FormalPara Case Vignette

An infant is born to a 21-year-old mother at 27 weeks post conception with a birthweight of 900 g. His serum creatinine concentration increases over the coming days from 0.8 to 1.0 mg/dL and then improves over the coming weeks to 0.3 mg/dL at 3 weeks of age. The infant develops apnea and bradycardia, sepsis and has necrotizing enterocolitis by abdominal X-ray. His enteral feeding is discontinued and interventions are started including multiple blood product transfusions, intravenous vasoactive medication infusions, and invasive mechanical ventilator support. Over the following week, he develops progressive abdominal distension, hypoalbuminemia, decreased urine output, and edema. His serum creatinine concentration has increased to 1.3 mg/dL and he now weighs 1350 g.

This chapter provides insights to understand the physiology and management of neonates at risk for acute kidney injury (AKI) during the neonatal intensive care units (NICU) course. We will review renal physiology and evaluation of the glomerular filtration rate in the neonate. We will also provide a review of the assessment, diagnosis, risk factors, and outcomes of neonatal AKI. Medical management and strategies for renal support, including a review of novel machines designed for neonates, will be discussed. Finally, outcome data on the long-term consequences of kidney disease in NICU graduates are presented.

Extraordinary advancements in neonatal care have markedly reduced the mortality rates of infants hospitalized in NICU. Over the last decade, studies show that neonatal AKI is common and those with AKI have higher mortality and prolonged length of stay. Premature infants are born with low nephron numbers which predisposes them to AKI and chronic kidney disease (CKD). Despite recent insights that substantiate the impact of poor kidney health on outcomes in sick neonates, significant critical gaps in our understanding of the antenatal and postnatal factors exist. The global burden of AKI and CKD in NICU graduates need to be better understood. Fortunately, progress is being made as investigators are performing large observational studies, and randomized clinical trials that evaluate risk factors, outcomes, and interventions. In addition, novel devices designed specifically to provide renal support for neonates are currently in use in a few centers around the world.

1 Neonatal Renal Physiology

Nephrogenesis starts in the 5th and ends in the 34th–36th gestational week, although kidney maturation continues in the postnatal period (until the 40th postnatal day) [1].

Renal Blood Flow (RBF) in neonates during the first week of life is only 10% of the cardiac output (2.5–4% on birth), and it reaches the adult rate (25%) by 2 years of age. This increase of RBF occurs due to a combination of an increase of renal perfusion pressure, increase of systemic vascular resistance, and a reduction of renal vascular resistance via angiotensin II (AT II), prostaglandin, and other physiologic changes [2].

Glomerular filtration rate (GFR) is a measure of kidney function. In a healthy term newborn, the GFR at birth is 10–20 mL/min/1.73 m2, rises to 30–40 mL/min/1.73 m2 in the second week, and reaches the adult clearance of 100–120 mL/min/1.73 m2 by the age of 2 years. In the first weeks of life, GFR is even lower in preterm neonates compared with term neonates as a result of renal immaturity, difference in renal blood flow, and distinct vascular resistance. Rigorous studies to estimate serum creatinine (SCr) clearance using SCr in neonates have not been performed recently. Older studies suggest that using SCr measured by Jaffe enzymatic reaction, that eGFR could be estimated using the following equation [3, 4].

Estimated creatinine clearance (mL/min/1.73 m2) = k * BH/sCr

(k—constant—0.33; BH—height in cm; sCr—serum creatinine in mg/dL (mg/dL × 88.4 = μmol/L)).

Studies in children suggest that when using the enzymatic SCr to estimate GFR, the coefficient is about 10% lower than when using the Jaffe reaction. More data are needed, but it is possible that the correct coefficient when using the enzymatic reaction should be 10% lower than above, or 0.3 [5].

The maximal urine concentration capacity in the term neonate is 500–700 mOsm. The adult level (1400 mOsm) is reached between the 6th and 12th month of life. This relatively lower urinary concentration ability increases the risks of fluid loss, decreased reabsorption of substances and electrolyte imbalance.

In newborns, water makes 80% of the body weight (BW) of the newborn. Soon after birth, there is a redistribution of body fluids, with an early postnatal BW loss of up to 5–10% in healthy term infants. Very low birthweight (VLBW) infants need intravenous support of fluids to prevent dehydration. As VLBW infants may have very high insensible losses due to very thin skin, their fluid loss may be significantly higher than healthy term neonates. It is not uncommon for VLBW infant to lose up to 15% in VLBW neonates due to isotonic contraction of extracellular water through disposal of excess sodium and water through kidneys. Studies to optimize fluid delivery in this vulnerable cohort are greatly needed.

