. Obstruction occursat an early stage, leading to a non-functioningkidney. On ultrasound examination, multiple,smooth, non-communicating cysts of variable sizesare seen. Unilateral disease is more common inmales; females are, however, twice as likely tohave bilateral MCDK and associated non-renalanomalies, as well as four times more likely to haveabnormal chromosomes. Bilateral disease causesanhydramnios. A positive family history may bepresent since 9% of relatives of index cases withbilateral renal agenesis or dysplasia have renalmalformations.50 Associated non-renal abnor-malities are frequent in both unilateral (26%) andbilateral (67%) diseases; this makes aneuploidymore likely, so karyotyping should be considered.51

Polycystic kidney disease

In polycystic kidney disease (PKD), cysts arise inotherwise normally developing kidneys, whicheventually leads to the destruction of normaltissue. There are two forms of PKD: autosomalrecessive (ARPKD) and autosomal dominant(ADPKD). In ARPKD, these cysts arise from thecollecting ducts, whereas in ADPKD, they arisefrom the nephron as well as the collecting ducts.Although both types may be present ante-natally, ADPKD generally manifests much later inlife, usually after the third decade. Antenatally,PKD usually presents with oligohydramnios andlarge hyperechogenic kidneys that maintain theirshape and smooth surface. The cysts usually are toosmall to resolve on ultrasound, and thus the kidneyssimply appear echogenic. The diagnosis can some-times be made early in the second trimester,although serial scans are often necessary to detectthe progression of both echogenicity and oligo-hydramnios. The two forms cannot be differ-entiated by ultrasound scanning, but ADPKD rarelypresents in utero and can be suspected from thefamily history or confirmed by genetic testing.

ARPKD can present as perinatal, neonatal,infantile or juvenile-onset disease. The spectrum inthe age of onset is the result of the variable expres-sion of mutations of the same gene, as well asthe effects of modifier genes and environmentalfactors, rather than the effect of mutations indifferent genes. Prenatally detected PKD, usuallyARPKD, has a poor prognosis, with a mortality rateof almost 90% in the early neonatal period becauseof lung hypoplasia and renal failure.50 Children who

survive the neonatal period have a 56–67% prob-ability of survival up to the age of 15 years withoutend-stage renal failure, and prolonged survivaltill the age of 55 years has been reported. Liverinvolvement with intrahepatic cholangiodysplasiaand fibrosis is, however, always present. Renalcysts in these children are rounder and larger thanin the perinatal form and may resemble ADPKD.This can be excluded by linkage analysis or renalultrasonography of the parents. The gene for ARPKDis located on chromosome 6.52

Two gene mutations have been identified inADPKD. In 85% of cases, there is a mutation of thePKD1 gene, which encodes a transmembrane glyco-protein (polycystin 1), located on the short armof chromosome 16. The remaining cases have amutation in the PKD2 gene for polycystin 2 onchromosome 4, although, in a small number ofcases, the gene locus has not been identified.50,53

ADPKD is fully penetrant, that is, virtually 100%of individuals who inherit a mutated PKD genewill develop renal cysts that can be detectedsonographically by the age of 30.52

Postrenal causes of fetal renalimpairment

Dilatation of the renal pelvis and calyces is theprincipal sign of impaired urinary flow on ultra-sound scanning, which can often reliably identifythe level of obstruction. Urine collection proximalto the obstruction leads to dilatation of the renalpelvis and an increase in hydrostatic pressure,which transmitted to the renal parenchyma,decreases renal blood flow. Severe longstandinghydronephrosis can cause renal tissue damage,leading to obstructive renal cystic dysplasia. Thedegree of renal tissue damage is determined by thedegree (complete or incomplete) and durationof obstruction. In a fetal rhesus monkey model,urinary tract obstruction induced in the secondtrimester resulted in a defect in the ureteric ductbranching and altered vascularization of theglomerulus. These abnormalities are also noted inhuman renal dysplasia, which is associated withcompromised postnatal renal function and couldthus be predictive of postnatal renal function.54 Aswith other anomalies, the management, counselingfor and prognosis of a fetus with hydronephrosis areinfluenced by the exclusion or detection of otheranomalies.

