Acute kidney injury frequently develops following the term perinatal hypoxia-ischaemia. Quantifying the degree of acute kidney injury is difficult, however, as the methods currently in use are suboptimal. Acute kidney injury management is largely supportive with little evidence basis for many interventions. This review discusses management strategies and novel biomarkers that may improve diagnosis and management of renal injury following perinatal hypoxia-ischaemia.
Following perinatal hypoxia-ischaemia, acute kidney injury is common. Management of neonatal acute kidney injury is largely supportive. Novel acute kidney injury biomarkers may play a role in optimizing new categorical definitions of renal injury. Studies are needed to investigate the impact of neonatal acute kidney injury on long-term outcome.
- Renal dysfunction and acute kidney injury frequently complicate term perinatal hypoxia-ischaemia.
- There are challenges surrounding diagnosis of acute kidney injury in asphyxiated infants.
- Management of renal dysfunction following perinatal hypoxia-ischaemia is largely based on expert advice, and few of the management strategies are evidence based. Acute kidney injury biomarkers may improve diagnostic accuracy, help guide therapy and aid management of acute kidney injury following perinatal hypoxia-ischaemia.
The reported incidence of acute kidney injury (AKI) amongst asphyxiated term infants is high (50–72%) [1-7]. However, different definitions of AKI are used in each study which makes comparison challenging and true incidence rates difficult to calculate. Hypoxia-ischaemia is one of the most commonly identified conditions causing neonatal AKI accounting for 30–40% of AKI in this period [8, 9]. Poor condition at birth, defined by a 5 min Apgar score <6 in asphyxiated infants, appears to correlate with increasing incidence of AKI [2-4] as does severity of asphyxia [5, 6, 10]. A hypoxic-ischaemic aetiology is also common to renal injury secondary to complex cardiac surgery in the neonatal period where AKI is quoted to affect 64% (170/264) of neonates in one study .
The mortality rates of neonatal AKI following perinatal hypoxia-ischaemia are difficult to define accurately; Agras et al. reported a rate of 22.2%. Many infants with perinatal hypoxia-ischaemia require mechanical ventilation (MV), and the requirement for MV is significantly associated with up to a 17-fold increase in mortality in infants with AKI [9, 12]. The mortality and morbidity of newborns with AKI is exacerbated in the presence of multiorgan failure [2, 5, 9, 13, 14], and oliguric AKI also portends a poorer outcome [1, 3, 5, 8, 10, 12, 13]. It has been proposed that renal involvement following perinatal hypoxia-ischaemia may correlate with severity of long-term neurological injury [5, 14]. Caution is warranted, however, when interpreting outcome ascribed to neonatal AKI as the true association between AKI and mortality after perinatal hypoxia-ischaemia is not known [1, 2, 5, 6, 14].
The majority of AKI following perinatal hypoxia-ischaemia is prerenal in origin . An asphyxial insult triggers a protective reflex whereby blood flow is preferentially maintained to the heart, brain and adrenal glands with reduction in blood flow to the kidneys and consequent renal hypoperfusion, ischaemia [15, 16] and risk of prerenal failure.
The kidneys of newborn infants are particularly susceptible to hypoperfusion due to high renal vascular resistance, high plasma renin activity, low glomerular filtration rate (GFR), decreased intercortical perfusion and decreased reabsorption of sodium in the proximal tubules [9, 15]. Together these factors result in neonates being very vulnerable to acute tubular necrosis (ATN) or cortical necrosis following renal hypoperfusion due to a hypoxic-ischaemic event. AKI secondary to perinatal hypoxia-ischaemia can persist for many days with tubular dysfunction (measured by fractional excretion of sodium, renal failure index and urinary myoglobin) remaining significantly deranged . Less frequently progression to complete anuria can occur. Supportive care is the main treatment modality available for AKI post-perinatal hypoxia-ischaemia.
Diagnosis of AKI
Elevated serum creatinine (SCr) (>132.6 μmol/L), lowered GFR and reduced urinary output are commonly used markers of AKI in neonates, however all are sub-optimal . Firstly, SCr may not increase until about 25–50% of renal function is lost  as the relationship between GFR and SCr relies on a steady state where the endogenous generation of creatinine from muscle is in equilibrium with renal clearance. In the first 48 h of life, neonatal SCr reflects maternal not neonatal renal function. Neonatal GFR also steadily increases during the first few months of life [1, 10, 16], which may influence the interpretation of early changes in SCr. Lastly, the method of SCr quantification utilized can affect the result, as the older Jaffe assay is influenced by medications and hyperbilirubinaemia .