The most significant site of sodium exchange is distal tubule. If the term neonate can successfully feed from the breast or formula, he/she can maintain the positive sodium balance. However, it is not uncommon for a newborn less than 35 weeks gestation to be at risk for negative sodium balance in the first 3 months of life which could lead to hyponatremia. This is due to increased delivery of sodium and its reduced absorption in the distal tubule. This is important to consider when testing for prerenal azotemia. In term newborns, the fractional excretion of sodium (FENa) is highest during the first 10 days of life and can decrease to below 0.4% by 1 month of age, an ultimate threshold that is similar to adults. Due to inability to reabsorb sodium avidly during prerenal azotemia states, VLBW infants will have much higher FENa than term counterparts. FENa can also be increased in hypoxia, respiratory distress, hyperbilirubinemia, and in neonates taking diuretics and increased intake of fluids and salts [3, 6, 7].

The renin-angiotensin-aldosterone system (RAAS) is responsible for regulation of blood pressure, renal hemodynamics, and maintenance of fluid and electrolytes balance. Plasma renin activity is increased after birth and stays increased during infancy until it is reduced to adult levels by the age of 6–9 years. Angiotensin stimulates the secretion of aldosterone which regulates the reabsorption of sodium in kidneys. The fetal response to secrete aldosterone is less than that seen in adults [8].

On the other hand, prostaglandins represent the most important counter-regulatory molecules in neonatal period and they lead to the dilation of afferent arteriole. Atrial natriuretic peptide (ANP) is a vasodilator and is significantly increased in the first days of life and then reduced by the 2nd week. It also reduces extracellular volume [3].

2 Evaluation of Neonatal Acute Kidney Injury (AKI)

Acute kidney injury (AKI) is a complex pathology characterized by a sudden reduction of kidney function caused by a heterogeneous group of underlying causes. Clinically, AKI is manifested by minimal kidney damage up to the complete kidney failure that requires renal replacement therapy.

2.1 History and Physical Examination

Previous reports suggest that neonatal AKI is due to: inadequate renal blood flow (previously referred to as prerenal—85%), intrarenal pathology—11%, and then obstruction of the urinary tract—3%. In newborns with AKI, it is very important to examine all the possible causes that could have led to AKI. The most causes of inadequate renal perfusion are: hypovolemia, hypotension, hypoxemia, heart failure, dehydration, septicemia, hypoalbuminemia, perinatal asphyxia, respiratory distress syndrome, congenital heart disease, cardiac surgery, polycythemia, and nephrotoxic drugs (e.g., indomethacin, captopril, vasodilators). The most common causes of intrarenal AKI are: acute tubular necrosis, corticomedullary necrosis, renal venous and arterial thrombosis, acute pyelonephritis, disseminated vascular coagulation, isoimmune hemolytic disease, congenital renal anomalies, systemic infections, intrauterine infections, and nephrotoxic drug exposure (aminoglycosides, radiocontrast agents). The most common causes of postrenal AKI are: posterior urethral valves, bilateral obstructive uropathy, neurogenic bladder, and blockade from fungal collections [9].

The clinical history should include data regarding gestational age, birthweight, antenatal ultrasound (renal anomalies, abdominal mass, oligohydramnios), the mother’s exposure to nephrotoxic drugs during pregnancy (nonsteroidal anti-inflammatory drugs, ACE inhibitors and antibiotics), birth history (vital parameters on birth, reanimation, Apgar score), as well as the usage of nephrotoxic drugs in newborns during birth, hypotension, sepsis, congenital heart disease, ECMO support, and vasoactive drug use.

The physical examination should include the assessment of volume status. Signs of dehydration include tachycardia, hypotension, sunken fontanelle, and dry mucous membranes; signs of fluid overload include tachypnea, edemas, elevated blood pressure, escalating oxygen requirement and ventilator support. Careful attention to vital signs, daily weights, intake and output and cumulative fluid balance will help direct fluid provision goals and rates.

2.2 Laboratory and Radiology Examination

Laboratory parameters including electrolytes calcium, magnesium, phosphorous, complete blood cell count, blood urea nitrogen, serum creatinine, albumin, blood gas, and urinalysis should be measured in infants with or at risk of AKI. A random urine sodium and creatinine to calculate FENa can determine if there is intact tubular function in context of a rising creatinine caused by poor renal perfusion.

Based on the parameters given in Table 5.1, prerenal AKI can be differentiated from the acute tubular necrosis (ATN).

