Oral absorption modeling of weakly basic drugs – Combining in vitro and in silico models
Katharina Krollik (1), Andreas Lehmann (1), Christian Wagner (1), Holger Kubas (1)
(1) Merck KGaA, Darmstadt, Germany
Introduction:
During formulation development of weakly basic drug candidates, special attention should be paid to potential drug precipitation during gastrointestinal transit due to a pH shift from acidic to neutral values. Intestinal drug precipitation may cause reduced drug absorption. Therefore, several biopharmaceutical precipitation assays have been developed to assess the supersaturation and precipitation behavior of weakly basic drugs and their risk for drug precipitation in the small intestine. However, the translation of experimental data to the in vivo environment is still challenging. In general, drug precipitation can be mainly described by two phenomena, nucleation from supersaturated solutions and subsequent crystal growth [1]. The population-balance based mechanistic modeling tool gPROMS formulated products (a PBPK software by Siemens) makes use thereof and provides an environment to estimate precipitation kinetics from experimental data and translating them to in vivo predictions.
Objectives:
Predicting intestinal precipitation of dipyridamole (DPM) and ketoconazole (KTZ) and its impact on the fraction absorbed (Fa) by deriving drug specific precipitation kinetics from experimental data and connecting the obtained parameters to a physiologically based absorption model.
Methods:
The precipitation behavior of DPM and KTZ was investigated in a micro-scale precipitation assay (PA). The drugs were dissolved in simulated gastric fluid pH 2 and transferred to fasted state simulated intestinal fluid (FaSSIF-V1). Precipitation was monitored by in line UV analytics [2]. The experiment was repeated with different drug doses in the artificial stomach, varying concentrations of solubilizers in the FaSSIF-V1 medium, and different transfer times (in total 13 concentration vs. time profiles). Next, the PA was recreated in the oral absorption tool from gPROMS (gCOAS model library, Siemens Process Systems Engineering Limited) to fit crystal growth and nucleation kinetics. The precipitation kinetics obtained for DPM and KTZ (plus experimental solubility and permeability data) were then used in gPROMS to predict drug absorption of an oral solution in the fasted state. Further, global sensitivity analyses (GSA) were performed for selected parameters.
Results:
Based on the in vitro input, it was possible to estimate nucleation and crystal growth kinetics for DPM and KTZ. After implementing these parameters to the oral absorption models, 88% (30 mg dose) and 52% (90 mg dose) Fa were predicted for DPM and 92% (100 mg dose) and 71% (300 mg dose) for KTZ. These results can be compared to a study in fasted adults, in which the amount of precipitated drug was measured in the intestine after administration of DPM and KTZ oral solutions. The authors observed only minimal amounts of precipitated DPM (≤7%) for 30 and 90 mg and limited precipitation (≤16%) for 100 and 300 mg KTZ [3]. For the lower doses, the model predictions of high Fa values and consequently low amounts of precipitated drug are in good accordance with the in vivo results. However, for the higher doses, precipitation seems to be overestimated. GSAs for the combined effect of dose and crystallization kinetics, revealed a strong impact of the nucleation kinetics on the predicted Fa. As nucleation occurs spontaneously and can also be initiated by external seeds (heterogenous nucleation) [1], this process can be associated with high variability. This hampers the extrapolation from the in vitro results to in vivo. Furthermore, the drug permeability was identified to play a crucial role in the prediction of precipitation. Highly permeable drugs can be absorbed very quickly after entering the small intestine, which leads to decreasing drug concentrations and therefore lower supersaturations compared to the thermodynamic solubility. This would yield in lower nucleation rates [1]. Consequently, uncertainty in the in vitro estimated permeability (via Papp/Peff correlation) should also be considered.
Conclusion:
In vitro drug precipitation data was successfully translated into in vivo Fa predictions. While the predictions for lower doses correlated well with available in vivo data, for higher doses the model overpredicted drug precipitation. Model uncertainties like the fitted nucleation rates and the drug permeability were identified as main impact factors for the predicted Fa. Consequently, more emphasis should be put on these values for precipitation modeling in the future.
References:
[1] D. Erdemir, A.Y. Lee, A.S. Myerson, Crystal Nucleation, in: A.S. Myerson, D. Erdemir, A.Y. Lee (Eds.), Handbook of Industrial Crystallization, Cambridge University Press, 2019, pp. 76–114.
[2] A. Lehmann, K. Krollik, K. Beran, C. Hirtreiter, H. Kubas, C. Wagner, Increasing the Robustness of Biopharmaceutical Precipitation Assays - Part I: Derivative UV Spectrophotometric Method Development for in-line Measurements, Journal of pharmaceutical sciences 111 (2022) 146–154.
[3] D. Psachoulias, M. Vertzoni, K. Goumas, V. Kalioras, S. Beato, J. Butler, C. Reppas, Precipitation in and supersaturation of contents of the upper small intestine after administration of two weak bases to fasted adults, Pharmaceutical research 28 (2011) 3145–3158.