Drug-induced acute kidney injury in children

Abstract

Acute kidney injury (AKI) is a serious problem occurring in anywhere between 8 and 30% of children in the intensive care unit. Up to 25% of these cases are believed to be the result of pharmacotherapy. In this review we have focused on several relevant drugs and/or drug classes, which are known to cause AKI in children, including cancer chemotherapeutics, non-steroidal anti-inflammatory drugs and antimicrobials. AKI demonstrates a steady association with increased long term risk of poor outcomes including chronic kidney disease and death as determined by the extent of injury. For this reason it is important to understand the causality and implications of these drugs and drug classes. Children occupy a unique patient population, advocating the importance of understanding how they are affected dissimilarly compared with adults. While the kidney itself is likely more susceptible to injury than other organs, the inherent toxicity of these drugs also plays a major role in the resulting AKI. Mechanisms involved in the toxicity of these drugs include oxidative damage, hypersensitivity reactions, altered haemodynamics and tubule obstruction and may affect the glomerulus and/or the tubules. Understanding these mechanisms is critical in determining the most effective strategies for treatment and/or prevention, whether these strategies are less toxic versions of the same drugs or add-on agents to mitigate the toxic effect of the existing therapy.

Introduction

Acute kidney injury (AKI) can be described as the sudden decrease in renal function over a variable period of time (hours to days) and can be caused by numerous aetiologies 1. While overall not well defined in children, the incidence of AKI in the paediatric intensive care units is reported to range from 8% to 30% and is thought to occur in roughly 7% of the general paediatric population 2,3. Pharmacotherapy is one of the major causes of AKI and may play a causative role in as many as 25% of all cases 4,5. Drugs including antimicrobials, chemotherapeutic agents and non-steroidal anti-inflammatory drugs (NSAIDs), among others, have all been implicated in drug-induced renal injury in children 2.

We chose to focus this review on drug-induced toxicity in children as drugs that are renally toxic in children may, but not always, differ from those that are toxic in adults. There are several physiological differences between adults and children that may account for variability in renal toxicity including volume of distribution, glomerular filtration rate (GFR), clearance and expression of hepatic and renal cytochrome P450 enzymes. Variations in these parameters can affect the pharmacokinetics of a drug ultimately resulting in different systemic concentrations of a possible renal toxin among distinct populations. Therefore in some cases this will result in increased exposure to a renal toxin in children increasing their susceptibility to injury, or it may decrease their exposure as compared with adults therefore protecting them 6,7.

The extent to which drugs play a role in AKI may be explained by the increased vulnerability of the kidney to damage by drugs 4. The kidney receives nearly 25% of resting cardiac output and is a major organ of drug excretion. It is exposed to a larger proportion and higher concentration of toxic drug than most other organs (due to the concentrating role of the kidney), putting it at greater risk for injury 1,8,9. Additionally, toxicity in the proximal tubule specifically may be increased due to its role in both secretion and reabsorption, allowing for accumulation of toxic species within the cell 8. Adding further to the disadvantage of the kidney compared with some other organs is its major role in drug biotransformation during which electrophilic toxic metabolites and reactive oxygen species (ROS) may be produced; increasing levels of oxidative damage and possible injury 8,10.

The pathophysiology of AKI is complicated. It can generally be grouped into pre-renal, intrinsic (glomerular, interstitial, vascular and tubular injury) and post-renal causes. The gold standard to diagnose these injuries is kidney biopsy. However, it is not always feasible and the risk might out-weigh the benefits, as it will not always alter the management. By contrast, some tests might help differentiate the causes of renal injuries. For example, nephrotic range proteinuria, haematuria and red blood cell casts are signs of glomerular injury, whereas urinary electrolytes, glucose and protein wasting may be a sign of tubular injury. For example, proximal tubular injury like Fanconi syndrome can lead to sodium, magnesium, phosphate, glucose, uric acid, amino acid and bicarbonate wasting. Other urinary biomarkers, such as β₂- microglobulin and neutrophil gelatinase-associated lipocalin (NGAL) have been investigated extensively in the area of acute tubular necrosis (ATN). Renal injury caused by drugs is complex and may often be a combination of both glomerular and tubular toxicity. It is important to identify the disease early because it might be reversible by stopping the offending agents. Patients who developed AKI are more likely to have increased risk of mortality and increased incidence of chronic kidney disease, and the poor prognostic factor for chronic kidney disease is related to the magnitude and the recurrence of AKI 11. One of the challenges in AKI is to define the disease and recognize it early.

