**2.2 Human autopsy studies**

(Keijzer-Veen et al., 2010a).

There have been three published human autopsy studies (apart from case studies) that have examined the effect of preterm birth on postnatal nephrogenesis (Rodriguez et al., 2004; Faa et al., 2010; Sutherland et al., 2011b). Since non-uniform portions of the kidney are usually collected at autopsy, stereological methods cannot be accurately employed. Under these circumstances, the medullary ray glomerular generation counting method (Hinchliffe et al., 1992b) (also referred to as radial glomerular count or glomerular generation count) is a useful technique to provide insight into renal maturity and potentially nephron endowment. The method involves counting all developed glomeruli along one side of clearly distinguishable medullary rays in histological renal sections. Glomeruli are counted from the inner to outer renal cortex. Importantly, in our studies we have found a strong correlation between glomerular generation number and nephron number, which supports the validity of the technique (Sutherland et al., 2011a).

In one of the first autopsy studies conducted, the number of radial glomerular counts in kidneys from extremely preterm neonates (56 neonates) was compared to 10 full-term infants (Rodriguez et al., 2004). Radial glomerular counts were found to be significantly reduced in preterm infants; however, since many of the preterm infants were also intrauterine growth restricted (IUGR) it is difficult to determine the effects of preterm birth *per se* from this study. In a smaller study, Faa *et al* (2010) have reported significantly reduced radial glomerular counts and marked inter-individual variability in the number of glomerular generations among the kidneys from preterm neonates compared to term newborns. In that study, 8 human fetuses, 12 preterm neonates and 3 full-term neonates were examined; it is unknown whether any of the neonates were also IUGR.

As follow on to these studies, we have recently undertaken a study examining kidneys obtained at autopsy from 28 preterm infants and 32 still-born gestational controls (Sutherland et al., 2011b); the preterm group included 6 infants that were also IUGR. Importantly, analyses comparing growth restricted and non-growth restricted kidneys demonstrated no significant differences, although the findings are limited by the small sample of growth restricted neonates. In contrast to the studies described above, we found accelerated nephrogenesis in the preterm group demonstrated by an increase in the number of glomerular generations, a decreased nephrogenic zone width (suggesting

Effects of Preterm Birth on the Kidney 67

Fig. 2. A representative photomicrograph of a histological renal section from a preterm baboon showing abnormal glomeruli in the outer renal cortex; these morphologically immature glomeruli exhibited a shrunken glomerular tuft and enlarged Bowman's space.

long-term (by reducing the functional reserve of nephrons).

Similar to the findings from the human autopsy studies, morphologically abnormal glomeruli were also found in kidneys from preterm baboon neonates; with up to 18% of glomeruli affected (Figure 2). High proportions of abnormal glomeruli were only found in those kidneys from preterm baboons, whereas in gestational controls the proportion of abnormal glomeruli was negligible. The observed abnormal glomeruli were only present in the superficial outer cortex of the preterm kidney suggesting that it is the glomeruli that are recently formed (possibly those formed in the extrauterine environment) that are 'at risk.' Further immunohistochemical analyses demonstrated that the abnormal glomeruli were poorly vascularised (lack of endothelial cell marker, CD31 immunostaining). In addition, immunostaining with the podocyte marker, WT-1, revealed that the abnormal glomeruli were in a relatively immature stage of development since the glomerular tuft contained WT-1 positive cells surrounding a mass of relatively undifferentiated cells. Importantly, there were a large number of parietal epithelial cells surrounding the Bowman's capsule; previous human studies have reported a similar morphology in atubular glomeruli (Gibson et al., 1996; Bariety et al., 2006). If the abnormal glomeruli in the preterm kidneys are atubular, then they will never be functional. Further studies are required to determine whether this is the case. Certainly, a large proportion of non-functional glomeruli in the preterm kidney is likely to have adverse consequences on renal function both in the neonatal period and in the

Scale bar = 100 m

earlier cessation of nephrogenesis postnatally) and a decreased proportion of glomeruli in the immature V (vesicle) -stage of maturation compared to still-born gestational controls. Furthermore, mean renal corpuscle cross sectional area was significantly larger in the preterm kidneys. Of particular concern, kidneys from preterm infants had a higher percentage of structurally abnormal glomeruli compared to the gestational controls with up to 13.7% of glomeruli affected. These abnormal glomeruli exhibited a dilated Bowman's space and shrunken glomerular tuft. The factors associated with the development of abnormal glomeruli are yet unknown and this is an important area of future research.

