EUROPEAN JOURNAL OF DRUG METABOLISM AND PHARMACOKINETICS 2000, Vol. 25, No. I, pp.18-24
Dissolution and in vivo evidence of differences in reference products: impact on development of generic drugs MICHAEL SPINOI,2, YU CHUNG TSANG! and RADU POpl I Apotex Inc., Weston, Ontario, Canada 2Faculty of Pharmacy, University of Toronto, Toronto, Ontario, Canada
Keywords: Bioequivalence, dissolution, absorption, generic, WHO
SUMMARY The WHO List of International Comparator Pharmaceutical Products (CPP) For Equivalence Assessment of Interchangeable MultiSource (Generic) Products will address an important issue in developing new generic drugs because it will identify the 'correct' reference product. This list will reduce unnecessary clinical studies in jurisdictions requiring new generics to be compared with brand products sold locally. Eventually, by employing the CPP, there will be a world-wide standard for brand and generic drugs, assuring the same level of quality internationally. The strategy of a single global reference is meritorious, but there are several hurdles to overcome. Most important is that the same brand may differ in dissolution and/or bioavailability in various jurisdictions, including some drugs with a narrow therapeutic index like phenytoin. Several examples are provided in this manuscript. This issue of regional differences has relevance, not only to the WHO list, but also to the matter of how safety and efficacy was established for that product in the first place. Normally, phase III clinical studies are conducted on a product manufactured in a single site, set to one standard. If the product differs in bioavailability in different jurisdictions, one is left with the question: 'which product has remained true to the original formulation?' Alternatively, if safety and efficacy is maintained with all formulations, then one is faced with the question: 'are the criteria currently employed for bioequivalence unnecessarily restrictive?'
INTRODUCTION The rising cost of health care has motivated governments around the world to examine methods of decreasing costs without compromising health care services. One of the methods to reduce costs, employed by many countries, is the passing of regulations that encourage the use of generic pharmaceuticals. While this can reduce the costs of pharmaceutical care by about 50%, it is critical that these savings not be accrued at the expense of the quality of
Correspondence to: Dr Michael Spino, Senior VP Scientific Affairs, Apotex Inc., 150 Signet Drive, Weston,Ontario M9L IT9, Canada.
health care. Thus systems must be in place to ensure that generic products will have the same level of safety and efficacy as the brand products which they replace. For this reason, a number of countries have implemented stringent regulatory guidelines. However, some countries do not have the resources to develop nor implement such guidelines. To address this matter, the World Health Organization (WHO) published a document entitled Marketing Authorization of Pharmaceutical Products with Special Reference to Multi-Source (Generic) Products: A Manual for a Drug Regulatory Authority (I). For those that have guidelines, or those that use the WHO document, a new problem emerges. There is no universal standard against which to compare new generic drugs. To address this matter, WHO prepared a document entitled
M. Spino et aI., Dissolution and in vivo evidence of differences in reference products
List of International Comparator Pharmaceutical Products (CPP) For Equivalence Assessment of Interchangeable Multi-Source (Generic) Products (2). As described in the document: This guidance provides a list of comparator products for equivalence assessment of interchangeable multi-source (generic) products. The list contains information on comparator pharmaceutical products collected by WHO from drug regulatory authorities and co-operating pharmaceutical companies. The list has been developed to assist regulatory authorities and pharmaceutical industries in deciding on appropriate comparator products in the context of multi-source (generic) marketing authorisation. This is a much needed and laudable step taken by the WHO. It is perceived that, initially, the value of this document and the extent to which it finds acceptance by regulatory agencies will vary among jurisdictions and will depend on several commercial and regulatory factors. Its adoption in jurisdictions that consider it a net advantage needs to be encouraged. However, WHO will need to work with the remaining jurisdictions in determining how it can ultimately be implemented because of many potential complicating factors that need to be addressed. The current regulatory situation for generic drug approval in different countries is likely to be a major determinant in deciding which jurisdictions are likely to adopt the WHO list. Criteria for generic drug acceptance in highly regulated jurisdictions such as Europe (3), USA (4,5), Canada (6,7), and Australia usually include a comparative bioavailability study compared to the market leader ('brand') sold in that jurisdiction. On the other hand, many countries do not have strict requirements compelling the generic manufacturer to conduct a bioequivalence study of the to-be-marketed product with an existing brand or reference product marketed in that country. In fact, in some jurisdictions, there is no 'brand' product. In some jurisdictions, concern for the viability of the small, local generic manufacturer may influence the extent to which there will be regulatory support for the WHO list, because of the potential local economic impact of identifying a world-wide single reference product. This concern might be particularly relevant in those jurisdictions where there is a regulatory policy that requires the new generic to be compared to a locally marketed product. Because of such a policy, some foreign generic firms may not consider such jurisdictions to be sufficiently desirable markets if they must conduct additional bioequivalence studies against locally marketed products, although such
19
decisions will be influenced by the size of those markets and the prices of the generic products in question.
