Measurement of 4-hydroxylation of ifosfamide in human liver microsomes using the estimation of free and protein-bound acrolein and codetermination of keto- and carboxyifosfamide

Abstract

Purpose. The aim of the present study was to determine the turnover (4-hydroxylation and N-dechloroethylation) of ifosfamide in a total of 25 human liver microsomal preparations in which the codetermination of keto- and carboxyifosfamide as well as the calculation of free and protein-bound acrolein was carried out for the first time.

Methods. The 4-hydroxylation of ifosfamide was estimated by using acrolein (free and protein-bound) and a newly developed procedure involving the codetermination of keto- and carboxyifosfamide (LC/MS). The ifosfamide N-dechloroethylation was determined as the sum of 2- and 3-dechloroethylifosfamide (LC/MS).

Results. Using the usual estimation of liberated free acrolein in 25 human liver microsomal preparations, the 4-hydroxylation of ifosfamide amounted to 0.28±0.16 nmol/min · nmol_P450. However, after calculating the 4-hydroxylation as the sum of free and protein-bound acrolein and keto- and carboxyifosfamide, a ninefold higher activity (2.40±0.73 nmol/min · nmol_P450) was found. The percentage of the inactive metabolites keto- (25/25) and carboxyifosfamide (5/25) in the 4-hydroxylation amounted to only 0.79–5.25% (mean 2.90%). The ifosfamide N-dechloroethylation (mean 0.21±0.11 nmol/min · nmol_P450) determined as the sum of 2- and 3-dechloroethylifosfamide was estimated as 8.3±4.3% of the total ifosfamide turnover. The application of the relative substrate-activity factor (RSF)-approach and the calculation of the contribution of various isoforms in the ifosfamide 4-hydroxylation yielded the following results: CYP 3A4: 58±31%, CYP 2A6: 25±15%, and CYP 2C9: 5±2% of the total measured 4-hydroxylation. A correlation between 4-hydroxylation and the N-dechloroethylation rates of ifosfamide and the activities of isoenzymes indicates the involvement of both CYP 3A4 (P=0.026) and CYP 2C9 (P=0.012) in the 4-hydroxylation reaction and of CYP 3A4 (P<0.01) in the N-dechloroethylation reaction.

Conclusions. The estimation of protein-bound acrolein should be included in the calculation of the ifosfamide 4-hydroxylation besides liberated free acrolein. Because of the small amounts of the inactive metabolites keto- and carboxyifosfamide, the exclusive determination of acrolein only (free and protein-bound) seems to suffice for the calculation of total ifosfamide hydroxylation. Using this method the hepatic in vitro turnover of ifosfamide was estimated as 92% for 4-hydroxylation (CYP 3A4 and CYP 2A6 mediated) and 8% for N-dechloroethylation (CYP 3A4 mediated), and in this way, a relative overestimation of the N-dechloroethylation of ifosfamide on the whole metabolism is avoided.