Jonas Langeraert 1, Pieter-Jan De Sutter 1, An Vermeulen 1
1 Ghent University (Ghent, Belgium)
Introduction: Gut-targeted compounds present an opportunity to locally affect the gastrointestinal (GI) tract while simultaneously benefitting from a more desirable safety profile due to limited systemic absorption [1,2]. In order to predict drug concentrations in the intestinal and colonic tissue, a more elaborate colonic absorption model was released in the latest update to GastroPlus™ (version X.2, Simulations Plus, Inc, Lancaster CA, USA). The model provides a description of additional layers of the intestinal wall including mucus layers in the gut lumen, enterocytes, mucosa, submucosa and the muscularis layers. Additionally, the model expands the number of compartments to also represent the transverse, descending, sigmoid colon and rectum in the human gastrointestinal tract.
Objectives: To validate this new model using a gut-targeted compound (JNJ-68179280) developed by Johnson & Johnson for which intestinal and colonic tissue biopsies from a phase I clinical trial are available.
Methods: A PBPK model was developed using a middle-out approach to describe the plasma pharmacokinetics (PK) of JNJ-68179280. The distribution of JNJ-68179280 was described using a full PBPK model and hepatic uptake and efflux kinetics were modelled together with metabolic clearance and renal glomerular filtration to describe elimination, mealtimes and composition were incorporated into the simulations. Absorption was modelled using both the standard ACAT and the ACATPlus model separately to be able to compare intestinal and colonic tissue predictions of both models. Plasma concentrations of fasted cohorts of a single ascending dose study were used to optimize intestinal permeability and the Vmax for hepatic uptake and efflux. The PBPK model was verified using plasma concentrations from the fed cohort of the single ascending dose study and the plasma concentrations from all cohorts of the multiple ascending dose study. Duodenal tissue biopsy concentrations were available for two doses from the multiple ascending dose study, while sigmoid colon and rectal tissue concentrations were available across all doses. The average fold error (AFE) and the absolute average fold error (AAFE) between the predicted intestinal and colonic tissue concentrations and the observed tissue biopsy concentrations was used to compare the predictive performance of both models. Enterocyte concentrations in the ascending colon were compared against both sigmoidal and rectal tissue biopsy concentrations since the standard ACAT model does not include these compartments.
Results: Plasma PK parameters such as Cmax, Tmax and AUC0-last were predicted within 2-fold of the observed geometric means across all doses in the single and multiple ascending dose studies. Duodenal tissue biopsie concentrations were underpredicted by the standard ACAT model (AFE = 0.01, 0.06 & AAFE = 10.2, 18.2) and the ACATPlus model (AFE = 0.05, 0.03 & AAFE = 20.0, 35.9). Here, it was observed that the standard ACAT model overpredicted the sigmoidal and rectal concentrations (AFE values ranged from 5.32 to 20.3 and from 4.62 to 28.7 and AAFE values ranged from 5.46 to 20.3 and from 4.80 to 28.7 for sigmoid colon and rectal tissues, respectively). Conversely, the ACATPlus model provided predictions which were much more in line with the observed sigmoid colon and rectal tissue concentrations (AFE values ranged from 0.26 to 0.93 and from 0.17 to 0.81 and AAFE values ranged from 1.68 to 4.13 and from 1.94 to 5.88 for sigmoid colon and rectal tissues, respectively).
Conclusion: A full-body PBPK model was built which successfully described the plasma PK of JNJ-68179280 across multiple doses and dosing regimens. The ACATPlus model was able to describe local tissue disposition of JNJ-68179280 in the sigmoid colon and rectal tissue, while the standard ACAT model overpredicted the observed tissue concentrations despite plasma PK being well captured by the model. Duodenal tissue concentrations were not well described by both absorption models. Additional gut-restricted and non-gut-restricted compounds should be included in the validation to inform further modelling efforts and to continue to improve predictions of intestinal and colonic tissue concentrations. Furthermore, the ACATPlus model could prove valuable in predicting tissue concentrations in patients with GI diseases when the relevant physiological changes are incorporated.
References:
[1] Filipski K, Varma M, El-Kattan A, Ambler C, Ruggeri R, Goosen T, Cameron K. Intestinal Targeting of Drugs: Rational Design Approaches and Challenges. Curr Top Med Chem 2013;13:776–802. https://doi.org/10.2174/1568026611313070002.
[2] Dorel R, Wong AR, Crawford JJ. Trust Your Gut: Strategies and Tactics for Intestinally Restricted Drugs. ACS Med Chem Lett 2023;14:233–43. https://doi.org/10.1021/acsmedchemlett.3c00001.
Reference: PAGE 34 (2026) Abstr 12237 [www.page-meeting.org/?abstract=12237]
Poster: Drug/Disease Modelling - Absorption & PBPK