Dissolution Tests for ER Products

Barbara Sievert and Dr. Martin Siewert, Hoechst Marion Roussel, Frankfurt, Germany

Abstract

This presentation was made in March 1998 during the workshop 'The Role of the Biopharmaceutical Classification System and In Vitro - In Vivo Correlations in the Approval of Oral Drug Product' held in Frankfurt, Germany. The aim was to take inventory of the status quo of in vitro dissolution testing for extended release (ER) products and to identify topics which are still open or which may represent points of conflict.

The aspects of dissolution tests for ER products discussed during the workshop are were the test procedures of dissolution testing, the setting of specifications and the verification of specifications (IVIVC). Furthermore, important aspects such as changes/variations and food effects were described.

Introduction

During the last decade many workshops have been devoted to in vitro dissolution testing of immediate as well as modified release formulations. Several guidelines have been created during the years and all key pharmacopeias now include the method of dissolution testing. It seems that the methods have been mostly settled and that real news on the theme is becoming rare.

Looking at the Biopharmaceutical Learning Curve (figure 1), the development of biopharmaceutical test procedures began in the early sixties. Throughout the eighties special problems were discussed as principles of correlation or food effects. During the nineties several guidances have been created by the FDA and by FIP covering most of the relevant aspects of in vitro dissolution testing. Nevertheless, there are topics which are still open or which may represent points of ongoing debate.

Figure 1: Biopharmaceutical Learning Curve (Swt 221-1 E)

Dissolution Concepts

Different concepts of dissolution testing exist for different purposes in the pharmaceutical industry. During drug development, in vitro dissolution testing is an important tool in the evaluation of the 'best' formulation. Dissolution testing is also utilized to define the biopharmaceutical characteristics and to identify possible risks such as potential food effects on bioavailability or interaction with other drugs. The simulation of specific gastrointestinal conditions which may lead to potential therapeutic failure or adverse events seems to be another point of interest during drug development. At this stage various models and test conditions for in vitro dissolution studies are applied to learn about the biopharmaceutical characteristics of early formulations and about the discriminating power of the test system. During development, in Quality Control and for post approval changes, in vitro dissolution testing is an important part of stability studies. After the approval of a new drug, commonly, only one test model is utilized for routine in vitro quality control dissolution testing. The quality control dissolution method assures the consistency of the manufacturing process, the stability of the product in regard to the established shelf life and the quality of finished product in relation to pivotal batches. Post approval, a drug product often undergoes changes with regard to the production site, the manufacturing process or the formulation. At this stage, and depending on the level of in vivo-in vitro correlation available, different sets of dissolution tests may be used (17) to prove equivalence between the changed product and the previous product.

Test Procedures Dissolution Testing

General Aspects

For extended release formulations, in principle, similar considerations are to be made with regard to testing conditions as for immediate release products. Solubility characteristics ('sink') and relation to physiologic environment have to be considered (16). Ranges are more important for ER products more than for IR products in order to achieve special pharmacokinetic target profiles, e. g., for drug substances with a narrow therapeutic range. Due to the different concepts of slow release formulations and the various therapeutic indications, 'standard' specifications for extended release formulations cannot be set (16).

Test Medium

An aqueous test medium is preferred. The instructions on pH differ slightly between the various Pharmacopeias (see figure 2). Water as dissolution medium is allowed by some Guidances (16,18,19) but always needs justification. Additives (enzymes, salts, surfactants) are to be justified (16). A permissible additive (18) is sodium lauryl sulfate (up to 1%). Regarding deaeration a product-by-product validation should be performed (16). For the flow-through cell deaeration is mandatory (19).

Figure 2 Test Medium (Swt 216-1 E)

pH of test medium

In Quality Control usually only one pH-level is used for dissolution testing. Exceptions are only made for delayed-release formulations. Different pH-levels are preferred in comparison to a pH-gradient. In general the decision of pH-levels versus gradient is related to the apparatus, i.e., gradient is used for the Bio Diss (USP app. 3) or the flow-through cell (USP app. 4).

