Biorelevant Dissolution Tests with the Flow-Through Apparatus?

Eleftheria Nicolaides(1), John M. Hempenstall(2) and Christos Reppas(1)
(1)School of Pharmacy, University of Athens, Athens, Greece
(2)Glaxo Wellcome R&D, Pharmaceutical Sciences, Hertfordshire, England

Introduction
Efficient simulation of the in vivo dissolution process of oral dosage forms serves two purposes: 1.) identification of those dosage forms for which the in vivo performance is limited by the dissolution process, and 2.) prediction of the in vivo performance of a dosage form with dissolution limited absorption.

Establishment of physiologically relevant test conditions in vitro requires consideration of several issues, including media compositions and volumes, hydrodynamics, duration of the test, and analysis/interpretation of the data. Decisions on these issues are primarily limited by the knowledge of the conditions in the gastrointestinal (GI) lumen (1).

The importance of medium composition in forecasting the in vivo performance of immediate release dosage forms of drugs with dissolution limited absorption has been recently demonstrated (2,3) utilizing a closed in vitro system, i.e. the rotating paddle apparatus (USP 23, Type II Apparatus) (4). In the present report we address some technical and physiologically relevant issues of an open system, i.e. the flow-through tester (USP 23, Type IV Apparatus) using an immediate release product of troglitazone. Troglitazone, an orally active antidiabetic (5), is a weak acid (pKa1 6.1, pKa2 12.0), with an aqueous solubility of the crystalline form of about 1.93 mg/ml (water, 37 °C) and a logP of 2.7 (3).

Materials and Methods
Troglitazone tablets (200 mg, batch# D157/155D) were provided by GlaxoWellcome, UK. Sodium Taurocholate 98% pure, lot # 15H5001 was purchased from Sigma-Aldrich, St. Louis, USA. Egg ­Phosphatidylcholine (99.1% pure) was purchased from Lipoid GmbH, Germany (Lipoid E PC, lot # 76H8380). Potassium dihydrogen phosphate, potassium chloride and, sodium lauryl sulfate were purchased from Fisher Scientific, Leicestershire, UK.

The dissolution behavior of troglitazone tablets was tested in water (flow rates: 4ml/min and 8ml/min), Fasted State Simulated Intestinal Fluid (FaSSIF; flow rates: 2.5 and 4ml/min), Fed State Simulated Intestinal Fluid (FeSSIF; flow rates: 4ml/min and 6ml/min) and Fasted State Simulated Gastric Fluid (FaSGF; flow rates: 4 ml/min and 12ml/min). Compositions of FaSSIF, FeSSIF and FaSGF have been reported elsewhere (1,2). Selection of appropriate flow rates was based on lumenal flow rates in the upper GI tract (1).
A Sotax flow-through dissolution tester (Sotax Ltd., Basel, Switzerland) was used for all dissolution tests. According to the USP 23 specifications (4), the tablet can be mounted initially on a holder. Also, in addition to the 5mm-size bead which is used to prevent back flow of solid material, 1mm-size beads can be added to the bottom of the cell.

Dissolution experiments in each medium were run in triplicate, with the exception of FeSSIF at 6 ml/min (n=1). Drug concentrations were measured at 5, 10, 15, 30, 45, 60, 75, 90, 120, 150 and 180 min after the beginning of the experiment. To assess the dissolution behavior of a single tablet in FaSGF and FaSSIF, one further experiment was performed by initially testing the dissolution in FaSGF (flow rate: 12 ml/min). At 30 minutes the medium was changed to FaSSIF (flow rate: 4ml/min) and dissolution continued to be monitored for an extra of 180 minutes.