A high proportion of post-asphyxial AKI is non-oliguric and rates are quoted as high as 60–78% [3-5]. However, if the definition for AKI used relies on the presence of oliguria then these infant's AKI may not be identified resulting in delayed or missed diagnosis. Oliguria is not a sensitive marker of AKI, and if serial SCr measurements are not monitored, AKI may not be recognized.
The standardization of AKI diagnosis has also been hampered by lack of an internationally recognized definition. Historically AKI definitions have relied frequently upon the identification of a variable combination of rising SCr [e.g. >132.6 μmol/L ] with absolute cut-off values for the presence or absence of AKI, oliguria [e.g. 1 mL/kg/h ] and elevated blood urea nitrogen (BUN) [e.g. >14.28 mmol/L ]. Therefore, the ability to pool data and draw conclusions across studies of neonatal AKI is limited.
However, recently three similar definitions of AKI have been derived in adults: Kidney Disease Improving Global Outcomes (KDIGO) Guidelines , Risk Injury Failure Loss End-Stage Kidney Disease (RIFLE) criteria  and Acute Kidney Injury Network (AKIN)  staging. The paediatric RIFLE (pRIFLE) [19, 20] and AKIN have been modified for the use in children but it is important to recognize that these definitions were developed based on adult data and therefore do not automatically translate to neonatal AKI. AKI in hospitalized children using both the pRIFLE and AKIN criteria is common and associated with poor short-term outcomes, for example, mortality/length of stay .
Efforts have been made recently to validate these classification systems in a neonatal population: AKIN [6, 11, 21] and RIFLE . Kaur et al. employed AKIN criteria to stage AKI in 36 term neonates with birth asphyxia. AKI was significantly more common in the severe asphyxia group (12/25), 56%, compared to the moderate asphyxia group (1/11), 9.1%, giving an overall incidence of AKI using AKIN criteria of (15/36) 41.7%. However, if the diagnosis of AKI had been based solely on SCr > 132.6 μmol/L, then only (10/36) 27.8% of the infants would have met the diagnostic criteria for AKI, highlighting the insensitivity of an absolute SCr cut-off value as a marker of AKI.
The Future of Neonatal AKI Diagnosis and Monitoring
Intensive research has been undertaken in recent years in the field of AKI biomarkers both in paediatric and in adult populations. Urinary and serum neutrophil gelatinase-associated lipocalin (NGAL) shows particular promise as a marker of ischaemic renal injury as occurs following both cardiopulmonary bypass (CPB) and perinatal asphyxia. Urinary and serum NGAL have been shown to identify AKI in paediatric patients following CPB much earlier than SCr [21, 23]. Urinary and serum NGAL were excellent predictors of AKI following CPB in 374 children (35 neonates) with area under receiver operating characteristic curves (AUROC) of 0.95 and 0.95, respectively . Serum and urinary NGAL and urinary cystatin C levels were also significantly increased in an asphyxiated group of term infants (n = 13) compared to controls (n = 22) on Day 1 and showed excellent ability to predict AKI with AUROCs of 0.94, 0.90 and 0.93, respectively .
Urinary interleukin-18 (IL-18) [23, 25], kidney injury molecule-1 (KIM-1) [23, 24], liver-type fatty acid-binding protein (L-FABP) [23, 26] and cystatin C [24, 25, 27] all show promise as candidates for entry onto a panel of neonatal AKI biomarkers. An AKI biomarker panel could potentially identify neonatal AKI at a much earlier time point following a perinatal hypoxic-ischaemic insult allowing earlier institution of supportive treatment. Whether such interventions could improve long-term outcome, however, requires further study in a large cohort of infants who have well-defined AKI based on AKIN criteria, AKI biomarker levels and complete follow-up data.