Table 5.1 Differential diagnosis of prerenal and renal AKI
Full size table

An ultrasound of the bladder and kidneys should be performed if there is a suspicion for congenital renal abnormality and to rule out obstruction. Doppler assessment of renal vessels (to evaluate the blood flow) should be considered if renal vein or artery thrombosis is suspected. Additional test may include chest X-ray to assess lung volumes and heart size and a voiding cystourethrogram in infants with hydronephrosis demonstrated on ultrasound.

2.3 Neonatal AKI Definition

The most common accepted definition of neonatal AKI is based on a rise in SCr and/or decrease in urine output (UO). The worse of the two parameters is used to make the AKI diagnosis and classify the AKI stage (Table 5.2) [10, 11].

Table 5.2 SCr and UO staging system for AKI diagnosis and severity
Full size table

It is important to recognize the limitation of this definition and staging system. Unfortunately, SCr is a suboptimal biomarker for AKI as: (1) SCr is a marker of kidney function, not injury, (2) SCr may not change until 25–50% of the kidney function has been lost, (3) at a lower GFR, SCr will overestimate kidney function due to tubular secretion of creatinine, (4) SCr varies by muscle mass, hydration status, sex, age, and gender, (5) once a patient receives renal replacement therapy, SCr can no longer be used to assess kidney function since it is easily dialyzed, and (6) certain medications and bilirubin can affect SCr measurements by the Jaffe method. A most pertinent neonatal limitation is that the SCr level is high after birth, as it reflects maternal SCr concentration. After 36–96 h of life, the SCr concentration gradually decreases. In neonates with a lower gestational age, the initial values of SCr are higher and the subsequent decrease is more gradual. SCr levels do not differentiate causes of AKI: prerenal causes, timing of the kidney insult, nephrotoxic drug exposure, and ischemic acute tubular necrosis [10].

Most cases of AKI in neonates are nonoliguric, which may be due to the inherent poor tubular function of premature infants who will be challenged to hold on to fluid in states of decreased vascular volume. Also, newborns, especially preterm, have significantly higher total body water content than older children and adults. This, combined with immature tubules, can explain why the urine output may be higher in newborns with AKI. Although there are not tremendous data available to determine the optimal cutoff for UOP, more recent studies suggest that a UO < 1.5 mL/kg/h is associated with poor outcomes [12,13,14].

2.4 Novel Biomarkers for AKI

Due to insufficient reliability on SCr for early AKI diagnosis, numerous studies to improve the ability to identify neonates with kidney injury have been performed in neonates. The ideal biomarker should rise early in the disease course, is noninvasive and sensitive indicator of AKI, and can specify different etiologies of neonatal AKI.

The most promising noninvasive early biomarkers of neonatal AKI are serum and urinary neutrophil gelatinase-associated lipocalin (NGAL), urinary interleukin-18, kidney injury molecule-1, serum cystatin C, osteopontin (OPN), and beta-2 microglobulin. For some of them there are established normal values in dependence of GA and BW. These biomarkers promise to improve our ability to identify AKI early in the course of disease and help to differentiate the etiology of a rising SCr and/or drop in UO [15].

2.5 Incidence and Outcomes of Neonatal AKI

While the study of AKI in critically ill neonates has lagged behind studies in pediatric and adult populations, the last 5 years have seen an intensification of research in this area. Small, single-center studies in select patient groups such as those with congenital heart disease [16], sepsis [17], hypoxic ischemic injury [18,19,20], infants who receive extracorporeal membrane oxygenation [21], and very low birthweight infants [11, 22,23,24] suggest that AKI is common, and that those with AKI have worse outcomes.

In 2014, a group of neonatologists and nephrologists formed the Neonatal Kidney Collaborative. The inaugural project of the Neonatal Kidney Collaborative, the Assessment of Worldwide Assessment of Kidney Epidemiology in Neonates (AWAKEN) was a retrospective cohort study that screened 4273 infants who were admitted to level 2–3 NICU across 24 sites in 4 countries. 2022 infants met inclusion and exclusion criteria (most were excluded due to not being on intravenous fluid for 48 h). Neonatal AKI occurred in 30% of those enrolled with differences in AKI rates in those who were born at less than 29 weeks (46%), 29–36 weeks (18%), and >36 weeks (46%). Those with AKI had about 10% mortality rate compared to 1.5% in those without AKI. Even after adjusting for potential confounders, those with AKI had 4.6 times higher odds of death and 8.8 more hospitalized days compared to neonates without AKI. These associations remained when these analysis were performed for individual GA groups [25].

