Busulfan Target Concentration Intervention Audit and Model Evaluation

Background: Intravenous busulfan is used prior to haematopoetic stem cell transplantation. All patients undergoing this procedure in New Zealand since 2012 have been treated in Auckland. A web based target concentration intervention tool (NextDose) has been used by the clinical laboratory. Demographic, dosing and concentration details and desired target concentration or target area under the curve (AUC) are requested from clinical staff. This information is entered by laboratory scientists with busulfan concentrations. A report is generated with a proposed dose calculated using both AUC and Bayesian methods and sent to clinical staff.
Objectives: 1. To conduct an audit of NextDose in a realistic clinical setting.
2. To evaluate the effectiveness of 3 pharmacokinetic models for Bayesian dose prediction
Methods: The NextDose database was used to identify all patients who had a report generated by laboratory staff. Reports were audited by an experienced pharmacometrician. The demographic, dosing and concentration information was used to evaluate two published pharmacokinetic models (1, 2). The NextDose tool used an adaptation of (2) which was expected to provide better predictions for infants. The 3 models were evaluated using the NONMEM objective function, visual external evaluation and prediction errors for clearance derived from the first blood sampling occasion. The ‘true’ clearance was obtained by fitting data from all blood sampling occasions.
Results: Audit revealed one serious dose prediction error attributable to using concentration units (mg/L) with a 24 h AUC (umol/L*min) value, 8 reports of appropriate doses with inappropriate dosing intervals, and minor missing data (lack of clinical consultant details) in 27 reports. Model evaluation favoured the McCune 2014 model in 5 out of 8 evaluation tests.
Conclusion: The NextDose tool can be used effectively in a clinical hospital environment without specialist pharmacometric intervention. The McCune 2014 pharmacokinetic model is preferable for busulfan target concentration intervention. Future versions of NextDose will provide better warnings for users when apparently inappropriate values are entered and implement the McCune 2014 model.

1. McCune JS, Bemer MJ, Barrett JS, Scott Baker K, Gamis AS, Holford NHG. Busulfan in Infant to Adult Hematopoietic Cell Transplant Recipients: A Population Pharmacokinetic Model for Initial and Bayesian Dose Personalization. Clin Cancer Res. 2014;20(3):754-63.
2. Booth BP, Rahman A, Dagher R, Griebel D, Lennon S, Fuller D, et al. Population pharmacokinetic-based dosing of intravenous busulfan in pediatric patients. J Clin Pharmacol. 2007;47(1):101-11.

busulfan-TCI-audit-and-model-evaluation pdf

Nick Holford