Apparatus

The types of apparatus used for ER-formulations are the paddle and the basket method, respectively. In addition, the flow-through cell, the reciprocating cylinder and others are used in testing of ER-formulations. Full harmonisation of the specifications for the dimensions of the paddles, baskets and vessels is still an important goal for the future. Automation of the process, including a technical modification (such as a sampling valve), is possible but has to be validated product by product. If alternative apparatus or other modifications is intended to be used, evidence of superiority to pharmacopeial specifications is the only acceptable justification.

Agitation

Different rotation speeds are specified in the various Pharmacopoieas and Guidances. For basket / paddle 50 - 100 (150) rpm are described in the FIP guidelines (16) and the European Pharmacopeia (19), whereas 100 rpm for basket and 50 - 75 rpm for paddle are recommended by the FDA (18). A final scientific decision regarding the velocity of rotation neither exists nor is to be expected other than on a product-by-product basis.

Sinkers

With few exceptions, such as the Japanese Pharmacopeia, no exact specifications regarding sinkers are currently established. Flexibility should be applied and justification of choice should be given (16).

Test Duration

At least 80 % dissolution shall be reached within the test period. A direct reference of test duration to the dosage interval is only scientifically justified when time axes in vitro and in vivo are in a 1:1 relationship. In special cases, an in vitro dissolution of < 80 % may be accepted if the test duration was at least 24 hours. In these cases during development a recovery control has to be performed, i.e., in establishing dissolution profiles at different test conditions.

Setting of Specifications

For ER-formulations, the FIP -Guideline and European Pharmacopeia demand at least 3 specification points (figure 3), the first after 1-2 hours (around 20-30% drug release) to provide assurance against premature drug release. The second specification point has to be around 50 % drug release to define the dissolution pattern. At the last point, the dissolution limit should be at least 80 % drug release to ensure almost quantitative release. Alternatively, a dissolution of <80% has to be justified and should be supported by a test duration of at least 24 hours (16). As can be seen, there are slight differences with regard to the United States Pharmacopeia (20), where only > 2 test points are demanded considering the individual monograph.

Figure 3 Dissolution Specifications (Swt 213 E)

The purpose of establishing dissolution specifications is to ensure batch-to-batch consistency within a range which guarantees acceptable biopharmaceutical performance in vivo and to distinguish between 'good' and 'bad' batches. Specification limits therefore have to be defined based on experience gained during the drug development stage especially regarding clinical development and/or bioavailability / bioequivalence studies. The capability of the manufacturing process and the commonly accepted range of 95% to 105% of stated amount for average content of drug substance has to be taken into consideration (16).

The USP approach for the acceptance criteria of ER formulations seems to be practicable. Stage 3 needs to be allowed, especially with regard to stability testing/ end of shelf-life considerations.

Verification of Specifications (IVIVC)

For pharmaceutical development especially of extended release formulations the relation between in vitro drug release and the biopharmaceutical performance in vivo needs to be known as a valid and therapeutically relevant acceptance criteria. In vitro dissolution tests without any relationship to in vivo data would be overestimated for controlling of critical production parameters and would represent a misunderstanding of biopharmaceutics. Therefore, the deduction of specification limits requires in vitro-in vivo comparison studies. Relevant guidances demonstrate increasing consensus on in vitro-in vivo comparison techniques, however, some approaches are still significantly different. Agreement exists that any in vitro test system is developed to distinguish between 'good' and 'bad' batches. The test specimen/batches need not be full production scale. For Controlled/Modified Release (C/MR)-formulations with a Level A correlation (independent drug release) at least 1 batch has to be tested. All other cases need at least 3 batches to be tested with the following conditions: Profiles of at least 12 individual dosage units from each lot with a coefficient of variation of not more than 10 % are required by the FDA (18). The number of volunteers to be included in the bioavailability study should be at least 12 according to the FIP guidelines(16), whereas 6 to 36 are accepted by the FDA (18).