RELEVANCE TO THE INTERNATIONAL MARKET AND HARMONIZATION Because we live in a global village, there may be offsetting factors to consider. Over the last decade, several large generic companies have evolved, manufacturing products for sale in hundreds of countries. The large sales' volumes, even at low generic prices, facilitate the investment required to produce a high quality product with extensive quality control and quality assurance programs, capable of meeting the most stringent GMP guidelines in the world. Even though some small local manufacturers may not have the resources, financial and human, to implement such costly programs for small local markets, eventually, harmonization will drive such changes, irrespective of the implications it may have for the viability of local companies. The impetus to enhance the quality of drugs world-wide is increasing, and that includes assurances that the new generic drug has provided the requisite evidence that it is equally as safe and effective as the brand product it replaces. In light of these developments, some of the multinational companies, developing generic drugs, encounter the wasteful exercise of repeating bioequivalence studies for their product several times for different countries, due to country-specific reference product requirements. Not only does this result in unnecessarily increasing drug development costs, but it raises the question of ethics. If the answer to a research question is already known, it is inappropriate to repeat studies in human subjects. This statement, of course presumes that one has already demonstrated that the generic product is equivalent to the brand and those results apply to the jurisdiction of interest. This is a matter that will be discussed in more detail subsequently, as it is not always the case.
A WORLD-WIDE REFERENCE PRODUCT The impact of establishing a single world-wide reference product will have a different effect depending on the situation in each local market where the reference is marketed, and three scenarios can be envisioned. In the case where there is only one reference product in that jurisdiction, the concept is likely to be acceptable, as long as the reference chosen is manufactured by the same
20
Eur. J. Drug Metab. Pharacokinet. 2000, No.1
brand company that markets the reference in that country, and it is identical to the reference sold in that jurisdiction. However, if the international reference is not bioequivalent, or has a different dissolution profile from the brand product in that jurisdiction, it raises concerns that need to be addressed. A second scenario can be envisioned in a jurisdiction where there is more than one brand product on the market, particularly where the two brand products are not bioequivalent with each other. If the CPP is bioequivalent with one of the brand products, but not the other, how does the regulatory agency address this matter? The third scenario relates to a jurisdiction where there is no brand product to serve as the reference. In such circumstances, it is obvious that the identification of a CPP is an unequivocal benefit. Not only would the CPP enable the establishment of a standard, but it would facilitate the development of a safe and effective new drug product for that jurisdiction.
INTERNATIONAL EXPERIENCE IN THE COMPARISON OF REFERENCE PRODUCTS It becomes evident that a major factor contributing to the successful implementation of the WHO list is the comparability of the CPP to the brand marketed in any given jurisdiction. Disclosure of the experience of Apotex, which markets pharmaceuticals in about 100 countries, will help to provide insight into this matter, as the company has extensive experience regarding regional differences in brand products. Apotex's policy is to develop a generic product for the Canadian market first. Subsequently, it considers marketing that product in other jurisdictions and, to do that, it may conduct a comparative dissolution study with the branded reference in the 2 jurisdictions, and, if necessary, follow that up with a comparative bioavailability study. Compared to the Canadian product, most brand products marketed in other jurisdictions studied to date show only minor differences in dissolution or bioavailability compared to the Canadian reference product. A particularly good example of the identicality of brand products in different jurisdictions is alprazolam. The dissolution of 1 mg alprazolam tablets purchased from the Australian and Canadian markets, employing a standard dissolution method, revealed dissolution profiles that were indistinguishable (Fig. 1). In this case, the similarity of dissolution profiles was translated into indistinguishable serum concentration-time profiles in a randomized crossover bioequivalence study in 27 normal healthy volunteers (Fig. 2).
125.0 "C 100.0 .
. ==1Of.
CI)
~
oIn
75.0
C
50.0
In
?fl.
•
..
+-+-~~_ _-~~~~~----.-----
_._---~ Canada
--0-
-
--- Australia •
-----.•
S1 Limit
25.0
_._---,_.~
0.0
o
10
20
30
50
40
Time (min) Fig. I : Comparative dissolution of alprozolam I mg tablets manufactured by Upjohn and marketed in Canada (Lot No. XC452A) and Australia (Lot No. DR63). Dissolution USP Apparatus #1 at 100 rpm in 500 ml aqueous solution, buffered to pH 6.0 with potassium phosphate, and employing a wavelength of 254 nm for detection (HPLC).
