Renal development in the fetus and premature infant

SUMMARY

Congenital abnormalities of the kidney and urinary tract (CAKUT) are one of the leading congenital defects to be identified on prenatal ultrasound. CAKUT represents a broad spectrum of abnormalities, from transient hydronephrosis to severe bilateral renal agenesis. CAKUT is a major contributor to chronic and end stage kidney disease (CKD/ESKD) in children. Both genetic abnormalities and the fetal environment contribute to CAKUT. Monogenic gene mutations identified in human CAKUT have advanced our understanding of molecular mechanisms of renal development. Low nephron number and solitary kidneys are associated with increased risk of adult onset CKD and ESKD. Premature and low birth weight infants represent a high risk population for low nephron number. Additional research is needed to identify modifiable factors to enhance nephron development in premature infants and biomarkers and appropriate follow-up of premature and low birth weight infants into adulthood.

1. Introduction

Renal anomalies are a frequent form of congenital abnormality and are often diagnosed prenatally. In addition, perinatal insults and premature birth may impact renal development. Here we review mechanisms of abnormal renal development and discuss the evaluation of fetal and premature infant kidneys.

2. Normal renal development

Human kidney development begins in the first trimester [1]. There are three stages of mammalian kidney development: the pronephros, mesonephros, and metanephros. The pronephros and mesonephros form and then essentially involute [2]. The metanephros develops into the final functional mammalian kidney.

The pronephros consists of simple tubules and forms at three weeks of gestation. Just caudal to the pronephros, the mesonephros forms at four weeks. The mesonephros consists of filtering units (glomeruli with tubules) and as they degenerate the mesonephros forms the mesonephric (or Wolffian) duct. The ureteric bud is an outgrowth of the mesonephric duct that invades the surrounding metanephric mesenchyme during the fifth week of gestation. The ureteric bud undergoes branching, establishing the radial structure of the kidney and the nephron number [2,3]. Signaling from the tips of the branching ureteric bud induces nephron formation, with conversion of metanephric mesenchyme to renal epithelia (renal vesicle). In turn, reciprocal signaling from the metanephric mesenchyme to the ureteric tree stimulates radial renal branching. The ureteric tree forms the renal collecting ducts, pelvis, and ureters. The renal vesicle becomes a comma, then an S-shaped body, and ultimately forms the glomerulus, proximal tubule, loop of Henle, and distal tubules.

The first glomeruli form at 9–10 weeks of gestation [4]. During the late second and third trimester, there is an exponential increase in nephrons between 18 and 32 weeks [1,5]. Nephron development is complete between 32 and 36 weeks [5]. Fetal urine becomes a major contributor to amniotic fluid at about 20 weeks, with production of 300 mL/kg fetal weight/day [6,7].

3. Congenital abnormalities of the kidney and urinary tract (CAKUT)

Congenital abnormalities of the kidney and urinary tract are one of the most frequent major birth defects, representing up to 20–30% of all major birth defects [8]. Whereas the prevalence of abnormalities often depends upon the setting (e.g. tertiary care centers often have a higher prevalence of anomalies), some estimates place the frequency of renal anomalies at about one in 500 live births [9,10]. There is a broad spectrum of renal defects, from simple hydronephrosis to bilateral renal agenesis. In one large population-based study of more than 20,000 fetuses and newborns, about one-third of renal anomalies were detected by prenatal ultrasound [9]. These anomalies were on average detected late in pregnancy, in the third trimester [9]. Time of diagnosis depended upon severity of the abnormality; for example, mean time of diagnosis of bilateral renal agenesis was 24 weeks of gestation, whereas mean time of diagnosis of hydronephrosis was 30 weeks [9]. The most frequently occurring renal anomaly diagnosed on prenatal ultrasound is hydronephrosis [11]. Fetal hydronephrosis may be transient or related to upper or lower urinary tract obstruction or vesicoureteral reflux (VUR) (discussed in detail below). The second most frequent anomaly is renal cysts (either bilateral or unilateral), followed by renal agenesis (unilateral > bilateral) [12]. Renal dysplasia may be observed, associated with hydronephrosis or cysts [12].

