Optimal Design of Drug–Drug Interaction Studies for the GLP-1 Receptor Agonist Drug Class for Weight Management: Closing a Potential Data Gap

Abstract

The glucagon-like peptide-1 receptor agonist (GLP-1 RA) drug class, originally developed for glycemic control in patients with type 2 diabetes mellitus (T2DM), has recently expanded its indication to weight management. Many members of this class are peptides which do not have mechanism-based drug–drug interaction (DDI) liabilities per se. However, the potential for interactions arises due to their pharmacologic effect of delaying gastric emptying. This delay can impact the absorption of concurrently administered oral medications, necessitating evaluation of the DDI potential during the drug development process. The clinical relevance of a DDI varies by object and risk/benefit such that some objects may have more dire consequences of an interaction. Oral contraceptives, for example, could have clinically significant outcomes if efficacy were to be reduced as a result of gastric-emptying-mediated effects of the GLP-1 RA reducing systemic exposure.

Tirzepatide is a subcutaneously (SC) administered, once-weekly dual GLP-1/glucose-dependent insulinotropic polypeptide (GIP) RA with label restrictions regarding concomitant use with oral contraceptives. Specifically, women of reproductive potential are advised to switch to a non-oral method of contraception or add a barrier method for 4 weeks after initiating/escalating the dose of tirzepatide.¹ This recommendation was likely based on the results of a combined oral contraceptive (COC) DDI study with ethinyl estradiol/norelgestromin, whereby AUC and Cmax of the progestin component decreased by 22% and 55%, respectively, following a single 5 mg dose of tirzepatide.² According to the FDA clinical pharmacology application review, this interaction magnitude may significantly reduce the contraception effectiveness of the progestin component.² The FDA also noted that the studied dose, a 5 mg single dose, is not the maximum therapeutic dose, and therefore the effect at the maximum dose is unknown. Recently the FDA released a guidance on the conduct of DDIs with oral contraceptives recommending that “The investigational drug should be given at the highest proposed dose for labeling and should be dosed for a sufficient duration to ensure maximal modulation effect of the drug on metabolizing pathways of COCs.”³ For perspective, the maximum dose of tirzepatide for the indication of weight management is 15 mg QW.

It is notable that upon initiation or dose escalation, the COC label restriction is only for 4 weeks. We believe this limited restriction is supported from DDI data with another orally administered medication commonly used to characterize gastric-emptying effects, acetaminophen,⁴ showing that the apparent reduction in exposure observed after the first dose of tirzepatide is no longer evident after 4 weeks at the same dose. For example, when comparing Cmax and AUC of acetaminophen with and without tirzepatide (5 mg QW), Cmax and AUC geometric mean ratios of acetaminophen were 0.5 and 0.75, respectively, after the first dose but increased to 0.92 and 1.05, respectively, following the fourth dose.² These observations align with the known tachyphylaxis or ablation of gastric-emptying delays with this class following repeat dosing. Similar results were observed in acetaminophen studies with the GLP-1 agonist dulaglutide, where the gastric-emptying-mediated DDI observed after the first dose essentially normalized with four weeks of dosing.⁵

SC semaglutide was originally approved at a maximum dose of 1 mg QW for glycemic control in T2DM,⁶ with a subsequent increase of the maximal dose to 2 mg QW for this indication.⁷ It has also been approved for weight management at a maximal dose of 2.4 mg QW.⁸ Unlike tirzepatide, semaglutide does not have label restrictions specifically regarding COC use but does advise caution when oral medications are concomitantly administered with semaglutide. The lack of a specific label restriction with regard to concomitant use of COC is likely based on the results of two DDI studies. A study with the COC ethinyl estradiol/levonorgestrel demonstrated that at steady state of 1 mg SC semaglutide (i.e., after the fifth dose), the exposures of the estrogen or progestin component were not impacted to a clinically significant magnitude.^{9, 10} Another DDI study evaluating the gastric -emptying delay impact on acetaminophen showed that at steady state with 2.4 mg SC semaglutide, semaglutide did not have an significant impact on the exposure, AUC (1.05 [0.99-1.12]) or Cmax (0.90 [0.79-1.04]), of acetaminophen measured over a 5 h period after adjusting for body weight.^{11, 12} However, DDI effects in these studies were not characterized after the single/first administration of a new dose and were only characterized after the fifth dose of SC semaglutide. As normalization of the gastric-emptying effects after the fourth dose of a given dose level were observed in studies with dulaglutide and tirzepatide, these negative DDI results with semaglutide are not unexpected. Given current knowledge, the magnitude of a DDI following the first dose of the maximum dose of SC semaglutide (i.e., 2.4 mg) is unknown. This data gap is particularly concerning given the numerous mainstream media reports highlighting anecdotal stories of unexpected pregnancies while taking GLP-1 RAs.^13-15 This gastric-emptying-mediated DDI may not be definitively at the root of unexpected pregnancy, as it is known that restoration of hormonal imbalances occurring with weight loss boosts fertility¹⁶ and these anecdotal reports do not seem to be corroborated with a dramatic increase in unexpected pregnancies within the FDA FAERS database.¹⁷ However, we are suggesting that this data gap can be avoided in future development programs and label recommendations be further enhanced by implementing these considerations within the contraceptive DDI study.

Clinical pharmacology studies evaluating the gastric-emptying-mediated effects for developmental therapies for weight management should be carefully designed to characterize potential DDI effects across the entire dose range. Prior to the concomitant administration with a weight management therapeutic, a single dose baseline exposure of the oral contraceptive and orally administered acetaminophen could be obtained (staggered baselines to avoid a confounding interaction with these therapies). As weight management therapeutics typically require several steps of up-titration to reach the “maximum modulation,” evaluation of the gastric-emptying-mediated DDI using single doses of oral contraceptives/acetaminophen could be determined throughout the up-titration period. Given the short half-life of these substrates, the risk of carryover and impact to the next assessment is deemed low/none. By utilizing these design elements, gastric-emptying-mediated DDI can be monitored, including achievement of tachyphylaxis, following a single dose and multiple doses across an extended period. While this DDI study may be longer in duration and more complex in design relative to the standard design, these proposed design elements leverage the use of two different oral probe substrates sensitive to this interaction and incorporate a duration to achieve the maximum dose and “maximum modulation.”

In summary, this commentary highlights critical considerations when designing DDI studies with this class of medications. First, it is important and in alignment with regulatory guidance to characterize the maximal potential DDI effects, which includes characterizing the DDI following the first dose of an initial or titrated dose of the precipitant, and, if possible, at the maximal dose. We recognize the challenge that conduct of these studies at the maximal dose may not be feasible as characterization of the risk/benefit profile is ongoing and, therefore, the maximal dose may not be known. However, this medication class often requires titration due to known gastrointestinal adverse effects, such as nausea and vomiting, offering a unique opportunity to characterize multiple levels of dosing throughout the titration period. Second, it is important to characterize the onset and duration of the DDI effect, as the mechanism of the DDI with this class and modality (i.e., peptides) of gastric emptying may diminish to non-relevance following multiple doses of the same dose. Ultimately, application of these DDI study considerations will allow for robust characterization of a potential DDI across a wide dose range, over time, that will inform dosing recommendations within the label for oral contraceptives and, through extrapolation, other orally administered drugs.

Elijah Weber, Ashit Trivedi, and April M. Barbour are all employees of AstraZeneca and own stock in AstraZeneca.