GLP‐1RA‐induced delays in gastrointestinal motility: Predicted effects on coadministered drug absorption by PBPK analysis

Abstract

Background

Glucagon‐like peptide‐1 receptor agonists (GLP‐1RAs) are breakthrough medicines for obesity treatment and have rapidly gained widespread clinical application. Although GLP‐1RAs are generally not associated with drug–drug interactions (DDIs) via drug metabolism or transporter pathways, their effects on reduced gastrointestinal (GI) motility could influence the pharmacokinetics of coadministered oral medications.

Objectives

This study uses physiologically based pharmacokinetic (PBPK) modeling to evaluate the DDI potential of GLP‐1RA‐induced GI motility delays.

Methods

Using Certara's Simcyp™ Simulator V23, we modeled the pharmacokinetics of atorvastatin, metformin, metoprolol, ethinyl estradiol, and digoxin in a virtual cohort of obese adults (n = 1000). GLP‐1RA‐related gastric emptying delays were simulated based on capsule endoscopy data from liraglutide‐treated patients. Results were compared with clinical data from semaglutide and liraglutide users. Additionally, exploratory analyses were conducted on frequently coadministered drugs identified from the 2022 Medical Expenditure Panel Survey, including rosuvastatin and dabigatran.

Results

GLP‐1RA‐induced gastric emptying delays led to increased area under the concentration–time curve (AUC) and prolonged time to maximum concentration (Tmax) for several medications. The model outputs for rosuvastatin, valsartan, and dabigatran indicate increases in AUC by 64%, 90%, and 205%, respectively. Dabigatran, a narrow therapeutic index anticoagulant, exhibited the most significant changes, raising potential concerns of higher drug exposure.

Conclusions

PBPK modeling suggests that GLP‐1RAs can influence the pharmacokinetics of oral medications by delaying gastric emptying, potentially leading to clinically relevant DDIs. While further clinical validation and pharmacovigilance is needed, these findings highlight the importance of PBPK tools in predicting and potentially mitigating risks associated with GLP‐1RA use.

1. INTRODUCTION

The glucagon‐like peptide‐1 receptor agonists (GLP‐1RA) are breakthrough drugs to treat obesity. GLP‐1RAs aid in weight management by slowing gastric motility, suppressing appetite, and increasing satiety, all of which help reduce caloric intake and promote weight loss. High drug cost and access are current barriers to their broad therapeutic use. ¹ However, increased investments in manufacturing, shifts in health policy, expanded medical insurance coverage, and affordability through competition are expected. Since their approval by the United States Food and Drug Administration (FDA), GLP‐1RA prescriptions have quadrupled over the past 4 years, with 4.7 million individuals in the United States using these agents in 2022 (Figure S1). ² This surge coincides with the introduction of newer and more potent incretin mimetics into clinical practice. ² If current trends continue, policy shifts could further accelerate this growth, with projections suggesting a fourfold increase in usage within the next decade. ³ Although this expansion marks a new era in the clinical management of chronic diseases such as obesity and diabetes, it also underscores a critical gap in our knowledge. Several GLP‐1RA product labels indicate that drug‐induced delays in gastric motility may impact the absorption of coadministered oral medications. ⁴ , ⁵ However, specific guidance is not available to healthcare providers on how to manage this potential drug–drug interaction (DDI). Likewise, conflicting findings exist for the effects of semaglutide and tirzepatide on ethinyl estradiol exposure. For example, tirzepatide lowered the area under the curve (AUC) of ethinyl estradiol by 21%, whereas semaglutide increased the AUC by a similar fraction. ⁴ , ⁵ This uncertainty in DDI potential is particularly problematic when initiating GLP‐1RA therapy in patients taking narrow therapeutic index medications, such as oral anticoagulants, anticonvulsants, and hormonal contraceptives.