4. Cellular mechanisms leading to CAKUT

The most severe renal anomaly, renal aplasia (agenesis), results when the ureteric bud either fails to form or fails to reach/induce the metanephric mesenchyme leading to apoptosis (Figure 1). Unless it receives growth factors released by the ureteric bud (e.g. glial-derived neurotrophic factor or GDNF), the metanephric mesenchyme will undergo apoptosis, leading to renal agenesis [13]. Bilateral renal agenesis results in oligohydramnios, leading to severe oligohydramnios (formerly Potter’s) sequence (pulmonary hypoplasia, club feet, micrognathia) and fetal or perinatal death [14,15].

Fig. 1.

Genetic, epigenetic, and environmental factors interact to modify renal developmental pathways and lead to a spectrum of congenital abnormalities of the kidney and urinary tract (CAKUT). VUR, vesicoureteral reflux; UPJ, ureteropelvic junction; UVJ, ureterovesical junction.

Inhibitory signals restrict and guide the site of ureteric bud outgrowth to a single location [16]. Failure to restrict the site of outgrowth may lead to multiple ureteric buds and duplication of the kidney and/or collecting system. In the most severe cases, there is complete duplication, with an upper pole and lower pole ureter. Duplication is frequently associated with other ureteral anomalies, including VUR and ureterovesical junction (UVJ) obstruction [17]. The classic findings in complete duplications are that the upper ureter connects to the bladder in an ectopic location and is obstructed, whereas the lower pole ureter has VUR [18]. Even in the absence of duplications, defects in ureteric bud outgrowth and branching are associated with VUR, UVJ as well as ureteropelvic junction (UPJ) obstruction [19].

Experiments in fetal sheep have demonstrated that ureteral or urethral obstruction impairs fetal kidney development [20–22]. Postnatal ureteral obstruction also induces tubulointerstitial fibrosis in rodent models [23]. Whereas in humans it is not possible to eliminate the possibility that factors that affected the ureteral development independently lead to defects in renal development, it is clear that both UPJ and UVJ obstructions may result in obstructive uropathies and CKD in childhood [24]. UPJ and UVJ are rarely bilateral and thus do not typically lead to childhood ESKD unless there are associated contralateral renal anomalies [25].

In contrast, urethral obstruction during fetal development (as with posterior urethral valves) is one of the more frequent causes of ESKD in childhood [26,27]. The pathophysiology of PUV is not fully understood, and there are several competing theories on its mechanism [28]. One theory is that the “valves” (which are not truly valves) are mucosal membranes that form and fail to involute during urethral development. Alternatively, they may represent an overgrowth of normally present mucosal folds. A final theory is that they result from abnormal development of the mesonephric or Müllerian duct. A variant of this is that the mesonephric duct has an abnormal anterior insertion into the cloaca. When the rectal–urethral septum forms, the abnormal insertion prevents proper migration of the duct cranially and posteriorly and leads to obstructing anterior–lateral folds. Obstruction of the urethra during bladder development leads to permanent defects in bladder smooth musculature differentiation, with excess fibrotic tissues. PUV is frequently associated with VUR and renal dysplasia, and, as with urethral obstruction, it remains unclear whether the dysplasia results from obstruction/VUR or separate individual defects in renal development (or both).

High grade VUR is associated with renal parenchymal damage, and bilateral VUR contributes to childhood CKD and ESRD [27]. Urinary tract infections in the setting of high grade VUR have significant risk of pyelonephritis [29]. Chronic pyelonephritis and inflammation likely contributes to scarring of renal parenchyma and loss of kidney function [29]. Some studies using mouse models suggest that infection is a key component of renal damage with high grade VUR [30]. However, it remains a source of controversy as, in some cases, high grade VUR in the absence of infection may also lead to progressive renal damage [31]. As with obstructive uropathies, it is also possible that defects that result in VUR may independently lead to renal developmental defects.