Vesico-ureteric reflux

An important cause of pre- and postnatal renalimpairment is vesico-ureteric reflux (VUR). This is


responsible for 10–30% of the cases antenatallydetected hydronephrosis and is bilateral in almosthalf of the cases.55 Postnatally, VUR is associatedwith recurrent urinary tract infection, which cancause renal parenchymal scarring. An animal modelof fetal VUR showed that it was associated withaltered renal growth, structural maldevelopment,excess matrix deposition and impaired excretoryfunction as well as persistent bladder instability.56

The results in this model suggest that even a sterileurinary reflux can cause an impairment of renalfunction and that the early prevention of urinaryreflux in neonatal life may be helpful. In a fewcases, the antenatal diagnosis of VUR caused by anureterocele may be amenable to in utero therapy;although controversial, fetal intervention may beconsidered if there is associated bladder outletobstruction causing oligo- or anhydramnios.57

Lower urinary tract obstruction

Posterior urethral valves (PUV) account for 50% ofcases antenatally diagnosed for LUTO and presenton ultrasound with oligohydramnios and a large,thick-walled bladder (Fig. 3A) with a keyhole sign

and bilateral hydronephrosis (Fig. 3B). A keyholesign represents the enlarged bladder and dilatedproximal urethra (Fig. 4). The remaining causes ofLUTO are prune belly syndrome and urethralatresia.58 The prognosis is worse (95% mortality) inthose diagnosed antenatally when mid-trimesteroligohydramnios is present. Poor prognostic factorson ultrasound examination include dilatation of theupper tract, increased bladder wall thickness,oligohydramnios and evidence of renal dysplasia(echogenic renal cortex and macrocystic renalchange), especially before 24 weeks.59,60

As nephrogenesis progresses, the concentrationof sodium in the urine falls, and that of creatinine,calcium and ammonia rises significantly. Fetalurine analysis assists the prognostication ofurinary obstruction. Good prognostic features infetuses with obstructive uropathies, generally,include a urinary sodium of less than 100 Meq/l,chloride less than 90 Meq/l, osmolality below200 milliosmol/l and �2-microglobulin below6 mg/l. More accurate, however, are gestationallyreferenced values.2,61–63 Urinary sodium and cal-cium levels are significantly higher in fetuses withrenal dysplasia (Fig. 5). Analysis of serial bladderaspirations may be more representative of renalreserve than the analysis of a single urine samplefrom an obstructed fetal bladder.2 Johnson et al.64

proposed three sequential urine samples at 48–72 hintervals because the first urine sample, which mayhave been present in the bladder for a long timeand thus have equilibrated with the serum, mayotherwise falsely predict poor outcome. The se-cond urine sample reflects urine from the dilated

Fig. 3 (A) Longitudinal ultrasound view of a fetus at 15 weekswith an enlarged bladder (arrow). (B) Kidney from the samefetus with severe hydronephrosis (arrow).

Fig. 4 Transverse view of the lower fetal abdomen on ultra-sound at 16 weeks, showing a dilated bladder (*) with a keyholesign (arrow).


upper urinary tract that has drained into thebladder. The third sample, however, is recentlyproduced urine, better reflecting the renal func-tion expected after a relief of the obstruction.Sequential urine sampling improves the sensitivityof predicting renal function by urinary electrolytesand minimally reduces specificity, although serialaspiration for electrolytes is only necessary, inpractice, if the first sample predicts poor outcome.