Although GLP‐1RAs generally do not cause drug metabolism or transporter‐mediated DDIs, their effect on gastrointestinal (GI) motility may alter the pharmacokinetics of orally administered medications, including the extent and rate of absorption, overall drug exposure, and the time to reach maximum concentration (T _max). The existing research on how GLP‐1RAs affect the absorption of oral medications is limited but indicates potential impacts on maximum concentration (C _max), total drug exposure defined by AUC, and time to C _max (T _max) of coadministered oral drugs. ⁶

Addressing the effects of GLP‐1RA‐induced delays in gastric motility on oral drug administration is complex and typically requires costly clinical trials, which may only evaluate a limited number of medications. This approach is suboptimal, particularly because patients indicated for GLP‐1RA therapy often have chronic comorbidities, such as hypertension and cardiovascular disease, which necessitate the use of multiple medications. Physiologically based pharmacokinetic (PBPK) modeling has emerged as an alternative, noninvasive method for assessing DDI potential and is often used as a surrogate for clinical trials. ⁷

The objective of this study was to evaluate the DDI potential of GLP‐1RA on common comedications based on a PBPK workflow that incorporated capsule endoscopy data of GLP‐1RA‐induced shifts in GI motility (Figure 1). We leveraged real‐world prescription data from the 2022 Medical Expenditure Panel Survey (MEPS) to identify common co‐administered medications and simulated their pharmacokinetic profiles using Certara's Simcyp™, an FDA‐approved biosimulation software. A comprehensive understanding of potential GLP‐1RA DDI risk can aid the design of pharmacovigilance and dedicated clinical studies to improve the safe use of these medications.

FIGURE 1.

Physiology‐based pharmacokinetic workflow for study analysis. Figure created with Biorender. GLP‐1RA, glucagon‐like peptide 1 receptor agonist; PBPK, physiology‐based pharmacokinetic; PK, pharmacokinetic.

2. METHODS

2.1. Substrate selection

Utilizing Certara's Simcyp™ Simulator V23, we performed PBPK modeling to examine the impact of GLP‐1RA‐associated delays in GI motility on the pharmacokinetics of atorvastatin, metformin, metoprolol, ethinyl estradiol, and digoxin. ⁸ These drugs were selected based on available pharmacokinetic data in semaglutide and liraglutide drug labels and validated compound models in Simcyp's compound repository.

Additionally, exploratory analyses were conducted on medications identified as being frequently co‐administered with GLP‐1RAs based on responses from the 2022 MEPS (Figure S2). ² The high frequency of co‐administered medications identified was then cross‐referenced with Simcyp's compound repository to determine substrates with validated PBPK profiles. This analysis included rosuvastatin, valsartan, rivaroxaban, bupropion, and dabigatran.

2.2. Simulation parameters

GLP‐1RA‐associated delays in gastric emptying and small intestine transit time were incorporated by modifying the built‐in virtual obese population model, using capsule endoscopy measurements in patients with diabetes taking liraglutide. ⁹ , ¹⁰ , ¹¹ Key validation and exploratory substrate simulation parameters are described in Tables S1 and S2, respectively. An overview of simulation parameters and study design is displayed in Table 1. No substrate or population values other than gastric motility were altered during model refinement; all parameters utilized were sourced from Certara SimCyp™ V23. ⁸ GLP‐1RA GI motility values selected from the literature extended the gastric emptying time to 4 h and mean small intestine transit time to 9 h in the GLP‐1RA treatment group. Control GI mean residence and small intestine transit time parameters were fixed as the default fed state values in the Simcyp obese population. Simulations were performed on independent virtual cohorts of obese adults (n = 1000, 100 individuals in 10 trials) over a 24‐h period following a single dose of the selected substrates in the fed state.

TABLE 1.

Simcyp V23 modeling parameters for validated substrates.

	Compound build	Absorption model	Dose
Validated substrates with clinical correlates
Digoxin	SV	ADAM	0.5 mg
Ethinyl estradiol	SV	First order	0.035 mg
Metoprolol	SV	First order	100 mg
Metformin	SV	First order	390 mg
Atorvastatin	SV	ADAM	10 mg
Validated substrates without clinical correlates
Dabigatran	SV	ADAM	150 mg
Bupropion	SV	ADAM	130.2 mg
Rivaroxaban	SV	ADAM	10 mg
Valsartan	Sim	ADAM	160 mg
Rosuvastatin	SV	ADAM	20 mg

Population parameters	Control	GLP‐1RA
Population	Healthy obese adults	Healthy obese adults
Population size	1000	1000
Number of trials	10	10
Trial size	100	100
Study duration	24 h	24 h
Female proportion	0.49	0.49
Age (years) a	18–85	18–85
Height (cm) b , c	148–178	148–178
Weight (kg) d , e	76–110	76–110
Fed gastric emptying (h)	1 (38%)	4 (38%)
Fed small intestine transit time (h)	3.32 (39.93%)	9 (39.93%)
Colon transit time (h)	24 (30%)	24 (30%)

Abbreviations: ADAM, Advanced Dissolution, Absorption, and Metabolism model; First order, First‐order absorption kinetics; Sim, files that have been developed using a “bottom‐up” approach; SV, files that have been optimized using in vivo data.