In an effort to prevent AKI in children an obvious solution would be to avoid the use of renally toxic drugs altogether, particularly in those at high risk of developing renal injury. The answer, however, is not so simple. Many of these drugs are used in the treatment of life-threatening conditions for which no alternative therapies exist. The benefits of these drugs to the patient often outweigh the potential risk and while dose adjustments can be made in an effort to prevent future or continued renal insult other strategies should also be considered. Given the increasing understanding of the underlying mechanisms involved in the toxicity of many of these drugs, one has the ability to develop more effective treatment and prevention strategies, some of which are available or are currently being researched.

This paper reviews some of the relevant drugs and/or drug classes whose adverse effects in children include AKI. We discuss known mechanisms of toxicity, as well as current or experimental treatments and/or prevention strategies.

Cancer chemotherapeutics

Cancer chemotherapeutics are among the most common offenders in drug-induced kidney toxicity 8. The use of these agents in combination has the potential to potentiate the dangerous effects of typical renal toxic drugs 9,12. While the list of chemotherapeutic drugs that have been implicated in renal injury is lengthy, this review will focus on those most relevant in children including ifosfamide, cisplatin and methotrexate 13,14.

Ifosfamide

Ifosfamide (IFO) is used in the treatment of solid tumours in both children and adults 15–18. While not necessarily first line, it is included in many paediatric chemotherapy protocols 19–24. It is particularly relevant in this discussion due to the high incidence of renal impairment seen in children but less commonly in adults, with 30% of children treated with IFO suffering some degree of renal impairment. Generally the proximal tubule is main site of toxicity (Table 1), explaining why the most severely affected children suffer from Fanconi syndrome 25–28. The glomerulus can also be affected with 30% of children developing reduced GFR 26,27,29,30. These impairments may lead to clinical manifestations including hypophosphataemic rickets, renal tubule acidosis (both proximal and distal), diabetes insipidus and hypokalaemia 31–34. Treatment may include renal supportive care such as dialysis and/or renal transplant and in the case of hypophosphataemic rickets supplementation with phosphate and bicarbonate 25. In the case of IFO renal toxicity, tubulopathies and glomerular toxicity have shown little to no improvement over time 26,35. Risk factors include age, with those younger than age 5 years being at the greatest risk, cumulative dose, prior unilateral nephrectomy and importantly, prior or current platinum therapy 9,12,36,37.

Table 1.

Classification of acute kidney injury

Drug	Type of injury
Ifosfamide	Direct tubular toxicity
Cisplatin	Direct tubular toxicity
Methotrexate	Obstructive crystal nephropathy and direct tubular toxicity
NSAIDs	Pre-renal and acute interstitial nephritis
Aminoglycosides	Direct tubular toxicity
Amphotericin	Direct tubular toxicity

Nephrotoxicity caused by IFO is generally believed to be the result of oxidative stress caused by the metabolite chloroacetaldehyde (CAA) produced during its biotransformation in the kidney. Younger children are at a greater risk of kidney toxicity due to the renal ontogeny of those enzymes responsible for IFO metabolism. Higher levels of CYP 3A, which metabolizes IFO to CAA, have been observed in an animal model whose age corresponds to toddlerhood 38,39. It appears that IFO enters the proximal tubule via the blood stream through the organic cation transporter 2 (OCT2) resulting in toxic CAA concentrations (50 μm) 40–43. Toxic levels of CAA have detrimental effects on renal proximal tubule cells likely through the generation of oxidative stress. Both CAA and reactive oxygen species (ROS) may damage cellular proteins and DNA, result in changes in sodium and calcium intracellular concentrations leading to impaired solute reabsorption and necrosis, respectively, affect ATP synthesis and critically, result in a depletion of glutathione 44–48.