Renal dysplasia may also be indicated by �2-microglobulin concentration in the fetal urine orserum. This microprotein has a molecular weight of11 800 Da, which renders transplacental passageunlikely.65 One study investigated the predictivevalue of fetal serum �2-microglobulin for neonatalrenal function and found that values above 5.6 mg/lprovided high sensitivity and specificity.66 Morerecent studies question the role of these markers asthey have improved neither prediction of outcomenor the selection of fetuses for fetal therapy.67 Thismay be due to the fact that, in early gestation (<20weeks), fetal urine is still very isotonic as a result oftubular immaturity, and thus electrolyte levels arefound to be largely independent of renal function.Since early diagnosis is thought to be crucial for thesalvage of renal function, the role of urine analysisin predicting renal function in the first half ofpregnancy is unclear.3 Further research is neededto identify more accurate predictors of long-termrenal function.

Another method of analyzing the composition offetal urine is the use of high-field proton nuclearmagnetic resonance spectroscopy (H NMR) toquantify low molecular weight metabolites. Onestudy analyzed urine samples obtained by

ultrasound-guided bladder aspiration in fetuseswith obstructive uropathy and compared this withthe first neonatal urine sample that is produced inutero, in normal controls. LUTO was characterizedby glycosuria, amino- and organic aciduria,68 withlevels up to 10 times higher in fetuses with obstruc-tive uropathy with a poor outcome than in normalcontrols.69 Although promising, this method is notwidely used in the evaluation of kidney functionin fetuses with LUTO because it faces the samerestriction as the biochemical analysis of fetalurine: not being representative in the first half ofpregnancy.

Ultrasound-guided renal biopsies

Bunduki et al.70 assessed the feasibility and value ofprenatal renal biopsies in detecting renal dysplasia.Ultrasound-guided renal biopsy, as well as urinesampling, was performed on 10 fetuses with severeurinary tract obstruction, with a success rate of50%. In five cases where the kidneys appearedfunctional on ultrasound, three biopsies wereobtained, all showing normal renal histology andhaving a good neonatal outcome. Renal biopsiesin the poor prognostic group showed a dysplasticpattern in one and normal histology in the other,although the fetus had mild renal impairment onfollow-up. Although this approach may have somevalue in predicting renal histology, a renal biopsy isonly obtained in 50% of cases, and a focal needleaspiration is not representative of the wholekidney parenchyma since renal dysplasia is patchilydistributed in renal parenchyma.

Fetal therapy for lower urinary tractobstruction

The rationale for in utero therapy is that restora-tion of the amniotic fluid volume prevents lunghypoplasia, while decompressing the renal tractrelieves back pressure on the fetal kidneys beforeirreversible renal damage occurs. This concepthas been demonstrated in numerous animalmodels.47,63

Intrauterine therapy includes vesico-amnioticshunting8 and fetal cystosopy.9 The standard treat-ment of vesico-amniotic shunting for LUTO involvesplacing a pigtail shunt between the fetal bladderand amniotic cavity under ultrasound guidance. Areview of five large series involving 169 successfullyplaced shunts showed that the overall peri-natal survival after intervention was only 47%.Shunt-related complications occurred in 45%of cases and included shunt blockage and/or

Fig. 5 Urinary sodium in fetuses with lower urinary tractobstruction with (+) and without renal dysplasia (B) plottedagainst the reference range (reprinted from reference 2, withpermission from Elsevier Science).


migration, preterm labor, urinary ascites, chorio-amnionitis, iatrogenic gastroschisis and limbcontracture.71 However, even in successful cases,end-stage renal disease occurred in 40% of thesurvivors. If oligohydramnios was present beforethe shunt was placed, 56% of these babies dieddespite intervention. Failure to restore amnioticfluid volume was associated with 100% mortality.When intervention did improve the chances of post-natal survival, it did not appear to alter the renaloutcome.72

