Bio-relevant dissolution tests are highly desirable and are often referred to. It appears that such tests are commonly understood as intuitive expectations rather than a clearly defined scientific objective. This creates confusion, misunderstandings and leads to a lack of progress in this regard. To avoid such confusion, the following may be considered as an appropriate definition:
“A bio-relevant dissolution test is a test which is conducted using experimental conditions representing GI tract environment, in particular intestinal, and capable of predicting in vivo response comparable to the actual blood concentration-time profiles obtained from the bioavailability/bioequivalence studies.”
It is essential to understand that a bio-relevant test must be conducted using appropriate and bio-relevant experimental conditions. Use of non-physiological test conditions such as de-aeration of dissolution medium, product dependent experimental conditions, inefficient or lack of stirring/mixing within dissolution vessels, etc., are to be avoided. Furthermore, the test conducted with appropriate test conditions must also be able to predict in vivo (blood) drug concentration-time profiles. For example, the predicted profiles should at least be able to differentiate between fast (e.g., IR) and slow-release (ER) products, as these would be from an in vivo study. Such a test should then be used to evaluate the potential drug release characteristics of a test product. For further discussion of the subject and detailed methodology to calculate drug concentration-time profiles from dissolution results, please see the links (1, 2).
This is the most commonly referred terminology and appears to originate from US FDA guidances. It defines the IVIVC as “A predictive mathematical model describing a relationship between an in vitro property of a dosage form (usually the rate or extent of drug dissolution or release) and a relevant in vivo response, e.g., plasma drug concentration or amount of drug absorbed”. Common interpretation: A point-to-point relationship is usually established between the in-vivo and in-vitro parameters of the same time points (e.g., in-vivo fraction drug absorbed vs in-vitro fraction drug dissolved) by applying a mathematical technique of deconvolution. Practical limitations: Both in vitro and in vivo data are required for the same product. If successful, the procedure does not provide predictability but reflects the relationship between in vitro (dissolution) and in vivo (fraction absorbed/dissolved). As results are compared/related using a single product, the approach may not be used for the comparison of products for their formulation/manufacturing attributes.
IVIVM: (In Vitro-In Vivo Matching).
I often use this terminology and discussed it in some of my publications. IVIVM reflects an unconvincing, perhaps misleading, interpretation and practice of IVIVC. This would require one or more products having different in vivo (blood drug concentration-time) profiles and then a number of in vitro dissolution profiles, using different sets of experimental conditions. If one set of experimental conditions provides a matched ranking between dissolution and in vivo profiles, then it is considered as achieving IVIVC. The dissolution test would be called bio-relevant. If none of the prior dissolution methods provide such matching, then a new set of experimental conditions may also be developed to match the ranking. It is clear to see that this approach seeks to match, thus would NOT reflect a relationship or predictability aspect, which is the requirement of an IVIVC.
IVIVP: (In Vitro to In Vivo Profiling).
Considering limitations in defining or interpreting IVIVC, a more appropriate and direct definition and interpretation is desirable. For this, one needs a clear and practical objective for IVIVC, which is to link or relate the in vitro (dissolution) and in vivo (blood drug concentration-time, or C-t) profiles. In this regard, the main purpose of conducting a dissolution test is to establish a dissolution profile and then predict/determine a C-t profile to assess potential in vivo characteristics of the test product. Therefore, it can be said that in reality, the purpose of commonly referred practices of IVIVC is to transfer a dissolution (in vitro) to a C-t (in vivo) profile or simply in vitro-to-in vivo profiling. The mathematical technique to transfer in vitro profile to in vivo profile is known as convolution. Convolution is relatively simpler than de-convolution as the former can be applied using simple spreadsheet software, e.g., MS Excel. For further discussion of the subject and detailed methodology to achieve IVIVP, please see the links (1, 2).
It is often considered to be highly unlikely, if not impossible, to have a universal dissolution test. Let us explore whether this is a myth or reality?
The most commonly used dissolution methods described in the USP monographs (500+) and the FDA database (500+). Examined closely, one observes that the majority (~76%) of the tests employ a paddle apparatus at 50 rpm (see US FDA database). Evaluation of USP monographs most likely would also lead to the same conclusion.
