Risk Factors for Ifosfamide Nephrotoxicity in Children

Title: Risk Factors for Ifosfamide Nephrotoxicity in Children
Authors: Skinner, MRCP, Roderick; Pearson, FRCP, Andrew D.J.; English, MRCP, Martin W.; Price, RSCN, Lisa; Wyllie, RSCN, Ruth A.; Coulthard, FRCP, Malcolm G.; Craft, FRCP, Alan W.
Publisher: Lancet
Date Published: August 31, 1996
Reference Number: 16
Risk factors for long-term nephrotoxicity after ifosfamide for childhood cancers are not fully known. We have studied patient-related and treatment-related risk factors for chronic ifosfamide nephrotoxicity.

A group of 23 children who had received ifosfamide at age 2.1 - 16.2 years (median 6.9) for various cancers were assessed for nephrotoxicity, at 1-28 (2) months after the end of treatment, by renal function testing, laboratory values, and a grading score (none, mild, moderate, severe). No patient had received cisplatin or undergone nephrectomy. 13 children were reassessed at 10-26 (23) months; eight had died and two were not evaluable. The median total ifosfamide dose was 100.8 (9.0-160.4) g/m2 over a median of 15 courses every 3 weeks as a 48-72 h continuous intravenous infusion (in 22 cases), with mesna and hydration.

Glomerular filtration rate was below normal in ten (45%) of 22 evaluable children; their rate was 61-85 mL/min per 1.73 m2. proximal tubular toxicity led to hypophosphataemic rickets and/or renal tubular acidosis in six children, and distal tubular toxicity caused nephrogenic diabetes insipidus in one. Of the risk factors analysed by multiple regression, only total ifosfamide dose was associated with proximal tubular toxicity. Only two of ten evaluable patients who received under 100 g/m2 developed moderate nephrotoxicity, whereas six of ten who received over this dose had moderate or severe nephrotoxicity.

High total ifosfamide dose was the only risk factor we identified. Although inter-patient variability was high, cumulative doses of 100 g/m2 or higher should be avoided in children with cancer.
The following article is reproduced with permission and The Lancet retains copyright. No part of this article may be reproduced in any way without permission in writing from the publisher. For more information on The Lancet, please see http://www.thelancet.com.

Cyclophosphamide is used extensively in several childhood cancers but the isomer ifosfamide may, although the issue remains contentious, have advantages, especially in sarcomas1,2 but with the risk of substantial nephrotoxicity.3-11 The frequency of severe chronic nephrotoxicity in children ranges from 1.4 to 30%,12 but there are no reliable data on long-term outcome. Most well-documented severe cases have occurred in younger children or after higher cumulative doses,12 but some studies have highlighted the importance of previous cisplatin treatment or nephrectomy.9 However, there is no clear information about the relative importance of these risk factors. Therefore, we studied a cohort of children treated with ifosfamide for the frequency, nature, and severity of chronic nephrotoxicity and for the relevance of patient-related and treatment-related risk factors.

Patients and methods
23 children and adolescents treated with ifosfamide in Newcastle upon Tyne between July, 1988, and August, 1991, were studied. No child died from nephrotoxicity. Median (range) age at treatment was 6.9 (2.1 - 16.2) years. Eight had rhabdomyosarcoma, eight Ewing's sarcoma, six a soft-tissue sarcoma, and one non-Hodgkin's lymphoma. Before ifosfamide treatment, all children had normal serum creatinine and phosphate concentrations and alkaline phosphatase activity, and urinalysis; none had radiological evidence of renal involvement by tumour. The study received ethical approval. Informed oral consent was obtained from the parents and, where appropriate, from the patient.

Each patient underwent a baseline study at a median of 2 (1-28) months after completion of ifosfamide. Glomerular and proximal tubular and distal tubular function were assessed, and nephrotoxicity was graded with a standard protocol.10,12,13 Nephrotoxicity grading was re-evaluated at a median of 23 (10-26) months after the baseline study in 13 children; eight children had died, and two were not evaluable (one received cisplatin after the baseline study; the other received bicarbonate replacement treatment, which precludes accurate grading).