Typically, the comparative bioavailability data for drugs studied by Apotex are similar, but not quite as close as those seen for alprazolam. In fact, one would expect some differences to exist among different lots of the same marketed product, and/or differences in results of bioequivalence due to biological variability, for certain compounds at least. Table I provides data for captopril and sotalol marketed in different jurisdictions and compared to the Apotex product developed for those jurisdictions. The brands sold in these countries appear to be comparable products and within the range expected
o
10
20
30
40
50
Time (h) Fig. 2 : Mean serum concentration-time curves for alprazolam I mg tablets (Upjohn) obtained from the Australian and Canadian markets following a randomized single dose cross-over study in 27 fasting normal volunteers.
21
M. Spino et al., Dissolution and in vivo evidence of differences in reference products
120 "0 Q)
100
> en en
60
0~
40
80
0
(5
20 10
20
40
30
50
60
Time (min) Fig. 3 : Comparative dissolution of indapamide 2.5 mg tablets manufactured by Servier and marketed in Canada, New Zealand and Australia. Dissolution USP Apparatus #1 at 100 rpm in 900 ml simulated gastric juice.
for lot-to-lot and study-to-study comparisons of marketed products within a given country. Unfortunately, this is not always the case. A major complicating factor is seen when there is a lack of equivalence of the brand products in different countries. Sometimes analysis reveals that the same brand drug has different dissolution profiles in various countries, suggesting different manufacturing processes or formulae. Figure 3 illustrates an example of 3 different dissolution profiles for the brand product of indaparnide 2.5 mg tablets in 3 different countries, Australia, Canada and New Zealand. The Q level was 75% in 45 min. Because it is possible to have products exhibiting substantially different dissolution profiles, without having different bioavailability, it cannot necessarily be assumed that these differences will translate into bioequivalence differences, and in this case, the data are not yet available. However, data like these create a dilemma for manufacturers of generic drugs because generally their products are developed in their home country against that country's
reference product. When the dissolution profiles of products from other jurisdictions differ from those used to develop any given product, what does it mean and what should be done regarding that drug's development? In the case of cimetidine, substantively different dissolution profiles did not translate into meaningful differences in bioavailability. The Smith-Kline Beecham product sold in France exhibited a rapid dissolution profile, meeting the S1 specification of 80% in 15 min. However, the Australian brand product exhibited much slower dissolution (Fig. 4). In this case, the decision was made to proceed with an assessment of bioequivalence, notwithstanding the remarkable differences of the in vitro dissolution profiles of the 'same' brand product from the two different countries. 125.0 100.0 "0
~
75.0
is
50.0
0
Table I : Typical examples of comparative bioavailability studies of the same brand products marketed in different countries Product Captopril Apotex versus UK Apotex versus Switzerland UK versus Switzerland Sotalol Apotex versus Australia Apotex versus France Australia versus France
%AUC ratio
90% confidence interval
100 98 98
93-107 92-105 92-105
::
•
- - 0 - Australia 25.0 +---f----r-----------1-- France
0.0
98 107 93
95-104 102-112 89-98
0
10
20
30
40
TlmeCmln)
Fig. 4 : Comparative dissolution of cimetidine 800 mg tablets marketed by Smith-Kline Beecham in France (Lot #215) and Australia (Lot #46567). Dissolution USP Apparatus #1 at 100 rpm in 900 ml water 80% in 15 min, and employing a wavelength of 218 nm for detection (HPLC).
22
Eur. 1. Drug Metab. Pharacokinet. 2000, No. 1 4.5
I
4 t 3.5
~ CJ
.5. CJ c 0
0
+-
-
---~-~~-------
~""-~~-
- ! - • - Australia
1---- France
--l
A 0
c: 0 0
3
e
.-:::s
2.5 2
cr
1.5
ell
CD
8.0
5
10
15
2.0
c::
0.0 0.0
20
Time (h)
Fig. 5 : Mean serum concentration-time curves for cimetidine 800 mg tablets (Smith-Kline Beecham) obtained from the French and Australian markets following a randomized single dose cross-over study in 24 fasting normal volunteers.
