Acute kidney injury (AKI) is a common complication, affecting almost one-third of critically sick children and also noncritically ill children admitted to wards [1,2]. In the last decade, there has been a better understanding of outcomes in the field of pediatric AKI, which include higher morbidity, increased length of stay, duration of ventilation, and mortality [3,4]. There are newer studies on pediatric AKI epidemiology, clearly delineated definitions, newer biomarkers, and new criteria for risk stratification of children admitted in emergency situations. Additionally, definitions and the understanding of neonatal AKI have undergone a drastic change due to recent studies [5]. There is now novel research on machines made especially for smaller children with smaller extracorporeal volume [6–8]. This review includes the major advances in the field of pediatric AKI in the last decade that have made a significant impact on learning and practice in this field (Fig. 1). This is a particularly important area of nephrology, where the clinical and translational advances have been performed first in pediatrics, much before the adult nephrology field.

There is increasing evidence that the incidence and awareness of pediatric AKI is rising. In infants and children undergoing cardiac surgery, the incidence varies from 30% to 50% [9–12]. Additionally, it is common in pediatric intensive care units (ICUs) and has an incidence of 10% to 35% [13–15]. The rate is higher in children who are ventilated and are on inotropes [16]. AKI is also common in wards, especially in children receiving aminoglycosides and multiple nephrotoxins during their hospital stay [17,18].

The first prospective study on pediatric AKI, the Assessment of Worldwide Acute Kidney Injury, Renal Angina and Epidemiology (AWARE ) study, was done over a 3-month observational period and included 4,683 children [19]. This study showed that AKI was seen in 26.9% of children, and severe AKI was seen in 11.6% of children within 7 days of ICU admission. This increase in AKI severity was associated with a stepwise increase in mortality. Additionally, cardiovascular and respiratory disorders had a higher association with severe AKI.

Among the neonate subgroup, the largest retrospective study in the neonatal population, known as Assessment of Worldwide Acute Kidney Injury Epidemiology in Neonates (AWAKEN), was performed in 2017. This study included more than 2,000 newborns in four different countries admitted to the neonatal ICU before 14 days of life, who received intravenous fluids for at least 48 hours. AKI was seen in 30% of all newborns and was differently stratified per gestational age, with a higher incidence in extreme preterm birth infants. AKI was also associated with mortality and increased length of stay after adjusting for confounding variables [5].

Additionally, the etiology of AKI varies based on the geographic setting. In the developed world, the setting of AKI has shifted from primary glomerular disorders to hospital-acquired AKI, with common causes being nephrotoxins, critically ill status, postsurgical, posttransplantation, and malignancy [20,21]. In the developing world, especially in rural regions, the etiological factors remain as dehydration, sepsis, and hemolytic uremic syndrome [22].

The ability of serum creatinine (SCr) to accurately estimate kidney function in a sick child has been problematic. This has resulted in the use of more than 35 definitions of AKI in clinical studies, ranging from changes in SCr to dialysis requirement. Earlier studies employed nonstandard AKI definitions without any grading (defining AKI as the doubling of SCr), thereby excluding early-stage AKI. Since there was no consensus in definitions, comparisons among studies were difficult, resulting in a wide range of quoted epidemiology, morbidity, and mortality rates within the pediatric AKI literature [23].

The Kidney Disease: Improving Global Outcomes (KDIGO) definition and staging system is the most recent and preferred definition even in pediatric AKI literature [24]. Other classification systems include pRIFLE (pediatric Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease) and a subsequent modification proposed by the Acute Kidney Injury Network (AKIN) [25,26] (Table 1). Each definition confers its own set of advantages and disadvantages. For example, pRIFLE can diagnose a greater number of mild AKI cases that are usually missed by the other two systems but requires patient height and baseline SCr value, which might not be readily available. In pRIFLE, the estimated creatinine clearance (CCl) is based on the original Schwartz formula to quantitate the change in glomerular filtration rate (GFR) rather than absolute changes in SCr used in the adult RIFLE criteria. Furthermore, the pRIFLE classification has outperformed the AKIN, KDIGO, and conventional grading criteria in predicting AKI in several pediatric patient populations. Zapitelli et al. [27] found that AKI prevalence increased when changes in estimated GFR (eGFR) (pRIFLE) were accounted for rather than changes in SCr (AKIN) in pediatric inpatients. Additionally, Sutherland et al. [23] recently demonstrated notable differences in incidences and substantial disparities in staging resulting from the use of these three definitions on the same cohort of hospitalized children. The AKIN definition appears more specific and does not require height and baseline SCr values; however, it has the most restrictive diagnostic timeframe. The AKIN system, which defines AKI as a ≥0.3-mg/dL increase in SCr within a restrictive 48-hour period, eliminates the need to estimate CCl. However, the AKIN criteria have not been adequately validated for use in children and the restricted diagnostic timeframe of 48 hours for a rise in SCr may limit its utility. Although pRIFLE shows a greater sensitivity in detecting AKI than AKIN and KDIGO, all three definitions correlate highly with outcomes (mortality, length of stay in ICU) and enable outstanding inter-stage differentiation [23]. The KDIGO classification is the only one that applies to both children and adults and has a less restrictive diagnostic timeframe than AKIN. Although it does require patient height data (eGFR calculation) for a complete assessment, it is still the preferred definition. The KDIGO AKI criteria have been validated in hospitalized children with both critical and noncritical illness [23,28].