Freedman et al.73 reviewed the long-term out-come of 34 fetuses that underwent vesico-amnioticshunt placement. Thirteen died, six because ofpulmonary hypoplasia. In the survivors, the ulti-mate diagnosis was mainly prune belly syndrome orPUV. Less than one-third of the children neededventilatory support in the neonatal period, andnone of the children had pulmonary restriction onnormal activities on follow-up. These childrenunderwent a total of 62 surgical procedures, ofwhich 47 were major (bladder augmentations andrenal transplants). Six children have maintainednormal renal function, three have renal insuffi-ciency, and five have progressed to renal failure. It,therefore, appears that fetal therapy influencesonly the outcome of fetuses with LUTO that leads tomid-trimester oligo- or anhydramnios by preventingpulmonary hypoplasia but does not seem to amelio-rate the underlying kidney disease. This is import-ant as parents need to be aware that there is still a33–50% chance of renal failure in childhood.74 Theyalso need to be counseled that although affectedchildren are expected to have normal cognitiveabilities, participate in normal activities andachieve a socially acceptable continence, they willneed considerable medical and surgical attention.

Fetal cystoscopy which can be both diagnosticand therapeutic, is currently undergoing evaluationas a management option in LUTO.58 It allowsexploration of the bladder and the proximal partof the urethra under endoscopic vision (Fig. 6).Obtaining a clear view of the bladder neck andurethra is facilitated by the use of flexible scopes,which may overcome the problem of negotiatingthe urethrovesical angle. PUV, if present, canbe disrupted by laser therapy although this maydamage the adjacent bowel and other structures.Alternatively gentler techniques, such as the inser-tion of a guidewire or simple pressure saline injec-tion, can be used.75 In our own center, withexperience of 13 cases, a therapeutic procedurewas attempted in 10 and was successful in at leastsix cases. Of these six cases, five had good prognos-tic values of the electrolytes in fetal urine analysis

prior to therapy, and on follow-up two of thoseinfants have normal renal function, one has poorrenal function, and three pregnancies ended inintrauterine death.76

A retrospective review from the University ofCalifornia, San Francisco, of fetuses with PUV whounderwent fetal therapy (mainly vesico-amnioticshunting) with favorable urine analysis results, con-firmed a high fetal mortality of 42%.77 Chronic renaldisease was diagnosed in 62% of the survivors, 25%of whom underwent renal transplantation. Thisstudy, like that of Freedman et al.73 suggests thatfetal therapy does not cure fetuses with PUV butmay improve bladder function and decrease themorbidity of incontinence and recurrent infections.

Reversibility of renal changes

Haecker et al.78 studied the histology of kidneys ofchildren with PUV (mean 29 months) who under-went nephrectomy, to determine whether renaldysplasia was a primary or secondary malforma-tion. In 80% of specimens, primary dysplastic mal-formations (mesenchymal or fetal cartilage tissueor dysplastic glomeruli and tubuli) were noted inthe presence of well-developed renal parenchyma.All the specimens also showed secondary pathol-ogy. The presence of these primary dysplasticchanges in well-differentiated kidneys makes asingle infravesical cause of renal dyplasia inpatients with PUV questionable. If correct, thismeans that secondary renal damage would beirreversible by 20 weeks and arguably not amenable

Fig. 6 Fetal cystoscopic view of a fetal bladder and the proxi-mal urethra (*).


to fetal therapy.78 Schwarz et al.79–81 also sug-gested that the primary embryologic abnormality—an abnormally lateral origin of the uretericbud—results in both urethral valve and renal dys-plasia. If true, this suggests that the evidence ofcurability from animal models of LUTO might havebeen oversimplistic.

Conclusions

FRI is an important cause of fetal and neonatalmorbidity and mortality, especially when associ-ated with mid-trimester oligohydramnios. Thereare many causes, most of which are not amenableto either in utero or postnatal therapy. LUTO maybenefit from in utero treatment, but the results arepoor and further research into prognostic factorsand new techniques are necessary.

References

1. Muller F, Dommergues M, Bussieres L et al. Development ofhuman renal function: reference intervals for 10 biochemi-cal markers in fetal urine. Clin Chem 1996;42(11):1855–60.

2. Nicolini U, Fisk NM, Rodeck CH et al. Fetal urine biochemis-try: an index of renal maturation and dysfunction. Br JObstet Gynaecol 1992;99(1):46–50.