#### **2.3 Nonhuman primate animal studies**

We have shown that the baboon is an ideal model to study human kidney development, as the ontogeny of the kidney very closely matches that of the human (Gubhaju and Black, 2005). Similar to the human, nephrogenesis in the baboon commences at approximately 30 days of gestation (Hendrickx et al., 1971) and ceases prior to term by 175 days gestation (Term = 185 days gestation) (Gubhaju and Black, 2005). Similar to the wide range in nephron number found in human kidneys, in the kidneys we examined total nephron number ranged from 193,983 to 334,316 in baboons delivered at term.

In collaboration with researchers at the Southwest Foundation for Biomedical Research (San Antonio, Texas, U.S.A) we have examined the kidneys from fetal baboons that have been prematurely delivered and ventilated after birth in a neonatal intensive care unit (NICU) in a similar manner to human preterm babies (Gubhaju et al., 2009). These appropriate weight-for-gestational age baboons were delivered extremely preterm (125 days of gestation); equivalent to approximately 27 weeks gestation in humans. After birth, all preterm neonates were intubated, administered 100 mg/kg surfactant (Survanta; donated by Ross Products, Columbus, OH), and ventilated with pressure limited infant ventilators (InfantStar; donated by Infrasonics, San Diego, CA). All preterm neonates were also treated with ampicillin and gentamicin for the first 7–10 days of life (Thomson et al., 2004). Further doses of antibiotics were only administered in cases of clinically suspected infection. Following birth, the baboon neonates were ventilated in the NICU for a maximum period of 21 days.

In this model, kidney volume, nephron number and size of the renal corpuscle were estimated using unbiased stereology, the gold standard method for the determination of nephron number (Bertram, 2001; Sutherland et al., 2011a). One of the most significant findings from the nonhuman primate studies was the clear evidence that nephrogenesis was on-going in the extrauterine environment following preterm birth. There was structural evidence of on-going nephrogenesis in the outer renal cortex (branching of the ureteric bud, metanephric mesenchyme and Comma and S-shaped bodies) and this was accompanied by a significant increase in the number of glomerular generations and nephron number in the postnatal environment by postnatal day 21 (Gubhaju et al., 2009). Furthermore, kidney weight and volume relative to body weight were significantly higher in the preterm baboon neonates compared to gestational age-matched controls; a finding that has been previously reported in human studies (Huang et al., 2007). There was a significant decrease in glomerular density (glomeruli/gram of kidney) in the kidney from preterm baboon neonates compared to gestational controls suggestive of altered renal growth and potentially an increase in tubular mass.

earlier cessation of nephrogenesis postnatally) and a decreased proportion of glomeruli in the immature V (vesicle) -stage of maturation compared to still-born gestational controls. Furthermore, mean renal corpuscle cross sectional area was significantly larger in the preterm kidneys. Of particular concern, kidneys from preterm infants had a higher percentage of structurally abnormal glomeruli compared to the gestational controls with up to 13.7% of glomeruli affected. These abnormal glomeruli exhibited a dilated Bowman's space and shrunken glomerular tuft. The factors associated with the development of abnormal glomeruli are yet unknown and this is an important area of

We have shown that the baboon is an ideal model to study human kidney development, as the ontogeny of the kidney very closely matches that of the human (Gubhaju and Black, 2005). Similar to the human, nephrogenesis in the baboon commences at approximately 30 days of gestation (Hendrickx et al., 1971) and ceases prior to term by 175 days gestation (Term = 185 days gestation) (Gubhaju and Black, 2005). Similar to the wide range in nephron number found in human kidneys, in the kidneys we examined total nephron number