2.6 Risk Factors of Neonatal AKI

The risk for neonatal AKI can be attributed to four broad factors [3]. The first is the state of the infant’s kidneys at the time of birth. If an infant is born with a paucity of nephron numbers, the kidneys will lack the potential reserve necessary to overcome a stressful event, leading a reduction in kidney function. Reasons that can lead to a paucity of nephron numbers prior to birth include maternal disease such as diabetes, maternal exposure to nephrotoxic and teratogenic substances, prematurity as renal development continues until 34 weeks gestational age, and congenital anomalies of the kidneys.

The second set of risk factors for the development of neonatal AKI are centered on the hemodynamic and metabolic physiologic demands which occur around the time of birth. Studies consistently show that Apgar scores, receipt of interventions around the time of birth, birth lactate levels, and CRIB 2 scores are risk factors for AKI [26]. Depending on the degree of kidney damage, disruption of the normal physiologic process can result in transient or permanent kidney damage.

The third set of factors of neonatal AKI are secondary to events that may occur during the neonatal time frame. These include episodes of shock (cardiogenic, hypovolemic, or ischemic shock) that can occur during cardiopulmonary bypass surgery, sepsis, and other neonatal conditions. The risk factors for neonatal AKI during cardiopulmonary bypass include preoperative factors (such as degree of hypotension), intraoperative factors (such as aortic bypass time), and post-op factors such as cardiac performance. The risk factors to develop AKI during episodes of sepsis, necrotizing enterocolitis, or wide patent ductus arteriosus have not been fully examined.

The fourth set of risk factors for neonatal AKI development are the iatrogenic medications that are used to treat the infant. Studies in premature infants suggest that most infants admitted to the NICU receive multiple nephrotoxic medications. Indeed the potential impact of nephrotoxic medications on the development of AKI is likely very substantial [11, 24]. Strategies to limit high nephrotoxic exposure and reduce AKI in those who are exposed to nephrotoxic medications could have a tremendous impact on the long-term outcomes.

Strategies to identify AKI in the most premature infants, around the time of birth in infants with difficult delivery, during high-risk events, and while nephrotoxic medications are being given may help prevent AKI and its consequences. The need for a more comprehensive evaluation of the risk factors associated with AKI during these four broad risk factors will be helpful in developing clinical guidelines for clinicians [3].

3 Interventions to Prevent/Treat AKI

Several therapeutic options have been used to prevent and/or mitigate AKI. In patients with perinatal asphyxia, four randomized clinical trials show that theophylline given in the first hours of life to neonates with asphyxia decrease the rates and severity of AKI. Theophylline is an adenosine receptor agonist which can prevent AKI by inhibiting adenosine-induced vasoconstriction [27]. Dopamine leads to vasodilation of the renal vasculature. However, regardless of encouraging results on animal models, it is not proven that there is a benefit in prevention or treatment of AKI with it [28]. On the other hand, fenoldopam, high selective dopamine type 1 receptor agonist which leads to the dilation of renal vasculature, showed some modest benefit in a small, single-center study of infants undergoing cardiopulmonary bypass for congenital heart disease repair [29].

Diuretics are commonly given in order to maintain urine output in the neonates with AKI. There are only a few of studies reporting the use of diuretics in neonates with oliguric AKI, and long-term outcomes have not been reported. The loop diuretics should not be used to prevent AKI, although in cases of fluid overload with oliguria/anuria they do provide a reasonable therapeutic option [30]. The studies haven’t shown a better outcome in adults who have been receiving diuretics [31]. The use of recombinant urate oxidase (rasburicase) in the context of hyperuricemia has been shown to reduce SCr levels and improve UOP in a case series of term infants with AKI [32].

4 Supportive Medical Management

Supportive care to help achieve electrolyte and fluid homeostasis should be started as soon as possible in order to prevent the development of sequelae. Close attention to detail and serial monitoring of urine output and kidney function in neonates is paramount. The keys to supportive management include: (a) identification and correction of risk factors when able, (b) identification and treatment of the cause, (c) prevent further kidney injury by maintaining kidney perfusion with adequate oncotic pressure, intravascular volume, and cardiac contractility, (d) avoidance and unnecessary and appropriate dosing of nephrotoxic drugs, (e) prevention of fluid overload, (f) maintenance of electrolyte balance, and (g) placement of a urinary catheter if obstruction is documented or suspected.

