(W0930-09-58) Evaluating the Biopredictivity and Robustness of In Vitro In Vivo Correlation (IVIVC) for Acetaminophen Absorption Using TIM Advanced In Vitro Gastrointestinal Models

Purpose: The application of advanced physiologically-relevant in vitro gastrointestinal (GI) systems is growing for the study of GI transit and in situ drug release, dissolution and absorption. Results can be used to correlate with deconvoluted in vivo data as part of the modelling strategy to accelerate development timelines. Growth of this approach may be hampered by the lack of validation data that demonstrates repeatability and reproducibility of biorelevant in vitro systems, compared to conventional United States Pharmacopeia (USP) apparatus [1. The purpose of this work is to investigate the degree that inter- and intra-lab variability in paracetamol absorption measured in the tiny-TIM and TIM-1 in vitro GI models impacts predicted paracetamol absorption when projected through in vitro in vivo correlation (IVIVC) [2].

Methods: Replicate-level TIM in vitro absorption profiles (not presented in this work) were collected from 3 labs for TIM-1 and from 5 labs for tiny-TIM, and the most divergent absorption profiles were identified. These profiles represented the upper (A) and lower (Z) extremes of TIM results spanning the complete inter- and intra-lab range of variability. The A and Z TIM absorption profiles were then plotted against deconvoluted in vivo paracetamol PK data from Souliman et al. [3] in the form of in vivo paracetamol absorption profiles. This facilitated two separate IVIVCs to be established for the absorption of oral paracetamol solution. As a means of testing the impact of reproducibility and repeatability of TIM in vitro data on biopredictivity of in vivo paracetamol absorption, all TIM replicate profiles (13 runs from tiny-TIM and 10 runs from TIM-1) were used as an input to the A and Z IVIVCs. The resulting range of predicted in vivo acetaminophen absorption was evaluated in terms of prediction error (PE) and bandwidth. The former representing biopredictivity of the TIM systems and the latter to quantify how robust the biopredictivity is robust to inter- and intra-lab variability of the TIM systems.

Results: The tiny-TIM and TIM-1 in vitro paracetamol absorption profiles were biopredictive of fasted state in vivo paracetamol absorption profile with level A IVIVCs (R² = 0.99, Figure 1) established for runs A and Z of each system. The average prediction error ranged from 3.21 (A) – 6.48 (Z) % for TIM-1 and 7.24 (A) – 8.20 (Z) % for tiny-TIM (Figure 2). PE was highest for the early absorption phase and reduced to 1.99 - 4.92 % and 4.62 - 5.05 %, respectively, considering only data where > 40 % paracetamol was absorbed. The high reproducibility of TIM in vitro results translated to a highly reproducible prediction of in vivo paracetamol absorption, demonstrated by the relatively small bandwidth of predicted in vivo paracetamol absorption, spanning 4.14 and 3.15 % of the central predicted in vivo data for TIM-1 and tiny-TIM respectively.

Conclusion: This study demonstrated that GI transit and absorption of paracetamol in TIM-1 and tiny-TIM in vitro gastrointestinal (GI) models was biopredictive of in vivo paracetamol absorption in adults through generation of level A IVIVCs. When used in this approach, the variability attributed to the in vitro test method is low due to the high intra-lab repeatability and inter-lab reproducibility of tiny-TIM and TIM-1.

References: [1] S. Mondal, S. Kollipara, M. Chougule, A. Bhatia and T. Ahmed, "Biopredictive dissolutions for conventional oral IR, MR and non-oral formulations - current status and future opportunities," Journal of Drug Delivery Science and Technology, vol. 97, 2024.
[2] M. Minekus, "The TNO Gastro-Intestinal Model (TIM)," in The Impact of Food Bioactives on Health, pp. 37-46.
[3] S. Souliman, S. Blanquet, E. Beyssac and J. M. Cardot, "A level A in vitro/in vivo correlation in fasted and fed states using different methods: applied to solid immediate release oral dosage form," vol. 27, no. 1, 2006.

Acknowledgements: The authors would like to acknowledge the contributors to the TIM Ring Study performed within the OrBiTo project that received support from the Innovative Medicines Initiative Joint Undertaking (www.imi.europa.eu) under Grant Agreement No. 115369, resources of which were composed of financial contribution from the European Union’s Seventh Framework Program and in kind contributions from EFPIA companies.

Linear level A IVIVCs generated with upper and lower bounds of replicate data for the tiny-TIM (left) and the TIM-1 (right) for oral paracetamol absorption [3].

Predicted versus observed human paracetamol absorption profiles [3] using linear IVIVC models from the tiny-TIM (left) and TIM-1 (right). Each prediction band represents the range of intra- and inter-lab reproducibility. The grey shaded region represents 80 – 125 % of the clinical data for visual purposes since variability measured in the clinical study was not available.