^a Age distribution Weibull (α = 2.81, β = 61 & α = 2.87, α = 60.91, male & female, respectively).

^b Male height (HT) = 177.07 + 0.0704 * age − 0.002 * age².

^c Female HT = 159.82 + 0.185 * age − 0.003 * age².

^d Male weight (WT) = 2.643 + 0.0099 * Male HT.

^e Female WT = 2.7383 + 0.0091 * Female HT.

2.3. Analysis

Simulated geometric means for AUC and C _max were utilized to determine the parameter ratios and respective 95% confidence intervals between the control and GLP‐1RA GI motility groups. Computed parameter ratios were then compared to available clinical data in adults with diabetes on semaglutide and liraglutide. A similar analysis was performed for the exploratory compounds to assess the impact of GLP‐1RA‐induced delays in GI motility on AUC, C _max, and T _max parameter ratios. Observed changes in the investigated PK parameters were further explored via sensitivity analyses using the Simcyp™ software. Local sensitivity analyses were performed using Simcyp's gradient analyzer.

3. RESULTS

In this analysis, the Simcyp™ Simulator V23 predicted the mean expected AUC and C _max ratios of atorvastatin, metformin, ethinyl estradiol, and digoxin combined with semaglutide within 20% of the package insert values (Table 2). The exception to this concordance was metoprolol, where simulations underpredicted the mean AUC and C _max ratio when combined with semaglutide by 30% and 37%, respectively (Table 2).

TABLE 2.

Mean (90% CI) predicted (simulated) and observed (package insert value) pharmacokinetic parameter ratios for oral medications coadministered with semaglutide. ⁵

Co‐administered medication	AUC ratio	C max ratio
Atorvastatin
Simulated value	1.19 (1.14–1.24)	0.76 (0.72–0.79)
Package insert value	1.01 (0.95–1.08)	0.66 (0.45–0.86)
Metformin
Simulated value	1.01 (0.98–1.04)	1.03 (1.00–1.06)
Package insert value	1.03 (0.97–1.08)	0.93 (0.86–0.99)
Metoprolol
Simulated value	0.89 (0.81–0.98)	0.95 (0.91–0.99)
Package insert value	1.19 (1.11–1.28)	1.32 (1.20–1.45)
Ethinyl estradiol
Simulated value	1.01 (0.97–1.05)	0.99 (0.89–1.10)
Package insert value	1.07 (1.05–1.10)	1.02 (0.99–1.07)
Digoxin
Simulated value	1.14 (1.12–1.16)	0.73 (0.71–0.74)
Package insert value	1.01 (0.98–1.06)	0.95 (0.88–1.02)

Abbreviations: AUC, area under the curve; C _max, maximum concentration; 90% CI, 90% confidence interval.

For reference, PBPK models generally achieve AUC ratio predictions within 20% to 50% error for well‐characterized DDIs. ¹²

3.1. Exploratory analysis substrate selection

In our analysis of the MEPS 2022 prescription drug survey, a total of 6,072,000 individuals prescribed GLP‐1RAs were identified, with the most common GLP‐1RAs prescribed being semaglutide (54%), followed by dulaglutide (33%) and liraglutide (13%) (Table S3). ² Medications selected for further analysis had a high frequency of coadministration with GLP‐1s and were available in the Simcyp compound repository (Figure S2). These medications included oral anticoagulants (dabigatran and edoxaban), bupropion, rivaroxaban, valsartan, and rosuvastatin.

3.2. Exploratory analysis

In the exploratory analysis of medications identified from the MEPS data set, the model predicted notable effects of peak GLP‐1RA delays on pharmacokinetic parameters, including the AUC_0–24h, C _max, and T _max for several coadministered medications (Figure 2). The model outputs for rosuvastatin, valsartan, and dabigatran indicate increases in AUC_0–24h by 64%, 90%, and 205%, respectively, suggesting enhanced overall exposure to these drugs over 24 h.