Figure 5 provides the mean serum concentration-time profiles for the Australianversus French cimetidine 800 mg tablets in 24 volunteer subjects and illustrates that the dissolution differences were only marginally reflected in the in vivo bioequivalence study. Based on extensive experience in bioequivalence studies, these were not considered to be meaningful differences between the two brand products. Unfortunately, this is not always the case, as substantial differences in bioavailability of the same brand product have been detected for some products sold in different countries. Figure 6 illustrates the remarkable differences in mean serum concentration-time profiles for both estrone and equilin in a group of 9 male volunteer subjects, administered Canadian and US formulations of Premarin (conjugated estrogens) 2 x 0.625 mg tablets made by Ayerst. The study was conducted as a 3 way cross-over study with all subjects getting the Canadian product and the US product as well as a test formulation developed by Barr Laboratories (8). Figure 6 presents the data for the Ayerst products only. Independent in vitro dissolution testing revealed that the US tablets were much slower dissolving than the Canadian tablets (9), supporting these in vivo data. Such substantive differences in, what is reported to be, the two major estrogens contained in Premarin need to be considered in light of the therapeutic claims. It would be of particular interest to determine which of these two formulations (or another) reflects the products that were used to obtain the original data to support the claims of safety and efficacy and whether or not supplemental data were submitted in support of the other formulation(s).
-+-
Canada
4.0
0 0
~\--~~~~---/
6.0
E en
ell
0.5
10.0
20.0
40.0
60.0
Time (h)
B u
e 0
u Q)
c:
..2 eQ)n
ell
E en
ell
a::
8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0
-+- Canada -0-- US
0.0
20.0
40.0
60.0
Time (h) Fig. 6: Mean serum concentration-time curves for (A) equilin and (B) estrone following administration of two 0.625 mg Premarin tablets (Ayerst) obtained from the Canadian and US markets following a randomized single dose cross-over study in 9 volunteer subjects.
The above examples, while of potential concern, do not represent drugs with narrow therapeutic indices. Data published on phenytoin, however, are potentially worrisome (10). Phenytoin, used in the prophylaxis of seizure disorders, is manufactured by Parke Davis and is sold throughout the world. As a result of parallel import laws, it is possible for a patient in any European country to get a brand product manufactured in a different European country,even though it has not been studied in that country for equivalence to the local brand product. In a study published by Farinha et al (11), 37 different phenytoin products, manufactured throughout EU countries were studied in collaboration with the PIP Official Laboratories & Medicines Control services section. While differences were noted among brands, particular concern was raised by the observation that dissolution results of 13 phenytoin products, manufactured by Parke Davis, and sold in these various European countries, exhibited remarkably differ-
M. Spino et aI., Dissolution and in vivo evidence of differences in reference products
~~~==;-==~==.
100
~testPrOduCtl ~rGroup' 1
80
mean ·1
,-:'Group' 2 tneaJ1]
60
Ulowest product]
40 20
o o
10 20 30 60 90 120 180
Time (min) Fig. 7 : Comparative in vitro dissolution profiles of phenytoin products manufactured by Parke Davis and marketed in 13 different European countries, illustrating the mean profiles of the fast and slow dissolving products and the fastest and slowest dissolving products in their respective groups. Adapted from Farinha et al (10).
23
substantive differences in both dissolution profiles and bioavailability of the 'same' brand product in different countries, it raises the larger question of the clinical meaning of these differences. One possibility is that products in some jurisdictions are ineffective or unsafe. However, generally, there are no data supporting such a view. The alternate view is that the current criteria for bioequivalence, required by some regulatory bodies, are unnecessarily stringent, even for NTI drugs. While bioequivalence is an extremely exacting approach to assuring comparability of brand and generic products, it may be so restrictive that clinically equivalent products are kept off the market because of small, and clinically irrelevant, differences in serum concentrations. This is a matter that requires further exploration, particularly in the case of highly variable drugs.