With an indication of the mechanisms involved being oxidative damage and glutathione depletion, anti-oxidants have been assessed as a preventive measure in animal models including l-histidinol, taurine, thymoquinone, resveratrol, melatonin, glycine and N-acetylcysteine (NAC) 45,47,49–53. NAC stands out among the potential therapeutic agents given its clinical use in children for paracetamol (acetaminophen) overdose, providing a wealth of safety data, as well as its documented use to prevent IFO nephrotoxicity in patients 54. Early in vitro and in vivo animal studies have shown the effectiveness of NAC in preventing IFO toxicity, as well its lack of interference on the antineoplastic effects of IFO. These findings have been supported by three clinical experiences 45,55–59, although randomized controlled trials (RCTs) are required to assess the role NAC played in protection of the kidney in the published cases. Given the promise of anti-oxidants, and in particular NAC, further research needs to be carried out to determine which potential treatments are the most clinically useful.

Cisplatin

With a cure rate of 90% for some cancers, cisplatin (CIS) is an effective paediatric chemotherapeutic agent that has clear benefits 60–62. However, AKI, present in 20%–80% of children treated with CIS, is an adverse effect that may limit its tolerable dose levels 61,63. CIS affects the S3 segment of the proximal tubule and nephrotoxicity is generally marked by increased serum creatinine and a decreased GFR, along with hypomagnesaemia and hypokalaemia 13,63–65. Risk factors include cumulative dose, dehydration, concomitant use of other nephrotoxic drugs and hypoalbuminemia. Renal damage caused by CIS is typically persistent in children 13,66. Treatment includes general supportive care for AKI and magnesium supplementation in the case of hypomagnesaemia 67.

The study of CIS renal injury has revealed a complex interaction of mechanisms affecting cell structure and cell signalling pathways 61. Simply put, CIS is taken up by basolateral OCT2, resulting in production of ROS and activation of signalling pathways, mitogen-activated protein kinase (MAPK), P53 and possibly P21, leading to renal tubular cell death. An inflammatory response also occurs, likely through activation of tumour necrosis factor-α (TNF-α) receptor 2 by intrinsically produced TNF-α. Oxidative stress appears to be both the driving force and end result of some of these changes 61,68.

Identifying these targets has allowed for evaluation of a number of compounds in preventing this nephrotoxicity. Cimetidine has been assessed in an attempt to prevent renal transport and accumulation of CIS, as has erythropoietin to block apoptosis and quercetin to prevent inflammation 61,68. However, the most exploited mechanism in attempts to prevent CIS nephrotoxicity, is oxidative stress. Numerous anti-oxidants have been assessed to thwart the effects of oxidative stress caused by CIS including, but not limited to, amifostine, NAC, sodium thiosulfate and theophylline 69–73. While NAC and sodium thiosulfate have not been assessed in RCTs, both amifostine and theophylline have. Theophylline was effective in some trials but not in others 69,74. Similarly, amifostine was effective in reducing nephrotoxicity from 30% to 10% in a trial of women with ovarian cancer but not so in a trial assessing CIS-induced nephrotoxicity in children with osteosarcoma 70,71. Guidelines of the American Society of Clinical Oncology support the use of amifostine in ovarian or non-small cell lung cancers treated with CIS, but not for other types of cancer 75.

Methotrexate

The nephrotoxic drug methotrexate (MTX) has a wide range of use owing not only to its anti-proliferative but also its immunomodulating effects. It is particularly important for children as an effective treatment for acute lymphoblastic leukemia 2,76–78. It is used in either low or high dose regimens, with only high dose regimens resulting in AKI. The overall incidence of AKI caused by high dose MTX is around 2% 79,80.

Mechanistically AKI caused by high dose MTX is similar to CIS and IFO in that direct drug toxicity is also thought to occur via ROS resulting in cellular damage 78,80,81. However MTX is distinct in that it may also result in crystal nephropathy. MTX and its metabolites can precipitate within the tubules due to their low solubility in an acidic pH resulting in tubular obstruction, leading to decreased GFR and tubular cell death 8,80,82,83. Risk factors include hydration status and urinary pH, with those who are dehydrated or have an acidic pH being at greatest risk. Unilateral nephrectomy and concomitant use of renal toxic drugs are also risk factors 8,14.

With respect to long term consequences, MTX AKI is completely reversible and therefore the primary concern with its use may not be its transient renal effects but rather the effect of decreased GFR on MTX renal clearance rate and serum concentrations. High MTX concentrations reduce folate concentrations within normal cells resulting in toxicity 8,78. 90% of MTX is excreted renally and therefore decreased GFR diminishes the capacity to excrete MTX, resulting in toxic systemic concentrations. Prolonged transient effects increase the risk of organ damage. However, the long term effects on the kidney have yet to be properly evaluated and may be cause for concern, as well 8,78.

