Tightening up buccal testing

Dr Lyn Hughes, from Rohm & Haas, US, discusses new techniques for buccal and gastrointestinal dissolution testing

In recent years there have been a number of advancements in dissolution testing, both in the type of apparatus used and in the technology required to operate it. These advances, along with an increasing focus on biorelevancy and in vivo-in vitro correlation (IVIVC), have led to a much greater confidence in the quality of the drug formulations, but have done little to promote the use of dissolution testing as a tool for formulations development.

Methods of dissolution testing vary and there has recently been an increased interest in liquid and dissolve-in-the-mouth dosage forms to improve patient compliance. However, buccal dissolution testing systems have very different requirements from gastro intestinal (GI) testing methods. For example, current methods of GI testing do not provide information about solids transfer (with the exception of USP apparatus 3), absorption or clearance. In addition, they require the manipulation of complex mathematical models, which tend to have limited applicability so that changing the dosage form (e.g. immediate release to extended release) requires the use of a different model. These problems seriously limit the use of dissolution testing for predictive purposes.

taste issues

Dissolution testing for dissolve-in-the-mouth dosage forms presents a different range of problems to GI dissolution testing and there is an increased need for appropriate systems to characterise this type of dissolution. If a patient is able to taste a drug, even when it is only partially dissolved in the mouth, and it is unpleasant, the whole aim of improved compliance is lost.

Even when taste is not an issue, current methods, such as variants of the USP dissolution test, cannot easily be used to characterise these dosage forms. The reason for this is the speed of disintegration and dissolution compared with typical dosage forms designed to be swallowed whole.

Reported here is a brief outline of two new methods, one for GI testing and the other for buccal, which resolves many of these limitations. The method for GI testing incorporates disintegration, solids transfer, dissolution, changing pH, composition of fluids, absorption and clearance under biorelevant conditions.

No mathematical model is needed and the data from the equipment is directly comparable to blood plasma concentration-time profiles. The technique is capable of giving excellent Level A IVIVC and appears to be predictive across different dosage forms of the same drug. The method for characterising buccal dissolution has been found to correlate well with human taste perception as well as being rapid and reproducible with applications to common liquid and dissolve-in-the-mouth solid dosage forms.

GI dissolution testing

The equipment comprises three continuous, stirred cells (a gastric cell, an intestinal cell and a systemic cell) connected in series and operated at a constant volume (figure 1).

The tablet dosage form is added to the gastric cell with simulated gastric fluid that allows disintegration to take place under biorelevant conditions, unlike tests, which use intestinal-type fluids. A dip tube allows solution and fine solids to leave this cell and transfer to the intestinal cell, simulating the in vivo transfer of undissolved solids from the stomach to the intestine.

It is mixed with simulated intestinal fluid and sodium hydroxide (to neutralise the gastric acid) before entering the intestinal cell.

In the intestinal cell, a typical intestinal pH is maintained while the drug continues to dissolve, although a pH probe linked to the supply of sodium hydroxide allows the pH to be maintained at any desired value.

Fluid must pass through a filter to exit the cell, so only dissolved drug can go on to the systemic cell, producing an overall effect of removal of the drug equivalent to intestinal absorption. The rate of removal will be the first order in drug concentration - a very good approximation of the in vivo absorption kinetics.

drug evaluation

Finally the dissolved drug enters the systemic cell which is simply a constant volume stirred cell. In this cell, the removal kinetics will also be first order, and can be considered equivalent to clearance from blood plasma.

The method has been evaluated using four different drugs: ibuprofen (Advil), pseudoephedrine (Sudafed), paracetamol (Tylenol) and diclofenac, which represent a wide range of absorption, clearance and solubility characteristics. All of the results showed excellent Level A IVIVC as classified by the FDA system, even across dosage forms.

Figure 2 shows the graphical data for the ibuprofen drug. The data points show the published in vivo data and the line represents the experimental data directly from the test results. The only mathematical manipulation used is the conversion of UV absorption data to concentration data and linear scaling of the concentrations so that the two datasets can be shown together.

Test conditions defined using the IR form of diclofenac were used to test the ER version and proved to be predictive for both (figure 3). Considering the fundamental principles of this method, these results are likely to be general.

The method has been extended to include non-disintegrating dosage forms and there are several other possible extensions including altering the pH and composition in the cells during a test to incorporate the effects of pH change in the intestinal tract. Fluids other than USP dissolution fluids can be used, such as different surfactants, ions, enzymes etc., to produce fluids that are more biorelevant.