3. Nicolini U, Spelzini F. Invasive assessment of fetal renalabnormalities: urinalysis, fetal blood sampling and biopsy.Prenat Diagn 2001;21(11):964–9.

4. Cuckow PM, Nyirady P, Winyard PJ. Normal and abnormaldevelopment of the urogenital tract. Prenat Diagn 2001;21(11):908–16.

5. Christianson C, Huff D, McPherson E. Limb deformationsin oligohydramnios sequence: effects of gestational ageand duration of oligohydramnios. Am J Med Genet 1999;86(5):430–3.

6. Harrison MR, Ross N, Noall R et al. Correction of congenitalhydronephrosis in utero. I. The model: fetal urethralobstruction produces hydronephrosis and pulmonary hypo-plasia in fetal lambs. J Pediatr Surg 1983;18(3):247–56.

7. Laudy JA, Wladimiroff JW. The fetal lung. 1 Developmentalaspects. Ultrasound Obstet Gynecol 2000;16(3):284–90.

8. Laudy JA, Wladimiroff JW. The fetal lung. 2. Pulmonaryhypoplasia. Ultrasound Obstet Gynecol 2000;16(5):482–94.

9. Kilbride HW, Yeast J, Thibeault DW. Defining limits ofsurvival: lethal pulmonary hypoplasia after midtrimesterpremature rupture of membranes. Am J Obstet Gynecol1996;175(3 Pt 1):675–81.

10. Harrison MR, Golbus MS, Filly RA et al. Management of thefetus with congenital hydronephrosis. J Pediatr Surg 1982;17(6):728–42.

11. Rotschild A, Ling EW, Puterman ML et al. Neonatal outcomeafter prolonged preterm rupture of the membranes. Am JObstet Gynecol 1990;162(1):46–52.

12. Atiyeh B, Husmann D, Baum M. Contralateral renal abnor-malities in patients with renal agenesis and noncystic renaldysplasia. Pediatrics 1993;91(4):812–5.

13. Douglas-Denton R, Moritz KM, Bertram JF et al. Compensa-tory renal growth after unilateral nephrectomy in the ovinefetus. J Am Soc Nephrol 2002;13(2):406–10.

14. Hill LM, Nowak A, Hartle R et al. Fetal compensatory renalhypertrophy with a unilateral functioning kidney. Ultra-sound Obstet Gynecol 2000;15(3):191–3.

15. Peters CA, Gaertner RC, Carr MC et al. Fetal compensatoryrenal growth due to unilateral ureteral obstruction. J Urol1993;150(2 Pt 2):597–600.

16. Nicolaides KH, Peters MT, Vyas S et al. Relation of rate ofurine production to oxygen tension in small-for-gestational-age fetuses. Am J Obstet Gynecol 1990;162(2):387–91.

17. Takeuchi H, Koyanagi T, Yoshizato T et al. Fetal urineproduction at different gestational ages: correlation tovarious compromised fetuses in utero. Early Hum Dev 1994;40(1):1–11.

18. Bassan H, Trejo LL, Kariv N et al. Experimental intrauterinegrowth retardation alters renal development. PediatrNephrol 2000;15(3–4):192–5.

19. Battista MC, Oligny LL, St-Louis J et al. Intrauterine growthrestriction in rats is associated with hypertension and renaldysfunction in adulthood. Am J Physiol Endocrinol Metab2002;283(1):E124–31.

20. Kingdom JC, Hayes M, McQueen J et al. Intrauterine growthrestriction is associated with persistent juxtamedullaryexpression of renin in the fetal kidney. Kidney Int 1999;55(2):424–9.

21. Suranyi A, Streitman K, Pal A et al. Fetal renal artery flowand renal echogenicity in the chronically hypoxic state.Pediatr Nephrol 2000;14(5):393–9.

22. Suranyi A, Retz C, Rigo J et al. Fetal renal hyperechogenicityin intrauterine growth retardation: importance andoutcome. Pediatr Nephrol 2001;16(7):575–80.