With respect to prevention and treatment strategies, AKI caused by MTX can be reduced, although not completely mitigated with hydration and urine alkalinization strategies 80,84. These strategies prevent the concentration of MTX from becoming too high in the tubules, as well as increase its solubility, which is pH dependent. Toxic systemic concentrations of MTX, brought on by AKI must also be addressed. Currently leucovorin is routinely administered approximately 24−36 h after MTX, allowing for rescue of normal tissues from MTX toxicity by replenishing folate concentrations 79,80. Glucarpidase has also recently been approved for use in patients with elevated MTX plasma concentrations. This agent cleaves MTX into two non-toxic metabolites and can decrease its concentration by more than 95% in a mere 15 min. It does not, however, have any effect on intracellular concentrations of MTX 79. While the current means of reducing AKI and systemic toxicity caused by MTX are quite effective, future research is necessary to address both the possible long term outcome of MTX therapy on the kidney, as well as possible strategies allowing for complete mitigation of its renal toxicity.

NSAIDS

NSAIDS are commonly used in children as analgesics, antipyretics and anti-inflammatory medications. Indomethacin, in particular, is an important treatment for patent ductus arteriosis in infants 85,86. NSAIDS are widely available putting large numbers of people at risk for their adverse effects. One such adverse effect is AKI, with an incidence of 2.7% in hospitalized children and adolescents 87.

There are two mechanisms through which NSAIDs cause AKI, changes in haemodynamics and acute interstitial nephritis (AIN). The former is thought to account for 78% of cases of AKI, while the latter accounts for 22% 87. NSAIDs may alter haemodynamics by acting as cyclo-oxygenase (COX) inhibitors. COX is responsible for producing prostaglandins (from arachidonic acid), which have a vasodilating effect on the afferent glomerular arteriole. While it is unlikely that prostaglandins have much impact on children with normal circulating volume, in individuals with volume depletion, their role as vasodilators becomes more important to maintain adequate renal perfusion. Blocked prostaglandin synthesis leads to unchecked vasoconstriction of the afferent arteriole, resulting in reduced GFR and eventually renal ischaemia and acute tubular necrosis 2,87–89.

AIN it thought to occur as a result of COX inhibition shunting the metabolism of arachidonic acid to the pathway producing leukotrienes, which are involved in the activation of the inflammatory response 90. Along with AKI, patients may also, although rarely, present with manifestations of systemic hypersensitivity reactions. Children receiving NSAIDs may also present with nephrotic syndrome 91,92. Treatment of haemodynamically driven AKI includes normal supportive care, while corticosteroid treatment may be used in AIN. While patients generally recover, one study has shown approximately 30% of children have persistent mild chronic kidney damage 91. Preventative strategies include avoiding NSAIDs and/or monitoring those at high risk, including children with volume depletion, pre-existing renal disease, or concomitant use of other nephrotoxic drugs 85. While attempts have been made to develop NSAIDs that are less nephrotoxic, to our knowledge no pharmacotherapy is being evaluated as a preventative measure.

Antimicrobials

A variety of different antimicrobials can cause AKI, in particular aminoglycosides and amphotericin B.

Aminoglycosides

Aminoglycosides (AGs) are a class of antibiotics used in the treatment of Gram negative and Staphylococcus aureus infections 93,94. They include antibiotics such as gentamicin and amikacin. In children they are implicated in AKI at a rate of 10% to 30% 1,95. There are conflicting reports as to whether neonates are at an increased or decreased risk for AG nephrotoxicity 2,96. Patients may present with oliguric AKI, along with proximal tubule dysfunction, resulting in loss of enzymes, proteins, glucose, calcium, potassium and magnesium. Most patients may recover but some progress to chronic interstitial nephritis 97. Risk factors for AKI resulting from AGs are underlying renal insufficiency, duration of therapy greater than 10 days, trough concentrations greater than 2 μg ml⁻¹, concomitant liver disease and hypoalbuminaemia 98.

