Nephrotoxicity of Cancer Treatment in Children

Roderick Skinner


Pediatr Health. 2010;4(5):519-538. 

In This Article

Nephrotoxicity Due to Specific Cytotoxic Agents


Ifosfamide is an important agent in the treatment of many pediatric solid tumors, including rhabdomyosarcoma, soft tissue sarcomas and Ewing's sarcoma. Unfortunately it may cause any combination of glomerular, proximal or distal tubular toxicity (Box 3), with considerable variability in the time of onset (in relation to ifosfamide treatment), clinical manifestations, severity and reversibility of renal damage.[23]


Acute nephrotoxicity is well documented after ifosfamide treatment in children and indeed reversible subclinical acute tubular damage is the most common initial manifestation of toxicity.[24] Although reported in adults,[25] acute renal failure (ARF) is rare in children. Acute tubular toxicity may become more persistent and severe with further treatment, and may be accompanied subsequently by progressive glomerular damage in some children even after discontinuation of ifosfamide.[26] The frequency of ifosfamide nephrotoxicity in children reported in the literature shows considerable variability, probably owing to differences between the patient groups investigated (and hence the treatment received) and perhaps just as importantly, differences between the methods used to investigate renal function in different studies.[27] Glomerular toxicity appears to be uncommon when evaluated by the measurement of serum creatinine concentration, as exemplified by an incidence of 1.4% in a large study from the USA,[28] but it is well known that serum creatinine is an insensitive indicator of glomerular impairment, becoming elevated only when glomerular function has already reduced by approximately half.[201] By contrast, a reduction in the glomerular filtration rate (GFR; a much more sensitive and accurate marker of glomerular function) was reported in 50% of 123 children studied for a median of 6 months after treatment with ifosfamide.[7] Chronic proximal tubular toxicity is even more common. Small single-center studies in the early 1990s found that clinically significant toxicity manifested by hypophosphatemic rickets (HR) and/or proximal renal tubular acidosis (RTA) was present in approximately 25% of children who were studied up to 4 years following completion of ifosfamide,[29,30] whilst glycosuria due to impaired proximal tubular reabsorption was detected in approximately 90% of children in a large UK multicenter study.[7] However, a larger multicenter study performed in the German Late Effects Surveillance System used more stringent criteria (ongoing hypophosphatemia and proteinuria) for the definition of tubular toxicity, which was found in only 5% of 593 patients at a median of 19 months post-ifosfamide treatment.[31] Although apparent reductions in urinary concentrating ability, suggesting distal tubular toxicity, may occur in up to 30% of children after ifosfamide, they are of doubtful clinical significance except in the very rare child with nephrogenic diabetes insipidus (NDI).[30]

Clinical Features

Proximal tubular damage is usually the predominant feature of chronic ifosfamide nephrotoxicity in children, in terms of both subclinical and clinically significant toxicity. Chronic subclinical proximal tubular toxicity may lead to glycosuria, aminoaciduria and increased urine excretion of low-molecular-weight proteins, including retinol-binding protein (RBP), α1-microglobulin and β2-microglobulin (β2-M), or of the proximal tubular antigen adenosine deaminase-binding protein.[24,32,33] More severe failure of proximal tubular reabsorption leads to a Fanconi syndrome, with an excessive urinary loss of phosphate, bicarbonate and potassium and, more rarely, calcium and magnesium.[34] Tubular damage may be accompanied by chronic glomerular impairment[30,35] and in some patients, particularly adolescents, glomerular toxicity may be the major feature of toxicity.[26] CRF has been reported occasionally.[36,37] Impairment of distal tubular acidification and concentration of urine is well described but seldom clinically significant.[30,38] Less specific features of ifosfamide nephrotoxicity include raised urinary excretion of renal tubular enzymes (RTEs)[24,39] following damage to proximal or distal tubular cells, and proteinuria,[7,31] which may be due to either increased glomerular permeability resulting in increased filtration and urinary loss particularly of albumin, or impaired tubular reabsorption of low-molecular-weight proteins, or both (see earlier). Reduced urinary excretion of Tamm–Horsfall protein has also been reported, presumably due to impaired tubular manufacture.[40] Although the importance of the magnitude of proteinuria as a component of ifosfamide nephrotoxicity has not been evaluated, it may have considerable long-term significance since it predicts (and indeed may be the cause of) progressive renal functional impairment in a wide range of renal diseases, and it is conceivable that antiproteinuric treatments may offer a potential renoprotective management strategy in patients with established renal damage.[41,42]

The clinical consequences of chronic ifosfamide-induced nephrotoxicity in children include HR, RTA and NDI,[34,43] whilst hypertension has been reported in 5% of survivors.[7] CRF, HR and RTA may cause growth impairment in children.[44]

Natural History

The degree of reversibility of chronic ifosfamide nephrotoxicity appears to be unpredictable. Although some recovery may occur after acute nephrotoxicity, both glomerular[26] and tubular function[45] may deteriorate after ifosfamide has been stopped. Partial[46] or even complete[47] recovery from severe chronic renal damage has rarely been reported. However, two LTFU studies have been published recently, demonstrating that ifosfamide nephrotoxicity may persist for at least 10 years after completion of chemotherapy. A longitudinal study of 25 patients revealed that although none had a significantly reduced GFR (<60 ml/min/1.73m2) at the end of treatment, 13% did 10 years later.[48] By contrast, none of the 28% patients who had clinically significant tubular toxicity at the end of treatment (requiring electrolyte supplementation) still exhibited this severity of damage 10 years later.[48] A larger cross-sectional study evaluated 183 patients once at a median of 10 years following ifosfamide treatment and found persistent mild glomerular impairment (GFR <90 ml/min/1.73m2) in 21% and tubular toxicity (manifest by reduced phosphate reabsorption) in 24%, whilst proteinuria was observed in 12%.[49]