In collaboration with researchers at the Southwest Foundation for Biomedical Research (San Antonio, Texas, U.S.A) we have examined the kidneys from fetal baboons that have been prematurely delivered and ventilated after birth in a neonatal intensive care unit (NICU) in a similar manner to human preterm babies (Gubhaju et al., 2009). These appropriate weight-for-gestational age baboons were delivered extremely preterm (125 days of gestation); equivalent to approximately 27 weeks gestation in humans. After birth, all preterm neonates were intubated, administered 100 mg/kg surfactant (Survanta; donated by Ross Products, Columbus, OH), and ventilated with pressure limited infant ventilators (InfantStar; donated by Infrasonics, San Diego, CA). All preterm neonates were also treated with ampicillin and gentamicin for the first 7–10 days of life (Thomson et al., 2004). Further doses of antibiotics were only administered in cases of clinically suspected infection. Following birth, the baboon neonates were ventilated in the NICU for a

In this model, kidney volume, nephron number and size of the renal corpuscle were estimated using unbiased stereology, the gold standard method for the determination of nephron number (Bertram, 2001; Sutherland et al., 2011a). One of the most significant findings from the nonhuman primate studies was the clear evidence that nephrogenesis was on-going in the extrauterine environment following preterm birth. There was structural evidence of on-going nephrogenesis in the outer renal cortex (branching of the ureteric bud, metanephric mesenchyme and Comma and S-shaped bodies) and this was accompanied by a significant increase in the number of glomerular generations and nephron number in the postnatal environment by postnatal day 21 (Gubhaju et al., 2009). Furthermore, kidney weight and volume relative to body weight were significantly higher in the preterm baboon neonates compared to gestational age-matched controls; a finding that has been previously reported in human studies (Huang et al., 2007). There was a significant decrease in glomerular density (glomeruli/gram of kidney) in the kidney from preterm baboon neonates compared to gestational controls suggestive of altered renal growth and

future research.

**2.3 Nonhuman primate animal studies** 

maximum period of 21 days.

potentially an increase in tubular mass.

ranged from 193,983 to 334,316 in baboons delivered at term.

Fig. 2. A representative photomicrograph of a histological renal section from a preterm baboon showing abnormal glomeruli in the outer renal cortex; these morphologically immature glomeruli exhibited a shrunken glomerular tuft and enlarged Bowman's space. Scale bar = 100 m

Similar to the findings from the human autopsy studies, morphologically abnormal glomeruli were also found in kidneys from preterm baboon neonates; with up to 18% of glomeruli affected (Figure 2). High proportions of abnormal glomeruli were only found in those kidneys from preterm baboons, whereas in gestational controls the proportion of abnormal glomeruli was negligible. The observed abnormal glomeruli were only present in the superficial outer cortex of the preterm kidney suggesting that it is the glomeruli that are recently formed (possibly those formed in the extrauterine environment) that are 'at risk.' Further immunohistochemical analyses demonstrated that the abnormal glomeruli were poorly vascularised (lack of endothelial cell marker, CD31 immunostaining). In addition, immunostaining with the podocyte marker, WT-1, revealed that the abnormal glomeruli were in a relatively immature stage of development since the glomerular tuft contained WT-1 positive cells surrounding a mass of relatively undifferentiated cells. Importantly, there were a large number of parietal epithelial cells surrounding the Bowman's capsule; previous human studies have reported a similar morphology in atubular glomeruli (Gibson et al., 1996; Bariety et al., 2006). If the abnormal glomeruli in the preterm kidneys are atubular, then they will never be functional. Further studies are required to determine whether this is the case. Certainly, a large proportion of non-functional glomeruli in the preterm kidney is likely to have adverse consequences on renal function both in the neonatal period and in the long-term (by reducing the functional reserve of nephrons).

