The pharmacokinetics/pharmacodynamics of theophylline in premature neonates during the first few days after birth.

Abstract

Theophylline is one of the few preparations available for the treatment of apnoea of prematurity. Currently little data is available on the pharmacokinetics and the pharmacokinetic/pharmacodynamic relationships of theophylline for premature neonates during the first few days of life, a time when neonates undergo profound physiological changes and when the drug is most often used. Furthermore, the influence of theophylline on hypoxaemic episodes has not yet been quantified. The study aimed to investigate optimal theophylline dosing in this group by establishing pharmacokinetic parameters, assessing the effectiveness of the drug in abolishing apnoea and hypoxaemic episodes and investigating the concentration/effect relationship. The project was conducted in the neonatal wards of King Edward VIII Hospital, Durban, South Africa. The study group comprised a total of 105 Black, apnoeic, premature neonates, with respiratory distress syndrome, who were receiving intravenous theophylline. Serum samples (263), collected from patients during routine care, were analysed for theophylline. Forty-six patients were monitored before and after theophylline therapy with a neonatal capnograph linked to a data acquisition. Apnoea incidents were classified into total (all apnoea <_5 seconds) and pathologic (all apnoea >_20 seconds) and a hypoxaemic episode was defined as a >_10% fall for >10 seconds in peripheral oxygen saturation. Within each of these groups patients were assessed as responders (>_50% reduction in the clinical effect from baseline to the last recording) and non-responders. Patient characteristics were identified as possible markers of non-response to theophylline therapy. The Nonlinear Mixed Effects Model (NONMEM) was used to derive population pharmacokinetic models and parameters for theophylline as well as to assess the concentration-effect relationship. The pharmacokinetic analysis estimated a low clearance and volume of distribution, with oxygen support enhancing clearance. Relatively high inter-individual and residual variability values were obtained prompting testing for inter-occasion variability. This resulted in a decrease of inter-individual variability for clearance and volume of distribution as well as in residual variability. In the theophylline doses used, a significant reduction in total and pathologic apnoea but not in hypoxaemic episodes occurred over the first three days after birth. The most positive improvement was seen on the first day of treatment after the loading dose. A statistically significant increase in the average pulse rate and a decrease in episodes of bradycardia from baseline to all three days of monitoring were recorded. Most patients responded at serum theophylline concentrations of 3 to 9 mg/L. Most serum theophylline concentration measurements were also in this range and it was not possible to clearly define a concentration-effect relationship. The cumulative percentage of non-responders was relatively high for total apnoea (48%) and hypoxaemic episodes (45%), but low for pathological apnoea (13%). Being one of a set of twins was identified as a marker of poor response for both total apnoea and hypoxaemic episodes. Other possible markers for poor response, in terms of total hypoxaemic episodes, were being born by caesarean section and having more than the 75th percentile pathologic apnoea per hour at baseline. It was interesting to note that, with regard to total apnoea, there were some features that seemed to predict a favourable response to theophylline. These were birth weight and 5 minute Apgar score below the 25th percentile, and patients with baseline total apnoea counts above the 75th percentile. The cumulative graphs of the responders and non-responders resembled the fixed effect model, which is the simplest model to explain drug-effect relationships. More sophisticated analysis of the concentration-effect relationship, using NONMEM and the count model proved difficult. None of the models tested were found to be satisfactory, but that which included the influence of a hypothetical respiratory depressant factor gave the most realistic value of EC50. It is suggested that further even more complex modelling may be required to accurately define the concentration-effect relationship (and hence the therapeutic range) for theophylline in neonatal apnoea.