The median total ifosfamide dose was 100.8 (9.0-160.4) g/m2, over a median of 15 courses every 3 weeks, administered intravenously as a continuous (22 children) or 3 h (one child) infusion at 3 g/m2 per day for 2-3 days per course, with equidose mesna and hydration fluid. No child received cisplatin or carboplatin, and none underwent nephrectomy. A small area of kidney was included in the radiotherapy fields in two children. Other potentially nephrotoxic treatment (aminoglycosides, vancomycin, acyclovir, amphotericin B) was given to 21 children, for a median cumulative duration of 24 (5-98) days.

Glomerular filtration rate (GFR) was measured by 51Cr-labelled ethylenediaminetetraacetic acid- (51Cr-EDTA-) plasma clearance. This method has been used widely in children and validated by comparison with inulin renal clearance.13 To assess proximal tubular function, we measured electrolytes, creatinine, calcium, magnesium, phosphate, glucose, and aminoacids in corresponding blood and urine samples, with calculation of fractional excretions (FEs) and the renal tubular threshold

Figure: Scatterplot of renal tubular threshold for phosphate (Tmp/GFR) against
total dose of ifosfamide received
for phosphate (TMp/GFR) with standard formulas.13 A high FE indicates reduced tubular reabsorption, although (except for FEglucose this was considered abnormal only in the presence of a reduced blood concentration. TMp/GFR is low when phosphate reabsorption is impaired. We assessed distal tubular function by early-morning urine pH and osmolality. After grading (0-4) of GFR, TMp/GFR, serum bicarbonate, and early-morning urine osmolality, the individual scores were summed to give a nephrotoxicity score, potentially ranging from 0 to 16 (0=no toxicity, 1-3=mild, 4-7=moderate, 8-16=severe).12

Reference ranges were obtained from investigation of 105 otherwise healthy children (aged 0.1-16.6 years, 27 male) attending hospital for investigation of a previous urinary tract infection, in whom renal and urinary tract investigations and imaging were normal. Established age-related and sex-related reference ranges13 were used for serum creatinine, phosphate, and magnesium.

Multiple linear regression was used to evaluate total ifosfamide dose, age at start of treatment, sex, and duration (in days) of other potentially nephrotoxic drugs as predictors for nephrotoxicity measured by GFR, serum phosphate, serum bicarbonate, and TMp/GFR. Initially all predictor variables were included simultaneously in multiple regression to determine which had a significant influence on the indicators of nephrotoxicity. Only total dose was significant; therefore linear regression was performed with total dose only.

Renal function and grading of nephrotoxicity
GFR was low (61-85 mL/min per 1.73 m2) in ten (45%) of 22 children. Of 22 evaluable children, all had raised FEglucose despite normoglycaemia, seven (32%) a high FEphosphate, and nine (41%) a low TMp/GFR (0.23-0.79 mmol/L). No child had increased FE for sodium, potassium, calcium, or magnesium (evaluated in 14-22 children). Hypophosphataemia (0.34-0.99 mmol/L) was observed in ten of 23 children (43%), hyperchloraemia (110-115 mmol/L) in six of 22 (27%), and a reduced serum bicarbonate concentration (18-19 mmol/L) in three of 22 (14%). Of 23 children, three (13%) had a slightly reduced serum concentration of sodium, eight (35%) of potassium, three (13%) of magnesium, and two (9%) of calcium. Ten of 11 children evaluated had generalised aminoaciduria despite normal plasma concentrations. Early-morning urine was adequately acidified (pH W5.4) in only four (19%) of 21 patients, and concentrated (osmolality D600 mOsmol/kg) in 15 (68%) of 22 evaluable children.