FIGURE 2.

Predicted relative exposure ratio and 95% CI for commonly co‐administered medications identified in MEPS 2022 survey data during peak GLP‐1RA effect on GI motility. Where AUC_0–24h, C _max, and T _max are denoted by bolded squares, triangles, and diamonds, respectively. AUC_0–24h, area under the curve from 0 to 24 h; C _max, maximum plasma concentration; T _max, time to reach maximum concentration; 95% CI, 95% confidence interval.

In addition to these AUC_0–24h changes, the simulations show a consistent delay in T _max across dabigatran, valsartan, and rosuvastatin, indicating a slower absorption rate. This T _max shift, combined with the relatively smaller increases in C _max, suggests that while peak concentrations were delayed, their magnitudes did not increase as dramatically. For instance, rosuvastatin and valsartan displayed only modest increases in C _max compared to their AUC changes, whereas dabigatran showed a relatively larger C _max increase.

3.3. Sensitivity analyses of dabigatran

Sensitivity analyses were conducted to investigate the impact of gastric emptying and small intestinal transit times on the pharmacokinetics of dabigatran (Figure 3). Dabigatran was selected for this analysis due to its highly variable absorption profile and clinical significance as an anticoagulant with a narrow therapeutic index. Additionally, similar analyses were performed for rosuvastatin and valsartan (Figures S3 and S4) to assess the broader effects of GLP‐1RA‐induced delays in GI motility on oral drug pharmacokinetics.

FIGURE 3.

The four panels illustrate the relationship between mean gastric emptying time and small intestinal transit time on key pharmacokinetic parameters of dabigatran using the ADAM absorption model. (A) Fraction absorbed (fa) as a function of transit parameters. (B) Maximum plasma concentration (C _max, mg/L) under varying transit conditions. (C) Area under the concentration–time curve (AUC, mg/L·h) showing drug exposure. (D) Time to reach maximum concentration (T _max, h), influenced by transit parameters. Each panel presents a 3D surface plot with color gradients representing different parameter values.

Local sensitivity analysis confirmed that the fraction of drug absorbed was primarily dependent on small intestinal transit time, as indicated by the positive correlation between small intestinal transit time and fraction absorbed (Figure 3A). In contrast, gastric emptying time had no observable impact on the fraction absorbed.

AUC was also primarily influenced by small intestinal transit time (Figure 3C). The 3D surface plot demonstrates a positive correlation between increasing small intestinal transit time and AUC, indicating that prolonged intestinal exposure significantly enhances overall drug absorption. In contrast, gastric emptying time had only a minor effect on AUC.

Conversely, T _max was strongly dependent on gastric emptying time, with minimal influence from small intestinal transit time (Figure 3D). The 3D surface plot illustrates that as gastric emptying time increases, T _max is delayed, suggesting that the time to peak concentration is primarily governed by the rate at which dabigatran leaves the stomach. Thus, slower gastric emptying leads to a prolonged T _max.

The relationship between GI motility parameters and C _max was more complex (Figure 3B). Increases in gastric emptying time were associated with a decrease in C _max, while reductions in small intestinal transit time also led to lower C _max values. This suggests that both delayed gastric emptying and faster small intestinal transit contribute to a reduction in peak plasma concentrations, highlighting a dynamic interplay between gastric and intestinal motility in determining C _max.

4. DISCUSSION

GLP‐1RAs have revolutionized obesity management, but their impact on GI motility introduces potential DDIs with orally administered medications, particularly those requiring precise dosing, such as anticoagulants and hormonal contraceptives. Despite their widespread clinical use, there is a lack of definitive guidance on managing these potential interactions. This study sought to bridge this gap by leveraging PBPK modeling with Certara's Simcyp™ software to quantify the effects of GLP‐1RA‐induced gastric motility delays on the AUC, C _max, and T _max of commonly coadministered oral medications.

The Simcyp™ Simulator V23 demonstrated reasonable predictive accuracy for AUC ratios across most coadministered medications; however, some simulated values deviated from the 90% confidence intervals reported in the semaglutide package insert (Table 2).

