Predicting bioequivalence (BE) for oral formulations with different dissolution profiles remains challenging; traditional empirical approaches (e.g., f₂) fail to capture gastrointestinal complexity and drug specific pharmacokinetic (PK) behaviour.
A semi mechanistic PK modelling framework provides a pragmatic alternative to full physiologically based pharmacokinetic (PBPK), enabling rapid and scalable BE risk assessment for immediate release formulations.
A novel Safe Space Designer module iteratively identifies dissolution profiles likely to remain bioequivalent to a reference product, based on predicted PK parameters.
Two BCS (Biopharmaceutics Classification System) class II case studies (neutral and weakly basic drugs) demonstrated that predicted dissolution safe spaces were consistent with clinical BE outcomes, even when f₂ failed.
Enables efficient, reliable dissolution specification setting, reducing the risk and cost of relative bioavailability studies during drug product development.

The development of oral drug products often requires predicting whether formulations with variable dissolution profiles will be bioequivalent (BE), necessitating the definition of a dissolution safe space. This safe space details the ranges of in vitro specifications for drug quality attributes (such as dissolution) where formulation variants are anticipated to remain bioequivalent to one another. Simple empirical methods, e.g., f2 factor, do not account for gastrointestinal (GI) tract complexity or specific drug’s pharmacokinetic (PK) properties [1]. Therefore, integrating mechanistic understanding of these processes within physiologically based pharmacokinetic (PBPK) models offers greater flexibility and improves BE predictions. Recently, we introduced COMPASSⓇ software, which allows rapid BE prediction for immediate-release (IR) oral dosage forms [2]. Using semi-mechanistic models, COMPASSⓇ evaluates formulations by focusing on the relative difference in predicted maximum plasma concentration (max. ΔCmax). A high BE probability is specifically defined if this difference is below 5.42%. Building on this capability, we created a novel "Safe Space Designer" module. It iteratively searches for the dissolution characteristics of potentially BE formulations, in relation to a reference product's experimental dissolution profile. In this work, we show two case examples of the Safe Space Designer application for establishing the dissolution safe space of two generic IR oral drug products. Case 1 describes a BCS class II neutral drug that is poorly soluble in FaSSGF [3]. Therefore, the dissolution profile was fitted to FaSSIF data only. Case 2 describes an etoricoxib example described previously by Mitra et al. [4]. For this BCS class II weakly basic drug, the formulation manufactured at the new site was shown to be BE in the clinical trial, despite a significantly different dissolution profile at pH = 6.8 (f2 = 44). In both case studies, the predicted dissolution safe spaces were aligned with the clinical BE study results. The Safe Space Designer module enables the efficient and dependable establishment of a dissolution safe space for immediate-release oral drug products. This can be utilized to mitigate risks associated with relative bioavailability studies during the drug product development program.

Omid Arjmandi-Tash

Principal Scientist - Biopharmaceutics

Pfizer