Consistent with its pharmacological effect of slowing gastric emptying, exenatide may reduce the rate of absorption of concomitantly administered oral drugs. Drug-drug interaction studies with digoxin [8], warfarin [9], lovastatin [10], and lisinopril [11] have demonstrated that concomitant exenatide treatment reduced the maximum plasma concentrations (C_max) and delayed the T_max for these drugs, both of which are consistent with slowing of gastric emptying. Reductions in overall exposure (area under the curve [AUC]) were only observed in the exenatide-lovastatin interaction study. However, given the known pharmacokinetic characteristics of exenatide, the potential for a CYP3A induction was considered unlikely and the observed results were considered to be related to incomplete characterization of the single-dose lovastatin AUC following exenatide dosing. Acetaminophen [12], a marker of gastric emptying, was studied with exenatide to understand how the relative timing of exenatide administration might change the magnitude of pharmacokinetic effects observed for orally administered drugs. In addition, this prior study provided information on the optimal timing for administration of other concomitant oral medications. Changes in the acetaminophen profile were not evident when acetaminophen was given 1 hour prior to the exenatide dose as the absorption process for acetaminophen had likely been completed before the onset of exenatide action. However, a reduced C_max and delayed T_max were observed when acetaminophen was given after exenatide administration; the magnitude of changes was greatest 1 to 2 hours after exenatide administration.

Drug-drug interaction studies with oral contraceptives are generally conducted to understand the potential for concomitantly administered drugs to induce or inhibit cytochrome P450 isoenzyme (CYP) mediated oxidative metabolism of EE [16]. While EE is metabolised by both CYP3A-mediated oxidative metabolism and Phase II metabolism, including glucuronidation and sulfation, the most clinically relevant metabolic pathway is induction or inhibition of CYP3A. Drugs that decrease EE bioavailability via CYP3A induction may potentially result in reduced OC efficacy. The observation that OC AUC concentrations were unaltered in this study confirmed that exenatide does not induce CYP3A. However, the clinical relevance of C_max reductions seen in the study (up to 46% for EE and 41% for LV) requires additional consideration.

Reports of large inter-subject variability in concentrations of OCs, with several-fold differences in serum concentrations likely due to inter-individual differences in first-pass metabolism, have been described in the literature. Goldzieher et al. [17] have reported that differences in EE concentrations have been shown to vary between ethnic groups, as well as across study sites and, even for a given individual, EE AUCs can vary by almost a factor of 4. This same study group has also reported the existence of high intra- and inter-subject variability in the pharmacokinetics of progestins such as LV [17]. Thus, the magnitude of C_max reduction observed in the present study was likely within the inherent PK variability of the OC components.

The current study did not measure OC pharmacodynamics (eg, follicle-stimulating hormone or luteinizing hormone concentrations); therefore, a direct within-study clinical relationship with the observed C_max reduction cannot be derived. Importantly, no break-through bleeding was reported. Break-through bleeding may be associated with low concentrations of estrogen-progestin [16,18].

A review of prescription labeling indicates that drug interactions with OCs are deemed to be clinically important, and dosage adjustments are thereby recommended, only when associated with a significant reduction in OC AUC. However, there does not appear to be a well-accepted minimum threshold concentration for pharmacological activity. Importantly, in this study OC AUC was unchanged. Furthermore, trough concentrations did not decrease in the presence of exenatide suggesting that sub-therapeutic concentrations were unlikely.

In the absence of conclusive literature on the exposure-efficacy relationships of OCs, other aspects of OC PK/PD were considered to help understand the possible clinical relevance of the C_max decrease observed in the present study. In food-effect studies, OC C_max reductions of up to 40% are commonly observed without changes in AUC. Despite this potential effect of food on C_max, OCs are generally recommended to be taken without regard to food [19], suggesting that PK changes observed in this study are not likely to be clinically relevant. In the current study, the effect of food consumption on the OC PK cannot be clearly differentiated from the effects of exenatide; however, these data reflect conditions under which the 2 drugs are likely to be co-administered, given that exenatide is to be administered within an hour of meals. Thus, in consideration of indirect evidence from food-effect studies, the large inherent variability in OC concentrations, and the fact that product labeling suggests changes in OC dosage only in the presence of large changes in OC AUC alone, we conclude that the PK changes observed in the present study are not likely to have clinical implications.