As for the dissolution medium, the majority (~60%) of the tests employ water or aqueous-based buffers in the pH range of 5 to 7.
Suppose one considers the two previous practices together. In that case, it appears that we already have a “sort of” universal test, which is paddle at 50 rpm using a medium having pH in the range 5 to 7, which may be water or a buffer. Furthermore, it should also be noted that these conditions are commonly applied for both IR and ER products. Then why do we not use these “universal” test conditions and have so many (hundreds) sets of experimental dissolution conditions?
One of the most common deficiencies in using this set of experimental conditions is that the results are often slower and more variable than expected. This is a normal and expected behavior of the paddle spindle as the spindle rotation speed of 50 rpm does not provide efficient product/medium interaction (figure) and the movement of solvent (medium) upwards. The issue in this lies with one of the suggestions to address this limitation, which is to increase the rpm. Therefore a significant number (~18%) of tests are conducted at higher rpm of 75-100. However, there are no established criteria available to select an appropriate rpm other than to obtain the expected dissolution characteristics of the product. Therefore, an analyst has to select a higher spindle rotation speed (rpm) arbitrarily. Interestingly, this practice of selecting the higher rpm is contrary to the popular belief/recommendation that a desirable characteristic of a dissolution test is that it should be discriminating. It is a well-established fact that at a higher rpm, the test becomes less discriminating. Therefore, the suggestion of changing the rpm lacks scientific merit and adds randomness to the test by selecting an arbitrary rpm or experimental condition.
On the other hand, it is very difficult to explain regarding the other variables of pH and the nature of the buffer used for the testing. At best, it can be said that choices are random and arbitrary as well, again the criteria of selecting these variables are to achieve expected dissolution behavior, perhaps by adjusting the solubility/wetting characteristics of the product. This choice is often made by a trial and error approach and without a clear and defined objective.
In short, it can be said that the commonly suggested or popular experimental conditions are paddle rotating at 50 rpm with a medium (water or buffer) having pH 5-7. However, analysts are free to select a different condition when one does not achieve expected dissolution results, which is often the case. This practice creates randomness and arbitrariness in the testing, which is currently reflected in the availability of many dissolution test conditions.
On the other hand, if one would use a different stirrer, crescent-shaped spindle is one such example, which provides improved stirring and mixing, then issues using the paddle would be addressed. Indeed, it has been demonstrated with the use of crescent-shaped spindle, even at a lower rpm of 25, many products can be analyzed using a single set of test conditions. The suggested “universal” set of experimental conditions is rpm=25 with water as a medium.
In conclusion, a universal dissolution test approach is already being used/suggested. However, its success has not been materialized because of the spindle’s poor mixing/stirring aspect. If this flaw is addressed, such as with the use of the crescent-shaped spindle, then the development of a universal dissolution test appears a genuine possibility.
If an analyst requires such a large array of tricks to describe how much drug is dissolved in water (or in an aqueous buffer), commonly known as drug dissolution testing, then there has to be a problem! The potential source of the problem=Paddle/Basket apparatuses.
In response to a query, I have provided the following opinion on the above-mentioned topic. I thought this topic might have general interest, thus posting it on the blog as well. Comments/critiques are always appreciated.
Thanks for your interest in my work and the website.
Concerning your question regarding the dissolution test for Nicotine Polacrilex Lozenges, I must say that I have not worked with such products. Therefore, consider my advice as a theory (quick thoughts) that may be workable with some experimental work.
You stated that the release of drugs (nicotine) from the lozenges occurs in 30 minutes in humans. However, in vitro test requires 8 hours for the release. Obviously, somewhere, there is a mismatch between in vitro and in vivo environments. Considering that there is a higher pH (6.2 to 7.4) in the oral cavity than in the intestinal tract, the suggested use of a higher pH of 7.4 for dissolution appears appropriate. My preference would be around 6.0
To me, however, there appear to be two issues here. First, there may be a lack of needed stirring and mixing within the dissolution vessel. I am of the view that the use of the basket/paddle apparatus would be the least efficient choice here. Nicotine Polacrilex is an ion-exchange complex, thus it may require some modest “shake” to pull the drug out of the polymer. Thus, it may require a different mechanism of stirring, may be a crescent-shape spindle as I have proposed, or something of that nature that should be good and efficient in moving the medium from the surface and providing mobility to the lozenges.