Defects in nephron differentiation may also result in CAKUT. In particular, defects in the mesenchyme to epithelial transition results in renal dysplasia [32]. Histologically, renal dysplasia is characterized by immature tubules surrounded by perivascular cuffs [33]. Metaplastic tissues, such as cartilage, may also be present. On renal ultrasound imaging, renal dysplasia most often appears as small echogenic kidneys with poor corticomedullary differentiation [34,35].

Nephrons are induced at the tips of the branching ureteric bud. Thus, the number of ureteric bud branches will determine the nephron number [36,37]. Severely impaired branching manifests as renal hypoplasia [38]. Additional potential mechanism resulting in renal hypoplasia is depletion of renal epithelial progenitors [39–41]. The cap mesenchymal cells that surround the ureteric bud tips represent a unique population of progenitor cells that are programmed to allow for differentiation into renal epithelia cells but also have capacity for self-renewal. There is a tight balance between differentiation and maintenance of the progenitor population [40]. In mouse models, premature differentiation and depletion of nephron progenitors results in renal hypoplasia [41]. Histologically, hypoplastic kidneys have few nephrons but differ from dysplasia in that they are normally formed [42]. Whereas it is difficult to reliably differentiate hypoplasia from dysplasia on ultrasound, classically renal hypoplasia manifests as small kidneys (<2 standard deviations below the mean for age) [42]. Together, defects of renal aplasia, dysplasia and hypoplasia along with reflux and obstructive uropathy are the leading cause of chronic and end stage kidney disease in childhood, accounting for about 40% of children with CKD and 30% of children with ESKD [27].

Whereas severe CAKUT leads to childhood CKD and ESRD, milder variations in ureteric bud branching likely contribute to the broad variation of nephron number observed in humans, from 200,000 to 2,000,000 [43]. Populations with decreased nephron number (such as Australian aboriginals) have higher rates of proteinuria and chronic kidney disease in adulthood (see below) [44]. It is thought that the decreased renal reserve contributes to risk for renal disease. Brenner famously developed the hypothesis that low nephron number leads to hyperfiltration of glomeruli [45]. Hyperfiltration increases glomerular pressures, inducing podocyte damage and loss. This additional glomerular damage exacerbates the nephron deficit and potentially systemic hypertension, leading to a vicious cycle of glomerular damage and scarring and progression to ESRD. In support of Brenner’s hypothesis, experimental models of reduced nephron number (5/6 nephrectomy) in rats lead to progressive renal disease [46,47].

The potential long term effects of nephron deficits are particularly relevant in low birth weight and premature infants. Low birth weight is strongly associated with low nephron number [48]. Prematurity may also result in a low nephron number, as discussed in greater detail below [49]. Multiple epidemiologic studies have identified associations of low birth weight with risk for renal disease, although a few have failed to identify such an effect [50–58].

5. Genetic defects leading to CAKUT

Advances in genomic sequencing and genetic manipulation in mouse and other animal models have led to major advances in our understanding of the gene networks that regulate renal development. The leading genetic syndromes associated with CAKUT are trisomies 21, 18, and 13 [59]. Large scale studies of cohorts of children with non-syndromic CAKUT have demonstrated that between 6% and 17% may have single gene defects [60,61]. Another 10–15% have large copy number variants (CNVs) that may contribute to CAKUT [62–64]. Even in mouse models, the phenotype from genetic deletion varies considerably, suggesting that gene–gene and gene–environment interactions contribute to CAKUT [65–67]. Thus, it is therefore likely that many cases of CAKUT are polygenic, with multiple gene variants contributing. Genetic causes of CAKUT have been reviewed recently [68–72], therefore we will focus here on a subset of mutations identified in humans.