Notably, metoprolol exhibited the largest discrepancies, with the simulated AUC ratios underestimating the observed values (0.89 vs. 1.19). Additionally, C _max predictions showed significant variability, particularly for metoprolol, where the simulated C _max ratio (0.95) was markedly lower than the reported value (1.32). These deviations suggest potential limitations in the model's ability to capture drug disposition dynamics of a cytochrome P450 family 2 subfamily D member 6 (CYP2D6) substrate, which is associated with significant interindividual variability. ¹³ These findings overall were within the regulatory acceptance margin of error, which is 20%–50% for these types of DDI assessments. ¹² Future model improvements should consider enhancing gastric emptying kinetics, refining hepatic metabolism assumptions, and incorporating transporter interactions to improve predictive accuracy for drugs with complex pharmacokinetics.

Our exploratory analysis of coadministered oral medications identified from the 2022 MEPS database revealed that GLP‐1RA‐induced delays in GI motility had the most pronounced effects on Biopharmaceutics Classification System (BCS) class II drugs, such as dabigatran and rosuvastatin, which have low solubility and high permeability (Figure 2). These results align with existing literature, which suggests that BCS class II drugs are particularly susceptible to delayed gastric emptying and altered small intestinal transit times. ⁶

Clinicians should recognize that GLP‐1RA therapies can alter drug absorption and systemic exposure, as reflected in changes to AUC. However, these PK changes may not fully reflect the pharmacologic effects of the drug, particularly for drugs that have active metabolites. For example, although dabigatran showed an increase in AUC, we were not able to characterize its primary metabolite, dabigatran acylglucuronide, which also contributes to its anticoagulant effects. ¹⁴ As a result, changes in parent compound exposure alone may not fully predict therapeutic efficacy or bleeding risk. These considerations highlight the importance of evaluating the complete metabolic profile of affected medications when interpreting pharmacokinetic alterations.

Although the increased systemic exposure of certain drugs may not always lead to toxicity, it remains a concern for high‐risk medications such as direct oral anticoagulants. Rivaroxaban, which is primarily absorbed in the stomach, did not exhibit significant changes in AUC, T _max, or C _max, suggesting minimal interaction with GLP‐1RA‐induced motility delays. In contrast, apixaban, which is absorbed in the distal small intestine and colon, may be more vulnerable to prolonged transit times, potentially increasing drug accumulation and bleeding risk. ¹⁵ Given apixaban's twice‐daily dosing, clinicians should monitor for signs of altered drug exposure when coadministering GLP‐1RAs. A validated substrate profile was not available for apixaban in Simcyp™ V23 to test this potential.

Other drugs, such as rosuvastatin and valsartan, showed increased AUC without notable changes in C _max. Although this suggests enhanced overall exposure, the clinical implications differ. For rosuvastatin, dose adjustments may be necessary to prevent long‐term toxicity, especially in patients with renal impairment. In contrast, valsartan's pharmacokinetic changes may not warrant immediate intervention but could justify periodic monitoring to ensure blood pressure remains well controlled. These findings underscore the importance of considering individual drug characteristics when evaluating GLP‐1RA‐induced pharmacokinetic alterations.

Although much attention has focused on GLP‐1RA‐induced delays in gastric emptying and gastroparesis, there remains a significant gap in studies examining the impact of small intestinal transit time on the absorption of oral medications. Our sensitivity analysis shows that delays in small intestinal transit time can substantially increase both the bioavailability and overall drug exposure—an area often overlooked in regulatory evaluations. Given the significant influence of small intestinal transit time on drug absorption, these delays should be recognized as a critical factor in GLP‐1RA therapy and incorporated into future PBPK models and regulatory submissions assessing GLP‐1RA–drug interactions.

Our sensitivity analyses showed that delays in gastric emptying result in a prolonged T _max, which may impact the timing of peak drug concentrations and overall drug absorption. This effect has significant clinical implications, particularly for medications with unique absorption properties or sequestrant DDI. For instance, current guidelines for drugs with sequestration potential, such as levothyroxine, recommend taking them at least 4 h apart from interfering agents like phosphate binders or multivitamins to prevent reduced drug absorption. ¹⁶ However, under the influence of GLP‐1RA therapy, this interval may need to be adjusted, as delayed gastric emptying could alter drug binding dynamics and further reduce absorption, potentially leading to therapeutic failure. Given these pharmacokinetic changes, clinicians should consider staggering the administration of oral medications that are highly sensitive to sequestration in patients on GLP‐1RA therapy to maintain therapeutic efficacy.