Close attention to fluid status (measure weight, fluid intake and output, serum electrolytes twice a day) is imperative. In a case of oliguria or anuria, one strategy to determine fluid intake is to calculate and only replace the estimated fluid losses: diuresis + insensible losses + extra losses (chest tube losses + gastrointestinal losses, etc.) Depending on the environment of the neonate (incubator vs. open crib vs. warmer) insensible losses may vary. The full term neonate may have insensible losses of 300–400 mL/m2 = 25 mL/kg/day but this can be significantly higher for preterm newborns—40–100 mL/kg/day due to excess fluid losses via their very thin skin. In the first week after birth, >10% weight loss is excessive and should be avoided.

Strategies for fluid balance maintenance of in the neonate should be optimized based on the stage of fluid provision. In the resuscitative stage, a fluid challenge of 10–20 mL/kg should be provided, and possibly repeated depending on the hemodynamic changes seen. Once the resuscitative phase in complete, a strategy to prevent further fluid accumulation should be instituted. Maximizing concentration of fluids (including nutrition) and avoiding excess fluid provision will minimize fluid overload. Diuretics can be given to help achieve euvolemia, recognizing that a failed response to diuretic challenge suggests that AKI will progress. Repeated attempts to escalate diuretics can delay appropriate renal support therapy (dialysis). Surgical decompression of the abdomen in cases of high abdominal pressures should be considered.

Close monitoring of electrolytes is critical, electrolytes should be replaced as needed. Discontinuation of infusions of phosphorous and potassium-containing solutions may be critical. Metabolic acidosis can be due to premature tubular function or other reasons and for the most part should be corrected.

It is important to follow the serum level of drugs the patient is getting which could potentially damage the kidney function. If there is no vital indication, nephrotoxic drugs should be avoided. If, however they must be given, they should be given at the proper interval and dosage based on the estimated creatinine clearance [20, 33, 34].

5 Renal Support Therapy for Neonates

Renal support therapy in the form of hemodialysis, peritoneal dialysis (PD), or continuous renal replacement therapy (CRRT) is rarely used in the NICU, even in high volume NICUs in large tertiary hospitals. One of the reasons is that until recently, CRRT machines have not been designed for neonates, thus the risk of the procedure push the balance toward watchful waiting to initiate therapy. In the AWAKEN cohort, neonatal RRT was performed on 25/4273 (0.5%) of neonates admitted to 24 tertiary NICUs during the 3 months period. The types of RRT included peritoneal dialysis alone (N = 9), continuous renal replacement therapy (CRRT) (N = 4), CRRT + ECMO (n = 11), and peritoneal dialysis + CRRT (n = 1). No infants were dialyzed with intermittent hemodialysis or slow low efficiency dialysis. Of those who received RRT, 19/25 (76%) survived [25].

Indications for renal replacement therapy: Absolute indications include hypervolemia resistant to diuretics, congestive cardiac failure, severe hypertension with high intravascular volume, hyperkalemia (>8 mmol/L), metabolic acidosis (pH < 7.20, or HCO3 < 12 mmol/L), other symptomatic electrolyte disorders (hypo or hypernatremia, hypocalcemia, hyperphosphatemia), rapid increase of urea, and creatinine concentrations (uremic symptoms). Relative indications for dialysis are inability to provide adequate nutrition in context of fluid restriction, prevention of further fluid overload. In addition, HD or CRRT with high clearance rates are first line therapies for infants born with specific inborn errors of metabolism with high ammonia levels.

6 Peritoneal Dialysis (PD)

Peritoneal dialysis is a method of choice for kidney function replacement in newborns, and especially in ELBW newborns. Advantages of PD come from its technical simplicity without the need for vascular access or blood prime of an extracorporeal circuit, no need for systemic heparinization, and slow continuous fluid removal. In infants, the peritoneal surface area per unit weight is approximately twice that of an adult and, overall, it shows more efficiency in both urea clearance and ultrafiltration. Disadvantages of PD are: slower correction of metabolic parameters; lower clearance of small molecules; PD is less effective than other modalities in pulmonary edema, poisoning, or drug overdose, hypercatabolic states, and hyperkalemia. The main risks of the procedure are peritonitis, catheter/exit-site infection, and electrolyte abnormalities. Also, as compared to CRRT, PD does not allow precision in the net fluid removal rates, instead one can only increase or decrease the dextrose concentration in hopes of achieving ultrafiltration goals [35].

The use of PD after cardiopulmonary bypass surgery to prevent fluid accumulation has been shown to improve outcomes [36, 37] and is discussed fully in Chap. 15.