Analysis of all samples was made with HPLC (Hewlett Packard series 1050). The mobile phase was comprised of 60 : 40 : 0.08 acetonitrile : water : orthophosphoric acid, the flow rate was 1.4ml/min, and troglitazone was detected at 230nm. Solutions were protected from light (6). 50µl of appropriately diluted samples containing 9-acetylanthracene (internal standard) were injected onto a Spherisorb S5-ODS2 (150x4.6mm) column. Linear standard curves were constructed for every sample set. Coefficients of determination were at least 0.994, none of the intercepts was statistically significant, and the coefficient of variation of the slopes in a specific medium was less than 12%.

Results
The dissolution data are presented in Table 1.

*One measurement

Dissolution in water is minimal regardless of the use of beads and/or holder. The effect of tablet holder on the dissolution rate in water is shown in Figure 1. Mounting the tablet on the holder results in a decrease in dissolution rate when beads are present, but an increase in the absence of beads. When the tablet is placed directly on top of the beads and not on the holder, wetting of tablet occurs earlier, leading to an increased dissolution rate (Figure 1a).Placing the tablet in the dissolution cell without beads, results in immediate wetting of the tablet (from the bath fluid) prior to the onset of flow of the dissolution medium. In this case, the solids adhere to the sides of the lower part of the cell and dissolution is slower because of the smaller effective surface area of the solid (Figure 1b).

Figure 1: Mean±SD % cumulative dissolution profile of Troglitazone in water at a flow rate of 4ml/min. (a) In presence of beads; (b) In absence of beads.

Using the same data, the effect of beads on the dissolution rate can be better seen from Figure 2.

Figure 2: Mean±SD % cumulative dissolution profile of Troglitazone in water at a flow rate of 4ml/min. (a) Holder was used; (b) Holder was not used.

When the holder is used (Figure 2a), the presence of beads does not substantially affect the dissolution profile. However, without holder, the beads increase the dissolution rate (Figure 2b).
Dissolution in FaSGF was studied by initially placing the tablet at the bottom of the cell. The data show more complete dissolution than that observed in water (Table 1), i.e. the lower pH of FaSGF, compared to water, is more than compensated for by the wetting effects contributed by sodium lauryl sulfate (composition of FaSGF is given in Reference 1).

Figure 3 shows the dissolution of troglitazone in FaSSIF and FeSSIF at a flow rate of 4 ml/min with no beads present and the tablet initially mounted on the holder. As expected, the presence of lecithin/bile salt increases the dissolution rate of troglitazone substantially. Despite the lower pH of FeSSIF, dissolution is more complete in this medium than in FaSSIF due to the increased concentration of bile salts and lecithin.

Figure 3: Mean±SD % cumulative dissolution profile of Troglitazone in FaSSIF and FeSSIF. The flow rate was 4 ml/min. Tablet was mounted on the holder in absence of beads.

Figure 4 shows the effect of the flow rate on the dissolution of troglitazone in FaSSIF (2.5 and 4 ml/min) and FeSSIF (4 and 6 ml/min). As would be expected, the dissolution rate increases with the flow rate.

Figure 4: The effect of flow rate on the mean±SD % cumulative dissolution profile of Troglitazone in FaSSIF (a) and FeSSIF (b) Tablet was mounted on the holder in absence of beads.

Discussion
Dissolution data in water suggest that, at physiologically relevant flow rates, in vitro dissolution profiles generated using USP 23, Apparatus IV, can be affected by the use of holder and/or beads. Data from other products / drugs are needed before these results can be generalized and to confirm the direction of the changes. The same set of data also suggest that the presence of beads decreases variability. Regardless of the use of holder, the coefficients of variation (CVs) for the data at 4ml/min. estimated from the standard deviations presented in Table 1 ranged from 12.8 to 40.0%. The corresponding CV values in the presence of beads ranged from 0.7 to 7.4%. At flow rates utilized in the present study, the difference on the variability of the data cannot be attributed to a possible alteration of the type of flow [regardless of the presence of beads, the flow pattern is always laminar (7)] but rather to the consistency of the laminar pulse applied on the solid materials from the fluid stream (7).

