Coralie Tardivon (1), Florence Loingeville (2), France Mentre (1), Satish Sharan (3), Jing Han (4), Wanjie Sun (4), Stella Grosser (4), Liang Zhao (3), Lanyan (Lucy) Fang (3) and Julie Bertrand (1)
(1) Université de Paris, IAME, INSERM,F-75018 Paris, France, (2) Faculty of Pharmacy, Univ. Lille, EA 2694, Public Health: Epidemiology and Healthcare quality, 59000 Lille, France, (3) Office of Research and Standards, Office of Generic Drugs, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, 10903 New Hampshire Ave, Silver Spring, MD 20993, USA, (4) Office of Biostatistics, Office of Translational Sciences, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, 10903 New Hampshire Ave, Silver Spring, MD 20993, USA
Introduction: Bioequivalence (BE) studies are key to the development and approval of generic drugs. Traditionally, a two-period two-way crossover study is conducted and the two one-sided test (TOST) is performed using estimates of area under the concentration-time curve (AUC) and maximal concentration (Cmax) obtained by non-compartmental analysis (NCA).
However, for ophthalmic products, classical NCA is not possible. Indeed, after administration only one sample of aqueous humor in one eye can be obtained in each patient to measure concentration. Further, as crossover design studies can only include subjects with bilateral cataracts, parallel studies are more feasible to conduct.
In the present work, we compared, by clinical trial simulation, two different approaches:
- the non-parametric bootstrap NCA-based TOST proposed by Shen and Machado [1], which derives AUC from the mean concentrations at each time-point per treatment group and bootstraps subjects at each time point with replacement,
- the model-based (MB) TOST proposed by Dubois et al. [2], based on a nonlinear mixed effect model (NLMEM) of all concentrations with a treatment effect (and period and sequence effects for crossover designs) on all pharmacokinetic parameters using asymptotic standard errors from the observed Fisher information matrix,
to analyse BE crossover and parallel design pharmacokinetic studies with one-time point per subject.
Methods: The simulation design mimic usual BE trials for ophthalmic drugs with 500 subjects with one sample per subject. We simulated 500 datasets for each of the sixteen scenarios varying the following:
- simulation under two null hypotheses (AUC and Cmax GMR = 0.8 or 1.25) or two alternative hypotheses (AUC and Cmax GMR = 1 or 0.9),
- two vectors of sampling times to draw the one-time point from (of size 5 or 10),
- crossover or parallel designs.
The pharmacokinetic model and parameter values for the fixed effect and variabilities were inspired from the previous simulation study on MBBE by Dubois et al. [2]. NLME modeling was performed using Monolix 2018R2.
Results: Using MB approaches, all fixed effects, six for the parallel and twelve for the crossover design, were estimated with no bias and good precision for all sixteen scenarios.
For AUC, MB approaches conserved a type 1 error at the nominal level, whereas the non-parametric NCA-based TOST type 1 error was inflated on two scenarios with a vector of sampling times to draw the one-time point of size 10: i) parallel design with AUC and Cmax GMR = 0.8, and ii) crossover design with AUC and Cmax GMR = 1.25. All approaches obtained similar power.
For Cmax, MB TOST conserved a type 1 error at the nominal level except on two scenarios, whereas the non-parametric NCA-based TOST has a type 1 error systematically below the prediction interval around 5%. As for the power, the non-parametric NCA-based TOST obtain very low estimate on parallel design especially when the set of sampling times to draw the one-time point from was of size 10 (<25%).
Conclusions: This realistic simulation study proved that MB approaches are not only feasible but also efficient for BE studies with only one-time point measured drug concentration. Improvements on the Bootstrap NCA-based TOST approach are currently under study.
Acknowledgements: This work was supported by the U.S. Food and Drug Administration (FDA) under contract 75F40119C10111. The authors thank FDA for this funding. The views expressed in this abstract do not necessarily reflect the views or policies of the FDA.
References:
[1] Shen M. and Machado S. G. J Biopharm. 27: 257-264, 2017.
[2] Dubois et al. Stat in Med, 30:2582-2600, 2011.
Reference: PAGE () Abstr 9256 [www.page-meeting.org/?abstract=9256]
Poster: Oral: Methodology - New Tools