Artificial Stomach and Duodenum: A Dynamic Dissolution Approach


The characterization of dissolution and solubility for an oral dosage form is of significance to the development and optimization of drug delivery, especially for BCS Class II and IV drugs of low solubility, where absorption is substantially influenced by these factors. While it is necessary to develop an accurate in vitro approach to predict the drug dissolution behavior in the gastrointestinal (GI) tract, current standard USP dissolution apparatus with fixed volume and static environment are not able to represent such dissolution behavior.

The purpose of this study was to design an artificial stomach and duodenum (ASD) unit for dynamic dissolution testing of Class II and IV drugs that allows for dynamic in situ measurement of dissolution in the simulated GI tract, taking into account active changes of fluid pH, volume, and composition. This unit modifies a previous model designed by Carino et al. [1], by incorporating a cage filter on the UV-Vis probes, which fundamentally solves the issue of the light obscuration by suspended solids. This filtration design is straightforward and simplistic and can be conveniently installed in an ordinary laboratory without special equipment or machining requirements. In addition, this ASD unit employs biorelevant dissolution media to closely simulate the environment of human/dog GI tract, and therefore provides data that are more relevant for correlation to in vivo tests and is more useful for solid form selection and formulation development.



Four programmable peristaltic pumps are connected to perform all fluids transportation in the GI tract, including the addition of simulated gastric and intestinal fluids, fluid transfer from stomach to duodenum, and removal of waste fluid from duodenum. The pH and concentration of the dissolving dosage are continuously monitored with pH meters and UV-Vis transmission dip probes. An analytical model is established to validate the ASD unit and the results are compared with a quasi-static test based on salicylic aqueous solution. For suspension dosage solid forms, particulates and flocculation notably complicate the UV-Vis data and can make results impossible to interpret. In this study, we install a USP standard basket on the UV-Vis probe to inhibit solids blocking the lights transmission path. Biorelevant dissolution media, which have major advantages over common compendial test media, are used to simulate the in vivo physiological environment. A bath circulator is incorporated to stabilize the temperature in the stomach and duodenum chambers during the entire process of testing.

Configuration of the ASD unit

The filter prevents the interference from large solids, agglomerates and air bubbles.

UV-Vis data of carbamazepine (CBZ) with and without filter

The spectrum of high dosage CBZ aqueous suspension, without filter, contains much noise and the high baseline leads to saturation, and therefore, unanalyzable data.

A validation for the filter was performed on salicylic acid (SS) solution. The above results show that the diffusion was not limited by the filter.

A validation for the ASD was performed on SS solution. The experimental results were consistent with the data obtained from an analytical model described below.

ASD test on carbamazepine 3 forms:
The ASD tests were performed for CBZ form I, III and dihydrate. SIF powders were used to simulate the fasted state in the GI tract. 40 mg compound was contained in a hypromellose capsule. A metal sinker was used to avoid flotation. During tests, both chambers were continuously stirred to maximize consistency of concentration. Experimental conditions are listed below.

Above is a typical raw UV-Vis absorption profile for CBZ. Before the capsule is disintegrated the spectrum is zero. After CBZ is released from the capsule, dissolution starts and the absorption at characteristic wavelength is converted to concentration.

For AUC, CBZ Form I > Form III > Dihydrate, which is consistent with the reported in vivo results (high dosage tests on dogs) [2].

For C(max), CBZ Form III > Form I > Dihydrate, which agrees with the previous in vitro results [1].

In conclusion, this ASD unit can be used to generate dissolution and pH profiles in stomach (not shown here) and duodenum. The results are consistent with both in vivo and in vitro data in literatures.

ASD test on itraconazole (ITZ) and the solid dispersion
The ASD tests were performed for ITZ and ITZ-HPMCP 1:2 dispersion. The dispersion has significantly higher solubility in the duodenum than ITZ, which is consistent with the in vivo results reported in the literature [3]. The signal/noise ratio here is much lower than the data for CBZ and salicylic acid, due to the low solubility of ITZ that the UV absorption is close to the detection limit of the UV-Vis device.


Full computer-controlled automation is achieved for this ASD unit, including continuous in situ data collection for UV-Vis absorption and pH in both the stomach and duodenum chambers. The experimental data of dynamic tests based on salicylic acid agree with the analytical model. The cage filter does not significantly influence diffusion, while it efficiently helps UV-Vis data collection avoiding the interference from large amount of solids. Two well-known drugs in solid dosage forms are tested and the results are in good consistency with data found in the literature.


A modified and user-friendly ASD unit has been successfully designed using a dynamic in vitro approach to generate dissolution profiles for drugs in a simulated GI tract. The instrument was validated using an analytical model compared to a quasi-static test. Biorelevant media are used to closely represent the environment in human GI tract, and therefore the results have critical advantages in in vivo-in vitro correlation and are especially valuable in designing in vivo experiments. As a result, ASD can be used to understand solid form and formulation issues for BCS Class II and IV drugs where solid form changes play a role in pharmacokinetics.


The help and support from Yizheng Cao, Jing Teng, Jon Selbo, and Chris Seadeek is greatly appreciated.


[1] SR Carino, DC Sperry, M Hawley. Journal of Pharmaceutical Sciences, 95 (2005) 116-125
[2] Y Kobayashi, S Ito, S Itai, K Yamamoto. International Journal of Pharmaceutics, 193 (2000) 137-146
[3] D Engers,  J Teng, J Jimenez-Novoa et al.. Journal of Pharmaceutical Sciences, 99 (2010) 3901-22