151: Evaluating Clinical Impact of Hormonal Contraceptive Drug-Drug Interactions/Obesity on Ovulation Rate & Pearl Index Using Integrated Physiologically-based Pharmacokinetic Modeling & Model Based Meta-Analysis Approach

Statement of Purpose: Combined hormonal contraceptives (CHCs) are a common method of birth control used by women worldwide. CHCs can be subject to drug-drug interactions (DDI), especially CYP3A4 inducers, which can affect the efficacy and lead to unintended pregnancies. The major clinical endpoint for efficacy is currently Pearl Index (PI). However, studies examining this at subtherapeutic doses are unethical. We utilized model based meta-analysis (MBMA) to explore the possibility of linking PI and ovulation rate (OR) to overcome the data gap for lower doses of PI and establish the dose-response relationship of levonorgestrel (LNG) and drospirenone (DRSP) using the operational model of agonism (OMA). Once established, physiologically-based pharmacokinetic (PBPK) modeling was utilized to generate exposure data to connect to the dose-response relationships. Thus, this study aims to 1) establish a strictly quantitative link between OR and PI and 2) evaluate the impact of obesity and DDIs on the exposure-response relationship.

Description of Methods & Materials: A systematic literature search was conducted via PubMed for the CHCs of interest (LNG and DRSP) and their efficacy endpoints (PI and OR). MBMA was utilized to quantify CHC dose-response using OMA to account for differences in receptor affinity and intrinsic efficacy of the progestins. PBPK simulations were then utilized to quantify exposure based on the approved doses of progestin and ethinyl estradiol: LNG: 150 µg and EE: 30 µg, LNG: 100 µg and EE: 20 µg, DRSP: 3 mg and EE: 20, and DRSP: 3 mg and EE 30 µg. DDI simulations between two progestins and strong CYP3A4 inducers (carbamazepine (CBZ) 400 mg q.d. and rifampicin (RIF) 600 mg q.d.) were conducted in normal weight and obese women (BMI ≥ 30 kg/m2) population. The data generated in these simulations were then utilized to predict the impact of exposure change on the efficacy endpoints PI and OR.

Data & Results: The PBPK simulations indicated that the average concentration (Cavg) of progestin is decreased in normal weight and obese women for LNG and DRSP when co-administered CYP3A4 inducers CBZ/RIF. In addition, obese women have a greater reduction in progestin exposure, which translate to increased OR and PI for both LNG and DRSP. Results showed that RIF led to a greater decrease in exposure than CBZ in both normal and obese women for both LNG and DRSP. Therefore, when compared to CBZ, co-administration with RIF resulted in greater increases in both OR and PI for both progestins in both populations. The increase in efficacy endpoint OR was greater when compared to PI under CMZ/RIF DDI scenarios in both normal weight and obese women.

Interpretation, Conclusion or Significance: The results of our analysis showed that PBPK predictions demonstrated the impact of CYP3A4 inducers on the PK and efficacy endpoints OP and PI for both DRSP and LNG. MBMA simulations confirmed that co-administration of CYP3A4 inducers CBZ/RIF with CHCs increased the risk of OR and PI in normal weight and obese women (BMI ≥ 30 kg/m2), which may require clinical action. When co-administration of CYP3A4 inducers is unavoidable, progestins with a lower fraction metabolized by CYP3A4 may be a suitable alternative.

Disclosures: