Quantification of busulfan in saliva and plasma in haematopoietic stem cell transplantation in children

Background and objective: Busulfan pharmacokinetic studies suggest that an individual dosing strategy may be necessary to optimise systemic exposure in order to decrease toxicity and improve outcome in haematopoietic stem cell transplantation. Therapeutic and toxic effects of the busulfan/cyclophosphamide regimen have been related to the area under the busulfan plasma concentration-time curve. Because of practical limitations in obtaining blood from children, saliva was evaluated as an alternative matrix for therapeutic drug monitoring, offering the advantages of a non-invasive, rapid and easy sampling procedure. Another objective was to evaluate an easy and robust liquid chromatographytandem mass spectrometry method for plasma and saliva busulfan determination. Methods: An online extraction cartridge with column-switching technique, analytical liquid chromatography over a Chromolith RP 18e column, and tandem mass spectrometry were used to quantify busulfan concentrations in matched plasma and saliva samples. The study population consisted of ten patients, aged 1.3-19 years (median age 11.8 years, seven females, three males), undergoing haematopoietic stem cell transplantation. All patients received busulfan 0.8-1.3 mg/kg orally every 6 hours for a total of 16 doses, followed by two doses of cyclophosphamide (60 mg/kg/day). Results: The lowest limit of detection was 2 mu g/L and the lower limit of quantification was 10 mu g/L. Only 100 mu L of plasma/saliva was needed. The mean recoveries (SD) of busulfan were 97.2% (2.7) in plasma and 100.4% (1.3) in saliva. Intra- and inter-assay imprecision was 2-3% and 2-4% for plasma, and 1-2% and 2-4% for saliva (concentration range 30-1500 mu g/L). The bias was < 4% for both plasma and saliva. The correlation between the busulfan concentration in plasma and saliva was highly significant (r = 0.958; p < 0.0001; saliva/plasma ratio = 1.09 +/- 0.04; n = 69 sample pairs). The apparent plasma clearance was slightly higher than the apparent saliva clearance (202 +/- 31 mL/h/kg vs 189 +/- 28 mL/h/kg; p = 0.001). The mean elimination half-life was found to be 2.31 +/- 0.46 hours for plasma and 2.30 +/- 0.36 hours for saliva; these were not significantly different (p = 0.83). Conclusion: The present study demonstrated that busulfan analysis in saliva could be a valuable and reliable alternative to plasma analysis.