2007 - København - Denmark

PAGE 2007: Applications- Oncology
Dolors Soy Muner

Population Pharmacokinetic Modeling Of Busulfan In Patients Undergoing Autologous Stem Cell Transplantation (ASCT) For Multiple Myeloma

Soy D1, Clopés A2, Farré R3, Mangues MA3

1Pharmacy Service. Hospital Clínic Barcelona, 2Pharmacy Service. Institut Català d’Oncologia, 3Pharmacy Service. Hospital Santa Creu i Sant Pau. Barcelona, Spain.

Objectives: Steady-state busulfan (BU) concentrations and area under the curve (0-6h) above 900 ng/mL and 5400 ng*h/mL, respectively, have been related with higher efficacy and acceptable toxicity. However, a great interindividual variability (IIV) in oral BU pharmacokinetics (PK) has also been described. The goals of the study were: (i) to assess the pharmacokinetics of busulfan in ASCT patients and (ii) to look for relationships between covariates and BU-PK.

Methods: Twenty-three adult patients treated at the Hematology Service of Hospital Santa Creu i Sant Pau for advanced stage mieloma multiple, in first response and undergoing an ASCT were included. Initial BU dosage was 0.75 mg/kg/6h x 16 doses or 1 mg/kg/8h x 12 doses (total dose: 12 mg/kg). Blood samples were collected at 0.5,1,3,4 and 6h in the first dose and BU concentrations were quantified by HPLC. Bu pharmacokinetic parameters were estimated after first dose and subsequent doses were modified as necessary to achieve target exposure. Sparse samples (trough) along treatment were also available in some individuals. Later, all the data were analyzed on the basis of the population (pop) approach (NONMEM-VI). Demographic, clinical and biochemistry data were collected for each patient and tested as covariates. A validation was conducted in new individuals (N=21) by predicting and comparing the concentrations at the same time (IPRED) than those observed (OBS). Bias (MPE) and precision (MAPE) were computed.  Statistics were performed using S-Plus5.

Results: The basic pop-PK model selected was a one-compartment model with first-order absorption and elimination. The FOCE with INTERACTION method was applied for parameter estimation. Data did not support the estimation of Ka/F. Lognormal IIV in clearance (CL/F) and volume of distribution (V/F) was added resulting in values of 21 and 51%, respectively. Interoccasion variability was only retained for CL/F (19%). Measurement error variability follows a proportional normal distribution (16%). A model based on weight and age was identified as appropriate to describe part of the CL/F variability. Sex and weight significantly influenced V/F. Including these covariates in the final model a 4% and 7% reduction in unexplained IIV was found for CL/F and V/F, respectively. Results from validation showed there was good correlation between IPRED and OBS in the validation dataset. Bias and precision were 11.5% and 25%, respectively.

Conclusions: BU pop-PK parameters were consistent with those previously published. Body weight, age and sex were important determinants on CL/F and V/F. Results from this study could be used to optimize the initial and maintenance oral BU dose regimens in daily practice.

Reference: PAGE 16 (2007) Abstr 1114 [www.page-meeting.org/?abstract=1114]
Poster: Applications- Oncology
Click to open PDF poster/presentation (click to open)