Relative contraindications for acute PD are: recent abdominal surgery, pleuroperitoneal communication, diaphragmatic hernia, severe respiratory failure, life-threatening hyperkalemia, extremely hypercatabolic state, severe volume overload in a patient not on a ventilator, severe gastroesophageal reflux disease, low peritoneal clearance, fecal or fungal peritonitis, and abdominal wall cellulitis.

There are two types of peritoneal catheters. The semirigid acute catheter—the advantages of this kind of catheter are that they can be placed bedside under a local anesthesia and they do not need any surgical help. The disadvantages are: higher risk of infection, bowel perforation risk, and they cannot be left in place for more than 72 h. The Cuffed permanent catheter—the most popular and the most used one is a Tenckhoff catheter. The advantages are: lower risk of infection and bowel perforation; can be used immediately after insertion. The disadvantage is that it requires a surgical insertion, and ideally these catheters are left to heal and optimize the tunnel for 2–3 weeks prior to use.

Due to the limited space in the peritoneal cavity in neonates especially in ELBW neonates, it is very difficult to place a rigid peritoneal catheter, which is why the alternatives are exploited, such as the suction catheter tip, plastic catheter, angiocath, neonatal chest drain, IV cannula, femoral vein catheter, Wallace catheter, and Cook 5F catheter.

A downward pointing exit site (outside the diaper area and stomas) is recommended for peritoneal access. Catheters can be inserted either through the linea alba or laterally or paramedially through the rectus muscle directly into the peritoneum. For permanent catheters, the catheter is tunneled under the skin. For very small neonates, the PD catheter is inserted directly through the abdominal wall, without tunneling. If the insertion is done surgically, it is recommended to do an omentectomy, at the time of PD catheter insertion. After implantation, the catheters are flushed with 10 mL/kg of dialysis solution until the effluent is clear. Perioperative it is recommended to give an antibiotic prophylaxis, mostly given as a single dose of a first- or second-generation cephalosporin or vancomycin. After inserting the catheter, heparin in the dialysate fluid (250–500 units/L) is used to prevent clot formation [35, 38, 39].

6.1 Acute PD Prescription for Neonates

Continuous PD is performed in newborns. The initial volume of fluid should ideally be 10 mL/kg per exchange for at least 1 week. Exchanges can be done every 20–30 min, although most of the time, hourly cycles are sufficient, especially if the procedure is performed 24 h a day. After 1 week, the PD fluid volume is increased slowly over a course of weeks toward the maximum of 40 mL/kg. As the volume is increased, the total cumulative duration of the exchanges is decreased toward 8–12 h per day [35].

Potassium can be added to the PD fluid in hypokalemic patients. Usually 3–4 meq/L is added to maintain normal potassium levels. Fluid removal will be driven by the dextrose concentration in the fluids. A higher dextrose concentration will lead to a higher total fluid removal or ultrafiltration.

6.1.1 Complications of PD

Infectious complications include: exit-site infection (flare, suppurative secretion, granulation) or peritonitis. Staphylococcus aureus is the most common causative agent. Noninfectious complications include: migration of the catheter, perforation, blood in dialysate, dialysate leakage, respiratory insufficiency, extravasation of fluid in tissue compartments, hernias, and hydrothorax. Metabolic complications include fluid, electrolyte, and acid base disturbances. It is important to recognize that up to 1.5 g/kg/day of protein can be lost during peritoneal dialysis.

6.1.2 Bacterial Peritonitis

The main symptoms of peritonitis are: cloudy peritoneal fluid, feeding intolerance, irritability, pain, and fever although the latter two may be difficult to manifest in an infant on a warmer. It is diagnosed based on peritoneal fluid cell count, differential count, gram stain, and culture. A presumptive diagnosis of peritonitis can be made when there are more than 100 leukocytes/mm3, with more than 50% polymorphonuclear cells. Bacterial peritonitis often is caused by a tunnel infection or an inadvertent break in the sterile handling of the PD catheter tubing set or transfer set. Therapy of bacterial peritonitis includes either systemic or intraperitoneal antibiotics. For intraperitoneal antibiotics, two or three rapid PD exchanges are performed initially, followed by a loading dose of intraperitoneal antibiotics (vancomycin 500 mg/L, ceftazidime 250 mg/L) in the abdomen for 4 h. After that the antibiotic doses are reduced (vancomycin 30 mg/L, ceftazidime 125 mg/L) and the patient is maintained on continuous (24 h/day) dialysis, often with longer exchange times. Intraperitoneal antibiotics are continued for 2–3 weeks after obtaining fluid for cell count, differential count, Gram stain, and culture.