Effects of Preterm Birth on the Kidney 69

renal tubule (short length of the tubules, and changes in the density and structure of transporter proteins) (Jones and Chesney, 1992), or due to renal injury (Ueda and Shah,

Studies that have assessed FENa during the neonatal period have determined that sodium excretion is significantly higher in preterm neonates compared to term controls (Siegel and Oh, 1976; Aperia et al., 1981), and significantly decreases with increasing gestational (Gallini et al., 2000) and postnatal age (Ross et al., 1977; Sulyok et al., 1979; Aperia et al., 1981; Gallini et al., 2000; Giapros et al., 2007). Therefore, with increasing renal maturity a positive sodium balance (low FENa) is achieved, which is essential for the growth and development of the

Endogenous creatinine is the most practical and commonly used marker of renal function, with calculated creatinine clearance widely used as an estimate of glomerular filtration rate (GFR). In the clinical setting, repeated serum creatinine levels are used to gauge renal function in neonates; this is an easily obtainable measure via routine blood collection and does not rely on timed urine samples or additional invasive procedures. This method does, however, have significant limitations. Immediately following birth, serum creatinine levels are equivalent to the fetal levels, which during the third trimester of gestation rise from 42 µmol/L at 23 weeks to 47 µmol/L at term; the increase likely reflecting an increase in muscle mass (Moniz et al., 1985). In the first forty-eight hours following birth, however, serum creatinine levels significantly increase (Bueva and Guignard, 1994; Miall et al., 1999). This is considered to be due, in part, to tubular creatinine reabsorption, as has been evidenced in a neonatal animal model (Matos et al., 1998), and also due to the inadequacy of glomerular filtration during the early postnatal period (Miall et al., 1999). Peak serum creatinine levels are reached at postnatal day 2-4 of life, with the highest levels and most delayed timing of the peak creatinine level seen in neonates at the lowest gestational ages (Miall et al., 1999). During the first week of life following preterm birth, GFR is significantly lower in preterm neonates than in term-born controls (Siegel and Oh, 1976; Finney et al., 2000; Schreuder et al., 2009), and is significantly positively correlated with both gestational age at birth, and postnatal age (Clark et al., 1989; Gordjani et al., 1998; Iacobelli et al., 2009). Compared to term neonates, the rate of increase in GFR after birth is slower in neonates born preterm (Gordjani et al., 1998). Up until two months of age there are similar findings, with a number of studies observing an increase in GFR concurrent to increasing gestational and postnatal ages (Ross et al., 1977; Fawer et al., 1979; Sulyok et al., 1979; Aperia et al., 1981; Wilkins, 1992; Bueva and Guignard, 1994; Gallini et al., 2000; Cuzzolin et al., 2006; Thayyil et al., 2008). Although a number of studies have now been performed in this area, there is still a lack of clear definition regarding expected GFR values in the preterm neonate. Recently published standard curves of GFRs in neonates born at 27-31 weeks gestational age, from 7 to 28 days of life, will go some way in aiding in the clinical interpretation of renal function in

Given that age has been found to be a strong determinant of GFR, the low GFR observed in the preterm neonate after birth is likely the result of renal immaturity (a low number of filtering glomeruli), and it is also likely to be influenced by differences in renal blood flow and vascular resistance. It is essential that GFR is monitored in the postnatal period following preterm birth, as a very low GFR is likely to impair renal drug clearance, leading

neonate and the maintenance of fluid homeostasis (Engle, 1986).

this particular group of neonates (Vieux et al., 2010).

to nephrotoxicity.

2000; Bonventre, 2007).

**3.2 Glomerular filtration rate**

#### **3. Renal function in the preterm neonate**

There have been a number of studies that have examined the effects of preterm birth on renal function. However, it must be kept in mind when interpreting the data from these studies that the function of the immature preterm kidney is likely to be quite different to that of the term infant, which in turn is likely to be quite different to the adult. Hence, although the 'normal' levels of the standard markers of renal function (such as serum creatinine and urinary albumin) have been well-established for the adult population, the standard levels in the neonate, especially those of the preterm neonate, are not clearly defined. This often makes the clinical assessment of renal function in the preterm neonate difficult. In future research, it is necessary to establish the 'normal' levels of renal function in the preterm infant and to identify robust biomarkers for the early diagnosis of renal injury in the neonatal period, which may in turn prevent long-term renal dysfunction.