Different approaches concern the categories of in vitro-in vivo comparison. Whereas major agreement refers to the 'high' positions as a correlation level A, B or C, differences on what is acceptable 'below' are remaining. (see figure 4)

Figure 4 Verification of Specification vs. Level of Correlation (Swt 207 E)

Figure 5 FIP - Rank Order (Swt 206 E)

A rank order correlation (see figure 5) according to the FIP guidelines (16) is a sufficient verification of a specification under the assumption that no quantitative interpolation is necessary. In this case, the specifications have to be defined as experimentally obtained in vitro for a pair of batches which in vivo show bioequivalence.

Figure 6 FIP - Side Batches (Swt 205 E)

Another approach for defining dissolution specifications (non-level A cases) for batches which have been manufactured and experimentally tested in vitro and in vivo is the side batch approach. 'Side batches' (16) represent batches which are created by modifications of manufacturing variables (e g. during process validation studies) in a range expected to represent maximum variability during routine production. These batches are supposed to show differences in the biopharmaceutical characteristics and to represent the upper and lower limits of the dissolution range, an approach which is required for level B and C correlations by USP. According to USP it must be demonstrated that these batches are acceptable by performing a bioavailability / bioequivalence study but it is not ultimately required that the batches should be strictly bioequivalent. Another approach in case of correlation failure according to the EEC Note for Guidance (12) 'could consist of demonstrating bioequivalence of the proposed formulation to formulations with dissolution profiles at the upper and lower limits of the specification'. Bioequivalence is not strictly required between the 'side batches' but is required between the target profile and the profiles representing the upper and the lower specification limit.

Setting Dissolution Specifications without an IVIVC

Without an IVIVC (18) the recommended range at any dissolution time point specification is ± 10 % related to the mean dissolution profile obtained from the clinical / bioavailability lots. In certain cases deviations from the ± 10 % range may be accepted. The maximum deviation then would be 25 % under the condition that bioequivalence of the side batches is proven. A minimum of three time points (early, middle and late stages of the dissolution profile) are required with a dissolution >80 % at the last time point.

The similarity factor (f2 ) for several dissolution profiles representing the ± 10 % and ± 25 %, respectively, of the target profile was calculated according to the Guidance for Industry SUPAC-MR (figure 7).

Figure 7 Setting Dissolution Specifications - Without an IVIVC (FDA) (Swt 202 E)

Since a similarity factor between 50 and 100 proves similar dissolution profiles only the 90 % vs. 100 % and 110 % vs. 100 % are similar. In the other cases f2 is less than 50 which means that the dissolution profiles are not necessarily similar.

Changes and Food Effects

Post approval changes in site of manufacture, manufacturing process or in the components of a formulation take place during the whole life-cycle of a product. For several products, sequential changes are carried out. In every case reference is to be made to the biobatch (pivotal batch) regarding in vitro dissolution. But is the so called 'Gold standard' helpful in order to not lose track in case of several incremental changes? Reference to the biobatch regarding similarity or dissimilarity is to be questioned when changes are applied on long periods of time and shelf-life is exceeded for the biobatch.

After food intake, environmental conditions change in the gastrointestinal tract which may cause an alteration of drug performance in vivo. In vitro tests under various experimental conditions have been used to attempt to predict and evaluate potential risks during therapy. Effects on bioavailability of C/MR formulations can only partially be simulated in vitro and 'correlations' to the results of in vivo studies are relatively poor. Results of various studies (in vitro vs. in vivo) show that a prediction of a food effect is nearly impossible (see figure 8 and 9, representing different interpretations of the same study results).

Figure 8: In vivo/in vitro association of mean times for four theophylline controlled/modified release preparations after bioavailability studies and in vitro dissolution tests with respect to food effects (from (21))

Figure 9: In vivo/in vitro association of mean times for four theophylline controlled/modified release preparations after bioavailability studies and in vitro dissolution tests with respect to food effects (modified from (21))

As it is shown the in vivo / in vitro association of mean times depends of the amount of data points taken as a basis.

