Intraindividual variability in theophylline pharmacokinetics: Statistical verification in 39 of 60 healthy young adults

Abstract

After administering a single 300 mg dose of theophylline in oral solution to 12 healthy adults, the dose-normalized area under the plasma concentration-time curve was 97.2±20.1 % (mean±SD) of that after giving a 500 mg dose and statistically indistinguishable. Similarly, these areas multiplied by the individual's terminal disposition rate constant (β) were statistically indistinguishable between 300 and 500mg doses (99.1±10.3%), giving no evidence of dose-dependence for theophylline kinetics at the levels below 15 μg/ml observed in these individuals. After an intravenous dose, a shortlived distribution phase (t_1/2α) is sometimes seen. An a phase, however, is hardly discernible in over 250 profiles arising from oral doses administered during five single dose bioavailability studies. Almost all such profiles appear to follow single-compartment model predictions. With precautions to avoid a potential a phase, a terminal log-linear slope can be fitted by least squares analysis with a relative standard error in the slope determination almost always less than 6%. Covariance analysis confirms statistically that 39 of the 60 participating individuals varied in their β on the different occasions each was required to take a dose during the course of a crossover bioavailability trial. In one study, even though each individual was observed on only two occasions, 9 out of 12 showed statistically identifiable variation in β. Fluctuations in β of 60% can be seen. Changes of 30% or greater are common and can occur within 3 or 4 days. Thus real, large, and potentially frequent changes in β of theophylline have been identified in a majority of normal subjects. These changes do not appear to be confined to either sex, to smokers or nonsmokers, or to heavier or lighter individuals. No chronological pattern has, as yet, been recognized in the intraindividual variability in β.