Extra uterine development of preterm kidneys

There are more than 15 million preterm (birth before 37 completed weeks of gestation) births globally per year, and the number is increasing [1]. Advances in perinatal care and the use of antenatal glucocorticoids, antibiotics, surfactant, and improved ventilation strategies have contributed to improving outcomes for premature infants. Improved health care in developing countries has resulted in an increased number of premature neonates surviving to adulthood [2]. Prematurity is associated with long-term morbidities such as cognitive, psychological, neurological and visual deficits [3]. In recent years, there is increasing evidence to indicate that prematurity is an independent risk factor for chronic renal diseases (CKD) [4, 5]. The pathogenesis of CKD in premature neonates remains to be fully understood. The currently available evidence indicates that prematurity results in reduced nephron endowment by interrupting normal organogenesis [5]. Nephron number can be determined by histopathological examination [6]. Renal volume measurements have also been shown to have good correlation with nephron number [7, 8].

Currently, microalbuminuria is used as an early indication of glomerular pathology [9]. Measuring urine albumin over a 24-h period is often not clinically practical, particularly in neonates; hence, urine spot tests and measurement of urine albumin-creatinine ratio (ACR) have become popular in clinical practice. Nephrin is a 180-kD transmembrane protein expressed in glomerular podocytes. It forms an integral part of podocytes, which, together with endothelial cells and the basement membrane, form the glomerular filtration barrier. Podocytopathies lead to the detectable levels of nephrin in the urine. Both animal and human studies have demonstrated that nephrinuria occurs early in glomerular injury, often preceding albuminuria, and that there is a positive correlation between nephrinuria and the severity of renal diseases [10]. Urinary nephrin analysis thus has the potential to be an important biomarker of early glomerular injury [10].

We carried out a study to determine the impact of prematurity on renal development in a cohort of extremely premature neonates. The primary outcomes were nephrinuria and albuminuria; renal volume (a surrogate for nephron number) and glomerular filtration rate were secondary outcomes.

During the study, 131 premature neonates less than 28 weeks gestation were admitted to the neonatal unit. There were nine deaths. Consent was obtained in 59 neonates, and one infant was excluded because of hydronephrosis. Complete data sets were available in 53 neonates. During the same period, 31 term neonates were recruited as the control group. The median gestational age of the premature cohort was 26.4 [24.7–27.4] weeks, with a mean birth weight of 886 (179) g. At 37 weeks PMA, the mean weight of the premature group was 2416 (328) g. The mean gestational age of term neonates was 39.1 (1.2) weeks and the mean birth weight was 3406 (406) g.

There was a statistically significant decline in NCR from 32 to 37 weeks PMA (0.07 [0.04–0.18] to 0.04 [0.03–0.09] g/mol; P = 0.028). However, there was no significant difference between NCR at 37 weeks PMA and levels in term neonates (0.04 [0.03–0.09] vs 0.05 [0.03–0.14] g/mol; P = 0.42). There was also a statistically significant in decline of ACR from 32 to 37 weeks PMA (24.5 [14.5–43.5] to 6.7 [4.2–23.8] g/mol; P = 0.0009). ACR at 37 weeks PMA was however significantly higher compared to controls (6.7 [4.2–23.8] vs 5.1 [1.9–6.8] g/mol; P = 0.022).

The TKV almost doubled from 10.3 (2.9) cm³ at 28 weeks PMA to 19.2 (3.7) cm³ by 37 weeks PMA (P < 0.0001). TKV at 37 weeks PMA was significantly smaller compared to term control (19.2 (3.7) vs 26.3 (7.0) cm³; P < 0.0001). Figure 1 shows the changes in TKV in the early postnatal period. Figure 2 shows eGFR in the premature neonates at the different postmenstrual age in comparison to term. There was no significant change in the eGFR from 28 to 37 weeks PMA (one-way ANOVA; P = 0.07). There was also no statistically significant difference in the eGFR in premature neonates at 37 weeks PMA when compared with term infants (43.5 [39.7–48.9] vs 42.0 [38.2–50.0] mL/min/1.73 m²; P = 0.75). Table 1 shows the comparison between term control and premature neonates at 37 weeks PMA.

Premature neonates in our cohort have smaller renal volumes (hence reduced nephron number) compared to term neonates. Despite this, their eGFRs were the same as those measured in term neonates. We postulate that preterm babies, despite having fewer glomeruli, achieve a similar eGFR to their term counterparts by single nephron hyperfiltration. This sets them on a trajectory for chronic renal diseases in due course, as proposed by Brenner et al. [15].

Urinary NCR was elevated in preterm neonates in early postnatal life, but subsequently, it declined and by 37 weeks PMA, there was no significant difference between premature and term neonates in NCR. To the best of our knowledge, this is the first time urine nephrin-creatinine ratio has been measured in preterm and term neonates. Previous studies have investigated nephrin excretion in adults and in women with preeclampsia [10]. Nephrin is an integral part of the glomerular filtration barrier and any injury to podocytes results in nephrinuria. Evidence of glomerular injury as a result of prematurity has been previously demonstrated in both human and animal models using other methods [16]. Gubhaju et al. demonstrated in preterm baboons that prematurity was associated with a high prevalence of abnormal glomeruli (up to 18%) on renal histological examination. These abnormal glomeruli with a cystic Bowman’s space and shrunken glomerular tuft were present in the superficial renal cortex [17]. Similar findings have also been reported from autopsy studies conducted on deceased human preterm neonates [18] with preterm kidneys showing a greater percentage of morphologically abnormal glomeruli and a significantly larger cross-sectional area of the renal corpuscle, suggesting renal hyperfiltration.