Measure of nephrotoxicity Change associated
with increase in total
ifosfamide dose of 50 g/m2
dose (g/m2)
Age at
Sex Time off
Duration of
other potential nephrotoxins (days)
GFR (mL/min per 1.73 m2 -12.8 (3.0 to -28.5) -0.46
(-0.92 to 0.006)
(-2.62 to 3.29)
(-22.9 to 33.5)
(-4.98 to 0.54)
(-0.60 to 0.54)
Serum phosphate (mmol/L) -0.27 (-0.10 to -0.43) -0.005
(-0.01 to -0.00002)
(-0.027 to 0.038)
(-0.41 to 0.20)
(-0.03 to 0.03)
(-0.009 to 0.004)
Serum bicarbonate (mmol/L) -1.8 (-1.0 to -2.6) -0.04
(-0.07 to -0.02)
(-0.22 to 0.15)
(-1.48 to 1.88)
(-0.24 to 0.09)
(-0.030 to 0.037)
Tmp/GFR (mmol/L) -0.30 (-0.14 to -0.46) -0.008
(-0.014 to -0.003)
(-0.042 to 0.026)
(-0.50 to 0.11)
(-0.05 to 0.01)
(-0.006 to 0.006)
Column 2=linear regression, results expressed as mean (95% Cl); columns 3-7=multiple regression coefficients. Reference ranges: GFR=87-174, serum phosphate=0.90-1.75 (age-related), serum bicarbonate=20.0-26.0, Tmp/GFR=1.00-1.65.
Table: Linear and multiple regression analysis of severity of nephrotoxicity on potential risk factors

The nephrotoxicity score was evaluable in 20 children: seven had no, five mild, six moderate, and two severe nephrotoxicity. Five children had a nephrotoxicity score of 4 in at least one individual aspect of renal damage.

Ifosfamide was discontinued early due to nephrotoxicity in two patients (one with nephrogenic diabetes insipidus, one with acute renal failure); one of these received no further treatment, the other continued cyclophosphamide-based chemotherapy. Six children had severe chronic tubular toxicity over follow-up of 2 years (or until death from cancer): five developed hypophosphataemic rickets, five renal tubular acidosis, and four both. Renal tubular acidosis tended to develop insidiously during the first year after completion of ifosfamide. One child with hypophosphataemic rickets and renal tubular acidosis also developed severe nephrogenic diabetes insipidus; another also developed severe hypocalcaemia and tetany; four children had impaired growth with low height velocities, falling height standard-deviation-scores, and retarded bone ages.

Seven children were given oral phosphate supplementation, four received bicarbonate, and one received calcium and 1 a-hydroxyvitamin D3.

Multiple regression analysis demonstrated that only total dose had a significant effect on serum phosphate (p=0.045), serum bicarbonate (p=0.006), and TmpGFR (p=0.005). The other predictor variables had no independent effect. Linear regression revealed statistically significant relations between total ifosfamide dose and serum phosphate (p=0.004), serum bicarbonate (p<0.001), and Tmp/GFR (p=0.002) (figure), resulting in clinically important changes over the dose range used (table). In each case, higher ifosfamide dose was associated with greater toxicity. Only two of ten evaluable patients receiving under 100 g/m2 developed moderate nephrotoxicity, while six of ten receiving 100 g/m2 or more had moderate or severe toxicity.

Of 13 children with evaluable nephrotoxicity scores at baseline and follow-up, the nephrotoxicity grade (ie, no, mild, moderate, or severe) was increased in seven, unchanged in five, and decreased in one.

Our findings confirm previous reports of clinically important nephrotoxicity in children treated with ifosfamide.4,6,7 The widespread distribution and the persistence of damage, leading to proximal tubular, glomerular, and distal tubular impairment (in descending order of frequency), are both demonstrated clearly. The major clinical sequelae are due to proximal tubular damage and include hypophosphataemic rickets and renal tubular acidosis,5,7,9,14-18 which were seen in over 25% of patients in our study. Nephrogenic diabetes insipidus due to distal tubular toxicity is a rare but potentially severe complication.

All 23 patients showed some evidence of subclinical tubular nephrotoxicity with excessive glycosuria and/or aminoaciduria (suggesting a Fanconi syndrome), and increased excretion of low-molecular-weight proteins and renal tubular enzymes (data not shown).

Although common, glomerular impairment was usually mild (GFR>80 mL/min per 1.73 m2) unless clinical proximal tubular toxicity (low TMp/GFR or serum bicarbonate) was present. The close association between glomerular and proximal tubular impairment supports previous suggestions that glomerular dysfunction may be secondary to severe tubular damage.9,12

Our study provides convincing evidence of a correlation between total ifosfamide dose and the severity of proximal tubular nephrotoxicity. Despite previous suggestions that younger children are more susceptible,7,19 multiple regression revealed no significant independent effect for age. Although investigation of a larger number of patients might reveal some effect of age independent of dose, our results showed that dose was the more important risk factor. Our results do not suggest that other potentially nephrotoxic treatment was important in the development of chronic nephrotoxicity.