SCr is an indirect and unreliable marker of GFR that can be confounded by renal tubular secretion and numerous other factors such as fluid balance, muscle mass, and medications. Moreover, changes in SCr lag behind the changes in GFR, which can even take several days. In acute settings, it is estimated that SCr rises after a >50% decline in GFR [29]. Given the inherent shortcomings of the SCr method, alternative functional and damage biomarkers of AKI have been evaluated.

The use of serum cystatin C in children as an endogenous marker is well established. Cystatin C is a cysteine protease inhibitor protein, which is produced by all nucleated cells of the body at a constant rate, is freely filtered by the glomerulus, and is catabolized by the proximal tubule. The plasma cells are not affected by sex, age, diet, or muscle mass, and are identical in adults and children over 12 months of age [30]. It outperforms SCr in children for estimation of GFR. Moreover, there is now pediatric data to show that it is an early predictor biomarker of AKI [31].

Genomic and proteomic technologies have revealed novel biomarkers that appear in urine or plasma well before changes in SCr are detected [29]. The most widely studied and validated early biomarker in children is neutrophil gelatinase-associated lipocalin (NGAL) [32]. Most studies of NGAL have been done in children post cardiac surgery, showing that the urine and plasma levels are significantly high in children within 2 hours of cardiac bypass surgery in patients who subsequently develop AKI [33]. Another study done in infants and children undergoing cardiopulmonary bypass established cut-off thresholds and showed strong associations between early NGAL levels and length of hospital stay, duration, and severity of AKI [11,34]. A recent study done on 220 children undergoing cardiac surgery showed that urine NGAL increased in patients within 2 hours, while urine interleukin-18 (IL-18) and urine liver-type-fatty acid binding protein levels were elevated within 6 hours, and urine kidney injury molecule-1 increased at 12 hours. All markers correlated well with severity, clinical outcomes, and, additionally, improved risk prediction [35]. Thus, a panel of biomarkers may help to establish injury timely and plan appropriate timely interventions. Standardized clinical platforms for measurement of plasma and urine NGAL are now available globally.

Recent clinical data in children and adults support the utility and superiority of a new AKI biomarker test, NephroCheck (bioMérieux, Marcy-l’Étoile, France), which detects urinary tissue inhibitor of metalloproteinase 2 (TIMP2) and insulin-like growth factor binding protein 7 (IGFBP7) concentrations and uses their arithmetic product [36,37]. AKI-induced urinary TIMP2/IGFBP7 elevations are not due to stress-induced gene transcription. Rather, increased filtration, decreased tubule reabsorption, and proximal tubule cell TIMP2/IGFBP7 urinary leakage seem to be the most likely mechanisms [38].

There is recent interest in patients who are ‘biomarker positive; creatinine negative’ which means their urinary or serum early biomarkers are high while SCr is normal. Two recent studies enrolled more than 4,000 cardiac surgical, critically ill and emergency patients [39,40]. Both studies showed almost >20% of patients had only elevated NGAL in urine. These ‘subclinical AKI’ patients in fact had two- to three-fold higher risk of death and need for renal replacement therapy (RRT). Even in patients with very high creatinine, markedly high tubular markers in urine had a worse prognosis.

Hence, NGAL and a panel of urinary or serum biomarkers may help clinicians make an early diagnosis of AKI and plan supportive care early. Moreover, structural biomarkers may further help in reliably classifying AKI in a mechanistic manner. The various functional and structural markers are illustrated in Table 2.

Renal Angina Index (RAI) combines objective parameters of kidney dysfunction (change in SCr and percent change in fluid overload [%FO]) and patient characteristics (AKI risk factors) to ascertain renal angina and has been successfully validated as a functional risk stratification tool in critically ill patients with AKI. An RAI of ≥8 within the first 12 hours of ICU admission has shown to entail very high sensitivity and negative predictive value for AKI development or persistence at 72 hours of ICU admission in children [41,42]. RAI is a risk discrimination model that enhances the pretest probability of AKI. It renders context to biomarker measurement and significantly optimizes their predictive performance, akin to the cardiac angina-troponin relationship. RAI has been shown to correlate with an increased need for RRT, prolonged mechanical ventilation, higher oxygenation index, and a higher risk of mortality when compared to children with a negative index score [43,44]. RAI entails moderate discrimination for predicting severe AKI prediction, but it improves after incorporation of biomarkers [45].