#### **3.1 Fluid and electrolyte homeostasis**

An imbalance of fluid and electrolyte intake versus excretion is very common in premature neonates, and can lead to significant morbidity and mortality (Bhatia, 2006); hypernatraemia, for example, can result in severe neurological injury (Moritz and Ayus, 2005). Insensible fluid loss is a major factor (Bhatia, 2006), and is primarily transcutaneous due to the developmental immaturity of the skin and a high body surface area to body water mass ratio (Baumgart and Costarino, 2000). Equally, the delayed loss of extracellular fluid volume following preterm birth is also associated with an increased risk of morbidity, in particular bronchopulmonary dysplasia (Oh et al., 2005) and patent ductus arteriosus (Bell and Acarregui, 2008).

Three phases of fluid and electrolyte homeostasis have been observed in the immediate period following preterm birth; these phases occurred similarly in extremely low birth weight infants and those at older gestational ages (Lorenz et al., 1982; Lorenz et al., 1995). As described by Lorenz *et al.*, (1982; 1995) in the first 24 hours following birth, a period known as the pre-diuretic stage, urine output is minimal and sodium excretion is low. On postnatal days 2-3, termed the diuretic phase, sodium excretion and urine output significantly increase, which occurs independently of fluid intake. From approximately days 4-5 of life, the post-diuretic phase, urine output changes in response to fluid intake (Lorenz et al., 1982; Lorenz et al., 1995). Importantly, however, the postnatal time-point that these phases occur, and their duration, differ between individual neonates (Lorenz et al., 1995), as does the amount of insensible fluid loss; together, this highlights the need for an individualised approach to fluid therapy in preterm neonates.

Urine output is the most commonly and easily measured indicator of renal function in the preterm neonate. Urine output less than 0.5 ml/kg/h, known as oliguria, can be indicative of acute kidney injury (AKI). AKI, however, can also be non-oliguric, therefore urine output is not a very specific indicator of renal function. Furthermore, from the post-diuretic phase of fluid homeostasis urine output is highly dependent upon fluid intake; high intakes may artificially increase urine output, while not accurately reflecting renal functional capacity.

The most common measure of electrolyte balance in the neonate is the calculation of the fractional excretion of sodium (FENa), which is the percentage of sodium that is excreted and not taken up through tubular reabsorption. The calculation of FENa takes into account the levels of both serum and urine sodium, and it is corrected for serum and urine creatinine levels. Therefore, high urine sodium levels may be indicative of structural immaturity of the

There have been a number of studies that have examined the effects of preterm birth on renal function. However, it must be kept in mind when interpreting the data from these studies that the function of the immature preterm kidney is likely to be quite different to that of the term infant, which in turn is likely to be quite different to the adult. Hence, although the 'normal' levels of the standard markers of renal function (such as serum creatinine and urinary albumin) have been well-established for the adult population, the standard levels in the neonate, especially those of the preterm neonate, are not clearly defined. This often makes the clinical assessment of renal function in the preterm neonate difficult. In future research, it is necessary to establish the 'normal' levels of renal function in the preterm infant and to identify robust biomarkers for the early diagnosis of renal injury in the neonatal period, which may in turn prevent long-term renal dysfunction.

An imbalance of fluid and electrolyte intake versus excretion is very common in premature neonates, and can lead to significant morbidity and mortality (Bhatia, 2006); hypernatraemia, for example, can result in severe neurological injury (Moritz and Ayus, 2005). Insensible fluid loss is a major factor (Bhatia, 2006), and is primarily transcutaneous due to the developmental immaturity of the skin and a high body surface area to body water mass ratio (Baumgart and Costarino, 2000). Equally, the delayed loss of extracellular fluid volume following preterm birth is also associated with an increased risk of morbidity, in particular bronchopulmonary dysplasia (Oh et al., 2005) and patent ductus arteriosus

Three phases of fluid and electrolyte homeostasis have been observed in the immediate period following preterm birth; these phases occurred similarly in extremely low birth weight infants and those at older gestational ages (Lorenz et al., 1982; Lorenz et al., 1995). As described by Lorenz *et al.*, (1982; 1995) in the first 24 hours following birth, a period known as the pre-diuretic stage, urine output is minimal and sodium excretion is low. On postnatal days 2-3, termed the diuretic phase, sodium excretion and urine output significantly increase, which occurs independently of fluid intake. From approximately days 4-5 of life, the post-diuretic phase, urine output changes in response to fluid intake (Lorenz et al., 1982; Lorenz et al., 1995). Importantly, however, the postnatal time-point that these phases occur, and their duration, differ between individual neonates (Lorenz et al., 1995), as does the amount of insensible fluid loss; together, this highlights the need for an