Previous nephrectomy5,9 and additional cisplatin treatment8,9 have also been suggested as risk factors for ifosfamide nephrotoxicity. However, although they may be important in individual patients, their relevance in the UK is limited because most children treated with ifosfamide neither undergo nephrectomy nor receive cisplatin. Therefore our results are more applicable to the majority of children, who are treated uniformly in well-defined protocols.

Although efficacy should be balanced carefully against toxicity, cumulative ifosfamide doses of 100 g/m2 or higher should be avoided. This recommendation may allow continued use of ifosfamide while reducing the severity and frequency of nephrotoxicity. However, the ubiquity of subclinical tubular toxicity, even in patients receiving low total doses, suggests that inter-individual differences in the nature and extent of ifosfamide metabolism may explain at least part of the variability in toxicity between patients. Also currently documented risk factors such as dose may not allow accurate prediction of renal outcome in all children. Quantification of subclinical toxicity occurring early in the course of treatment might permit more accurate prediction of incipient chronic renal damage,20 but further study of this approach is required.

Nephrotoxicity persisted over 2 years of follow-up, although a few individual children showed partial improvement or progressive deterioration in some aspects. Despite previous reports6,16 of partial recovery, the long-term prognosis of chronic toxicity should remain guarded, pending long-term follow-up studies.


  1. Zalupski M, Baker LH. Ifosfamide. J Natl Cancer Inst 1988; 80: 556-66.
  2. Shaw PJ, Eden OB. Ifosfamide in paediatric oncology: tried but not tested? Lancet 1990; i: 1022-23.
  3. Arndt C, Morgenstern B, Wilson D, Liedtke R, Miser J. Renal function in children and adolescents following 72 g/m2 of ifosfamide. Cancer Chemother Pharmacol 1994; 34: 431-33.
  4. Ashraf MS, Brady J, Breatnach F, Deasy PF, O'Meara A. Ifosfamide nephrotoxicity in paediatric cancer patients. Eur J Paediatr 1994; 153: 90-94.
  5. Burk CD, Restaino I, Kaplan BS, Meadows AT. Ifosfamide-induced renal tubular dysfunction and rickets in children with Wilms tumor. J Pediatr 1990; 117: 331-35.
  6. Caron HN, Abeling N, van Gennip A, de Kraker J, Voute PA. Hyperaminoaciduria identifies patients at risk of developing renal tubular toxicity associated with ifosfamide and platinate containing regimens. Med Pediatr Oncol 1992; 20: 42-47.
  7. DeSchepper J, Hachimi-Idrissi S, Verboven M, Piepsz A, Otten J. Renal function abnormalities after ifosfamide treatment in children. Acta Paediatr 1993; 82: 373-76.
  8. Pratt CB, Meyer WH, Jenkins JJ, et al. Ifosfamide, Fanconi's syndrome, and rickets. J Clin Oncol 1991; 9: 1495-99.
  9. Rossi R, Godde A, Kleinebrand A, et al. Unilateral nephrectomy and cisplatin as risk factors of ifosfamide-induced nephrotoxicity: analysis of 120 patients. J Clin Oncol 1994; 12: 159-65.
  10. Skinner R, Pearson ADJ, Price L, Coulthard MG, Craft AW. Nephrotoxicity after ifosfamide. Arch Dis Child 1990; 65: 732-38.
  11. Smeitink J, Verreussel M, Schroder C, Lippens R. Nephrotoxicity associated with ifosfamide. Eur J Pediatr 1988; 148: 164-66.
  12. Skinner R, Sharkey IM, Pearson ADJ, Craft AW. Ifosfamide, mesna and nephrotoxicity in children. J Clin Oncol 1993; 11: 173-90.
  13. Skinner R, Pearson ADJ, Coulthard MG, et al. Assessment of chemotherapy-associated nephrotoxicity in children with cancer. Cancer Chemother Pharmacol 1991; 28: 81-92.
  14. Moncrieff M, Fott A. Fanconi syndrome after ifosfamide. Cancer Chemother Pharmacol 1989; 23: 121-22.
  15. Skinner R, Pearson ADJ, Price L, Cunningham K, Craft AW. Hypophosphataemic rickets after ifosfamide treatment in children. BMJ 1989; 298: 1560-61.
  16. Van Gool S, Brock P, Wijndaele G, et al. Reversible hypophosphatemic rickets following ifosfamide treatment. Med Pediatr Oncol 1992; 20: 254-57.
  17. Heney D, Lewis IJ, Bailey CC. Acute ifosfamide-induced tubular toxicity. Lancet 1989; ii: 103-04.
  18. Newbury-Ecob RA, Noble VW, Barbor RRH, Ifosfamide-induced Fanconi syndrome. Lancet 1989; i: 1328.
  19. Skinner R, Pearson ADJ, Price L, Coulthard MG, Craft AW. The influence of age on nephrotoxicity following chemotherapy in children. Br J Cancer 1992; 66 (suppl XVIII); S30-35.
  20. Skinner R, Pearson ADJ, Price L, et al. Ifosfamide nephrotoxicity in children: early detection and prediction of severity. Med Pediatr Oncol 1994; 23: 178 (abstr).