Conclusion

Overall, the following needs to be done. Industry and regulatory authorities need to accept the consensus achieved so far and not try to go beyond this consensus. Also, industry and regulatory authorities need to focus on harmonization of 'geometries' and 'acceptance criteria',

allow product specific flexibility and accept the need for product specific justification.

References

(1) J. G. Wagner and E. Nelson. Per cent absorbed time plots derived from blood level and/or urinary excretion data. J. Pharm. Sci. 52 (6): 610-611 (1963).

(2) G. Levy. Effect of dosage form on drug absorption - A frequent variable in clinical pharmacology. Arch. Int. Pharmacodyn. 152 (1-2) 59-68 (1964).

(3) J. C. K. Loo and S. Riegelman. New method for calculating the intrinsic absorption rate of drugs. J. Pharm. Sci. 57 (6) 918-928 (1968).

(4) Dissolution, USP XVIII, 934-935 (1970).

(5) Joint Report of the Section for Control Laboratories and the Section of Industrial Pharmacists of FIP: Guidelines for dissolution testing of solid oral products. Pharm. Ind. 43 334-343 (1981).

(6) F. Langenbucher. Numerical convolution/deconvolution as a tool for correlating in vitro with in vivo drug availability. Pharm. Ind. 44 (11) 1166-1172 (1982).

(7) D. Brockmeier, D. Voegele and H. M. van Hattingberg. In vitro - in vivo correlation, a time scaling problem? Arzneimittelforschung/Drug Res. 33 598-601 (1983).

(8) USP Subcommittee on Biopharmaceutics. In vitro / in vivo correlation for extended release oral dosage forms. Pharm. Forum 14 (4) 4160-4161 (1988)

(9) A. Karim. Importance of assessing 'food effects' in evaluating controlled-release formulations. In: A. Yacobi and E. Halperin-Walega (eds). Oral sustained release formulations: Design and evaluations. Pergamon Press New York (1988)

(10) V. E. Steinijans, R. Dietrich, H. Trautmann, R. Sauter and G. Benedikt. A novel approach to the specification of in vitro dissolution boundaries based on regulatory requirements for once-daily administration. Arzneimittelforschung/Drug Res. 38 1203-1209 (1988).

(11) M. Nicklasson and F. Langenbucher. Description and evaluation of the flow cell dissolution

apparatus as an alternative test method for drug release. Pharmacopeial Forum 16 (3) 532-540 (1990)

(12) EEC Note for Guidance: Quality of prolonged release oral solid dosage forms, III/3172/91-EN (1991)

(13) FDA Interim Guidance Immediate Release Solid Oral Dosage Forms - Pre- and Post- Approval Changes. (1994).

(14) Bio-International Conference (1994)

(15) <1088> 'In vitro in vivo evaluation of dosage forms', USP 23 (1995)

(16) FIP Guidelines for Dissolution Testing of Solid Oral Products (1997).

(17) Guidance for Industry SUPAC-MR: Modified-Release Solid Oral Dosage Forms Scale-Up and Post Approval Changes: Chemistry, Manufacturing, and Controls; In Vitro Dissolution Testing and In Vivo Bioequivalence Documentation (1997).

(18) FDA Guidance for Industry Extended Release Solid Oral Dosage Forms: Development, Evaluation and Application of In Vitro/In Vivo Correlations (1997)

(19) European Pharmacopeia

(20) USP 23 (1995)

(21) P. U. Maturu, V. K. Prasad, W. N. Worsely, G. U. Shiu and J. P. Skelly. Influence of a high fat breakfast on the bioavailability of theophylline controlled release formulations: An in vitro demonstration of an in vivo observation. J. Pharm. Sci. 75 1205-1206 (1986).