6.1.3 Leaking of Peritoneal Fluid Around the PD Catheter

Incidence of pericatheter leaks can vary from 0 to 40%. PD catheters with fluid leak pose a significant risk for peritonitis, and continuation of PD with a leak is not recommended. Management strategies include: temporary discontinuation of PD (2–7 days) in favor of hemodialysis, placement of a new PD catheter (rare), or decrease of the PD fill volume. Surgical glue has been used successfully in combination with the above maneuvers [40].

7 Continuous Renal Replacement Therapy (CRRT)—Novel Machines to Provide Renal Support for Neonates

As mentioned above, continuous renal replacement therapy (CRRT) is rarely used in the NICU, and when it is used, it is usually performed in neonates who have contraindications for peritoneal dialysis or have failed PD. CRRT is used over PD in most ICUs due to its more reliable access, and its ability to ultrafilter a precise amount of fluid; however it is not the primary modality in neonates because CRRT poses additional challenges and risks in this population. The technical challenges of traditional machines make CRRT initiation very difficult, even at experienced tertiary children’s hospitals. Until recently, the volume needed to prime the CRRT circuit varied between 92 and 165 mL. In the United States, the smallest CRRT circuit is around 92 mL (Prismaflex™ M60), and no circuits are FDA approved for use in children <20 kg. In Canada, Europe, and many other countries, the Prismaflex™ HF20 (extracorporeal volume (ECV) = 60 mL) is available. When used, these larger machines necessitate that small children receive CRRT with proportionally larger filters, higher blood flows, massive clearance rates, and big vascular catheters compared to bigger children [41]. The relative large ECV require larger vascular access and make the CRRT initiation procedure higher risk. For these reasons, many centers do not offer CRRT to any infant, especially very small infants and those who are too critically ill to tolerate CRRT initiation.

Initiation of neonatal CRRT therapy has historically been fraught with anxiety, and hemodynamic compromise when initiating the therapy in small infants. In our experience, the higher the % ECV and the more critically ill the patient is before initiation, the more likely problems will arise as a consequence to CRRT initiation. Using the pediatric prospective continuous renal replacement therapy (ppCRRT) registry, which included 13 centers and over 350 children on CRRT, we showed that children <10 kg have lower survival than children >10 kg (33 vs. 67%; p < 0.05) [41]. The lower rates of survival are perhaps due to patient selection, initiation of CRRT only as “last resort,” the added risks of circuit initiation, or a combination of these factors.

Priming the CRRT circuit with packed red blood cells (pRBC) reduces morbidity; however, it is important to recognize that compared to physiologic blood, pRBC are cold, very concentrated, acidic, hyperkalemic, and hypocalcemic. Most centers that perform neonatal CRRT have protocols in place to buffer the acidic environment, reverse hypocalcemia, and dilute the higher hematocrit of pRBCs. Even with these measures, blood primes are not without risk, as blood primes can cause hypothermia, acidosis, hypocalcemia, hyperkalemia, thrombocytopenia, hypotension, and coagulopathy. These risks increase exponentially with smaller sized infants, those who are hemodynamically unstable and with frequent repeated initiation. These complications and the need to alter the CRRT prescription stem from the markedly large ECV inherent to available machinery.

CRRT machines with smaller ECV could reduce the risks associated with the therapy and improve outcomes. For these reasons, several groups have developed or adapted dialysis and convective clearance devices which use a smaller size circuit. What is common to all of these machines is that the smaller ECV of these machines allows for adequate flows with much smaller diameter catheters. In addition, initiation of CRRT circuit with much lower ECV is for the most part uneventful, even in very small and very critically ill infants [42].

The Newcastle Infant Dialysis Ultrafiltration System (NIDUS) [43] has an ECV of around 10 mL and can provide continuous or intermittent dialysis with the use of a 4F single lumen catheter. The circuit uses automated syringe pumps to accomplish four separate steps: (a) remove a volume of blood from the patient, (b) send blood through a dialysis filter that has countercurrent dialysis running, (c) return the blood back to the initial syringe pump, and (d) return that volume of blood (now cleaned) back to the patient.

The Cardiac And Renal Pediatric Dialysis Emergency (CARPEDIEM™) has available circuits of 27, 34, and 45 mL [42] for filters of 0.075; 0.15, and 0.25 m2). By using smaller blood pump with a unique design, it is able to reduce the peak pressure for a given blood volume, reducing the need for a very wide catheter. A 4.5 F double lumen catheter can be used to accomplish the needed flows for this machine. The circuit has to be changed out daily. The blood flow can be titrated from 2 to 40 mL/min and it has very precise fluid scales available.