23. Oberg KC, Pestaner JP, Bielamowicz L et al. Renal tubulardysgenesis in twin–twin transfusion syndrome. Pediatr DevPathol 1999;2(1):25–32.

24. Mahieu-Caputo D, Dommergues M, Delezoide AL et al. Twin-to-twin transfusion syndrome. Role of the fetal renin–angiotensin system. Am J Pathol 2000;156(2):629–36.

25. Kilby MD, Platt C, Whittle MJ et al. Renin gene expression infetal kidneys of pregnancies complicated by twin–twintransfusion syndrome. Pediatr Dev Pathol 2001;4(2):175–9.

26. Mahieu-Caputo D, Muller F, Joly D et al. Pathogenesis oftwin–twin transfusion syndrome: the renin–angiotensinsystem hypothesis. Fetal Diagn Ther 2001;16(4):241–4.

27. Cincotta R, Oldham J, Sampson A. Antepartum and post-partum complications of twin–twin transfusion. Aust N Z JObstet Gynaecol 1996;36(3):303–8.

28. Reisner DP, Mahony BS, Petty CN et al. Stuck twin syndrome:outcome in thirty-seven consecutive cases. Am J ObstetGynecol 1993;169(4):991–5.

29. Barr M Jr, Sedman AB, Heidelberger KP. Renal tubulardysgenesis in twins. Pediatr Nephrol 1998;12(5):408–13.

30. Christensen AM, Daouk GH, Norling LL et al. Postnataltransient renal insufficiency in the feto-fetal transfusionsyndrome. Pediatr Nephrol 1999;13(2):117–20.

31. Hilgers KF, Norwood VF, Gomez RA. Angiotensin's role inrenal development. Semin Nephrol 1997;17(5):492–501.

32. Sedman AB, Kershaw DB, Bunchman TE. Recognition andmanagement of angiotensin converting enzyme inhibitorfetopathy. Pediatr Nephrol 1995;9(3):382–5.

33. Tabacova SA, Kimmel CA. Enalapril: pharmacokinetic/dynamic inferences for comparative developmentaltoxicity. A review. Reprod Toxicol 2001;15(5):467–78.

34. Buttar HS. An overview of the influence of ACE inhibitors onfetal-placental circulation and perinatal development. MolCell Biochem 1997;176(1–2):61–71.

35. Centers for Disease Control and Prevention. Postmarketingsurveillance for angiotensin-converting enzyme inhibitor


use during the first trimester of pregnancy—United States,Canada, and Israel, 1987–1995. JAMA 1997;277(15):1193–4.

36. Norton ME, Merrill J, Cooper BA et al. Neonatal complica-tions after the administration of indomethacin for pretermlabor. N Engl J Med 1993;329(22):1602–7.

37. Khan KN, Paulson SK, Verburg KM et al. Pharmacology ofcyclooxygenase-2 inhibition in the kidney. Kidney Int 2002;61(4):1210–9.

38. Peek MJ, McCarthy A, Kyle P et al. Medical amnioreductionwith sulindac to reduce cord complications in monoamniotictwins. Am J Obstet Gynecol 1997;176(2):334–6.

39. Sawdy R, Lye S, Fisk N, et al. A double blind randomisedstudy of fetal side effects during and after the short termmaternal administration of indomethacin, sulindac andnimesulide for treatment of preterm labour. Am J ObstetGynecol [in press].

40. Sawdy R, Slater D, Fisk N et al. Use of a cyclo-oxygenasetype-2-selective non-steroidal anti-inflammatory agent toprevent preterm delivery. Lancet 1997;350(9073):265–6.

41. Mitra SC, Seshan SV, Salcedo JR et al. Maternal cocaineabuse and fetal renal arteries: a morphometric study.Pediatr Nephrol 2000;14(4):315–8.

42. Mitra SC, Ganesh V, Apuzzio JJ. Effect of maternal cocaineabuse on renal arterial flow and urine output of the fetus.Am J Obstet Gynecol 1994;171(6):1556–9.