The second aspect is related to chemistry. The complex between nicotine and Polacrilex is ionic. Thus, you may need some form of competitive cation in the medium, which would facilitate the pull-out or replacement of nicotine from the ion exchange. Otherwise, ion exchange may keep the nicotine from releasing, which would result in an extra ordinary delay in release (8h?). It is to be noted that working at pH 7.4 (dissolution) is not very favorable to this dissociation, as the pKa of nicotine is 8.5. Having said that, please keep in mind that if you are going to try a competitive cation, this has to be a bio-relevant one. Otherwise, your dissolution test will lose bio-relevancy. Perhaps, CTAB may be tried, a commonly suggested solubilizer (surfactant) for dissolution testing.
The main operating principle of a paddle/basket (or vessel-based) apparatus is to provide a precise and controlled stirring and mixing mechanism at 37 C. In reality, from the operational aspect a beaker with a magnetic stirring bar may be considered equivalent to a dissolution tester if the rpm of the stirrer is precisely controlled and beaker content can be maintained at 37C. Therefore, it is important to note that the operation of a dissolution tester should be as simple as any other stirring device.
To achieve time savings and consistency in results, the current dissolution apparatuses come in units of 6 or more stirrers with appropriate mechanical and electronic controls. However, the operating principle remains the same whether the apparatus is based on a single or multiple stirring units.
Generally, stirrers (or stirring devices) come with their limitations. For example, a magnetic bar may not be suitable for mixing viscous or solid materials because bars lack sufficient torque and the stirred contents are not easily moveable. Similarly, basket/paddle stirrers cannot be used as mixers for solid materials (tablets/capsules). These settle at the bottom of the vessels where the liquid (medium) flow is low to negligible. The figure shows a lack of interaction between solid/liquid with such apparatuses, which is known as “cone formation.” Therefore, by its nature, basket/paddle stirrers (or apparatuses) cannot be used where thorough mixing (interaction) of solid/medium is required, such as what occurs in the GI tract.
Furthermore, the random settling positions of a product (tablet/capsule and their particles) at the bottom of a vessel add variability to this mixing. Thus dissolution results would be erratic and unpredictable.
To sum up, the operating principle of the basket/paddle apparatuses is based on a simple stirring device. However, the operational limitations of such stirrers are that they cannot provide efficient product/medium interaction or mixing, thus providing erratic (highly variable) and unpredictable results.
In a recent article titled “Evaluation of In-Situ Fiber Optics Dissolution Method for Compound A Extended-Release Tablets,” published in the March/April issue of American Pharmaceutical Reviews (Link), the authors described the usefulness of the technique for dissolution testing. The usefulness is described based on method validation parameters such as specificity, linearity, accuracy/recovery, precision, robustness, etc.
To me the precision is the most important parameter as almost all the other parameters depend on it, i.e., if the precision is not acceptable, then the other method validation parameters will also reflect poorly, thus the method itself.
It is very well accepted and common practice to describe the precision of a method based on Relative Standard Deviation (RSD) or Coefficient of Variations (CV) in percentages. However, in the article (method and intermediate precisions, Tables 5 and 6), the authors reported no %RSD (CV%) values. Therefore, the question is, how is the precision of the method established? Moreover, why would the authors omit the precision (%RSD) values?
The reason, in my opinion, is that the method is not as precise a method as one would like to see. The underlying method here is the drug dissolution method which is well known for its high variability in results. Therefore, the use of a detection/quantitation method, whether a simple UV-based or in situ fiber optics-based, to evaluate method precision will not help but would make the detector look bad. It is a similar situation, where drug dissolution results are reported based on individual (Q-) values, but without associated precision (%RSD values), because one can never get acceptable precision from a dissolution test using paddle and basket apparatuses.