Urine output is the most commonly and easily measured indicator of renal function in the preterm neonate. Urine output less than 0.5 ml/kg/h, known as oliguria, can be indicative of acute kidney injury (AKI). AKI, however, can also be non-oliguric, therefore urine output is not a very specific indicator of renal function. Furthermore, from the post-diuretic phase of fluid homeostasis urine output is highly dependent upon fluid intake; high intakes may artificially increase urine output, while not accurately reflecting renal functional capacity. The most common measure of electrolyte balance in the neonate is the calculation of the fractional excretion of sodium (FENa), which is the percentage of sodium that is excreted and not taken up through tubular reabsorption. The calculation of FENa takes into account the levels of both serum and urine sodium, and it is corrected for serum and urine creatinine levels. Therefore, high urine sodium levels may be indicative of structural immaturity of the

individualised approach to fluid therapy in preterm neonates.

**3. Renal function in the preterm neonate** 

**3.1 Fluid and electrolyte homeostasis**

(Bell and Acarregui, 2008).

renal tubule (short length of the tubules, and changes in the density and structure of transporter proteins) (Jones and Chesney, 1992), or due to renal injury (Ueda and Shah, 2000; Bonventre, 2007).

Studies that have assessed FENa during the neonatal period have determined that sodium excretion is significantly higher in preterm neonates compared to term controls (Siegel and Oh, 1976; Aperia et al., 1981), and significantly decreases with increasing gestational (Gallini et al., 2000) and postnatal age (Ross et al., 1977; Sulyok et al., 1979; Aperia et al., 1981; Gallini et al., 2000; Giapros et al., 2007). Therefore, with increasing renal maturity a positive sodium balance (low FENa) is achieved, which is essential for the growth and development of the neonate and the maintenance of fluid homeostasis (Engle, 1986).

#### **3.2 Glomerular filtration rate**

Endogenous creatinine is the most practical and commonly used marker of renal function, with calculated creatinine clearance widely used as an estimate of glomerular filtration rate (GFR). In the clinical setting, repeated serum creatinine levels are used to gauge renal function in neonates; this is an easily obtainable measure via routine blood collection and does not rely on timed urine samples or additional invasive procedures. This method does, however, have significant limitations. Immediately following birth, serum creatinine levels are equivalent to the fetal levels, which during the third trimester of gestation rise from 42 µmol/L at 23 weeks to 47 µmol/L at term; the increase likely reflecting an increase in muscle mass (Moniz et al., 1985). In the first forty-eight hours following birth, however, serum creatinine levels significantly increase (Bueva and Guignard, 1994; Miall et al., 1999). This is considered to be due, in part, to tubular creatinine reabsorption, as has been evidenced in a neonatal animal model (Matos et al., 1998), and also due to the inadequacy of glomerular filtration during the early postnatal period (Miall et al., 1999). Peak serum creatinine levels are reached at postnatal day 2-4 of life, with the highest levels and most delayed timing of the peak creatinine level seen in neonates at the lowest gestational ages (Miall et al., 1999).

During the first week of life following preterm birth, GFR is significantly lower in preterm neonates than in term-born controls (Siegel and Oh, 1976; Finney et al., 2000; Schreuder et al., 2009), and is significantly positively correlated with both gestational age at birth, and postnatal age (Clark et al., 1989; Gordjani et al., 1998; Iacobelli et al., 2009). Compared to term neonates, the rate of increase in GFR after birth is slower in neonates born preterm (Gordjani et al., 1998). Up until two months of age there are similar findings, with a number of studies observing an increase in GFR concurrent to increasing gestational and postnatal ages (Ross et al., 1977; Fawer et al., 1979; Sulyok et al., 1979; Aperia et al., 1981; Wilkins, 1992; Bueva and Guignard, 1994; Gallini et al., 2000; Cuzzolin et al., 2006; Thayyil et al., 2008). Although a number of studies have now been performed in this area, there is still a lack of clear definition regarding expected GFR values in the preterm neonate. Recently published standard curves of GFRs in neonates born at 27-31 weeks gestational age, from 7 to 28 days of life, will go some way in aiding in the clinical interpretation of renal function in this particular group of neonates (Vieux et al., 2010).

