Ifosfamide and its metabolites were the kind gift of Dr Jürg Pohl, ASTA Medica, Germany. All other reagents were obtained from Sigma, Poole, Dorset, UK, except where otherwise indicated, and were of an appropriate analytical grade.
In all, 19 patients (five female) aged between 2 and 19 years were recruited to the study. In each case, informed written consent was obtained from the patient and/or parent, as appropriate, and the study was approved by the ethical committee of the Newcastle Hospitals Trust. Details of patient characteristics, including the protocol that the patients received, IFO dose and schedule are given in . Of these patients, eight were studied after a second dose of IFO in order to assess the degree of intersubject variability in DNA damage. Treatment was the same on the two courses studied.
Ifosfamide was administered either as a continuous infusion over 3 days (n
=8) or as a 3
h infusion every 24
h for 3 days (n
=5). Three patients were treated with just 2 consecutive days of fractionated dosing and three received 3 consecutive days of IFO, but with a 1
h infusion time. No attempt was made to control concurrent medication, but a record was kept of potential interacting drugs ().
For those patients receiving a 3
h infusion of IFO, blood samples were taken for the determination of IFO and its metabolites before administration, then at 1.5 and 3
h during the infusion and at 0.5, 1, 2, 4, 6 12 and 24
h after the end of infusion. Patients with a 1
h infusion had a pretreatment and end of infusion samples taken together with the postinfusion samples as for a 3
h infusion. Sampling was repeated for each day of drug administration. For those patients receiving a continuous infusion of IFO, sample times were pretreatment, 3, 6, 12, 18, 24, 36, 48, 60 and 72
h during the infusion. Samples were also taken at 2, 4, 6, 12, 18 and 24
h after the end of the infusion.
Plasma was separated by centrifugation at 1500g
and stored frozen at −20°C until subsequent analysis. Ifosfamide, 2DCI, 3DCI and carboxyifosfamide were determined by a validated high-performance thin-layer chromatography method (Boddy and Idle, 1992
). A minor metabolite, 4-ketoifosfamide, was detected in some patients, but could not be consistently quantified. Noncompartmental methods were used to estimate pharmacokinetic parameters for parent drug as described previously (Boddy et al, 1995b
), and the linear trapezoidal rule was used to calculate area under the curve (AUC) for parent drug and metabolites. Area under the curve values for the continuous infusion data relate to all 3 days of administration, including data after the end of infusion.
In addition to samples collected for pharmacokinetic analysis, whole blood samples were taken prior to treatment, and at 3 and 24
h after each dose. A final sample was taken prior to the next course of treatment, that is, 3 weeks after the start of the initial course. The sampling times were based on previous data (Johnstone et al, 2000
). Patients who were studied on more than one course of treatment had a second set of blood samples taken for pharmacokinetic and COMET analysis.
The COMET assay was performed as described previously (Johnstone et al, 2000
). Peripheral blood lymphocytes were isolated at the time of blood sampling using Lymphoprep. Storage of PBLs was at −80°C, and analysis was always completed within 1 month of sampling. Stability studies indicated that there was no change in the measurement of DNA damage using the COMET assay during 1 month of sample storage. The degree of DNA damage in a given patient sample or in vitro
investigation was quantified by measuring the distributed tail moment (TM) in a minimum of 30 cells per slide. Images were analysed using the COMOS image analysis software (Biorad Laboratories, Hertfordshire, UK). The TM data were log normalised and a logarithmic mean and standard deviation (s.d.) determined. The number of highly damaged (HD) cells (log TM more than 2 × s.d. from the mean) per sample was calculated. To estimate the overall degree of DNA damage caused during a cycle of chemotherapy, the AUC of percentage of HD cells (AUCHD) per sample plotted against time was calculated. An internal standard of CCRF-CEM cells that had received a standard dose of radiation was included as a control in each COMET experiment.
In order to provide some background to the clinical studies, the effect of IFO and its active metabolites on the formation of COMET-detectable DNA damage in vitro
was investigated. CCRF-CEM leukaemia cells were exposed to varying concentrations of the precursor of the active 4-hydroxy metabolite (4-hydroperoxyifosfamide) or IPM for up to 72
h, although these unstable metabolites break down quickly in the medium. Growth inhibition was determined by counting the cells using a Coulter counter and was compared to the degree of DNA damage, as measured by the COMET assay. AUCHD was determined by measuring the percentage of HD cells at times 1, 24, 48 and 72
h after the start of drug exposure.
Statistical analysis was performed using Graphpad Prism, version 3 (Graphpad, San Diego USA). Differences between treatments were analysed using one-way analysis of variance and relationships between pharmacokinetic parameters and DNA damage measurements were analysed by linear regression.