BACKGROUND: Noncompartmental analysis (NCA) and model-based method (MBM) can be used to estimate the cumulative area under the concentration-time curve (AUCcum) during therapeutic drug monitoring. Understanding predictive differences among these techniques should assist in switching between them and interpreting their results. The aim of this study was to compare busulfan AUCcum prediction based on NCA technique (Kinetica) and MBM (NextDose) applied to the same concentration-time data from pediatric hematopoietic stem cell transplant (HSCT) recipients.

METHODS: Data on busulfan therapy administered once daily through intermittent infusion were obtained from 4 hospitals in Australia and New Zealand. Busulfan concentrations were measured at 3, 3.25, 4, 5, 6, and 8 hours after infusion initiation over 4 treatment days. Information on busulfan dose, pharmacokinetic profile, and patient covariate factors (if required) were supplied sequentially to Kinetica and NextDose to generate NCA-based and MBM-based predictions of busulfan exposure; differences in AUCcum estimates at the end of the treatment course were compared using Bland-Altman plots, box plots, and Wilcoxon signed-rank sum test.

RESULTS: Data from 90 HSCT recipients (2131 busulfan samples) were included. The median patient age and weight were 4.3 years and 17.0 kg, respectively. Median AUCcum estimated based on NCA and MBM were 78.0 mg.h/L [range: 51.7-107.0] and 85.5 mg.h/L [range: 60.6-120.8], respectively, with statistically significant difference in AUCcum values (P < 0.001). The mean percentage difference in AUCcum values between the 2 different methods suggested that if the AUCcum target using MBMs (NextDose) is 78-101 mg.h/L, then an equivalent target using NCA (Kinetica) would be reduced by 9.1%.

CONCLUSIONS: When switching between NCA and MBMs to estimate busulfan AUCcum in pediatric HSCT recipients, a change in the target AUCcum is likely required to maintain a similar drug exposure during therapeutic drug monitoring.