43. Mitra SC. Effect of cocaine on fetal kidney and bladderfunction. J Matern Fetal Med 1999;8(6):262–9.

44. Reuss A, Wladimiroff JW, van der Wijngaard JA et al. Fetalrenal anomalies, a diagnostic dilemma in the presence ofintrauterine growth retardation and oligohydramnios. Ultra-sound Med Biol 1987;13(10):619–24.

45. Fisk NM, Ronderos-Dumit D, Soliani A et al. Diagnosticand therapeutic transabdominal amnioinfusion in oligo-hydramnios. Obstet Gynecol 1991;78(2):270–8.

46. Sepulveda W, Stagiannis KD, Flack NJ et al. Accuracy ofprenatal diagnosis of renal agenesis with color flow imagingin severe second-trimester oligohydramnios. Am J ObstetGynecol 1995;173(6):1788–92.

47. Harrison MR, Nakayama DK, Noall R et al. Correction ofcongenital hydronephrosis in utero. II. Decompressionreverses the effects of obstruction on the fetal lung andurinary tract. J Pediatr Surg 1982;17(6):965–74.

48. Cameron D, Lupton BA, Farquharson D et al. Amnioinfusionsin renal agenesis. Obstet Gynecol 1994;83(5 Pt 2):872–6.

49. Stevens CA, McClanahan C, Steck A et al. Pulmonary hyper-plasia in the Fraser cryptophthalmos syndrome. Am J MedGenet 1994;52(4):427–31.

50. Winyard P, Chitty L. Dysplastic and polycystic kidneys:diagnosis, associations and management. Prenat Diagn2001;21(11):924–35.

51. Lazebnik N, Bellinger MF, Ferguson JE II et al. Insights intothe pathogenesis and natural history of fetuses withmulticystic dysplastic kidney disease. Prenat Diagn 1999;19(5):418–23.

52. Igarashi P, Somlo S. Genetics and pathogenesis of polycystickidney disease. J Am Soc Nephrol 2002;13(9):2384–98.

53. Friedmann W, Vogel M, Dimer JS et al. Perinatal differentialdiagnosis of cystic kidney disease and urinary tract obstruc-tion: anatomic pathologic, ultrasonographic and geneticfindings. Eur J Obstet Gynecol Reprod Biol 2000;89(2):127–33.

54. Matsell DG, Mok A, Tarantal AF. Altered primate glomerulardevelopment due to in utero urinary tract obstruction.Kidney Int 2002;61(4):1263–9.

55. Herndon CD, McKenna PH, Kolon TF et al. A multicenteroutcomes analysis of patients with neonatal reflux present-ing with prenatal hydronephrosis. J Urol 1999;162(3 Pt 2):1203–8.

56. Gobet R, Cisek LJ, Zotti P et al. Experimental vesicoureteralreflux in the fetus depends on bladder function and causesrenal fibrosis. J Urol 1998;160(3 Pt 2):1058–62 discussion1079.

57. Quintero RA, Homsy Y, Bornick PW et al. In-utero treatmentof fetal bladder-outlet obstruction by a ureterocele. Lancet2001;357(9272):1947–8.

58. Quintero RA, Johnson MP, Romero R et al. In-utero percu-taneous cystoscopy in the management of fetal lowerobstructive uropathy. Lancet 1995;346(8974):537–40.

59. Thomas DF. Prenatal diagnosis: does it alter outcome?Prenat Diagn 2001;21(11):1004–11.

60. Hutton KA, Thomas DF, Arthur RJ et al. Prenatally detectedposterior urethral valves: is gestational age at detection apredictor of outcome? J Urol 1994;152(2 Pt 2):698–701.

61. Dommergues M, Muller F, Ngo S et al. Fetal serum beta2-microglobulin predicts postnatal renal function in bilateraluropathies. Kidney Int 2000;58(1):312–6.