In my opinion, authors should have used only the last dissolution sampling time to describe the usefulness and precision of the in situ fiber optics approach. The reason being, given sufficient time, which is 24 hours in this case, dissolution testing would have been completed. The variation due to the dissolution testing would be eliminated and thus would not be a contributing factor. Therefore, as shown in the table below, the observed precision (%RSD) would be a true reflection of the quantitation method, which is quite acceptable, less than 4%. Otherwise, the %RSD of the method is as high as 16.6% from a single batch and can be even higher for inter-batch and/or inter-lab situations.
In short, for describing the precision of a method (dissolution) or technique (in situ fiber optics), one should provide %RSD (CV%) values. These values should reflect the property of the desired underlying method or technique.
It is often highly desirable to obtain bio-relevant results as such results increase the confidence and usefulness of the testing. In fact, one should always focus on achieving bio-relevant results as non-bio-relevant results are of limited use.
To obtain bio-relevant results, one should try to evaluate products using experimental conditions as close as possible to the conditions one would expect during the physiological testing and/or product use.
For evaluating in vivo drug dissolution, as conducted based on the bioavailability/bioequivalence studies, it is a common regulatory requirement that products be tested using a standard and common protocol. For example, the study protocol (physiological environment) remains the same for the evaluation of IR vs ER products and any release type in between. Therefore, if one wishes to achieve bio-relevant results, then one has to conduct in vitro testing using a common set of experimental conditions. These should be product-independent.
Conducting dissolution studies using product-dependent experimental conditions violates this principle, thus should not be considered bio-relevant. To obtain bio-relevant results, testing should be done using a common set of experimental conditions, which should also be product-independent.
One of the requirements for dissolution testing is a mixing mechanism to provide efficient product (tablet/capsule) and medium interactions. Within the GI tract, such a mixing mechanism is provided by peristaltic compressions and motions.
To evaluate dissolution characteristics in vitro, one needs to provide a mixing mechanism as well. The question is should the in vitro environment have this mixing/stirring based on the peristaltic mechanism as well? The answer is not necessarily yes. For any in vitro testing, one tries to simulate an in vivo environment or process but not to duplicate it. This is one of the basic underlying principles of conducting in vitro testing. There are numerous examples of such practices.
For example, in vitro cell growths (cultures) are usually achieved in simple media not in body fluids. Dissolution tests are conducted using simple media (e.g. buffers). Similarly, controlled temperature environments (baths or cabinets) for testing are maintained, including for dissolution testing, using electronic thermostats with heating elements with or without circulating gasses or water. None of these reflect in vivo environments. The mechanisms to simulate the in vivo environments do not require duplication of the physiological (feed-back) processes where such controls are achieved based on enzymatic-based chemical reactions and circulating physiological fluids. It is important to note that arguments (suggestions) of conducting tests by duplicating physiological environments usually reflect a lack of appreciation and experience in the physiological aspect of dissolution testing.
The reason for presenting such weak arguments is to support the continued use of paddle and basket apparatuses, which do not provide the needed stirring and mixing, thus product-medium interactions. One may argue that stirring and mixing based on peristaltic motion may be preferred or used, which does not mean that one needs to keep using a knowingly flawed approach in the absence of such a mechanism. Obviously, an irrelevant testing environment will result in irrelevant dissolution results no matter how such studies, and results thus obtained are presented. One should be watchful of such erroneous suggestions and practices.
On the other hand, it is possible to modify currently used stirrers (paddle/basket) with others, such as crescent-shaped (brush-based), to address the flaws to achieve a better product-medium interaction representative of the physiological process. A number of studies have demonstrated that the use of modified (crescent-shape) sinpdle can provide improved and physiologically relevant dissolution results compared to paddle and basket apparatuses.
One of the criteria to judge the physiological relevancy is that the apparatus must provide testing using a single or common set of experimental conditions, as in vivo products are tested using a single set of experimental conditions. The use of the crescent-shape spindle fulfills this requirement.