AG-induced renal toxicity occurs as a result of its uptake by the proximal tubule cells. With a positive charge at physiological pH AGs bind anionic phospholipids within the proximal tubule cell membrane. AGs are endocytosed in a megalin-dependent fashion resulting in accumulation within the cell. With accumulation primarily in the S1 and S2 segments 2,99, AKI occurs via tubular cytotoxicity caused by a variety of routes including increased lysosomal permeability, mitochondrial dysfunction, proteolysis and disrupted protein sorting leading to death via both apoptosis and necrosis 1. AGs have also been implicated in the production of ROS. Cytosolic AGs act on mitochondria directly and indirectly, and thus activate the intrinsic apoptotic pathways, interrupt the respiratory chain, impair ATP production and produce oxidative stress by increasing superoxide anions and hydroxyl radicals 100–102. Renal toxicity associated with various AGs is correlated with the affinity of each drug for the proximal tubule cell 103–105.

Prevention strategies for AG AKI include using extended dosing intervals, administering during active periods of the day, limiting the duration of therapy, monitoring serum drug concentrations and renal function two to three times per week and maintaining trough concentrations of less than or equal to 1 μg ml⁻¹ 98. While not currently used clinically, several compounds have been assessed in the prevention of AG renal toxicity. Anionic polyamino acids are able to interfere with AG’s binding to the proximal tubule cell, while several antioxidants including selenium, vitamin A and vitamin C have been successful in mitigating AG toxicity in experimental models. More research is necessary to determine the potential clinical usefulness of these strategies 106,107.

Amphotericin B

Amphotericin B (AmB) is an antifungal agent, which is the drug of choice in immuno-compromised children 2. While extremely toxic to the kidney, causing AKI via tubular cell toxicity, its common use for life-threatening fungal infections contributes to its mainstay in antifungal therapy 108. Some degree of decreased renal function has been reported in up to 80% of patients undergoing therapy and the reported incidence of AmB-induced AKI increases with the cumulative dose delivered 109,110. On the contrary patients receiving cumulative low doses of AmB of less than 1 g have been reported to develop renal failure less frequently. Risk factors for AmB-induced AKI include underlying renal insufficiency, rapid infusion, large daily dosing and prolonged duration of therapy 109.

The mechanism by which AmB exerts is toxicity is not well understood. AmB can increase cell permeability by inserting itself into membranes and forming pores. It may also increase permeability in the macula densa of the distal tubule resulting in dysregulation of the tubuloglomerular feedback system. Deoxycholate, a component of AmB is also thought to be responsible for direct toxicity to the tubules 98,111–113. Prevention strategies include saline hydration before and after drug administration, use of liposomal formulations, limiting the duration of therapy, and considering a continuous low dose infusion over a 24 h period rather than a large dose at set intervals 98.

Liposomal formulations include liposomal AmB and AmB lipid complex. Lipid formulations appear to be less renally toxic as they do not contain deoxycholate and are thought to preferentially target fungi within the endoplasmic reticulum, with less delivery to the kidney. The rate of nephrotoxicity by both AmB lipid complex and liposomal AmB is approximately 13%, which is considerably lower than formulations with deoxycholate 109,114. Continued research is necessary in order to mitigate fully renal toxicity caused by AmB and its lipid based formulations.

Conclusion

Drug-induced AKI is a serious problem. As outlined above, it remains a side effect for a number of both commonly used and important drugs for children. While ultimately avoidance of renally toxic drugs, especially in those most at risk, is best, it is also unlikely. Many of the drugs presented above are extremely effective agents for potentially life-threatening illnesses and unfortunately alternative therapies do not always exist. There are, however, other options to this dilemma. We can continue the research and use of other co-administered agents, which will protect against the toxic effects of some drugs. Alternatively, developing newer formulations with equal effectiveness but less toxicity may be the best approach. Examples of both have been described above. In either scenario, an understanding of the mechanisms underlying the toxicity caused by these drugs is our best hope in developing successful treatment and prevention strategies. While for some drugs we have begun to understand even the most complex mechanisms at play, others still remain unclear. Continued research in the area surrounding mechanisms of adverse drug reactions is imperative to our advancement in continuing to predict, treat and ultimately prevent these serious adverse events.

Competing Interests

All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare no support from any organization for the submitted work and no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years. LNH, MJR and GK have carried out and published work regarding the protective role of N-acetylcysteine for ifosfamide-induced AKI.

We would like give special thanks to nephrologist Dr Susan Huang for providing us with feedback for this paper.