The challenge of defining neonatal AKI accurately using novel biomarkers is significantly hampered by the reliance on historical gold standard definitions which as discussed previously have many disadvantages. To develop novel definitions of neonatal AKI in a systematic manner, normative values for AKI biomarkers in healthy term neonates and critically ill term newborns need to be established and compared to, perferably AKIN criteria for AKI. AKIN is superior to the pRIFLE categorical definition of AKI in neonates as it uses a combination of rising SCr and decreasing urinary output and avoids the use of estimated creatinine clearance which is not a reliable indicator of a newborn's true creatinine clearance . Although AKIN criteria are not perfect, they are superior to using SCr alone. Once biomarker cut-off values for the identification of AKI are established, they can be further validated and incorporated into a novel categorical definition of neonatal AKI.
Management of AKI Post-Perinatal Hypoxia-Ischaemia
Using evidence-based medicine is challenging in the management of post-asphyxial AKI due to the paucity of published studies in the literature. Therefore, many of the recommendations described here are based on expert advice.
History and clinical examination
The antenatal history may give vital information on the peripartum period. Catastrophic events such as placental abruption or vasa praevia are more likely to result in hypotension at birth with consequent renal hypoperfusion and risk of ATN development. Nevertheless, a hypoxic-ischaemic event such as uterine rupture or shoulder dystocia also puts the term infant at risk of AKI via hypoxia-mediated renal hypoperfusion. There may be additional risk factors for AKI such as family history, antenatal ultrasound scans, maternal illness, sepsis and drug usage such as Non-steroidal Anti-inflammatory Drugs. On examination, dysmorphic features should be identified and the infant's haemodynamic status should be assessed .
Serum urea, creatinine, electrolytes, calcium and phosphate may be measured at regular intervals (every 8–12 h). However, neonatal electrolytes may reflect maternal renal function for the first 24 h of life [1, 28]. Regular monitoring of the full blood count and arterial blood gases allows rapid diagnosis of anaemia and acid–base imbalance and facilitates prompt treatment, avoiding further hypoxic injury and anerobic metabolism. Following perinatal hypoxia-ischaemia, infants may be at risk of coagulopathy , so coagulation profiles should be monitored. Urinary creatinine, sodium and osmolality may be useful measurements in the setting of hyponatraemia secondary to syndrome of inappropriate anti-diuretic hormone (SIADH). Urinalysis and microscopy should be preformed to check for proteinuria and microscopic haematuria along with urinary casts. A renal and bladder ultrasound is essential to outline renal anatomy and exclude any congenital malformations. It is important to document blood flow to the kidney, to determine the presence of urine in the bladder and outrule bladder outlet obstruction [7, 28].
Importance of achieving appropriate balance between fluid intake and urinary output
There is much controversy surrounding the perscription of fluids to infants with perinatal hypoxic-ischaemic AKI. However, there is no doubt that close attention to fluid balance is vital with the aim of achieving an equilibrium between fluid intake and urinary output. Fluid input calculations should include drug volumes, blood products and continuous infusions concentrated where possible [16, 28]. Urinary output should be monitored closely, and bladder catheterization may be considered for all asphyxiated infants . Twice daily weighing is ideal but not always feasible if therapeutic hypothermia is ongoing [16, 28].
Risks of perscribing too little fluid
Fluid restriction (40–60 mL/kg/day) for term infants following perinatal hypoxia-ischaemia is commonly instituted . However, a recent Cochrane review of the practice of fluid restriction following perinatal asphyxia  found no randomized controlled trials examining this practice. There is little evidence basis for this practice, and extreme fluid restriction could lead to dehydration and hypotension, resulting in hypoperfusion of vital organs and decreased cerebral perfusion which potentially could worsen the neurological outcome. The rationale for fluid restriction of term infants post-perinatal hypoxia-ischaemia has been extrapolated from adult, paediatric and animal studies and is based on the premise that fluid restriction avoids fluid overload and exacerbation of cerebral oedema . This theory does not take into account the unique physiological attributes of neonates, that is, a much higher maintenance fluid requirement per kilogram bodyweight and the specific mechanisms of injury during and following perinatal asphyxia . Excessive fluid restriction may also contribute to prerenal failure and worsening oliguria. The degree of oliguria secondary to post-asphyxial AKI versus that secondary to iatrogenic fluid restriction has not been defined.
Malnutrition can also result from fluid restriction as the volume available for nutritional intake is limited. Indeed inability to meet nutritional requirements without increasing fluid overload remains an important indication to commence renal replacement therapy (RRT).
