Category: Clinical Pharmacology
Purpose: Medical countermeasures (MCMs) drugs include antimicrobial or antiviral drugs regulated by the U.S. Food and Drug Administration (FDA) that are used in a bio-terrorist attack or a naturally occurring emerging disease. Due to the strategic importance of applying MCMs in the public health emergency, a timely assessment and control of the critical quality attributes (CQAs) of these drug products are crucial. In the current study, we propose to develop physiologically-based pharmacokinetic (PBPK) absorption models for three different MCM drug products and explore the applications of these models to predict the effects of CQA changes such as particle size and dissolution on drug systemic exposure.
Methods: PBPK absorption modeling and simulations of three different orally administered antiviral MCM drugs - oseltamivir phosphate (OP), baloxavir marboxil (BM) and ciprofloxacin (Cipro) were developed using GastroPlusTM 9.6 (Simulations Plus, Inc.). PBPK absorption models for prodrugs were established using OP and BM as examples. OP and BM are ester prodrugs for the treatment of pandemic influenza. Oseltamivir carboxylate acid (OC), an active metabolite of OP was tracked using the enzyme/metabolism module with liver carboxylate esterase 1 as the major enzyme responsible for metabolic conversion. Whereas, for BM with limited information of metabolic enzymes and undetectable parent drug concentration in plasma, assumption of gut/liver first past extraction has been established to allow the development of metabolite (baloxavir acid, BXA) model using physicochemical properties of the parent BM as model inputs. For MCM drugs with sparse human PK data (e.g., BM), PBPK absorption model was originally constructed in animals and extrapolated to human following allometric scaling. Disposition and PK of MCMs were further simulated using either compartmental PK module (BM and Cipro) or mechanistic full PBPK module (OP). Parameter sensitive analysis (PSA) was conducted to identify the CQAs that affect the in vivo PK performance. For BM and Cipro with low solubility and immediate-release (IR) formulation, PSA shows that drug substance particle size is one of the key CQAs. Thus, with Johnson model incorporated as a dissolution model, virtual bioequivalence (BE) was performed to define the lower and upper bound of particle size using the target clinical batch as a reference. For all three MCMs, dissolution rate is another key CQA. In vitro dissolution data were further incorporated in the model to predict/simulate clinically relevant “safe space” for dissolution profiles (i.e., the upper and lower bound of a clinically relevant dissolution profile).
Results: PBPK absorption models of OP (IR capsules), BM-BXA (IR tablets) and Cipro (IR and XR tablets) have been developed (Fig. 1A, D, G). Purpose-dependent verifications were conducted using all available in vitro and in vivo datasets. The human PBPK absorption model for all 3 drugs accurately predict drug exposure (i.e., AUC0-t, AUC0-inf and Cmax) with the prediction error within ±20%. PSA suggested that particle size and dissolution rate are key CQAs that affect the in vivo exposure of BXA (BM is not detectable) and Cipro. Using virtual BE analysis compared with target clinical batch, safe space of particle size was set for BM (< 30 µm) and Cipro (0.1-1 µm). Though particle size is a not an absorption rate limiting parameter for BCS Class I drug OP, dissolution rate affects the in vivo drug exposure. Incorporating several theoretical dissolution profiles slower than those of bio-batch into virtual BE analysis, we predicted that the safe space of dissolution profile for OP is 10% slower than that of pivotal bio-batch in quality control (QC) medium. Dissolution profile safe space has also been identified for BM (15% slower) and Cipro (12% slower) using virtual BE analysis.
Conclusion: Current study summarizes the PBPK modeling strategies of multiple MCM drugs with variable physicochemical properties under different conditions (e.g., prodrug with active metabolites, drug with sparse human data). These PBPK modeling and simulation can be used to establish clinically relevant safe space for CQAs such as particle sizes and dissolution rates.
Lei Miao– Silver Spring, Maryland
Huong Moldthan– Postdoctoral, US Food and Drug Administration, Rockville, Maryland
Da Xu– US Food and Drug Administration, Silver Spring, Maryland
Liang Zhao– Supervisory Pharmacologist, Office Director, USFDA, Silver Spring, Maryland
Kimerly Raines– Silver Spring, Maryland
Paul Seo– Division Director, USFDA, Silver Spring, Maryland