In conclusion, this method allows data to be interpreted quickly and effectively 'at the bench' without the need for extensive mathematical manipulation. The optimum test conditions are determined during test development by adjusting the residence time in each cell until a good fit with the in vivo is obtained, which makes the task of defining conditions much easier. As a QC/QA method, it can give far more information than the standard USP-type tests and is a more useful tool in product development because it assesses disintegration, dissolution, absorption and clearance under bio-relevant conditions.

It has the potential to reduce both the amount of animal testing, and development time and cost. Use of surrogate dissolution parameters, such as fraction dissolved, time to 70% etc., can be eliminated.

buccal dissolution

When characterising buccal dissolution, there are a number of factors that differ from GI dissolution, such as small volume, short residence time, solids transfer, composition and incomplete dissolution. Most USP dissolution tests use large volumes of solution for complete dissolution of the active ingredient. However, in buccal dissolution, the volume of saliva is very small and the residence time in the mouth is short as most of the dosage is swallowed within a minute (with the exception of lozenges). In addition, complete dissolution is not usually required, or even desirable, as is the case for fast-melt tablets such as the Zydis system where complete disintegration is required. This means that the removal of finely divided solids from the test vessel is absolutely critical to get any type of biorelevant test.

simulated formulae

The new system comprises a single, stirred, continuous flow-through filtration cell with a dip tube to remove finely divided solid particles (figure 4).

The volume of liquid in the cell is very small (10 ml) and fluid is pumped through to give a short residence time with almost complete removal in about eight minutes. The cell is filled and flows are set up and allowed to reach steady-state before the dosage form (solid, liquid, suspension or powder) is introduced.

The filtered sample is analysed in-line (e.g. by UV flow-thru cell) or samples are collected in a fraction collector for later analysis. The dissolution fluid is simulated saliva formulated from published data, as there is no USP recommended simulated saliva.

The equipment has been tested using Claritin RediTabs, fast-melt tablets of loratidine, a liquid formulation of paroxetine and dissolve-in-the-mouth tablets of Drug X*.

Claritin Reditabs are not a taste-masked formulation but demonstrate the ability of the equipment to handle this dosage form. The results show a peak in the concentration-time curve (figure 5), which is characteristic of all the dosage forms tested and is an inevitable result of a continuous, flow-thru system where a compound is added in a single bolus.

For a particular dosage form the time at which this peak occurs is approximately constant at a constant flow rate, but the height (i.e. peak concentration) varies with the amount of drug that dissolves and the speed at which it dissolves. It is the peak concentration that has been found to correlate to taste.

Only about 13% of the loratidine actually dissolves (figure 5), which is still acceptable as complete dissolution is not necessary, but the amount that dissolves will affect the taste of the fast-melt tablet. Where taste is an issue for a formulation then clearly this test method could be used to quantify the amount dissolved for a variety of formulations at an early stage in the development programme and the best ones selected for further development.

The results of testing a non-taste-masked and two taste-masked formulations of paroxetine were also found to correlate with the published taste perception (figure 6). The two taste-masked formulations both had much lower peak concentrations, indicating that less drug was dissolved.

The time-to-peak concentration is longer for the taste-masked formulations because the paroxetine is released very slowly from the complexes and the test is detecting the release from the solid that remains in the cell the longest. This is relevant because there will inevitably be some material that stays in the mouth after the first swallow and leaves an after-taste.

masking efficiency

The formulations of Drug X were classified in a blind trial. The results correlated exactly with the results of human taste panel evaluations and clearly identified the non-taste-masked version from those that has been taste-masked. It is possible to use the data from these tests to quantify the taste-masking efficacy of a formulation as a percentage reduction in concentration compared to the non-taste marked version and relative to the same amount of drug added in solution in each case.

In conclusion, the in vitro test for buccal dissolution presents a method that correlates well with taste perception and characterises the amount of a drug that dissolves during passage through the mouth. The data from this method allows the prediction of the intensity of the taste of a dosage form or of a performance target.

The method is rapid, taking only about 20 mins per test, and is reproducible, which makes it particularly suited for evaluating taste masking. It would be ideal as a QC test to ensure dosage uniformity, and as a development tool to optimise formulations before human testing, thus reducing the amount of human testing needed.