The major challenges confronted by clinicians involved in the care of neonates with AKI stem from numerous factors; unique renal physiology in term and preterm neonates, lack of a standardized AKI definition, and weight- and gestational age-dependent baseline SCr value in the neonates. Moreover, neonates usually have nonoliguric renal failure, making oliguria an insensitive marker of AKI in this cohort [57]. Neonatal AKI is further confounded by the reflection of maternal SCr levels in neonates for the first 3 days postbirth and the variable decline over days to weeks depending on gestational age [58].

The formation of the Neonatal Kidney Collaborative (NKC) was a giant leap forward which accomplished the heretofore unmet need of neonatal AKI quantification at a global level. The AWAKEN study retrospectively evaluated 2,022 neonates from 24 centers across the globe, which formed the NKC. The group concluded that neonatal AKI is common, with an incidence of 29.9%, and is an independent risk factor for mortality and prolonged hospital stay, independent of demographics, severity of illness, and existing comorbidities. The incidence of AKI was 43% in patients <29 weeks gestation, 18% in those between 29 and 36 weeks gestation, and 37% in those >36 weeks gestation [5].

Since maternal SCr is transmitted across the placental barrier and its clearance is dependent upon the infant’s gestational age, the KDIGO definition was modified in such a manner that baseline SCr was assumed to be the lowest SCr level noted in each infant. Also, the SCr threshold for stage-3 AKI was reduced to 2.5 mg/dL rather than usual KDIGO threshold of 4 mg/dL [59].

A recent secondary analysis from AWAKEN also showed that caffeine administration in preterm neonates is associated with reduced incidence and severity of AKI. Further studies should focus on the timing and dosage of caffeine to optimize the prevention of AKI [60]. Other ancillary studies from the same group include a report on the association of AKI and hypertension [61], a study showing the association between AKI and mortality in those with severe neonatal encephalopathy [62], the association of AKI and intraventricular hemorrhage [63], and the association of AKI and chronic lung disease in premature and near term/term infants [64,65].

RRT modalities for pediatric AKI have expanded from peritoneal dialysis (PD), HD to CRRT and sustained low-efficiency dialysis (SLED). Advancements in use of RRT in children have led to a higher standard of care for young and critically ill patients [86]. Since no difference in survival outcomes has been seen with any dialysis method, the optimal RRT modality to be chosen for children with AKI is based on the patient’s size, overall clinical status, on the performance of the dialytic modality, and the availability of resources and expertise [87].

PD is the most common and simple method of providing solute and water removal in the ICU. It is easy to perform, can be easily learned, and does not require vascular access or anticoagulation. In a recent worldwide survey by Raina et al., 68.5% of respondents in developing countries preferred PD for treating infant AKI while only 29.1% of physicians in developing countries and 22.2% in developed countries favored PD to treat AKI [88]. Additionally, certain modifications to PD have been made recently to improve ultrafiltration, namely continuous equilibration PD, high volume PD, tidal PD, and continuous flow PD [89].

HD is the most efficient method of dialysis with rapid solute and fluid removal. It is ideal for managing pulmonary edema, hyperkalemia, intoxications, hyperammonemia, and acute tumor lysis syndrome [90]. However, patient hemodynamic stability is a must for a child to be put on HD. It does require a vascular access, careful evaluation of the extracorporeal blood volume (in the circuit and the dialyzer), and the need for anticoagulation [90]. Recently, the Pediatric Continuous Renal Replacement Therapy Foundation (PCRRT) gave recommendations on how to avoid intradialytic hypotension in children [91].

CRRT is the preferred modality for the management of AKI and fluid overload in critically ill children. It can be used with both or one of the diffusion or convection strategies. It is a complex dialysis modality that requires expertise and systemic heparin or regional citrate anticoagulation. The Prospective Pediatric CRRT Registry Group has published guidelines for dialyzing children with sepsis and multiorgan dysfunction in the last decade [51,92,93].

SLED is an alternative to CRRT in hemodynamically unstable pediatric patients with AKI. It utilizes conventional dialysis machines with low blood pump and dialysate flow rates for ≥6 hours daily. Recently, Sethi et al. [94,95] published a retrospective and prospective experience of SLED in unstable pediatric patients utilizing heparin-free dialysis and prefilter convective replacement fluid.

Management of AKI is challenging in critical infants and children. Over the past decade, revolutionary landmark studies and machineries have evolved, greatly improving the diagnosis, early detection, and management of renal support in this population. The pediatric nephrology community is working together closely to provide more scientific data to improve renal support in smaller critically sick children.