Risks of perscribing too much fluid
As discussed earlier, there is a theoretical risk of cerebral oedema if excessive fluid is administered to neonates with perinatal hypoxia-ischaemia . Hypervolaemia (defined as fluid overload >7%) has been identified as an independent risk factor for mortality in newborns with AKI (n = 154) from any cause [the majority had hypoxic-ischaemic injury (43.5%)], with a relative risk = 12.9 . The basis for the association between hypervolaemia and increased mortality in neonatal AKI is not explored in the above publication . Whether hypervolaemia potentially worsens pulmonary oedema, thereby increasing the requirement for mechanical ventilation and also risk of mortality, is open to further study.
In established neonatal AKI, many units advise giving fluids to a volume equal to insensible water losses (about 25–30 mL/kg/day) plus urinary losses (should be replaced mL for mL) [28, 29, 32] with adjustment according to clinical condition of the infant. Urinary losses could amount to 2 mL/kg/hr in a non-oliguric infant with AKI which gives 48 mL/kg/day, so in addition to insensible water losses, this equates to 75–80 mL/kg/day. This highlights the difficulty of managing fluid balance correctly in post-asphyxial neonatal AKI where guidelines are contradictory and there is potentially a two-fold difference between the two fluid volumes recommended. There are no published studies which have examined this controversial topic.
A fluid challenge of 0.9% sodium chloride 10 mL/kg could be considered to increase the circulating intravascular blood volume in infants with oliguric AKI. However, caution is advised when administering fluid boluses as this could exacerbate cerebral oedema, seen in post-asphyxial term infants secondary to loss of cerebrovascular autoregulation [30, 33].
Syndrome of inappropriate anti-diuretic hormone
Syndrome of inappropriate anti-diuretic hormone is common following perinatal asphyxia. The diagnosis is based on the presence of low urinary output, high urine osmolality (>100 mOsm/kg) which is always greater than serum osmolality, which is low (<280 mOsm/kg). Also present is a low serum sodium level (<135 mmol/L), high urinary sodium level (>40 mmol/L) and high urinary specific gravity ≥1.030 . Monitoring serum sodium levels, serum and urine osmolality and urinary output, and instituting fluid restriction is recommended, further complicating the management of asphyxia-related AKI [15, 30].
Management of fluid overload and oliguria
Intravenous furosemide in the dose 1 mg/kg bolus has been proposed for the treatment of oliguria and fluid overload . However, the evidence for this in neonates is lacking and specifically this treatment has never been studied in post-asphyxial AKI. Moghal et al. reviewed the role of furosemide in neonatal AKI and proposed that through a prostaglandin-mediated action, furosemide could potentially reverse vasomotor nephropathy. This may increase renal blood flow, especially after administration of a large bolus dose. Although data from controlled studies is lacking, many clinicians expect that furosemide may prevent the progression of AKI to established renal failure. There is no data to support this practice however, and indiscriminate use of furosemide in neonates is associated with not insignificant side effects, for example, electrolyte abnormalites, ototoxicity and nephrocalcinosis .
Despite the fact that oliguric AKI is associated with worse outcomes compared to non-oliguric AKI, studies have not shown that transforming oliguric to non-oliguric AKI improves outcomes . In an adult randomized, double-blind placebo-controlled trial (RCT), high-dose furosemide helped to maintain urinary output but made no impact on survival or renal recovery . Nevertheless, furosemide may have a specific role in converting oliguric to non-oliguric AKI thereby facilitating clinical management by allowing administration of fluids, electrolytes, nutrition and intravenous medications while limiting volume overload [10, 17, 28, 34].
Oliveros et al.  found that bumetanide therapy significantly increased urinary output with a concomitant transient increase in SCr within 24–48 h in preterm infants with oliguric acute renal failure. Bumetanide may show potential as a second-line agent to improve urinary output if furosemide is not effective. Bumetanide is the subject of several ongoing clinical trials, looking at the efficacy and safety of bumetanide as a treatment of neonatal seizures in neonatal encephalopathy. [The Pilot Study of Bumetanide for Newborn Seizures (NCT00830531), Dr. Janet Soul and the NEMO Trial (NCT01434225), Dr. Ronit Pressler]. Bumetanide is a potent diuretic, whose role in AKI in term infants is not well described and we await the trial results.
