Modulation of Celecoxib Pharmacokinetics by Food in Pediatric Patients*

To the Editor In the November 2002 issue of the Journal, we presented preliminary pharmacokinetic results of a clinical trial evaluating the use of celecoxib in conjunction with metronomic dosing of chemotherapy for antiangiogenic treatment in children with recurrent solid tumors.1 The purpose of this report is 2-fold: (1) to present new pharmacokinetic data for celecoxib in a larger group of children taking celecoxib at a dose of 250 mg/m2 twice daily and (2) to present the effect of high-fat food on celecoxib kinetics in a subset of children. Our data suggest a trend toward increased absorption and increased overall systemic exposure when celecoxib is taken with a high-fat meal or snack. In addition, it appears that food high in fat tends to modulate trough plasma levels such that patients spend a smaller proportion of the dosing interval below the targeted antiangiogenic concentration of 500 Μg/L (Dr Jaime L. Masferrer, oral communication, 2002). Data presented in our original report and data that have been accrued since that publication indicate that the trough concentrations of celecoxib in our patients were below the desired concentration that possesses antiangiogenic properties (500 Μg/L) (Dr Jaime L. Masferrer, oral communication, 2002). For safety and regulatory reasons, we opted to try to increase the systemic exposure to celecoxib by coadministering the drug with a high-fat meal or snack rather than increasing the dose. Early clinical studies investigating the pharmacokinetics of celecoxib in healthy adults showed that the maximum plasma concentration (Cmax) and area under the plasma concentration-time curve (AUC) were approximately dose-proportional across the dose range of 100 to 200 mg. However, at higher doses (400 mg, which is approximately equivalent to our study dose of 250 mg/m2), there appears to be a less than proportional increase in Cmax and AUC. This is thought to be a result of the low solubility of celecoxib in aqueous media.2 Therefore it was recommended that higher doses of celecoxib be coadministered with food to try to improve absorption.3 Our study was amended to include fatty food coadministration. Examples of appropriate foods included peanut butter on toast or a buttered muffin (Pharmacia, oral communication, 2002). In the children in our study, high-fat food has resulted in significant differences in the Cmax and AUC values both after a single dose and at steady state (Table I and Fig 1). We were particularly interested in the trough concentration at steady state because we were aiming to maintain plasma concentrations above the antiangiogenic concentration of 500 Μg/L (Dr Jaime L. Masferrer, oral communication, 2002). Although the difference observed for the trough concentration (Cmin) was not statistically different, the trend we were expecting has been demonstrated and may reach significance with an appropriate sample size. Without increasing the celecoxib dose, this seems to be a reasonable way to increase the plasma concentrations of celecoxib in an effort to achieve antiangiogenic therapeutic levels (Table I and Fig 1). Plasma concentration-time curve (mean ± SD) for 1 dose interval after single dose and at steady state in pediatric cancer patients who were administered 250 mg/m2 of celecoxib orally with and without food. An additional valuable piece of information that arose from this study concerns the safety of long-term use of celecoxib in children at a twice-daily dose of 250 mg/m2. The median duration of treatment was 10 weeks (range, 3 weeks to 18 months). In fact, 1 patient was treated for 16 months and another for 18 months, and none showed any signs of central nervous system or cardiovascular toxicity. There are several limitations of this study resulting from factors that were beyond our control given the vulnerability of the children involved. First of all, our study did not have a blinded, randomized crossover design. The data we present are opportunistic in nature and were collected as part of a phase I pilot study whose main objective was to determine the safety of a particular drug combination. In addition, we recognize the unbalanced sample sizes of our 2 groups (with versus without high-fat food), but this is a result of the original study design. Future studies of celecoxib at this dose should include the recommendation for high-fat food coadministration. Finally, although high-fat food intake was not recommended for the original group of children studied, fasting was not recommended either. However, experience with pediatric cancer patients eligible for phase I studies suggests that their eating habits generally include light meals, and because the pharmacokinetic studies were started early in the morning, in most cases the children had not yet eaten breakfast. For the patients who did take celecoxib with food, the high-fat meal or snack was not standardized because, again, this was beyond the scope of this study. Instead, we suggested several examples of foods that were considered to be high in fat, and the patients were asked to keep a food diary for the pharmacokinetic study days. Despite these limitations, the evidence for increased systemic exposure to celecoxib with high-fat food in children is quite convincing (Table I).