(W1230-02-11) Risk Assessment for Extrapolation of Bioequivalence Conclusion from a Male-Only Study Population to the General Population

Purpose: Per the revised Guidance for Industry: Bioequivalence Studies With Pharmacokinetic Endpoints for Drugs Submitted Under an ANDA (PK BE Guidance; August 2021), if a drug product is intended for use in both sexes, the applicant should include similar proportions of males and females in the study or provide a justification supporting the use of a single-sex population. However, many pharmacokinetic (PK) bioequivalence (BE) studies are conducted in only male subjects due to logistical reasons which cannot be used as acceptable scientific justification. The purpose of this research is to evaluate the risk of extrapolating a BE conclusion from PK BE studies that are conducted on male subjects only to the general population (i.e., males and females) when a drug product is indicated for both sexes.

Methods: We surveyed reference listed drug (RLD) labeling containing information from population pharmacokinetic (popPK) models as well as the literature to identify factors leading to PK differences between males and females.  We focused on differences between males and females because differences between males and females are theoretically larger than those between males and the general population and would thereby represent the worst-case scenario. We used PK modeling and simulation to evaluate a wide range of PK parameters identified from the survey [e.g., clearance (CL), volume of distribution (Vd), absorption rate constant (Ka), etc.] to identify combinations of PK parameters in one population (e.g., males) that yielded test/reference (T/R) ratios of Cmax and AUC within 0.80-1.25. By varying PK parameters (e.g., CL) in the aforementioned model, we simulated T/R ratio changes for Cmax and AUC after switching from one population (e.g., males) to another (e.g., females). In addition to the impacts of physiological differences between males and females on T/R ratios as evaluated by modeling and simulation, differences between test and reference formulations can also contribute to T/R ratio differences between males and females. In this regard, we searched the literature for inactive ingredients with known sex-specific impacts on PK profiles. Additionally, we surveyed ANDAs referencing different RLD drug products to identify differences in the T/R ratios of Cmax and AUC between males and females. The RLD products selected have known PK differences between the two sexes based on the labeling as well as the literature.

Results: Survey

Results: We identified only a few inactive ingredients with known sex-specific impacts on PK profiles, such as polyoxyethylated solubilizing excipients. Our survey results also indicated that differences in CL between males and females account for a large portion of PK differences. Specifically, we found that females usually have slower clearance compared to males. Simulation

Results: We applied a one compartment PK model using the PK parameter ranges identified from our survey to assess T/R ratio differences between males and females. Our simulation results showed that the chance was low ( < 1%) for T/R ratios of Cmax and AUC in females (population 2) to be out of the BE acceptance criteria of 0.80-1.25 if T/R ratios of Cmax and AUC in males (population 1) were within 0.80-1.25. The simulation results based on a two compartment PK model yielded a similar conclusion. Results for Comparison of Males vs. Females in ANDAs: The comparison of T/R ratios between males and females in ANDAs where both sexes were dosed (19 studies total) showed that the T/R ratios of Cmax and AUC in males were comparable to those in females, e.g., the ratio of the female Cmax T/R ratio over the male Cmax T/R ratio was 1.01 ± 0.11 for one drug product, which is consistent with the aforementioned simulation results. The variabilities of Cmax and AUC in males were also comparable to those in females, e.g., the within-subject variability was 0.48 ± 0.15 in females and 0.49 ± 0.09 in males for one drug product.

Conclusion: The risk of generalizing the BE outcome from male-only studies to the general population is generally low because 1) there are only a few known inactive ingredients with sex-specific impacts on PK profiles, and 2) the BE conclusion between the two sexes remains the same for most scenarios based on the simulation results and our ANDA survey. Research is ongoing to identify drug products and inactive ingredients which may lead to a high risk of an incorrect BE conclusion for the general population based on the results of male-only PK BE studies.

Acknowledgements: This abstract reflects the views of the authors and should not be construed to represent FDA’s views or policies