Low dose dopamine has been utilized in the past to improve urinary output in critically ill term and preterm infants. However, a systematic review of the role of low dose dopamine in Neonatal Intensive Care Units & Paediatric Intensive Care Units found no evidence that it improves renal function or urinary output in preterm infants, critically ill neonates or children . Asphyxiated term neonates were not included specifically in any of the studies included in the systematic review. There appears to be insufficient evidence at present to recommend the use of dopamine to improve renal function, in the absence of hypotension, in critically ill neonates and more specifically following perinatal hypoxia-ischaemia. More recently, a double-blind, randomized controlled crossover study of adult patients with AKI found that dopamine worsened renal perfusion with significant increases in median renal resistive index and pulsatility index and did not improve urinary output .
On Day 1, 10% dextrose solutions are recommended ; however, with the use in many centers of fluid restriction to 40 mL/kg/day for asphyxiated infants [30, 39], higher dextrose concentrations may be required. Hypoglycaemia should be avoided as this could worsen cerebral injury . Early hypoglycaemia within 0–6 h of birth has been associated with adverse neurodevelopmental outcome at 2 years following perinatal hypoxia-ischaemia (OR = 5.8) . In contrast, the safety of hyperglycaemia in the human newborn has to date, not been proven . Blood glucose levels should be carefully monitored and both hypo- and hyperglycaemia should be avoided.
Following an asphyxial insult, infants with AKI are in a catabolic state which in turn potentiates hyperkalaemia and acidosis . Extrapolation from paediatric studies  would suggest that it is important to ensure adequate caloric intake during the AKI phase to prevent catabolism, encourage anabolism, reverse protein–energy wasting and promote renal recovery following AKI [19, 28, 32]. In reality it is often not possible to achieve this goal without instituting RRT such are the limitations imposed by fluid overload on prescribed fluid intake. The benefits and risks of RRT need to be carefully balanced against the risk of malnutrition.
If indicated total parenteral nutrition (TPN) may be prescribed. Fluid restriction and biochemical disturbances need to be taken into account if prescribing TPN. Infants receiving RRT require, in addition to the recommended daily allowance for a sick infant, at least 1 mg/kg/day extra protein to account for the losses from RRT . Enteral feeding may be contraindicated in the initial phase if there is a concern about intestinal ischaemia or aspiration while the risk of necrotizing enterocolitis is also increased following perinatal asphyxia .
Acid–base homeostasis maintains optimal enzyme and cellular function. Arterial blood gases every 2–6 h are often required in the first 24–48 h. Respiratory acidosis can be manipulated by ventilation [28, 39]. A Cochrane review in 2005  concluded that there was insufficient evidence to determine whether infusion of sodium bicarbonate or fluid bolus reduced morbidity and mortality rates in preterm infants with metabolic acidosis. Berg et al.  examined the use of sodium bicarbonate in preterm infants within the first week of life and showed that sodium bicarbonate did not improve blood pH and may in fact be associated with increased intraventricular haemorrhage severity and mortality rates . Neither of these studies included asphyxiated infants, and recently there has been controversy over the efficacy and safety profile of using sodium bicarbonate in this group .
There are no published trials which have compared the use of sodium bicarbonate infusions versus placebo during the management of asphyxiated infants with regard to the outcomes of improvement in acid–base status, morbidity or mortality. However, Lokesh et al.  compared the use of sodium bicarbonate to placebo (5% dextrose) in the resuscitation of asphyxiated infants (n = 55) at birth with regard to the outcomes of survival and neurological status at discharge. The incidence of encephalopathy, cerebral oedema, need for inotropic support, intraventricular haemorrhage and the mean arterial pH at 6 h was similar between the two groups. The administration of sodium bicarbonate during neonatal resuscitation did not improve survival or immediate neurological outcome. This trial did not examine the effect of sodium bicarbonate on acid–base status or outcome of asphyxiated infants beyond delivery room resuscitation, and long-term outcomes are required. The results of these studies should nevertheless prompt careful consideration before routinely administering sodium bicarbonate infusions to term infants with AKI and acid–base derangements following an asphyxial insult.
Following an asphyxial insult, infants are at significant risk of developing hyponatraemia [3, 46] secondary to oliguric AKI, consequent water retention and decreased proximal tubular sodium reabsorption [3, 6]. SIADH is a risk following brain injury and can also contribute to hyponatraemia. Irrespective of the aetiology, hyponatraemia can be corrected with careful fluid management, usually fluid restriction. However, if serum sodium is <120 mmol/L or associated with symptoms such as seizures then prompt intravenous correction is required. Total sodium deficit is calculated with the formula: Sodium value desired (130–135 mmol/L) – Infant's sodium value (mmol/L) × Weight (kg) × 0.7 . In order not to correct the deficit too rapidly only 50% of this amount is given over 12–24 h and serum sodium levels need to be reassessed frequently . A maximum increase of 0.5 – 1 mmol/L per h in serum sodium is recommended . Infants with non-oliguric AKI often have very large urinary sodium losses of up to 10 mmol/kg/day requiring replacement [16, 32]. Hypernatraemia is rarely reported following perinatal hypoxia-ischaemia which is interesting given that the majority of these infants are kept fluid restricted for several days .
Hyperkalaemia may occur in the severely asphyxiated infant . Any fluids that contain potassium should be stopped and electrocardiographic monitoring may be required. Insulin/dextrose infusion and salbutamol infusion may be required to lower potassium levels [16, 28, 32, 47]. If electrocardiographic changes are present, calcium gluconate (10%) should be administered to decrease myocardial excitability. If enterally fed, low potassium formulas and breast milk are better options compared to standard formulas due to their lower potassium content. Oral and rectal ion exchange resins are not indicated for use in newborn infants due to the risk of bowel perforation . In cases of severe, intractable hyperkalaemia, dialysis may be considered. Hypokalaemia may require potassium supplementation once the renal function is adequate.
Hypocalcaemia can complicate perinatal hypoxia-ischaemia [46, 48]. Hypocalcaemia is defined as total serum calcium of <1.75 mmol/L or <1.1 mmol/L ionised calcium in term neonates . Basu et al.  found a strong positive correlation between Apgar scores of 50 asphyxiated infants and serum calcium levels. Serum albumin levels should be checked and if hypoalbuminaemia is present, the calcium level should be corrected for the albumin level using the formula: Corrected calcium (mmol/L) = Total calcium (mmol/L) + 0.02 [40(g/L) – Patient's albumin (g/L)] . Calcium supplementation may be required – intravenously, that is, calcium gluconate (10%) if symptomatic, orally if asymptomatic [16, 30, 32]. Hypo- and hypermagnesaemia can also occur  and may require either intravenous or oral supplementation to maintain serum levels >0.75 mmol/L .
Following perinatal hypoxia-ischaemia, infants should be evaluated for infection with a full septic screen and placental pathology should be requested. Antimicrobial cover should be strongly considered especially if no definite sentinel event can be identified in the antepartum history. There is mounting evidence that lipopolysaccharide exacerbates hypoxic-ischaemic neonatal brain injury  and asphyxiated infants who receive therapeutic hypothermia may have compromised host defences secondary to neutrophil dysfunction . Additionally, the risk of sepsis is thought to be increased in neonates with AKI [9, 16]. Empiric antibiotics provide antimicrobial cover during this critical period.
Careful attention to nephrotoxic medications such as aminoglycosides is warranted, and plasma levels should be monitored to ensure that the levels remain in the therapeutic range and do not exacerbate the AKI [16, 28, 39]. It is recommended that in the presence of oliguria or anuria, aminoglycosides should not be given. If given, the interval of drug dosage needs to be prolonged. Confirmation of adequate clearance by drug monitoring after a dose is necessary before a second dose is given .