One of the earliest genes to have mutations identified in association with non-syndromic human CAKUT was the tyrosine kinase RET receptor for GDNF. Biallelic inactivating genetic mutations in RET are associated with the most severe manifestation of CAKUT, bilateral renal agenesis [73]. The RET receptor is expressed on the tips of the ureteric buds, whereas GDNF is secreted by the metanephric mesenchyme [74]. GDNF/Ret signaling stimulates ureteric bud outgrowth, and thus defects in this signaling pathway result in failure of the ureteric bud to reach the metanephric mesenchyme, resulting in renal agenesis [75–77]. Some studies suggest that mutations in RET/GDNF are relatively rare in renal agenesis, although genetic variants in GDNF/Ret are associated with sporadic CAKUT [78,79].

Autosomal dominant mutations in HNF1B are the most frequent monogenic etiology of CAKUT [60,80,81]. HNF1B is a transcription factor that contributes to both ureter development and nephron segmentation [82–84]. Likely due to its multiple roles in renal development, HNF1B genetic variants are associated with a broad spectrum of CAKUT, from renal hypoplasia to non-functioning multicystic dysplastic kidneys [81,85–88]. HNF1B variants are also associated with maturity onset diabetes of the young (MODY), as in renal cysts and diabetes syndrome [89,90], as well as hyperuricemia with hypomagnesemia [91]. The hypomagnesemia is likely due to a direct effect of HNF1B on expression of sodium/potassium ATPase subunits that modify renal absorption of magnesium in the distal convoluted tubule [92,93].

Other autosomal dominant mutations that have been associated with sporadic CAKUT are in transcription factors in the renal cells that regulate early renal epithelial differentiation, including EYA1 [61,94,95]. EYA1 binds and regulates multiple other transcription factors that maintain the nephron progenitor population and stimulate GDNF expression (including SIX and PAX2 proteins) [96–99]. Mutations in EYA1/SIX1/SIX5 are associated with branchio-oto-renal syndrome [100–102]. PAX2 is expressed in the differentiating renal vesicle but also in the branching ureteric bud [103]. Autosomal dominant mutations in PAX2 are associated with renal–coloboma syndrome [104]. In large studies, identification of genetic mutations in EYA1 and PAX2 in what appeared to be sporadic cases of CAKUT led to more detailed eye and hearing examination, with identification of subtle deficits [60,61].

Autosomal recessive genetic mutations likely occur at a low frequency in sporadic CAKUT (for example, Kohl et al. identified autosomal recessive mutations in 12 genes in 2.5% of almost 600 children with CAKUT [105]). These included gene defects in FRAS/FREM, which are associated with Fraser syndrome (CAKUT) [105–107]. The protein products of these genes modify heparin sulfate, which bind the majority of growth factors (e.g. HGF, EGF, VEGF) that stimulate UB branching [108–111].

Studies of familial VUR have identified genetic defects in the SLIT-ROBO guidance signaling, and more recently genetic defects in TNXB, which encodes Tenascin XB [112–115]. Mouse models indicate that Slit-Robo signaling likely restricts interactions between the nephrogenic mesenchyme and ureteric bud to facilitate bud outgrowth at a single site [116,117]. Tenascin XB is an extracellular matrix glycoprotein that is expressed by uroepithelia of the UVJ, and thus may contribute to closure of the UVJ during voiding [115].

6. Adverse fetal environment effects on renal development

Together, chromosomal anomalies, copy number variants, and monogenic genetic abnormalities may account for up to 50% of CAKUT (Figure 2). However, environmental factors likely also contribute to CAKUT [118]. In-utero exposure to an adverse fetal environment is associated with CAKUT in humans. A variety of insults including hyperglycemia, vitamin A deficiency (in the developing world), and cocaine and alcohol exposures are associated with a higher risk of renal anomalies [119–121]. The mechanisms of these defects are not fully understood; however, hyperglycemia may affect ureteric bud branching [121]. Similarly, vitamin A (retinoic acid) is required for Ret expression [122]. Thus, vitamin A deficiency may impair ureteric bud branching due to decreased Ret expression [123,124].

