(M1130-01-05) Evaluating Influence of Critical Quality Attributes of Metronidazole Topical Gel on Bioequivalence Using the Dermal Virtual Bioequivalence (VBE) Module in Simcyp Simulator

Purpose: The purpose of this work was to virtually evaluate the bioequivalence of a generic metronidazole gel (applied to the skin) against its reference product under various study designs by using a dermal physiologically based pharmacokinetic (PBPK) modelling approach for population predictions. An additional aim was to use the Virtual Bioequivalence (VBE) module in the Simcyp simulator to evaluate the influence of differences in the Critical Quality Attributes (CQAs) of the metronidazole gel formulation on the BE outcomes for metronidazole exposure in systemic compartment (plasma) and in different skin layers.

Methods: PBPK models for each formulation, reference product (Prasco® topical gel, 0.75% w/w) and the generic product (Taro® topical gel, 0.75% w/w), were developed and verified using the Multi-Phase Multi-Layer Mechanistic Dermal Absorption (MPML MechDermA) within Simcyp simulator (V20) (Certara, NJ, USA) [1]. To carry out the VBE simulations, the compound files for the reference and the generic product were utilized. Table 1 enlists both VBE and formulation parameters including number of individuals, proportion of females and age of population details used in the simulations. Bioequivalence between these formulations was evaluated using various available BE design options such as crossover design, crossover design with within subject variability (WSV), crossover design with multiple-dosing scenario and parallel design. In the absence of literature to parametrize within subject variability for skin physiology parameters, an extreme case scenario using 30% CV for skin pH and dermis blood flow scalar has been explored here. Global sensitivity analysis was carried out on formulation quality attributes and apparent viscosity was identified as a critical parameter for drug permeation to the systemic compartment and local skin layers. Therefore, apparent viscosity was modified for the generic formulation to generate “microstructural (Q3) variants” and the simulations were conducted as explained above exploring the impact of this CQA on BE outcomes. Bioequivalence was assessed using the non-compartmental analysis and the BE modules in Phoenix WinNonLin version 8.3 (Certara USA Inc.). For BE evaluation, plasma PK parameters, maximum plasma concentration (Cmax) [mg/mL], Area Under the Concentration versus Time Curve extrapolated to infinity (AUCINF) [mg/mL×h], and local tissue PK parameters, maximum amount (Amax) (mg), AUCINF (mg*h) in stratum corneum (SC), viable epidermis (VE) and dermis, were compared between simulated generic (“Q3 variants”) and reference formulations by average BE.

Results: Results in Figure 1 show that mean percent ratio of test (generic) least square mean (LSM) to reference LSM values (Ratio_%Ref) and 90% confidence interval (CI) of metronidazole plasma Cmax and AUCINF, Amax in various layers of skin under various VBE trial design setups; the generic product was found to be bioequivalent to the reference product under the majority of the explored study designs with certain exceptions. The scenarios in which BE criteria were not met include: in the dermis under crossover design with 30% WSV (Cmax and AUCINF) and in the plasma and the SC in the parallel design with multiple dosing condition (AUCINF). A sensitivity analysis was carried out for plasma and local tissue exposure data using parallel design BE setup. The “safe space” for apparent viscosity, a critical parameter based on its impact on AUCINF and Cmax, was found to extent between 127 to 25558 cPs. For metronidazole products, pH was shown to be an insensitive parameter in the developed gel formulation model and skin surface pH was used to estimate fraction of unionized drug at skin surface. Figure 2 shows that apparent viscosity did impact BE of the generic formulation when it was lower than 10-fold or higher than 2-fold of the reference product. More specifically, plasma Cmax and SC AUCINF fell outside of 90% CI limit of BE criteria after lowering the viscosity by 10-fold. When the apparent viscosity of the Q3 variant product is 2-fold higher, Cmax and AUCINF of plasma and Amax of SC also falls outside of the 90% CI limit of BE criteria.

Conclusion: Using the VBE module with metronidazole gel formulations as a case example, model capabilities in terms of simulating different BE trial design setups and accounting for WSV for crossover study designs has been shown. Here, the VBE module in the Simcyp Simulator/Phoenix was used to evaluate VBE between a reference and a generic product applied on the skin for relevant PK metrics in the plasma and various skin layers. In addition, changes in apparent viscosity, a CQA, was explored via simulation for its impact on systemic and local tissue exposure leading to BE failures. Coupled with PBPK modelling, this methodology can be used to identify a ‘safe space’ for CQAs.

References: 1. Arora S et al. Modelling Dermal Drug Absorption from Complex Semisolid Formulations: Insights from Multi-phase, Multi-layer MechDermA Model. Topical Drug Development-Evolution of Science and Regulatory II Workshop. Virtual Meeting, Jul. 23 – Jul. 24, 2020. https://www.pharmacy.umaryland.edu/centers/cersievents/topical2/agenda/
2. Prasco Laboratories Prescribing Information, Metronidazole Topical Gel, 0.75%. 2019.

Acknowledgements: This project was supported by the Food and Drug Administration (FDA) of the U.S. Department of Health and Human Services (HHS) as part of a financial assistance award [FAIN] U01FD006522 totalling $250,000 with 100 percent funded by FDA/HHS. The contents are those of the author(s) and do not necessarily represent the official views of, nor an endorsement, by FDA/HHS, or the U.S. Government. This abstract reflects the views of the author and should not be construed to represent FDA’s views or policies.


Mean ± 90% Confidence Interval (CI) (n= 50) percent ratio of test least square mean (LSM) to reference LSM values (Ratio_%Ref) of test product PK parameter (a) Plasma AUCINF, Cmax (mg) parameters; (b) SC AUCINF, Cmax parameters; (c) VE AUCINF, Cmax parameters and (d) Dermis AUCINF, Cmax under various BE study designs. WSV (Within Subject Variability):, SC (Startum Corneum), VE (Viable Epidermis), AUCINF (Area Under the Concentration versus time Curve extrapolated to infinity), Cmax (maximum plasma concentration), Amax (maximum amount)


Mean ± 90% Confidence Interval (CI) (n= 50) percent ratio of test least square mean (LSM) to reference LSM values (Ratio_%Ref) of test product PK parameter for (a) Plasma AUCINF, Cmax (mg) parameters; (b) SC AUCINF, Cmax parameters; (c) VE AUCINF, Cmax parameters and (d) Dermis AUCINF, Cmax by varying apparent viscosity of generic formulation in BE testing. WSV (Within Subject Variability):, SC (Startum Corneum), VE (Viable Epidermis), AUCINF (Area Under the Concentration versus time Curve extrapolated to infinity), Cmax (maximum plasma concentration), Amax (maximum amount)


VBE trial Design Setup and the Formulation parameters.