Bioanalysis and pharmacokinetic analysis
Plasma was obtained by centrifugation at 3,000 rpm for 10 min and stored in polypropylene tubes at -70 ºC until concentrations were determined. Plasma concentrations of proparnolol and 4-OH-propranolol were determined with a validated liquid chromatography-tandem mass spectrometry (LC-MS/MS) method. The lower limit of quantification (LLOQ) for propranolol was 0.4 ng/mL with a linear calibration range of 0.4-100 ng/mL and that for 4-OH-propranolol was 0.3 ng/mL with a linear calibration range of 0.3-75 ng/mL. Intra- and inter-day accuracies for propranolol were 90.5-104.0% and those for 4-OH-propranolol were 98.8-101.3%; intra- and inter-day precisions varied with <9.4 CV% for propranolol and <6.67% for 4-OH-propranolol. The peak plasma concentration (Cmax) and time to Cmax (i.e. Tmax) were directly obtained from the observed values. The area under the concentration-time curve (AUC) from time 0 to the last measurable time (AUClast) was calculated using the trapezoidal rule. The terminal half-life and AUC extrapolated to infinity (AUCinf) were not determined because the pharmacokinetic blood samplings were not conducted until sufficient time after drug administration to evaluate elimination phase precisely. Non-compartmental analyses were performed using PK solutions pharmacokinetic software (SummitPK.com).
Statistical analysisDemographic characteristics and pharmacokinetic parameters were compared among the patient groups using the Kruskal-Wallis test. Post-hoc analysis was performed if there was a significant difference in the comparison of pharmacokinetic parameters. A multivariate regression with stepwise backward elimination was performed to determine the best model for predicting the Cmax or AUClast of propranolol. The initial variables used in the multivariate regression were age, sex, body weight, clinical laboratory test results, and H/L ratio. IBM SPSS Statistics™ 21 (Datasolution Inc., version 21, Seoul, Korea) was used for statistical analysis. The two-sided level of statistical significance was set at 0.05.