Given that age has been found to be a strong determinant of GFR, the low GFR observed in the preterm neonate after birth is likely the result of renal immaturity (a low number of filtering glomeruli), and it is also likely to be influenced by differences in renal blood flow and vascular resistance. It is essential that GFR is monitored in the postnatal period following preterm birth, as a very low GFR is likely to impair renal drug clearance, leading to nephrotoxicity.

Effects of Preterm Birth on the Kidney 71

renal dysfunction (Lavery et al., 2008; Parravicini, 2010); in particular, NGAL shows potential as a promising biomarker of late-onset sepsis (Lavery et al., 2008; Parravicini et al., 2010). Urinary NGAL levels also strongly correlated with gestational and postnatal age (Lavery et al., 2008; Huynh et al., 2009), perhaps reflecting the renal production of NGAL during nephrogenesis (Gwira et al., 2005) which is often still ongoing during the early postnatal period. Normative values for urinary NGAL in preterm neonates with uncomplicated clinical courses have also been published, with the results indicating a

Proteinuria, the presence of high levels of protein in the urine, may be of glomerular and/or tubular origin. The number of different proteins that have been identified in the adult urinary proteome is 1,543, and these are primarily of membrane, extracellular and lysosomal origin (Adachi et al., 2006). Despite this large number, unless renal function is impaired, proteins are normally only present at very low levels in urine, due to the function of the

Presence of high molecular weight (HMW) proteins in the urine, such as albumin traditionally indicates a disruption in the integrity of the glomerular filtration barrier. Recent debate, however, has suggested that the contribution of tubular reabsorption of albumin from the filtrate may be greater than previously considered (Comper et al., 2008). In general, albuminuria is a strong marker for renal and cardiovascular disease, and a risk factor for mortality (Matsushita et al., 2010; Methven et al., 2011). Normally, adults excrete less than 30 mg of albumin per 24 hours (Mathieson, 2004). Urinary albumin levels between 30 – 300 mg in 24 hours is considered microalbuminuria, with levels greater than 300 mg classified as macroalbuminuria (Mathieson, 2004). Traditionally, 24 hour urine samples were required for reliable estimates of urinary protein. However, single random spot samples with protein levels corrected for urine creatinine, have been shown to be significantly correlated with results from 24 hour collections, and are equally effective in the prediction of outcomes (Ralston et al., 1988; Methven et al., 2011). In neonates, 24 hour urine collection is difficult, therefore analysis of urinary protein levels are undertaken using spot

Low molecular weight (LMW) proteins, such as α1-microglobulin, β2-microglobulin and retinol binding protein pass freely through the glomerular filter and undergo reuptake via proximal tubule cells (Tomlinson, 1992). Megalin and cubulin have been identified as important receptors involved in tubular protein uptake, with mutations in the receptors resulting in proteinuria (Christensen and Birn, 2001). To date, LMW protein levels in the urine are not routinely measured in the clinical setting. Importantly, however, amongst the LMW proteins there may be potential novel biomarkers of tubular cell injury and this

In the preterm neonate, few studies have been conducted to examine urine protein excretion. In general, there is a high variability in urine albumin levels between individual neonates (Clark et al., 1989; Fell et al., 1997), with the highest levels exhibited by those with a low gestational age at birth and those that are clinically unstable (Galaske, 1986; Clark et al., 1989; Tsukahara et al., 1994; Fell et al., 1997; Awad et al., 2002b). The majority of studies have only been conducted during the first week of life following preterm birth. However, in a study by Tsukahara *et al.* (1994) urine albumin levels were assessed in preterm and term

greater variation in females than males (Huynh et al., 2009).

glomerular filtration barrier and tubular reabsorption capabilities.

urine samples obtained using urine collection bags.

requires further research (Rosner, 2009; Parikh et al., 2010).

**3.4 Proteinuria**