62. Qureshi F, Jacques SM, Seifman B et al. In utero fetal urineanalysis and renal histology correlate with the outcome infetal obstructive uropathies. Fetal Diagn Ther 1996;11(5):306–12.

63. Glick PL, Harrison MR, Golbus MS et al. Management of thefetus with congenital hydronephrosis. II: Prognostic criteriaand selection for treatment. J Pediatr Surg 1985;20(4):376–87.

64. Johnson MP, Corsi P, Bradfield W et al. Sequential urinalysisimproves evaluation of fetal renal function in obstructiverenopathy. Am J Obstet Gynecol 1995;173(1):59–65.

65. Lipitz S, Ryan G, Samuell C et al. Fetal urine analysis for theassessment of renal function in obstructive uropathy. Am JObstet Gynecol 1993;168(1 Pt 1):174–9.

66. Berry SM, Lecolier B, Smith RS et al. Predictive value of fetalserum beta 2-microglobulin for neonatal renal function.Lancet 1995;345(8960):1277–8.

67. Guez S, Assael BM, Melzi ML et al. Shortcomings in predictingpostnatal renal function using prenatal urine biochemistryin fetuses with congenital hydronephrosis. J Pediatr Surg1996;31(10):1401–4.

68. Foxall PJ, Bewley S, Neild GH et al. Analysis of fetal andneonatal urine using proton nuclear magnetic resonancespectroscopy. Arch Dis Child Fetal Neonatal Ed 1995;73(3):F153–7.

69. Eugene M, Muller F, Dommergues M et al. Evaluation ofpostnatal renal function in fetuses with bilateral obstructiveuropathies by proton nuclear magnetic resonancespectroscopy. Am J Obstet Gynecol 1994;170(2):595–602.

70. Bunduki V, Saldanha LB, Sadek L et al. Fetal renal biopsies inobstructive uropathy: feasibility and clinical correlations—preliminary results. Prenat Diagn 1998;18(2):101–9.

71. Chan FY, Borzi P, Cincotta R et al. Limb constriction as acomplication of intra-uterine vesico-amniotic shunt: feto-scopic release. Fetal Diagn Ther 2002;17(5):315–20.

72. Coplen DE. Prenatal intervention for hydronephrosis. J Urol1997;157(6):2270–7.

73. Freedman AL, Johnson MP, Smith CA et al. Long-termoutcome in children after antenatal intervention forobstructive uropathies. Lancet 1999;354(9176):374–7.

74. McLorie G, Farhat W, Khoury A et al. Outcome analysis ofvesicoamniotic shunting in a comprehensive population. JUrol 2001;166(3):1036–40.


75. Agarwal SK, Fisk NM. In utero therapy for lower urinary tractobstruction. Prenat Diagn 2001;21(11):970–6.

76. Welsh AW, Argawal AS, Smith RP, et al. Fetal cystoscopy inthe management of fetal obstructive uropathy. Br J ObstetGynaecol [in press].

77. Holmes N, Harrison MR, Baskin LS. Fetal surgery for pos-terior urethral valves: long-term postnatal outcomes.Pediatrics 2001;108(1):E7.

78. Haecker FM, Wehrmann M, Hacker HW et al. Renal dysplasiain children with posterior urethral valves: a primary orsecondary malformation? Pediatr Surg Int 2002;18(2–3):119–22.

79. Schwarz RD, Stephens FD, Cussen LJ. The pathogenesis ofrenal dysplasia. II. The significance of lateral and medialectopy of the ureteric orifice. Investig Urol 1981;19(2):97–100.

80. Schwarz RD, Stephens FD, Cussen LJ. The pathogenesis ofrenal dysplasia. III. Complete and incomplete urinaryobstruction. Investig Urol 1981;19(2):101–3.

81. Schwarz RD, Stephens FD, Cussen LJ. The pathogenesis ofrenal dysplasia. I. Quantification of hypoplasia anddysplasia. Investig Urol 1981;19(2):94–6.


Documents

Fetal renal impairment