Fig. 2.

Studies suggest genetic factors, including major chromosomal anomalies, copy number variants (CNV), and monogenic mutations/deletions likely contribute to up to 45% of congenital abnormalities of the kidney and urinary tract (CAKUT). Environmental factors, including an adverse fetal environment, and not yet discovered genetic factors likely contribute to the remainder.

Barker first hypothesized in 1990 that fetal programming has long term effects on risk for adult onset disease [125]. His study of the Dutch population exposed to famine during World War II demonstrated that children from mothers with inadequate caloric intake had higher incidence of cardiovascular disease and diabetes. As discussed above, Brenner extended his hypothesis to the kidney, focusing on the maladaptive effects of hyperfiltration in the setting of low nephron number [46]. Given that nephron number is established at birth, recent studies have focused on the role of the fetal environment on determining nephron number.

Intrauterine growth retardation is a frequent cause of low birth weight.

The effects of ureteroplacental insuffiency, as well as maternal caloric and protein deprivation on fetal renal development have been studied in rodent models [126]. Whereas not all models have had the same manifestations, likely due to differences in precise timing and nutrient content, caloric or protein restriction in utero induces low nephron number and hypertension and proteinuria in the adult offspring.

One of the mechanisms by which in utero exposures may result in long term effects is epigenetic modifications [70,127]. Epigenetic modifications include histone methylation and acetylation that regulate gene transcription by modifying the accessibility of chromatin to transcription factors [128]. Another form of epigenetic modification is DNA methylation. Highly methylated DNA in promotor regions is thought result in repression of gene transcription. Research into epigenetic regulation of renal development is an area of active research, and our understanding of how in-utero exposures may induce epigenetic changes in the kidney is limited. However, there are strong data suggesting that the epigenetic state of the fetal kidney is different than the adult kidney, and that shifts in the epigenetic state likely contribute to regulation of transcription of renal developmental genes [118,128,129]. In addition, human chronic kidney disease is associated with alterations in the epigenetic profiles [130]. Thus, epigenetic changes may provide a mechanistic link whereby in-utero exposures lead to long term increased risk of renal disease in adulthood.

Maternal medication exposures are another source of renal developmental defects. The renin–angiotensin system augments ureteric bud branching, and maternal use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers is associated with renal agenesis and anomalies [131]. Other medications, including glucocorticoids, mycophenolate mofetil, antiepileptic drugs, and cyclophosphamide may be associated with renal anomalies [132–136].

7. Prematurity effects on renal development

As discussed above, new nephrons form at an exponential rate between 20 and 28 weeks of gestation. Nephron formation ceases between 28 and 36 weeks. Thus, in addition to low birth weight, premature infants born before 36 weeks may not have completed nephron development. Limited data are available about the effects of prematurity on renal development. Unlike in term infants, nephrons may continue to form after birth in premature infants [137,138]. Studies of autopsies indicated that nephron formation may continue after birth in premature infants born prior to completion of nephron development [138,139]. However, premature cessation of nephrogenesis appears to be widespread, leading to a nephron deficit [138,139].

In addition to prematurity, postnatal exposures to nephrotoxins may affect renal development. A study of very low birth weight neonates in the neonatal intensive care unit determined that 90% of the neonates were exposed to at least one nephrotoxic agent and that the average neonate was exposed to almost two weeks of nephrotoxins [140]. Thirty percent of the premature infants received gentamicin after birth and 50% received amphoteracin. Fifty percent also received another nephrotoxin, indomethacin, as a non-surgical approach to close patent ductus arteriosus. Due to high renin–angiotensin levels in the preterm neonate, neonatal renal blood flow critically depends upon prostaglandin-induced vasodilation of afferent arterioles [141]. Thus, indomethacin can markedly reduce renal perfusion pressure and induce acute kidney injury (AKI) in a significant subset of neonates [140,142]. Of note, 26% of the cohort developed AKI, with the smallest infants being at highest risk.

