Dorota Danielak 1, Grzegorz Banach 1, Maciej Winiarski 1, Grzegorz Garbacz 1, Michał Romański 2
1 Physiolution (Wrocław, Poland), 2 Poznan University of Medical Sciences, Department of Physical Pharmacy and Pharmacokinetics (Poznań, Poland)
Introduction: The so-called dissolution safe space defines the specific boundaries of in vitro dissolution specifications within which different batches of a drug product are anticipated to be bioequivalent (BE). To establish these boundaries, physiologically based pharmacokinetic (PBPK) or, more specifically, physiologically based biopharmaceutics models (PBBM) are used. Safe Space Designer is a new add-on to the recently introduced COMPASSâ„¢ software [1]. It utilizes a semi-mechanistic PK model to describe drug dissolution, absorption, and disposition, and analyzes the influence of gastric emptying kinetics on bioequivalence, including physiological gastric emptying types [2]. The simulations produce numerical and graphical representations of the dissolution safe space in the selected dissolution test media.
Objectives: The main objective of this work was to evaluate the dissolution safe spaces of an immediate-release oral drug product administered under fasting conditions using the novel Safe Space Designer tool.
Methods: For evaluation, two case studies from the literature were selected, in which a dissolution-safe space was established by creating virtual dissolution profiles. Case 1 describes zolpidem (BCS class I), for which the reference product dissolved 85% of the dose within 10 min [3]. The authors created six virtual profiles (T1-T6), which released 85% of the dose at different time points: 15 min, 30 min, 45 min, 60 min, 90 min, and 120 min. PBPK simulations showed that the virtual profiles T1, T2, and T3 would be bioequivalent with the reference product. Case 2 describes gefapixant (BCS class IV weak base), for which the reference formulation released over 85% of the drug within 15 minutes [4]. Using PBBM, the authors established two cut-off thresholds that yielded 5% and 10% differences in Cmax relative to the reference, respectively. Digitized dissolution profiles were fitted in COMPASS using the z-factor model. PK profiles were simulated by entering basic PK parameters (Peff, V1/F, CL/F, V2/F, Q/F) from the literature. The bioequivalence estimations were done in two ways: by direct comparison in the Bioequivalence Assessment module and in the Safe Space Designer with the Detailed Analysis of the gastric emptying pattern option enabled. In both modules, bioequivalence between the two products is based on the comparison of the maximum absolute relative difference in Cmax (max. ΔCmax), calculated for various combinations of the gastric emptying rate constant (kGE) and the housekeeping wave time (GET). When this value is below 5.42%, the probability of bioequivalence is high, while values above 7.84% indicate the risk of bioinequivalence.
Results: For Case 1, agreement was reached for profiles T1 and T2, in which 85% of the drug was released within 30 minutes. When the virtual formulation T3 was evaluated in the Bioequivalence Assessment module, the predicted max. ΔCmax was 7.47%, exceeding the previously established 5.42% bioequivalence threshold. This profile was also slightly outside the predicted bioequivalence safe space. In Case 2, the 5% cut-off profile aligned perfectly with the lower boundary of the dissolution safe space defined by the Safe Space Designer. For this virtual profile, the max. ΔCmax predicted in the Bioequivalence Assessment module was 4.34%. In contrast, the 10% cut-off point was outside the dissolution safe space and also predicted to be nonbioequivalent in COMPASS (max. ΔCmax = 7.96%). All the nonbioequivalent profiles failed due to rapid gastric emptying (GET within 15 minutes). Such a pattern, caused by a very fast gastric emptying rate (up to 14 1/h) or the early phase III housekeeping wave, may occur in over 20% of the population. For typical or slow gastric emptying (kGE 3-14 1/h with GET 30 min, or kGE 2 1/h with GET 2 h), the COMPASS software indicated highly probable bioequivalence.
Conclusions: The dissolution safe spaces established in the novel Safe Space Designer module were highly comparable to the results reported in two literature cases. The major cause of predicted bioinequivalence and potential disagreements was rapid gastric emptying. It suggests that the novel software might be more sensitive, leading to more cautious bioequivalence predictions than the previously described models.
References:
1. Danielak D, Myslitska D, Winiarski M, Paszkowska J, Dobosz J, Staniszewska M, et al. Compare and PASS − Fast screening of oral dosage forms for bioequivalence probability with the COMPASS software. Int J Pharm. 2025;670:125123. https://doi.org/10.1016/j.ijpharm.2024.125123
2. Romański M, Staniszewska M, Dobosz J, Myslitska D, Paszkowska J, Kołodziej B, et al. More Than a Gut Feeling─A Combination of Physiologically Driven Dissolution and Pharmacokinetic Modeling as a Tool for Understanding Human Gastric Motility. Mol Pharm. American Chemical Society; 2024;21:3824–37. https://doi.org/10.1021/acs.molpharmaceut.4c00117
3. Paraiso RLM, Rose RH, Fotaki N, McAllister M, Dressman JB. The use of PBPK/PD to establish clinically relevant dissolution specifications for zolpidem immediate release tablets. Eur J Pharm Sci Off J Eur Fed Pharm Sci. 2020;155:105534. https://doi.org/10.1016/j.ejps.2020.105534
4. Wang M, Heimbach T, Zhu W, Wu D, Reuter KG, Kesisoglou F. Physiologically Based Biopharmaceutics Modeling for Gefapixant IR Formulation Development and Defining the Bioequivalence Dissolution Safe Space. AAPS J. 2024;26:69. https://doi.org/10.1208/s12248-024-00938-2
Reference: PAGE 34 (2026) Abstr 11887 [www.page-meeting.org/?abstract=11887]
Poster: Drug/Disease Modelling - Absorption & PBPK