Dialysis is rarely employed for the treatment of AKI secondary to perinatal hypoxia-ischaemia. Amongst clinicians, there is a reluctance to initiate RRT in a severely asphyxiated infant in view of the ethical issues posed particularly if there is thought to be a high likelihood of significant long-term neurological deficit . However, the indications remain the same as for non-asphyxia-related neonatal AKI and include the following: signs of uraemia, severe hyponatraemia, hyperkalaemia, refractory acidosis, hypertension, congestive heart failure, volume overload and the requirement for additional fluid space to allow for adequate nutrition and/or medication administration [10, 13]. Peritoneal dialysis is currently the preferred method for neonates in view of the difficulties of vascular access in small infants. With recent advances in the technology of continuous haemofiltration and haemodialysis, the feasibility of employing these techniques in the treatment of AKI during the neonatal period becomes ever more possible. In addition, the availability of local expertise, resources and physician preference can significantly influence the decision about which mode of dialysis to employ [10, 13, 28]. There is mounting data from adult and paediatric studies suggesting that early initiation of RRT is associated with improved outcome [12, 19]; however, the optimum timing for initiation of dialysis is unknown, specifically as it relates to the amount of fluid overload accumulated. There are no studies in newborns comparing the outcomes following different modes of dialysis in the treatment of neonatal AKI [12, 13]. The need for dialysis is recognized as a poor prognostic factor [13, 27] with infants who require peritoneal dialysis for anuric/oliguric AKI having a much higher mortality rate than infants with AKI who have adequate urine output.
Interaction between Therapeutic Hypothermia and AKI
The meta-analyses and systematic reviews of all the trials of therapeutic hypothermia for neonatal encephalopathy found no statistically significant difference in the rates of oliguria  or renal failure  in infants who underwent therapeutic hypothermia compared to controls. A much smaller study by Róka et al.  found that renal failure (defined by rate of diuresis and SCr levels) affected significantly less cooled infants (n = 3/12) than normothermic infants (n = 7/9). However, only a small number of infants were included in the study (n = 21), and therefore, the findings may not be generalizable until challenged in a larger study. Still the diagnosis of AKI in these trials was based on urinary output and rising SCr (not all the infants in the trials even had SCr measured) and therefore may underestimate AKI. The development of more sensitive and specific AKI biomarkers should allow a more focused evaluation of the interaction between therapeutic hypothermia and AKI.
Prophylactic theophylline, given early after birth (within 1 h), has beneficial effects on reducing renal dysfunction in asphyxiated full-term infants according to evidence from four RCTs . Further studies are required to fully elucidate the safety and efficacy of theophylline in the management of this cohort of infants before its use can be recommended.
Although not specific to perinatal hypoxic-ischaemic-associated AKI, other treatments, that is, fenoldopam and rasburicase, have been studied in neonates at risk of AKI. Fenoldopam is a synthetic benzazepine derivative which acts as a selective D1 receptor partial agonist. High-dose fenoldopam given to infants undergoing CPB significantly reduces levels of urinary NGAL and cystatin C when compared to placebo . Rasburicase is a recombinant version of urate oxidase which catalyses the conversion of uric acid to allantoin, a much more soluble metabolite of purine and therefore easily removed by the kidneys. Rasburicase was found to significantly decrease uric acid levels and improve urinary output in a small cohort of seven infants with AKI defined as SCr > 132.6 μmol/L . Table 1.
|History & clinical exam|| |
Risk factors for AKI 
Dysmorphic features 
Blood pressure, heart rate, peripheral pulses & perfusion 
Serum creatinine, urea and electrolytes every 6–8 h [16, 28]
Full blood count, regular arterial blood gases 
Urinary creatinine, sodium and osmolality & serum osmolality, urinalysis & microscopy [16, 32]
Renal & bladder ultrasound [7, 28, 32]
|Fluids & nutrition|| |
10% dextrose intravenous infusion on Day 1 but may need higher concentration dextrose infusion [28, 32]
Avoid hypoglycaemia 
Consider TPN if enteral feeding contraindicated [19, 29]
Additional electrolytes Day 2 depending on electrolyte status [16, 28, 32]
Precise fluid input & output documentation & twice daily weights if feasible [16, 28, 32]
Evidence for fluid restriction of infants post-perinatal hypoxia-ischaemia lacking – aim to avoid hypovolaemia or fluid overload [16, 28, 31, 32]
|Oliguria & fluid overload|| |
Consider fluid challenge 10 mls/kg 0.9% sodium chloride [16, 28]
