Simultaneous Assessment of Hepatic Transport and Metabolism Pathways with a Single Probe Using Individualized PBPK Modeling of ¹⁴CO₂ Production Rate Data

Abstract

Erythromycin is a substrate of cytochrome P4503A4 (CYP3A4) and multiple drug transporters. Although clinical evidence suggests that uptake transport is likely to play a dominant role in erythromycin's disposition, the relative contributions of individual pathways are unclear. Phenotypic evaluation of multiple pathways generally requires a probe drug cocktail. This approach can result in ambiguous conclusions due to imprecision stemming from overlapping specificity of multiple drugs. We hypothesized that an individualized physiologically based pharmacokinetic modeling approach incorporating 14 CO 2 production rates (iPBPK-R) of the erythromycin breath test (ERMBT) would enable us to differentiate the contribution of metabolic and transporter pathways to erythromycin disposition. A sevencompartmental physiologically based pharmacokinetic (PBPK) model was built for 14 C-erythromycin administered intravenously. Transporter clearance and CYP3A4 clearance were embedded in hepatic compartments. 14 CO 2 production rates were simulated taking the first derivative of by-product 14 CO 2 concentrations. Parameters related to nonrenal elimination pathways were estimated by model fitting the ERMBT data of 12 healthy subjects individually. Optimized iPBPK-R models fit the individual rate data well. Using one probe, nine PBPK parameters were simultaneously estimated per individual. Maximum velocity of uptake transport, CYP3A4 clearance, total passive diffusion, and others were found to collectively control 14 CO 2 production rates. The median CYP3A4 clearance was 12.2% of the input clearance. Male subjects had lower CYP3A4 activity than female subjects by 11.3%. We applied iPBPK-R to ERMBT data to distinguish and simultaneously estimate the activity of multiple nonrenal elimination pathways in healthy subjects. The iPBPK-R framework is a novel tool for delineating rate-limiting and nonrate-limiting elimination pathways using a single probe. SIGNIFICANCE STATEMENT Our developed individualized physiologically based pharmacokinetic modeling approach incorporating rate data (iPBPK-R) enabled us to distinguish and simultaneously estimate the activity of multiple nonrenal elimination pathways of erythromycin in healthy subjects. A new interpretation of erythromycin breath test (ERMBT) data was also obtained via iPBPK-R. We found that rate data have rich information allowing estimation of per-person PBPK parameters. This study serves as proof of principle that the iPBPK-R framework is a novel tool for delineating rate-limiting and non-rate-limiting elimination pathways using a single probe. iPBPK-R can be applied to other rate-derived data beyond ERMBT. Potential areas of application include drug-drug interaction, pathophysiological effects on drug disposition, and the role of biomarkers on hemodialysis efficiency utilizing estimated adjustment factors with correlation analysis.

Dose Optimization in Kidney Disease: Opportunities for PBPK Modeling and Simulation

“…We have developed a top-down plus bottom-up or "middle-out" PBPK approach, namely, individualized PBPK modeling of rate data (iPBPK-R) using a reduced-order model and applied it to 14 CO 2 production rate data. 87,88 Figure 1 contrasts the middleout iPBPK-R approach with general bottom-up PBPK approaches. 89 iPBPK-R enables the simultaneous evaluation of hepatic transport and metabolism pathways and exploration of uremic toxin-mediated alterations of drug disposition in individuals.…”

“…86 In another study, individualized PBPK model fitting was conducted to indirectly estimate multiple parameters on nonrenal elimination pathways utilizing reduced order models. 87 Sixth, PBPK modeling is a dynamic mechanistic approach that enables exploration of the association of uremic toxins as potential causes of altered drug disposition and parameter estimates related to drug elimination pathways. Uremic toxins can be treated as perpetrator drugs in PBPK DDI modeling.…”

“…r To conduct research on sampling times to optimize a PK study design in kidney disease using PBPK modeling, which may influence parameter estimates r Research to optimize sampling times for designing clinical trials in CKD may be conducted using Simcyp simulation similarly to Durmont et al 110 analysis combined with a middle-out PBPK approach is another method to explore the impact of uremic toxins on transporter-and metabolism-mediated pathways. 87 Last, to maximally benefit from PBPK modeling, clinical trial designs may need to be designed and coordinated with subsequent PBPK analysis in mind to optimize the clinical information available for subsequent modeling (ie, optimal sampling times to enhance PBPK modeling). 110 In summary, PBPK modeling and simulation provide exciting opportunities to support quantitative studies in pharmacology, PK, and drug dosing in the setting of kidney disease.…”

“…We have developed a top‐down plus bottom‐up or “middle‐out” PBPK approach, namely, individualized PBPK modeling of rate data (iPBPK‐R) using a reduced‐order model and applied it to 14 CO 2 production rate data 87,88 . Figure 1 contrasts the middle‐out iPBPK‐R approach with general bottom‐up PBPK approaches 89 .…”

“…Furthermore, PBPK modeling can be co‐implemented with in vitro cell culture experiments with an increasing concentration of uremic toxins as reviewed above 84 . Correlation analysis combined with a middle‐out PBPK approach is another method to explore the impact of uremic toxins on transporter‐ and metabolism‐mediated pathways 87 …”