Konstantinos Stamatopoulos, Shriram M. Pathak, and David Turner
Simcyp Limited (a Certara Company), Sheffield S1 2BJ, United Kingdom.
Objectives: Bile salts (BS) play a crucial role in the solubilisation and the absorption of lipophilic and poorly soluble drug compounds. The concentration of BS, however, varies significantly within the GI tract due to a number of factors such as cyclic motility patterns of gall bladder (GB) linked to the interdigestive migrating motor complexes (IMMC) [1,2], dynamic changes within the GI luminal fluids and most importantly the prandial state. As opposed to the fasted state, strong GB contractions empty a significant amount of BS in response to the feeding events. Thus, a novel mechanistic BS model was developed and coupled with a previously developed dynamic GI fluid volumes model [3], to predict the intestinal BS concentrations as a function of time and prandial status.
Methods: The distinctive characteristics of the model include the generation of individualised IMMC cycle patterns; IMMC associated GB filling and emptying phases; hepatic bile acid mass secretions rates and regional bile acid absorption kinetic parameters. The number of IMMC events within 24 hours, duration of the IMMC cycle based on its origin, either from stomach or duodenum, and their corresponding proportions is defined priori for every individual. The Model also accounts for the inter-individual GB volumes (GBV); Initial amount of bile acid in GB (mmol); and GB residual volume in the fasted and high- low-fat meals. The GB filling phase takes place during the 0–30% of the IMMC cycle followed by an emptying phase that terminates at 60% or 90% of the IMMC cycle depending on where it had a duodenal or antral origin respectively. Using this relationship, the individualised GB filling-emptying cycles could be generated according to the number of IMMC events taking place per 24 hour period. Within the model, the GB filling rate is linked to the hepatic bile secretion rate (mL/h). Thus, the hepatic bile flow rate is back-calculated from the slope of the linear regression of GBV vs IMMC cycle time during the filling period. In the fed state, the GB contracts strongly, emptying a significant proportion of its content (ejection fractions can reach 95% [4] compared to 30% in the fasted state [2]). Furthermore, the GB empties its content within an hour of the meal ingestion followed by a refilling period which can last up to ~6 h [4] after which the cyclic fasted GB motility profile is restored. The total bile mass secretion in the duodenum is assumed well mixed with the dynamic luminal fluid volume and thus can predict the time dependent BS concentration within the GI tract. The model includes re-absorption of BS from the luminal fluids and kinetic parameters defining passive (jejunum and colon) and/or active (ileum) components has been used. The performance of the proposed model was assessed by comparing the predicted BS concentration with those reported from in vivo studies [5].
Results: Twenty-four hours fasted motility profiles of the GB were generated together with a meal ingested at a random time during this period. The results of the simulations using 1000 virtual individuals showed cyclic fluctuations of BS concentration in the duodenum, with ranges of 0.25-37 mM (fasted) and 0.45-48 mM (fed) which is within the ranges reported in the in vivo studies (fasted: 0.03-36.18 mM; fed: 0.74-86.14 mM, [5]). Multiple minor peaks were also observed in the duodenal fluid volumes, following the discontinuous release of biliary secretions. However, these peaks are diminished as moving towards ileum compartment. Similar results have been reported from MRI studies, showing time-dependent fluctuations of intestinal fluid volumes which in part can be attributed to the bile fluid dynamics [6, 7].
Conclusions: Quantitative estimation of time dependent BS concentration within the GI tract is critical for the accurate prediction of oral drug absorption within the PBPK modelling framework. The developed dynamic BS model, coupled with GI fluid dynamics, could successfully describe the multiple BS peaks observed within the in vivo study [1] and also predicted the BS concentrations within the known physiological range [5]. This dynamic model would also form a basis of mechanistic enterohepatic re-circulation of drug and/or metabolites modelling within the PBPK framework.
References:
[1] Peeters TL, Vantrappen G, and Janssens J. Gastroenterology (1980) 79: 678-681.
[2] Stolk et al. Am J Physiol. (1993) 264: 596-600.
[3] Jamei et al. The AAPS journal (2009) 11(2): 225–237.
[4] Keszthelyi et al. Eur J Nutr. (2013) 52(4): 1417-1420.
[5] Riethorst et al. J Pharm Sci. (2016) 105(2): 673-681.
[6] Mudie et al. Mol Pharm. (2014) 11(9): 3039-3047.
[7] Yu et al. AAPS J. (2017) 19(6) :1682-1690.
Reference: PAGE 27 (2018) Abstr 8669 [www.page-meeting.org/?abstract=8669]
Poster: Drug/Disease Modelling - Absorption & PBPK