Recently studies have suggested that repeated or severe AKI may result in CKD, and it is unclear what the effect of nephrotoxins are on the developing kidney in premature infants [49]. Low birth weight and prematurity are associated with glomerular disease later in childhood, especially focal glomerulosclerosis [49,143].

Recently a multicenter retrospective chart review of neonatal acute kidney was performed [144]. This and other studies may improve our understanding of the impact of prematurity, low birth weight, and perinatal insults on renal outcomes.

8. Antenatal diagnosis of renal pathology

8.1. Normal prenatal ultrasound findings

Fetal kidneys appear lobular and can be identified in the paraspinal region on prenatal ultrasound as early as 9–12 weeks of gestation. At 12 weeks, kidneys are typically 1 cm in length, and they grow to an average 2.7 cm in length by 20 weeks. In addition, between 15 and 20 weeks corticomedullary differentiation occurs, and an echogenic cortex with the hypoechogenic, dark renal pyramids of the medulla should be apparent by 20 weeks. Typically, the renal cortex is less echogenic than the liver by 20 weeks. Glomerular filtration begins at about nine weeks, but does not contribute significantly to amniotic fluid until skin develops at 19–20 weeks. The bladder may be visualized by ultrasound at 10–14 weeks, and bladder emptying is visible from 15 weeks. Bladder capacity is 10 mL at 30 weeks and increases to 50 mL in term infants. Urine becomes a major contributor to amniotic fluid and increases from 20 weeks until term with an average of 300 mL/kg body weight/day.

8.2. Antenatal evaluation of pyelectasis

The most frequent renal abnormality to be diagnosed by prenatal ultrasound is pyelectasis. Whereas there is no complete consensus about the best definition and grading system for pyelectasis, measurement of the anterior–posterior diameter (APD) of the renal pelvis in the transverse plane is the standard approach for diagnosis [145]. An APD of >15 mm is strongly predictive of uretero-pelvic junction obstruction [146]. However, APD is affected by multiple factors, including gestational age and maternal hydration status (and hence fetal urine output), and may worsen or improve during later development [147]. An APD of >4 mm in the second trimester or >7 mm in the third trimester are indications for follow-up [145,148–150]. Generally, dilation in the second trimester is an indication for follow-up ultrasound in the third trimester. Grading systems incorporate whether the hydronephrosis involves dilation major – or, in more severe cases, minor renal calyces – and whether there is associated thinning of the renal parenchyma [145].

8.3. Other factors predicting outcomes in pyelectasis

Low amniotic fluid (or oligohydramnios) is a major factor that influences postnatal outcomes of pyelectasis [151]. Oligohydramnios is defined as <500 mL of amniotic fluid, and is typically assessed by the amniotic fluid index (AFI, measurements of amniotic fluid pockets ≥2 cm) [152,153]. AFI <5–6 indicates oligohydramnios [152]. Oligohydramnios is most often associated with the severest forms of CAKUT, such as PUVs or bilateral renal agenesis. Oligohydramnios leads to fetal compression and oligohydramnios (Potter’s) sequence [wide set eyes, beaked nose, low set and posteriorly rotated ears, micrognathia, talipes equinovarus (club feet), and pulmonary hypoplasia] [154,155]. Pulmonary hypoplasia leads to increased risk of pneumothorax after birth and is the major determinant of postnatal survival [156]. In the setting of PUV, unilateral VUR may be protective and decrease the risk of mortality and renal failure [156].

There have been attempts to use urinary indices to predict renal outcomes, especially in cases of severe pyelectasis. Cutoffs exist for predicting better renal prognosis for urinary electrolytes (sodium <100 mEq/L and calcium <8 mg/dL) and urinary concentration (osmolarity <200 mOsm/L), and markers of renal damage (β2 microglobulin <4 mg/L and protein <40 mg/dL) have been proposed, but are not without controversy [150,157]. Various measures have been attempted to predict long term outcomes of fetal hydronephrosis. Serum creatinine at one year of life seems to be the best predictor of long term outcomes [158,159], although recently ultrasound assessment of renal parenchymal area has been proposed as a marker [160].

