Population pharmacokinetics of the BEACOPP polychemotherapy regimen in Hodgkin's lymphoma and its effect on myelotoxicity.

BACKGROUND AND OBJECTIVE: The BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine and prednisone) chemotherapy regimen for the treatment of advanced Hodgkin's lymphoma has a superior outcome, but its toxicity (mainly haematotoxicity) is pronounced and highly variable. The present study was conducted to address the role of pharmacokinetics in individual toxicity. STUDY DESIGN: Three plasma samples and a 24-hour urine collection for day 1 of the first three cycles of chemotherapy were analysed in 30 patients, and the pharmacokinetic parameters of the respective drugs were estimated by population pharmacokinetic methods (nonlinear mixed-effects model [NONMEM] software). Demographic data, doses and durations of infusion were also recorded. The effect of these parameters on platelet counts was estimated by analysis of covariance using a general linear model. RESULTS: The pharmacokinetic parameters and respective covariates were similar to the published data. The body surface area, peak concentrations of etoposide, urinary recovery of dechloroethylcyclophosphamide (formed by cytochrome P450 [CYP] 3A4) relative to the cyclophosphamide dose and number of cycles had a significant effect on toxicity. These factors explained 37% of the interindividual variability in the change in platelet counts from day 1 to day 8 of each cycle. CONCLUSION: The results show that the individual pharmacokinetics of BEACOPP drugs are an important link between dosage and toxicity. Accordingly, individualisation of treatment based on pharmacokinetics may result in more uniform toxicity. Individualisation may also allow escalation of the mean dose, which is probably related to better efficacy. As a consequence of the present study, infusion rates should be standardised, and the potential of a dose reduction in the first cycle and of CYP3A4 phenotyping should be addressed in clinical studies.