Risk Factors

Risk factors suggested for the development of nephrotoxicity after ifosfamide include higher cumulative ifosfamide dose, younger patient age at treatment, the administration schedule used to deliver ifosfamide, previous or concurrent treatment with other potential nephrotoxins (especially cisplatin), and pre-existing renal impairment or tumor invasion. Children receiving higher total doses of ifosfamide (>60–80 g/m2) appear to be at greater risk[7,50] but extensive damage may still occur after much lower doses.[7] However, the importance of high cumulative ifosfamide dose in causing very long-term nephrotoxicity (i.e., persisting at 10 years) is uncertain. Although the longitudinal LTFU study reported earlier (see 'Natural history' section) found a clear correlation between dose and several measures of tubulopathy (phosphaturia, hypophosphatemia and acidosis) at the end of treatment, no such relationship remained at 10 years.[48] However, the larger cross-sectional LTFU study described a relationship between higher-dose and increased phosphaturia, reflecting greater tubular toxicity.[49]

Although ifosfamide may cause severe renal damage at any age, most published reports of severe toxicity have been in infants and young children, who may be more vulnerable to both proximal tubular toxicity and its consequences, especially growth failure.[34] However, the importance of age as a risk factor for chronic nephrotoxicity remains unclear, since although several studies have found an increased risk in young children,[31,37,50,51] other large studies have not.[7,52] Moreover, young age does not appear to be a predictor of nephrotoxicity at 10 years,[53] and indeed it has even been suggested that older age at treatment may predict greater glomerular toxicity, although the magnitude of this effect was very small (RR: 1.08).[49]

There is no convincing evidence that the duration of intravenous administration of ifosfamide (bolus, short or continuous infusion) influences the risk of acute or chronic nephrotoxicity.[54,55] Ifosfamide nephrotoxicity appears to be exacerbated by prior cisplatin therapy,[28,52] but not carboplatin or abdominal radiotherapy.[31] However, the combination of ifosfamide with very high-dose carboplatin may cause severe toxicity.[56] Severe ifosfamide-induced renal damage has been reported in patients with prior unilateral nephrectomy, renal impairment or tumor infiltration,[43,52] but the RR of toxicity in such patients is unknown. However, these factors are of limited predictive value in individual patients. It is tempting to speculate that differences in ifosfamide pharmacokinetics or intrarenal metabolism (or both) may explain some of the interindividual variability in the severity of nephrotoxicity,[57] but there is no direct evidence to support this hypothesis.


The few reported renal biopsies have described focal proximal tubular or tubulointerstitial changes with no or relatively little glomerular damage.[36,58,59] The pathogenesis remains poorly understood although most evidence suggests that direct cytotoxicity is responsible. Studies in rats and in vitro proximal renal tubular cell cultures have demonstrated cellular damage caused by the drug or, more likely, a metabolite.[60–62] The nature of this toxicity is not clear, but disruption of tubular cell energy pathways and membrane function due to the production of reactive oxygen species have been implicated.[57,63] The metabolism of ifosfamide and its non-nephrotoxic structural isomer cyclophosphamide are qualitatively similar, but quantitative differences exist that offer possible clues about why ifosfamide is frequently nephrotoxic in clinical use whilst cyclophosphamide is not.[34]

Ifosfamide is activated by ring-hydroxylation to 4-hydroxyifosfamide/aldoifosfamide, which may either decompose to isophosphoramide mustard (the active alkylating metabolite) and acrolein, or undergo oxidation to inactive carboxyifosfamide. The magnitude of each of these metabolic routes varies greatly between individual patients, while the influence (if any) of ifosfamide dose and schedule on its metabolism is unclear.[54,64] However, whilst side-chain dechlorethylation to chloroacetaldehyde (CAA) may account for up to 50% of ifosfamide metabolism in some patients, very little CAA results from cyclophosphamide metabolism.[65] Both unchanged ifosfamide and its metabolites are excreted in urine.

Mesna is a synthetic thiol compound capable of detoxifying potentially nephrotoxic ifosfamide metabolites including CAA and acrolein. After intravenous injection, mesna is rapidly oxidized in plasma to dimesna. Both mesna and dimesna are then filtered at the glomerulus, following which they are either reabsorbed by renal tubular cells or excreted in urine. The amount of mesna in tubular cells available to detoxify ifosfamide metabolites probably varies according to the mesna administration schedule, but the clinical importance of this is not clear.[34]

Chloroacetaldehyde induces an experimental Fanconi syndrome in the LLCPK1 cell line (a proximal renal tubular cell culture model) and in the isolated perfused rat kidney, possibly by inhibiting active transport and increasing the permeability of tubular cell membranes.[66,67] The production of CAA may explain why ifosfamide, but not cyclophosphamide, causes nephrotoxicity, while quantitative differences in systemic or renal formation and urinary excretion of CAA may lead to the variable severity observed in clinical practice.[68] Although mesna should be capable of detoxifying CAA, this protective effect may not be complete[69] and may be reduced further by the short plasma half-life and hence tubular 'transit time' of mesna.[34]

Alternative explanations for ifosfamide nephrotoxicity have been proposed, including the production of a toxic metabolite not formed by cyclophosphamide metabolism (e.g., isophosphoramide mustard).[63]