To mitigate concerns posed by CRRT machines with large ECV in relation to blood volume size, Askenazi adapted the Aquadex™ machine (a machine designed for ultrafiltration in adults with heart failure) to provide convective clearance by using an independent IV pump to deliver the replacement fluid. The machine can ultrafilter up to 500 mL/h. It has an integrated hematocrit detector that guide ultrafiltration by detecting abrupt changes in blood volume, which is inversely proportional to hematocrit. This system can provide adequate clearance, adequate electrolyte balance and fluid ultrafiltration, with minimal need for interventions during circuit initiation. The biggest drawback to this system is that the replacement pump is not directly in communication with the blood flow pump [44].

Clearly, as these devices become more widely accepted, there will be lower risk in performing this procedure. The risk/benefit ratio will allow for more judicious use, earlier intervention and will likely serve as a profound improvement in critically ill neonates.

8 Follow-Up After Neonatal AKI

Long-term risk of CKD in premature infants—Nephrogenesis begins at the fifth week of gestation and continues until 34–36 weeks. Nephron endowment and kidney function is determined by genetics and the intrauterine environment and varies across individuals. In term neonates and healthy adults, nephron numbers vary from 200,000 to 1.2 million per kidney—which also correlates closely with kidney mass and function in an individual. Prenatal factors (i.e., prenatal delivery, maternal diabetes, and intrauterine growth restriction) are factors which lead to impaired nephrogenesis [45].

The epidemiology of CKD in NICU graduates was recently reviewed by our group [3] and others [46]. The exact incidence of CKD in NICU graduates is not known. Based on a meta-analysis published by White et al. [47] in 2009, moderately premature infants (those with birthweight <2500 g) have a 70% higher risk of CKD during childhood than term infants. This meta-analysis likely underestimated the current CKD risk of premature infants, as it studied infants with birthweight <2500 g who were born in the 1980s. Now that many NICUs successfully support and graduate many more extremely premature infants (<24 weeks GA, and with birthweights <500 g) compared with 20 years ago, the magnitude of the problem is likely more substantial.

Long-term risk of CKD after AKI in NICU graduates—Previously it was assumed that those who survived an episode of AKI would recover kidney function without long-term effects. Recent data from animals, critically ill children [48], and adults with AKI suggest that survivors are at risk for the development of CKD. The impact of prematurity, intrauterine growth failure, and AKI on nephrogenesis has not been fully delineated but small studies suggest that the extrauterine environment and AKI are detrimental to nephrogenesis and in turn lead to CKD in adult life. Recent studies in term and preterm NICU graduates that evaluate the impact of AKI on the development of CKD are listed in Table 5.3. Acknowledging that these were small single-center studies with a potential risk of selection bias due to relatively low follow-up rates, the prevalence of CKD in these studies in aggregate was 31% (range of 9–83%). Importantly, the studies found that those with neonatal AKI had higher composite CKD rates, low estimated GFR, smaller total kidney volume (TKV), and abnormal urine kidney damage markers. An adequately powered prospective cohort study using a formal GFR metric with iohexol is greatly needed to evaluate the impact AKI has on CKD in NICU graduates.

Table 5.3 Long-term studies on CKD in term and preterm NICU graduates
Full size table

Key Learning Points

  • The most common accepted definition of the neonatal AKI is based on a rise in serum creatinine (SCr) and/or decrease in urine output (UO).

  • Most cases of AKI in neonates are nonoliguric.

  • The most promising noninvasive early biomarkers of neonatal AKI are serum and urinary neutrophil gelatinase-associated lipocalin (NGAL), urinary interleukin-18, kidney injury molecule-1, serum cystatin C, osteopontin (OPN), and beta-2 microglobulin.

  • The keys to supportive management include: (a) identify and correct risk factors when able, (b) identify and treat the cause, (c) prevent further kidney injury by maintaining kidney perfusion with adequate and avoidance oncotic pressure, blood pressure and intravascular volume, and contractility, (d) avoid and appropriately dose nephrotoxic drug, (e) prevent fluid overload, (f) maintain electrolyte balance, and (g) placement of a urinary catheter if obstruction is documented or suspected

  • Peritoneal dialysis is a method of choice for kidney function replacement in newborns, and especially in ELBW newborns.

  • CRRT machines with smaller extracorporeal volume could reduce the risks associated with the therapy and improve outcomes. Currently, Newcastle Infant Dialysis Ultrafiltration System (NIDUS), cardiac And Renal Pediatric Dialysis Emergency (CARPEDIEM™), and the Aquadex™ are being used to provide CRRT in small children.