Adam Lucas
University of Manchester
Objectives: Bioequivalence (BE) studies are surrogate clinical trials which bring new products to market with reduced resource following assessment of relative bioavailability (FT/FR) [1]. Regulators [2,3] focus on one-size-fits-all approaches such as comparison of area under the concentration-time curve (AUC), despite literature reports of accounting for pharmacokinetic (PK) variability affecting AUC which would otherwise lead to failure [4]. AUC is a valid endpoint when clearance (CL) remains constant; yet this assumption cannot always be made. An area correction method (AUC·ke) was first proposed by Wagner [5] to account for such variation. Abdallah [6] suggested it is wrong to assume CL remains constant between dosing occasions or is equivalent between individuals. Whilst volume of distribution (V) also varies, it is plausible that it may be more conserved. Adjustment of AUC by the elimination rate constant (ke) to account for CL variation related to the individual subject represents a potential tool in the BE assessment of highly variable drugs (HVD) [7]. The aim of this work is to assess the benefit of AUC∙ke on BE trial outcome through scenario simulation. Methods: A Monte-Carlo simulation was performed in MATLAB to generate a large population of CL and V parameters for reference and test product in a crossover design (with a range of intra-occasion variability (IOV) and between subject variability (BSV)). Parameters were assigned from a bivariate distribution with weak CL-V correlation. Trials were conducted in MATLAB for a range of group sizes. Subjects were randomly selected from the population without replacement and returned for subsequent use. The conventional AUC and AUC·ke ratio test were then computed for trial subjects and an assessment of average bioequivalence (ABE) conducted for each trial using the Two-one sided test procedure [8]. A two-sample t-test in MATLAB was performed to determine the 90% confidence interval (CI) of each trial, and was assessed in relation to predefined BE limits to determine the proportion of trial outcomes (bioequivalent, non-bioequivalent, or inconclusive). Results: Scenario simulation demonstrated increasing subject numbers enhanced probability of a correctly assigned conclusive trial outcome, and reduced the likelihood of an inconclusive trial. Reduced subject numbers were required for equivalent trial outcome using AUC·ke compared to AUC alone when IOV in CL exceeded V. The data indicatives that when CL variability is approximately 3-fold greater than V, a 3-fold reduction in subject requirements can be expected when correction for ke is made. When IOV for CL and V were equivalent, use of AUC∙ke made no difference to the outcome. For a bioequivalent product (FT/FR=1) there is little chance of incorrectly determining non-BE. The outcome is likely to be bioequivalent or inconclusive depending on variability and subject numbers. The opposite is true for a non-bioequivalent product (FT/FR=1.35) where the outcome is most likely to be non-bioequivalent or inconclusive. However, for a bioequivalent product with a clinically meaningless difference (FT/FR ≠ 1), there is a risk of inconclusiveness or worse still incorrectly assigning non-BE. Increasing subject numbers reduces the risk of an indecisive outcome whilst at the same time increasing the danger of mistakenly assigning non-BE.
Investigation of the impact of differing CL and V proportionally, or in isolation determined a change in the outcome in terms of subject numbers and performance of AUC∙ke. The benefit of AUC∙ke remained, and appeared to be enhanced as CL and V increased in a proportional manner, yet subject requirements for a conclusive trial increased as these parameters became larger. Furthermore, as CL escalates relative to V, the benefit of AUC∙ke in terms of minimising subject requirements becomes increasingly evident. Conclusions: The potential use of AUC∙ke as a readily measurable covariate to enhance control of variability should be considered to re-evaluate current, one-size-fits-all approaches. Application in testing HVD could be considerable in reducing subject requirements for study power, or drug exposure in replicate designs. However consideration must be made to identify when CL variability exceeds V, and the reliability of ke measurement. Finally, compound specific PK characteristics, and the influence on bioavailability may dictate the applicability of this approach.
References:
[1] Chow, S. C. Bioavailability and bioequivalence in drug development. Wiley Interdiscip. Rev. Comput. Stat. 6, 304–312 (2014).
[2] Guidance for industry, Bioavailability and bioequivalence studies submitted in NDAs or INDs – General considerations. US Department of Health and Human Services Food and Drug Administration Center for Drug Evaluation and Research. https://www.fda.gov/media/88254/download (2014).
[3] Guideline, The investigation of bioequivalence. Committee for medicinal products for human use (CHMP). European Medicines Agency (EMA). https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-investigation-bioequivalence-rev1_en.pdf (2010).
[4] Yang, J., Ma, P., Bullman, J., et al. Adjustment of the area under the concentration curve by terminal rate constant for bioequivalence assessment in a parallel‐group study of lamotrigine. Br. J. Clin. Pharmacol. 85, 563–569 (2019).
[5] Wagner, J. G. Method of estimating relative absorption of a drug in a series of clinical studies in which blood levels are measured after single and/or multiple doses. J. Pharm. Sci. 56, 652–653 (1967).
[6] Abdallah, H. Y. An area correction method to reduce intrasubject variability in bioequivalence studies. J. Pharm. Pharm. Sci. 1, 60–65 (1998).
[7] Davit, B. M., Conner, D. P., Fabian-Fritsch, B., et al. Highly variable drugs: Observations from bioequivalence data submitted to the FDA for new generic drug applications. AAPS J. 10, 148–156 (2008).
[8] Schuirmann, D. J. A comparison of the two one-sided tests procedure and the power approach for assessing the equivalence of average bioavailability. J. Pharmacokinet. Biopharm. 15, 657–680 (1987).
Reference: PAGE () Abstr 9485 [www.page-meeting.org/?abstract=9485]
Poster: Methodology - Model Evaluation