8.4. Postnatal evaluation of fetal pyelectasis

Fetal pyelectasis may represent urethral/bladder outflow obstruction (as discussed above), VUR, UPJ obstruction, non-obstructed dilated ureters (mega-ureter), or transient hydronephrosis. Studies suggest that the majority (up to almost 90%) of patients diagnosed with fetal pyelectasis in the second trimester have mild hydronephrosis, and 80% of these patients will not need surgical intervention [149,161,162]. The majority of these will have transient hydronephrosis that will resolve. An estimated 10–20% of infants with fetal pyelectasis have VUR [163].

There remains some controversy about postnatal evaluation of fetal pyelectasis. In cases where urologic intervention may be required after birth, such as bilateral hydronephrosis, oligohydramnios and/or bladder abnormalities, a renal/bladder ultrasound should be obtained immediately (1–2 days, depending on the situation) after birth [164]. In addition, in cases of fetal pyelectasis where adherence to follow-up is a concern, postnatal ultrasound may be performed after birth. However, due to relatively lower urine output in the initial 24–48 h after birth, a postnatal ultrasound performed in the first days of life may under-represent the degree of hydronephrosis and be falsely reassuring [164]. Thus, except in select cases as above, optimal timing for a postnatal ultrasound to assess for persistence and severity of fetal pyelectasis is between 7 and 10 days after birth [164]. In cases where there is a normal postnatal ultrasound initially, it is recommended to repeat the ultrasound at 4–6 weeks of age [164].

The more severe the postnatal pyelectasis the greater the likelihood of a condition requiring surgical intervention, either UPJ obstruction or VUR [165]. Therefore, antibiotic prophylaxis and a voiding cystourethrogram (VCUG) is generally indicated for infants with persistent moderate or severe hydronephrosis after birth [153,166]. Unfortunately, the severity of dilation does not correlate well with severity of VUR, and even those with resolved hydronephrosis may have VUR. The consensus guidelines for fetal hydronephrosis recommend discussion with families about risk and benefits of antibiotics and the VCUG in these cases [145].

In the absence of reflux, the next step for evaluation of moderate or severe hydronephrosis is assessment for urinary tract obstruction with a Tc99 Mag3 renal scan with furosemide [167]. Infants require bladder catherization to ensure bladder emptying during the study, as bladder retention can delay emptying times. Tc99 Mag3 and furosemide are secreted by renal proximal tubules, and the immaturity of renal proximal tubules after birth can make a Mag3 scan non-diagnostic in the first few weeks of life. Generally, it is preferable to wait until at least 4–6 weeks after birth to obtain a Mag3 renal scan.

9. Conclusions

CAKUT are one of the leading congenital anomalies. CAKUT most often results from defects in ureteric bud outgrowth and branching or from epithelial differentiation from mesenchymal renal progenitors. Both gene and environmental factors likely contribute to CAKUT. Whereas single gene defects are likely associated with the minority of non-syndromic CAKUT, genetic studies of animals and humans have led to advances in the understanding of the cellular and molecular mechanisms regulating renal and genitourinary tract development. Prenatal ultrasound is useful for detection of CAKUT, although not all defects may be identified. Ultrasound findings and amniotic fluid parameters provide some metrics to help predict prognosis. Postnatal imaging is required to confirm the diagnosis and define the defect. Continued research into genetic and environmental influences on renal development, genotype–phenotype correlations, and predictors of fetal outcomes may lead to improved options for diagnosis and prognosis.

In addition, CAKUT represents a broad spectrum of disease. It includes “mild” defects, including low nephron number, to severe bilateral defects that lead to fetal or neonatal death. Even “mild” defects may result in long term risk for CKD in adulthood. Thus, long term prospective studies, for example, of the extremely preterm infants, will be important to better understand the impact of low nephron number on the life-course and risk for adult onset CKD.