Pharmacokinetics of low-dose doxorubicin and metabolites in patients with AIDS-related Kaposi sarcoma

Purpose: Systemic chemotherapy is the treatment of choice for AIDS-related advanced Kaposi sarcoma. One principal schedule combines adriamycin (doxorubicin), bleomycin, and vincristine (ABV). We analysed the plasma concentrations of low-dose doxorubicin (Dx) and its metabolites doxorubicinol, 7-deoxydoxorubicinone, doxorubicinone, doxorubicinolone, and 7-deoxydoxorubicinolone in AIDS-patients to define patient-group and dose-specific pharmacokinetic parameters.

Materials and methods: A previously described high-performance liquid chromatographic (HPLC) method and a population approach with non-linear mixed effects modelling (NONMEM) were used for analysis and subsequent modelling of the time-concentration data of low-dose Dx and metabolites in seven patients with AIDS-related advanced Kaposi sarcoma. Patients received Dx 20 mg m(-2), bleomycin 15 U m(-2) and vincristine 2 mg as a 30-min intravenous infusion each. Blood samples were collected up to 72 h after the start of Dx treatment. WinNonlin software version 4.1 was used for non-compartmental analysis and NONMEM software version V for compartmental analysis. Covariate analysis was performed for various clinical and laboratory parameters.

Results: Non-compartmental analysis yielded an area under the plasma concentration-time curve (AUC) for Dx of 566 mug h L(-1), a maximum plasma concentration (c(max)) of 599 mug L(-1) and an elimination half-life (t(1/2)) of 30.8 h. Compartmental analysis resulted in a two-compartment model with first-order elimination, which best fitted the concentration-time data. Model estimate for Dx clearance was 61.8 L h(-1), for intercompartmental clearance (Q) 112 L h(-1), for the volume of the central compartment (V(1)) 23.3 L, and for the volume of the peripheral compartment (V(2)) 1,130 L. Metabolite data could adequately be estimated by NONMEM using single-compartment models. Graphical plots of residuals versus time for all metabolites yielded no evidence of non-linear pharmacokinetic behaviour. Laboratory parameters of liver and renal function were all in the normal range and their inclusion in the pharmacokinetic model did not improve data fit. A final jack-knife analysis was performed.

Conclusions: Concentration-time data for low-dose Dx and metabolites in the ABV-regimen are best described by a two-compartment model with first-order elimination. The results confirm that the aglycones doxorubicinone, 7-deoxydoxorubicinone, and doxorubicinolone can be reliably detected in the studied patient group and implemented into a common metabolic model. Model estimates suggest that pharmacokinetic parameters are similar for low-dose Dx and higher-dosed Dx. As the role of the aglycones is still poorly understood, despite their potential clinical relevance, their analysis should be implemented in future pharmacokinetic and pharmacodynamic studies of Dx.