Abnormal extra uterine glomerular development may be the result of ischaemia of the outer still developing nephrons, which are the most immature. Pappenheimer and Kinter proposed that cell-free blood is “skimmed” from the periphery of the intralobular arteries to enter the afferent arterioles of deeper glomeruli [19]. Animal studies have also demonstrated that the outermost part of the kidney receives only 20% of the blood perfusing the innermost nephrons. [20]. Thus, the most immature nephrons in the preterm kidney would be exposed to the lowest oxygen tension potentially causing ischaemic damage.

Renal toxicity from antibiotics is an additional potential cause of impaired renal function in preterm infants. In our department, a combination of beta-lactam antibiotic (penicillin) and aminoglycoside (gentamicin) is often used as first-line antibiotics for 5 to 7 days to treat early-onset sepsis. This combination is the preferred choice based on the microbial profile observed in our department [21]. The second line of preferred antibiotic, based on our microbial pattern for late-onset sepsis, is vancomycin and cephalosporin [22]. Therapeutic drug monitoring is carried out routinely on all neonates receiving aminoglycosides or vancomycin, and administration is adjusted to ensure optimal therapeutic levels. Despite this, it is possible that gentamicin and vancomycin could have contributed to glomerular injury in the early neonatal period. Patent ductus arteriosus (PDA) is a common neonatal condition in which non-steroidal anti-inflammatory drugs have been used in the past; however, we no longer use non-steroidal anti-inflammatory drugs (indomethacin and/or ibuprofen) in the management of PDA [23].

The presence of albumin in the urine could indicate either immaturity of renal glomerular function, tubular function or both. Studies designed to investigate this show that unlike their term counterparts, preterm neonates show high variability in urinary albumin levels between individuals [24]. Tsukahara et al. demonstrated that albuminuria decreased postnatally in term neonates, while it remained almost constant in preterm neonates [25]. These authors showed that glomerular permeability increased and proximal tubular protein reabsorption decreased with increasing degrees of prematurity. In another study, the same authors attempted to identify the relative contribution of tubular and glomerular dysfunction to albuminuria by measuring on days 1, 4, 7, 14, and 28 days urinary B2-microglobulin (B2M) concentration (an indicator of tubular reabsorption) in a cohort of premature neonates (gestational age 32 (2) weeks) [26]. They found that urinary B2M increased from day 1 to a peak at day 7, and then it declined. In term neonates, the B2M peaks earlier (day 4) and then falls. It is proposed that between 32 and 35 weeks of gestation, the proximal tubules mature and once this is complete, “glomerulotubular balance” is reached, and urinary albumin excretion declines. Awad et al. demonstrated that glomerular and tubular functions were also relatively impaired at birth among healthy term neonates and this corrected spontaneously by 72 h of life [24]. Urinary albumin excretion is elevated in premature neonates and shows a decline in the first 72 h of life. Normalization of glomerular function was delayed in sick premature neonates, with a slower rate of decline in urinary ACR [24]. Our findings suggest that the decline in ACR from 28 to 37 weeks represents maturation of both glomerular and tubular handling of albumin. Elevation of ACR in premature neonates at 37 week PMA when compared with term controls suggests that these babies may have residual glomerular damage. It is tempting to suggest that this is related to the higher single nephron GFR of these infants as studies in fetal and neonatal sheep have shown that hyperfiltration is associated with increased urinary protein excretion [27, 28].

The main limitation of our study is that our recruitment rate is approximately 50%. The predominant reason is that nearly half of our admissions are neonates from regional areas, who are often transferred back to health facilities nearer to home after 32 weeks PMA for ongoing care. This limits the availability of the neonates for full participation in our study. There is approximately 3 weeks in gestational age difference between term (control) and premature (at term corrected) groups. Term birth is defined at any birth after 37 weeks of gestation, up to 42 weeks. To recruit term neonates at the exact same gestation will be ideal, but more challenging in clinical practice. Furthermore, nephrogenesis is complete between 34 and 36 weeks of gestation [29]; hence, the difference in renal volume (hence nephron number) and renal function due to this disparity is likely to be minimum.

Despite having smaller renal volume (and reduced nephron number), premature neonates achieve a similar eGFR as term neonates through glomerulo-hyperfiltration. They also show evidence of glomerular injury in the early neonatal period as demonstrated by nephrinuria and albuminuria. This may be due to a combination of abnormal glomerular development, ongoing ischaemia and the use of potentially nephrotoxic antibiotics. By term equivalence, nephrinuria returns to normal but albuminuria remains abnormal. Calculated glomerular filtration rate in the preterm infants is also normal at term equivalence in spite of the relatively small kidneys, strongly suggesting the presence of single nephron hyperfiltration. The hyperfiltration in preterm born infants may predispose them to continuing loss of additional nephrons and a higher risk of earlier-onset renal failure.