Experiments in FaSSIF and FeSSIF were run without beads and the tablet was initially mounted on the holder. Although the test conditions (no beads/holder) were deliberately chosen to maximize the variability in the results, CVs in FaSSIF (1.3-10.4%) and FeSSIF (4.4-12.6%) are somewhat lower than the CVs for data with the same formulation obtained using the USP 23 Apparatus II [up to about 20% (3)].

Troglitazone is a drug with low solubility and high lipophilicity characteristics (3) and, therefore, appropriately designed dissolution tests can provide a picture of the in vivo performance of immediate release products of this compound (1,8). Open systems are more attractive than closed systems in simulating intestinal dissolution of drugs because, in principle, in vivo hydrodynamics can be better simulated. However, the total % dissolved in FaSSIF and FeSSIF at 3 hours (which corresponds roughly to small intestinal residence time) is less than one would expect from the oral absorption data of troglitazone from the product tested (GlaxoWellcome data on file). A single experiment, in which the tablet was first subjected in FaSGF before testing its dissolution in FaSSIF (Table 1), resulted in greater dissolution in the first 15 minutes in FaSGF than in 3 hours in FaSSIF. Considering the fact that troglitazone is a poorly soluble weak acid, these results are unlikely to be representative of the in vivo dissolution behavior.

These discrepancies are thought to be attributable to inherent design problems with the in vitro setup and to the inadequate simulation of the in vivo hydrodynamics and obviously need to be studied further. A problem with all the in vitro testers is that they do not account for radial loss (absorption) of the drug. In conjunction with the specific hydrodynamics utilized in this study ("slow" flow rates, vertical positioning of the cell, and possible sedimentation of the solids), absence of radial loss led to non-sink conditions in most cases. This was confirmed by comparing the measured concentrations with the solubility of troglitazone in the various media (3). With respect to the hydrodynamics, in the present study a piston pump was used. This pump does not allow for simulation of bidirectional movement of chyme or segmental mixing, and was operated at flow rates which represent the average net flow in the aboral direction. In order to match the actual hydrodynamics found in the small intestine, an alternative flow pattern may be required.

References
1. Dressman JB, Amidon GL, Reppas C, and Shah VP. Dissolution as a Prognostic Tool for Oral Drug Absorption: Immediate Release Dosage Forms. Pharm. Res. 15:11-22 (1998)
2. Galia E, Nicolaides E, Hörter D, Löbenberg R, Reppas C, and Dressmnan JB. Evaluation of Various Dissolution Media for Predicting In Vivo Performance of Class I and Class II Drugs. Pharm. Res. 15:698-705 (1998)
3. Nicolaides E, Galia E, Efthymiopoulos C, Dressman JB, and Reppas C. Forecasting the In Vivo Performance of Four Low Solubility Drugs from their In Vitro Dissolution Data. Pharm. Res. 16:1877-1883 (1999)
4. The United States Pharmacopeia (USP 23), pages 1792-1795, United States Pharmacopeial Convention, Inc., Rockville, MD (1995)
5. Johnson MD, Campbell LK, Campbell RK. Troglitazone: Review and Assessment of its Role in the Treatment of Patients with Impaired Glucose Tolerance and Diabetes Mellitus. Ann. Pharmacother. 32:337-348 (1998)
6. Fu Y, Shen C, Fujita T, and Foote CS. Photooxidation of Troglitazone, a New Antidiabetic Drug. Photochem. Photobiol. 63:615-620 (1996)
7. Papaioannou A. Flow through layers of solids (Greek) in: Mechanics of Fluids, Volume II, Chapter 5, National Polytechnic University Editions, Athens, 1996.
8. Amidon GL, Lennernas H, Shah VP, Crison JR. A Theoretical Basis for a Biopharmaceutic Drug Classification: The Correlation of In Vitro drug Product Dissolution and In Vivo Bioavailability. Pharm. Res. 12:413-420 (1995)0