CONCLUSIONS ent profiles. Figure 7, adapted from the Farinha paper, points out that the 13 brand products could be grouped into 2 very different types of profiles. The figure also portrays the profile for the most slowly dissolving tablets in the slow group and the fastest dissolving tablet in the fast dissolving group. Phenytoin is a drug with a 'high risk potential' for bioavailability problems because its absorption is dependent on particle size, shape, salt, and even certain excipients. Data were not presented on whether anyone had compared the in vivo performance of the various products manufactured by Parke Davis, but it would seem reasonable to assume there might be substantive differences. Both the regulator and the manufacturer of generic drugs have a dilemma to address with data like those presented for phenytoin, particularly when there are large differences in the bioavailability of the brand for a drug that has a known serum concentration-effect relationship. There are many reasons for such differences, including: (i) products manufactured in different sites using different processes; (ii) regional differences introduced in manufacturing process, equipment or formulation; (iii) planned changes over time to correct a poor formulation; and (iv) differing regulatory demands in various jurisdictions. Although not the case for phenytoin, differences in mean or individual bioavailability results can also be due to chance in the case of highly variable drugs such as verapamil (11) or propafenone (12,13), in addition to these manufacturing and regulatory reasons noted above. While it is important for the scientific, regulatory and industry communities to be aware that there may be
In conclusion, the WHO list of comparator products (CPP) is perceived to be invaluable for countries without a currently marketed brand product, as it enables the establishment of a specified standard of known quality to provide assurance of safety and efficacy. However, in considering the acceptance of a CPP as the reference product in any given jurisdiction, a major issue that must be addressed concerns the regulatory decision process when that CPP exhibits substantially different dissolution profiles or serum concentration-time profiles from the brand product marketed locally. Also, it is evident that the selection of a single world-wide reference product would facilitate the development of generic products for international markets and speed the introduction of quality products into many regulatory jurisdictions. While this may be most advantageous to large multinational generic pharmaceutical companies, it may, at the same time, have negative economic implications for some small local companies. Finally, after the resolution of such problems as those noted above, a common reference standard should lead to a higher level of quality of both brand and generic pharmaceuticals and should facilitate the seamless interchangeability of products among countries.
ACKNOWLEDGEMENTS The authors are most grateful to Paul Gordon and Elizabeth Kovacs for the data provided in the generation of comparative bioavailability and dissolution results of
Eur. J. Drug Metab. Pharacokinet. 2000, No. I
24
brand products presented in this paper. The authors also wish to thank Dr Barry Shennan and Dr Nick Cappuccino for their insight, helpful comments and support given in the preparation and publication of this research and manuscript.
6.
Z
REFERENCES 1. World Health Organization (1999) : Marketing Authorization of Pharmaceutical Products with Special Reference to MultiSource (Generic) Products: A Manual for a Drug Regulatory Authority. Regulatory Support Series No.5. Geneva: WHO. 2. World Health Organization (1999) : list of International Comparator Pharmaceutical Products (CPP) For Equivalence Assessment of Interchangeable Multi-Source (Generic) Products. In: report of informal discussions on the selection of comparator pharmaceutical product for equivalent assessment of interchangeable multi-source (generic) products. Geneva: WHO, 8-1Z 3. The European Agency for the Evaluation of Medicinal Products/Human Medicines Evaluation Unit. (1988) : Note for Guidance on The Investigation of Bioavailability and Bioequivalence. London. 4. Food and Drug Administration/Office of Generic Drugs/Division of Bioequivalence. (1992) : Statistical Procedures for Bioequivalence Studies using a Standard TwoTreatment Crossover Design. 5. Food and Drug Administration/Office of Generic
8.
9. 10.
11.
12.
13.
Drugs/Division of Bioequivalence. (1993) : Oral Extended (Controlled) Release Dosage Porms - In Vivo Bioequivalence and In Vitro Dissolution Testing. Drugs Directorate, Health Protection Branch. (1992) : Conduct and Analysis of Bioavailability and Bioequivalence Studies. Part A: Oral Dosage Formulations used for Systemic Effects. Ottawa, Ontario, Canada. Therapeutic Products Programme, Health Canada. (1996) : Conduct and Analysis of Bioavailability and Bioequivalence Studies. Part B: Oral Modified Release Formulations. Ottawa, Ontario, Canada. Kumar V. (1991) .In-house data supplied by Barr Laboratories. Conjugated Estrogens Pilot Bioequivalency Study. McGilveray I. (1992) : Director, Bureau of Drug Research, Health Canada. Personal communication. Farinha A, Nobrega S.D.,Paulo T. (1997) : Evaluation of pharmaceutical quality of phenytoin sodium capsules and tablets from multinational markets. Drug Dev. Ind. Pharmacol., 23, 47-61. Eichelbaum M., Somogyi A (1984) : Intra- and inter-subject variation in the first-pass elimination of highly cleared drugs during chronic dosing. Studies with deuterated verapamil. Eur.]. Clin. Pharmacol, 26, 47-53. Hollmann M., Brode E., Hotz D., Kaumeier S.,Kehrhahn O.H. (1983) : Investigations on the pharmacokinetics of propafenone in man. Drug. Res., 33, 763-770. LeeJ.T., Kroemer H.K., Silberstein DJ. et al. (1990) : The role of genetically detertnined polymorphic drug metabolism in the beta-blockade produced by propafenone. N. Engl. ]. Med., 322, 1764-1768.