Although the reasonable predictions of our model compared to literature values suggest that incorporating real‐world data from GLP‐1RA capsule endoscopy and colonic scintigraphy studies can effectively support PBPK model development for GLP‐1RA–drug interactions, several limitations must be acknowledged.

First, the transit times used in this model were derived from patients on liraglutide, one of the least potent approved GLP‐1RAs. More potent agents, such as semaglutide and dual GLP‐1/GIP receptor agonists, may exert a greater or synergistic impact on gastric motility. As such, our model may underestimate the effects of these agents on drug absorption and pharmacokinetics. Future PBPK models should incorporate data from these more potent incretin‐based therapies to improve predictive accuracy.

Additionally, the capsule endoscopy measurements used in this model were obtained from patients with diabetes mellitus, some of whom had documented gastroparesis, but did not include patients with obesity. Both conditions independently affect gastric motility, with diabetes often being associated with delayed gastric emptying. This model assumes independent effects of diabetes, obesity, and GLP‐1RA therapy on gastric motility, which may oversimplify gastric physiology. If these effects are found to be additive, the model may underestimate GLP‐1RA‐associated gastric hypomotility's impact on oral drug absorption; conversely, if they are redundant, the model may overestimate these effects. Since many patients with obesity initiating GLP‐1RA therapy for weight management do not have diabetes, the predicted changes in AUC for drugs such as rosuvastatin, valsartan, and dabigatran may differ in nondiabetic populations. Further research is needed to assess the accuracy of these predictions in individuals without diabetes.

Another key limitation is that although capsule endoscopy allows for direct comparisons with orally administered solid dosage forms and provides complete lower GI transit times, it does not replace radiolabeled scintigraphy studies, which remain the reference standard for measuring gastric emptying time. ¹⁷ However, scintigraphy typically does not capture small intestinal transit times, limiting its utility in evaluating the effects of altered motility on drug absorption. Given that small intestinal transit time was identified as a key determinant of AUC changes in our sensitivity analysis, future scintigraphy studies incorporating small intestinal motility assessments would provide additional clinical data to refine and validate these PBPK models.

Our model also has inherent limitations related to built‐in physiological assumptions, including a constant GLP‐1RA effect on gastric motility, without accounting for tachyphylaxis. This is a critical consideration, as GLP‐1RA treatment duration in patients with obesity is highly variable, ranging up to 24 months depending on weight loss goals. ¹⁸ The potential for diminishing effects on gastric motility over time remains largely unexplored and may influence long‐term drug absorption patterns.

Last, our model did not incorporate the effects of GLP‐1RA therapy on gastric pH. GLP‐1RAs have been shown to reduce gastric acid secretion, with some studies reporting up to a 20% increase in gastric pH. ¹⁹ This pH shift could alter the solubility and absorption of oral medications and may have synergistic effects when combined with other pH‐altering drugs, such as proton pump inhibitors and H2 receptor antagonists. Future PBPK models should integrate these factors to enhance pharmacovigilance and ensure the safe coadministration of GLP‐1RAs with other oral medications.

5. CONCLUSION

This study provides valuable insights into the pharmacokinetic interactions between GLP‐1RAs and coadministered oral medications, with a focus on how delays in GI motility—particularly small intestinal transit—affect drug absorption. Our PBPK simulations indicate that GLP‐1RAs can increase AUC and prolong T _max for several drugs, especially dabigatran, a narrow therapeutic index anticoagulant. These findings highlight the importance of PBPK modeling in predicting GLP‐1RA‐related DDI, with significant implications for improving drug safety and efficacy. Although further clinical validation is necessary, this study demonstrates the potential of PBPK tools to guide pharmacovigilance efforts amid the rapid expansion of GLP‐1RA treatment.

FUNDING INFORMATION

Levi Hooper is supported by an NIH T32 (T32TR004371) National Center for Advancing Translational Science (NCATS).

CONFLICT OF INTEREST STATEMENT

Dr. Pai is a member of the Pharmacotherapy Editorial Board. All other authors declare no conflicts of interest.

Supporting information

Appendix S1.

Hooper L, Liu S, Pai MP. GLP‐1RA‐induced delays in gastrointestinal motility: Predicted effects on coadministered drug absorption by PBPK analysis. Pharmacotherapy. 2025;45:211‐219. doi: 10.1002/phar.70007

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.