A Simplified Limited-Sampling Strategy for Accurate Prediction of Doxorubicin and Doxorubicinol Exposure

BACKGROUND: Cumulative dose-based management of doxorubicin (DOX) does not necessarily reflect the true systemic exposure to DOX or its cardiotoxic metabolite, doxorubicinol (DOXol). The aim of this study was to develop and validate a regression-based limited-sampling strategy to estimate the areas under the concentration-time curves (AUCs) of DOX and DOXol in patients with diffuse large B-cell lymphoma receiving CHOP chemotherapy. METHODS: Forty-five patients receiving CHOP therapy were enrolled in this study. Plasma samples were collected at 7 time points between 1.5 and 25.5 hours post-DOX administration. AUCs were calculated using noncompartmental analysis. Linear regression models were constructed using plasma concentrations from one or 2 time points. Predictive performance was assessed using 5-fold cross-validation and evaluated by mean error, mean absolute error, and root mean square error. RESULTS: All sampling time points showed significant correlations with the full AUCs for DOX and DOXol, with the strongest being observed at 13.5 hours. Two-point models incorporating 1.5 and 13.5 hours concentrations yielded the highest predictive accuracy (R2 = 0.968 for DOX and 0.993 for DOXol), with low %MAE and %RMSE values. These models enabled simultaneous, accurate AUC estimation for both analytes with minimal sampling. CONCLUSIONS: Regression-based limited-sampling strategy using one or 2 sampling time points is a feasible and accurate method for estimating the AUCs of DOX and DOXol in clinical settings. This strategy enables the practical assessment of cumulative exposure beyond conventional cumulative dose-based management and may support individualized dosing of DOX.