Clinical pharmacokinetics of irinotecan and its metabolites in relation with diarrhea

Our objective was to build population pharmacokinetic models that describe plasma concentrations of irinotecan (CPT-11) and its metabolites 7-ethyl-10-hydroxycamptothecin (SN-38) and SN-38 glucuronide (SN-38G) and to investigate the pharmacokinetic-pharmacodynamic relationships between drug exposure and diarrhea, the major dose-limiting toxicity.

Data were obtained from 109 patients (65 men and 44 women) who received 1.5-hour (range, 0.75- to 2.25-hour) intravenous infusions of irinotecan at doses that ranged from 100 to 350 mg/m(2); 44 patients had a second course. The population pharmacokinetic models were developed to describe plasma concentration-time profiles. The area under the concentration-time curve from time zero to 60 hours [AUC (0-60)] was used as a measure of drug exposure to model the probabilities of diarrhea with use of a logistic regression model.

A 3-compartment pharmacokinetic model best described the disposition of irinotecan, whereas SN-38 and SN-38G showed 2-compartmental characteristics. The population estimate of clearance for irinotecan was 31.6 L/h, and the volume of distribution at steady-state (V(SS)) was 263 L. The clearance divided by formation fraction (F(m)) was 712 L/h and 66.8 L/h for SN-38 and SN-38G, respectively. The V(SS)/F(m) was 72,000 L for SN-38 and 85.4 L for SN-38G. The frequencies of diarrhea scores in this study were 46% (grade 0), 28% (grade 1), 20% (grade 2), 4% (grade 3), and 2% (grade 4). Significant correlations between AUC(0-60) and diarrhea scores were found for irinotecan (P <.05) and SN-38G (P <.01) but not for SN-38 or the biliary index.

In this population analysis, irinotecan and SN-38G AUC values were appropriate predictors of the risk for diarrhea, and SN-38G AUC showed the stronger relationship of the two. The developed population models may be useful in further clinical development of this agent.