Consider trial of intravenous bolus 1 mg/kg furosemide [16, 34]
No evidence for use of low dose dopamine to improve renal function or urinary output [37, 38]
Bumetanide , also in clinical trials currently as a treatment for seizures in neonatal encephalopathy
Regular arterial blood gases every 4–6 h & manipulate ventilation 
Avoid hypocarbia as associated with worse neurological outcome long-term 
Consider addition of acetate to TPN 
No evidence sodium bicarbonate improves metabolic acidosis in asphyxiated infants [42-45]
|Electrolyte imbalance|| |
Hyponatraemia generally resolves with careful fluid management but if <120 mEq/L may require treatment [16, 28, 32]
Hyperkalaemia if severe may require insulin dextrose infusion, salbutamol infusion & calcium gluconate 10% infusion [16, 28, 32, 47]
Hypokalaemia can occur once renal function recovers, may need replacement[16, 28]
Hypocalcaemia, hypomagnesaemia – treatment if symptomatic[16, 28, 32]
|Antimicrobial cover|| |
Broad spectrum antibiotics e.g. benzyl penicillin +/- gentamycin [16, 49, 50]
Careful monitoring of plasma levels of aminoglycosides with dose adjustment according to drug levels & eCCL [19, 35, 40]
|Prophylactic theophylline|| |
Reduces renal dysfunction in asphyxiated term infants if given early
Requires further validation in a larger study, not routinely recommended at present [32, 54]
|Fenoldopam||At high dose reduces levels of AKI biomarkers, NGAL and Cystatin C post-CPB in infants |
|Rasburicase||Improves urine output & reduces uric acid levels in infants with AKI |
|Dialysis||Rarely used for AKI secondary to perinatal hypoxia-ischaemia [19, 35, 40]|
Following neonatal AKI, hyperfiltration in the surviving nephrons is thought to occur. Each individual remaining nephron increases filtration which eventually leads to glomerular changes, progressive nephron damage, vascular dropout and ultimately results in decreasing renal function and the evolution of chronic kidney disease (CKD) [32, 57]. Animal studies have shown that renal ischaemic injury causes permanent damage to peritubular capillaries contributing to a urinary concentrating defect and towards the development of renal fibrosis and abnormalities in long-term function . Recent follow-up studies of paediatric AKI quote rates of CKD in the range of 10–59%, depending on the definition of CKD used [58, 59]. Consistent long-term follow-up studies of post-asphyxial neonatal AKI are scarce and give contradictory reports of the incidence of CKD following AKI in the neonatal period. Gupta et al. remeasured BUN and SCr levels in the survivors of a cohort of 70 asphyxiated infants (n = 65) between 1 and 6 months of age and found that all biochemical parameters had normalized. Some older studies suggest an incidence of CKD in up to 40% of survivors following neonatal AKI .
It is reasonable to follow-up infants who have sustained AKI in the neonatal period with regular lifelong measurement of growth and nutrition, SCr, blood pressure and urinalysis for proteinuria and albumin/creatinine ratio [13, 28, 32]. Typically, the late development of chronic kidney disease will first become evident with the development of hypertension, proteinuria and eventually elevation of blood urea nitrogen and creatinine . At present, the long-term consequences of such an acute renal insult in the neonatal period is not known. Detailed follow-up studies are required to evaluate the impact of perinatal hypoxic-ischaemic AKI on later childhood and adult renal function.
Acute kidney injury is common in the setting of perinatal hypoxia-ischaemia. Clinicians are still forced to rely on rising SCr, falling GFR and oliguria to establish the diagnosis of renal injury because at present, there is no alternative definition of AKI. These markers are flawed, may miss a significant proportion of neonates with AKI and do not facilitate the rapid diagnosis of AKI to allow therapeutic interventions. Markers of AKI more specific to ischaemic pathophysiology and with better sensitivity, specificity and predictive values are at various phases of study, namely cystatin C, NGAL, L-FABP, IL-18 and KIM-1. At present, management of neonatal AKI post-perinatal hypoxia-ischaemia is supportive with little evidence base for some of the treatments utilized, for example, fluid restriction and sodium bicarbonate infusion. Adequate caloric intake is vital during the course of AKI as limiting protein–energy wasting is essential for recovery. The relationship between neonatal AKI and therapeutic hypothermia warrants further study, and AKI biomarkers may facilitate this. A large prospective, multicentre, cohort study with internationally accepted definitions of AKI incorporating serum and urinary AKI biomarkers and reporting long-term follow-up data would increase understanding of the incidence and impact of AKI on asphyxiated infants.
Conflicts of Interest
National Children's Research Centre, Crumlin, Dublin 12 and Children's University Hospital, Temple Street, Dublin 1, Ireland.