This translation by the NDI Foundation is to assist the lay reader. To provide a clear, accessible interpretation of the original article, we eliminated or simplified some technical detail and complicated scientific language. We concentrated our translation on those aspects of the article dealing directly with NDI. The NDI Foundation thanks the researchers for their work toward understanding and more effectively treating this disorder.
© Copyright NDI Foundation 2007 (JC)

This article describes a study undertaken by the authors to determine the risk factors for nephrotoxicity in children who have been treated with the anti-cancer drug ifosfamide. The authors studied a group of 23 children, ranging in age from two to 16 years. All had been treated with ifosfamide for their various cancers. The children underwent an initial evaluation approximately two months after completing their ifosfamide treatment in order to develop baseline measurements of their kidney function. Functions measured included glomerular filtration rate (GFR), proximal tubular function, fractional excretions, the renal tubular threshold for phosphate (Tm/GFR), and distal tubular function. The children were also measured for total ifosfamide dose, age at start of treatment, sex, duration of use of other potentially nephrotoxic drugs as predictors for nephrotoxicity measured by GFR. In addition, the researchers developed a grading score to measure the degree of nephrotoxicity, ranging from none, to mild, to moderate, to severe.

The children were again evaluated at a median of 23 months after the initial evaluation. In the interim, eight of the children had died and two could not be evaluated as they had received treatment which interfered with accurate measurement. This left a total of 13 children for the second phase of the study.

In the first phase of the study, of all the predictor variables analyzed, only total dose of ifosfamide showed a significant correlation with the nephrotoxicity score developed by the authors. In each case, higher infosfamide dose was associated with greater toxicity. Of the 20 children evaluable, seven had no, five had mild, six had moderate and two had severe nephrotoxicity. Five children had moderate nephrotoxicity in at least one individual aspect of renal damage. Six children had severe chronic tubular toxicity over follow-up of two years (or until death from cancer); five developed hypophosphataemic rickets, five renal tubular acidosis, and four both. One child with hypophosphataemic rickets and renal tubular acidosis also developed severe nephrogenic diabetes insipidus, another also developed severe hypocalcaemia and tetany, and four children had impaired growth.

The authors' study confirmed previous reports of nephrotoxicity in children treated with ifosfamide, clearly showing the widespread distribution and persistence of kidney damage. All 23 children showed some evidence of subclinical nephrotoxicity. Further, their study provides evidence of a correlation between total ifosfamide dose and the severity of kidney damage. They recommended avoiding cumulative ifosfamide doses of 100 g/m2 or higher as their study showed only two of the children in their study receiving under 100 g/m2 developed moderate nephrotoxicity while six of ten receiving 100 g/m2 or more had moderate or severe toxicity. The nephrotoxicity persisted over two years of follow-up, and the authors recommend long-term follow-up studies.