A Simple Dosing Scheme for Intravenous Busulfan Based on Retrospective Population Pharmacokinetic Analysis in Korean Patients

Abstract

Busulfan is an antineoplastic agent with a narrow therapeutic window. A post-hoc population pharmacokinetic analysis of a prospective randomized trial for comparison of four-times daily versus once-daily intravenous busulfan was carried out to search for predictive factors of intravenous busulfan (iBu) pharmacokinetics (PK). In this study the population PK of iBu was characterized to provide suitable dosing recommendations. Patients were randomized to receive iBu, either as 0.8 mg/kg every 6 h or 3.2 mg/kg daily over 4 days prior to hematopoietic stem cell transplantation. In total, 295 busulfan concentrations were analyzed with NONMEM. Actual body weight and sex were significant covariates affecting the PK of iBu. Sixty patients were included in the study (all Korean; 23 women, 37 men; mean [SD] age, 36.5 [10.9] years; weight, 66.5 [11.3] kg). Population estimates for a typical patient weighing 65 kg were: clearance (CL) 7.6 l/h and volume of distribution (Vd) 32.2 l for men and 29.1 L for women. Inter-individual random variabilities of CL and Vd were 16% and 9%. Based on a CL estimate from the final PK model, a simple dosage scheme to achieve the target AUC0-inf (defined as median AUC0-inf with a once-daily dosage) of 26.18 mg/l·hr, was proposed: 24.79·ABW0.5 mg q24h, where ABW represents the actual body weight in kilograms. The dosing scheme reduced the unexplained interindividual variabilities of CL and Vd of iBu with ABW being a significant covariate affecting clearance of iBU. We propose a new simple dosing scheme for iBu based only on ABW.

Population pharmacokinetics analysis of intravenous busulfan in Chinese patients undergoing hematopoietic stem cell transplantation

“…A plausible explanation is that physiological function was relatively well controlled in the patients undergoing HSCT, such variation in these covariates was in a narrow range which can't reflect its impact in the final population model. 43 Elevated serum creatine or low creatinine clearance was not related to CL and Vd after intravenous busulfan, as expected, which is consistent with the previous studies, 26,28,29,31,34 since renal elimination of busulfan is limited. 41 Antifungal prophylaxis is commonly used in HSCT patients.…”

“…Clearance was higher in Chinese patients (16.8 L/h for a person 70 kg) compared with Caucasian (10.46 L/h reported by Russell) and Korean patients (8.02 L/h reported by Ryu), which is inconsistent with our result. The inter‐individual variability values for CL and Vd following intravenous busulfan in adult patients were reported as 16% and 13% by Nguyen et al., and 14.7% and 25.6% by Boyoon Chio et al., 16% and 9% by Sangmin Choe et al . While those for CL and Vd following intravenous busulfan treatment of paediatric patients were reported as 23% and 11% by Booth et al .…”

“…The pharmacokinetic parameters were estimated with NONMEM subroutines ADVAN1 TRANS2, using the First Order Conditional Estimation with interaction (FOCE‐I) method. One‐compartment structural model with constant rate infusion was used as a basic pharmacokinetic model for busulfan . Initial pharmacokinetics estimates for NONMEM modelling were from the document by Takama et al …”

“…Several papers applying population pharmacokinetic analysis have been published using data from Caucasian, [19][20][21][22][23][24][25][26][27][28] Japanese, 29 and Korean 30,31 populations, but no data in Chinese HSCT population has been reported. A study on the pharmacokinetics (PK) of once-daily intravenous Bu in a Chinese population showed there were differences in PK parameters among Chinese, Caucasian and Korean patients.…”

“…The pharmacokinetic parameters were estimated with NONMEM subroutines ADVAN1 TRANS2, using the First Order Conditional Estimation with interaction (FOCE-I) method. One-compartment structural model with constant rate infusion was used as a basic pharmacokinetic model for busulfan 26,31. Initial pharmacokinetics estimates for NONMEM modelling were from the document by Takama et al34 The following combinations of drugs, demographic and physio-pathological data were considered in the analysis: age, gender, ABW, AIBW, BMI, serum albumin (ALB), creatinine, creatinine clearance (Clcr), alanine transaminase (ALT), aspartate transaminase (AST), alkaline phosphatase (ALP), gamma glutamyl transferase (GGT), lactic dehydrogenase (LDH), total bilirubin (TBTL), diagnosis, concomitant antifungal treatment (Micafungin sodium or voriconazole), concomitant 5-HT3 antiemetic treatment (tropisetron or palonosetron), concomitant immunosuppressive treatment (Cyclosporine A).…”