Regulation

Internationally, it is useful to start at two locations: either the Scientific Information Service (SIS) of ECVAM (the European Centre for Validation of Alternative Methods) or its USA counterpart, ICCVAM (Interagency Coordinating Committee on the Validation of Alternative Methods. Another useful source is the Secretariat of the Environmental Section of the OECD(Organization for Economic Development) in Paris.

Other relevant information on regulation of animal research in European countries can be found on the site of the European Biomedical Research Association:
http://www.uel.ac.uk/research/ebra/regulate.html
This site covers regulation in France, Germany, The Netherlands, Sweden, Switzerland and the United Kingdom.

Information on international harmonization of international toxicity testing guidelines is reproduced below from the EBRA website:
http://www.uel.ac.uk/research/ebra/news/nov97/harmon.html

Harmonisation of international toxicity testing guidelines for pharmaceuticals: Contribution to refinement and reduction in animal use

Background

Events such as the thalidomide tragedy in 1959-1962 catalysed the widespread introduction of medicines regulations, particularly those governing toxicity testing in animals. In the ensuing decades there was a tendency for national and international guidelines to become more specific which, in many cases, resulted in detailed guidelines that were treated as testing protocols from which companies hesitated to deviate, because of the potential expense involved in risking the rejection of a product licence submission.
Furthermore, national regulations evolved in parallel in relative isolation, resulting in disagreement between different regulators regarding the type and design of animal tests that should be required before a compound can be used ethically and safely in humans. Consequently, in order to achieve international marketing of a new medicine, toxicology studies have had to be carried out according to experimental protocols recommended by the most exacting authority and even repeated with slight variations to meet the idiosyncrasies of some regulations in specific countries. In order to resolve some of the differences between regions, in 1989 the idea was put forward of a joint industry/regulatory initiative, between the European Community (EC), the United States and Japan. This became established as the International Conference on Harmonisation (ICH). In addition to participation from the three regions, the WHO, Canada and the European Free Trade Association (EFTA) countries are Observers to the ICH process.

The topics for discussion within ICH are divided into three broad areas: Quality, Efficacy and Safety (which covers testing in animals). Expert Working Groups (EWGs) seek consensus on the issues, and the outcome of their discussions may be recommendations or draft harmonised guidelines. The ICH process is, however, only a way of facilitating and accelerating the process of consensus, and the translation of recommendations into action must follow the established consultation and statutory procedures in each region1. The progress to date has been reported at four formal conferences, ICH1, ICH2, ICH3 and ICH4 2,3,4,5.

One of the primary purposes of the discussions on safety topics has been to examine ways in which the process of safety testing could be made more efficient and economical, by rationalising the way in which tests in animals are carried out and by reducing the numbers of animals used, without compromising safety 1.

The main safety achievements are summarised in Table 1. The ICH discussions have had a major impact on at least two of the three R's of Russell and Burch, as the implementation of the harmonised agreements in the three regions has resulted in both the refinement of experimental designs and a reduction in animal numbers. This can be illustrated by the impact of the ICH agreements/guidelines for single and repeat-dose toxicity testing, reproductive and developmental toxicity testing and carcinogenicity testing.

Table 1. ICH Achievements on Safety Topics

Single dose toxicity testing

Prior to ICH1, there was a requirement in all three regions for single dose toxicity studies to be conducted to characterise the dose-response relationship of the approximate lethal dose6. Although there appeared to be a consensus that a precise LD50 test should not be required for pharmaceuticals, there were some discrepancies. For example, in Japan the conditions under which determination of the LD50 value was necessary applied to the majority of potent pharmaceuticals. In addition, there were differences in the numbers and species to be used for single dose toxicity testing, with a requirement in Japan for studies in rodents and non-rodents.

These single dose toxicity tests could inflict pain and distress and required large animal numbers. At ICH1 agreement was reached that the formal LD50 test would no longer be required in the three regions (USA, Japan and Europe), and that single dose studies in two rodent species would suffice. The single dose test should now involve a protocol which uses the smallest number of animals possible for the approximation of the highest non-lethal or the lowest lethal dose 2. As mentioned in the ICH guideline "Timing of Non-Clinical Safety Studies in Relation to Clinical Trials", a dose escalation study is considered an acceptable alternative to the single dose design. Agreement on the appropriate approach to single dose testing has resulted in a significant reduction in the numbers of animals and the abandonment of a test which has been widely criticised for ethical and scientific reasons as a waste of animals and resources providing unnecessary and unreliable results.

Repeat-dose toxicity testing

In the early 1980s, the maximum duration of repeat-dose toxicity studies (excluding those to assess carcinogenic potential) to support approval for marketing of medicines intended for chronic use in man differed between the three regions, with the USA and Japan requiring 12 month studies and the European agencies six month studies. This resulted in many pharmaceutical companies conducting two chronic repeat-dose studies in rodents and in nonrodents, one of six months duration in each species to support clinical trials, and 12 month studies to support marketing in the USA and Japan.

In order to address the issue of the appropriate duration of toxicity studies, databases were established by the Centre for Medicines Research (CMR) International 7,8, the US FDA and the Japan Pharmaceutical Manufacturers Association (JPMA) comprising detailed information from pharmaceutical companies on compounds evaluated in short- and long-term tests. Results from all three databases were presented at ICH1. The conclusion of the CMR and JPMA studies was that a six month period of dosing is all that is routinely required for evaluating the chronic toxic (excluding carcinogenic) potential of a human pharmaceutical product 9,10. The FDA reached a similar conclusion with regard to rodents 11. Based on these analyses, agreement was reached at ICH1 to reduce the duration of chronic repeat-dose rodent studies from 12 to six months 2. However, it was pointed out subsequently that the FDA had agreed to this change because their study showed that any late emerging toxicity in the rat would be identified in a two year carcinogenicity study, also required for all new medicines intended for chronic use, and therefore the 12 month rat study was redundant12.

The situation with regard to the non-rodent is less straight forward, as the chronic study is the longest test in this species (ie there is no requirement for a non-rodent carcinogenicity study). The European and Japanese regulatory 13 and industry representatives agreed that non-rodent studies should be limited to a maximum of six months, based on the findings of the CMR International 14 and JPMA 15 databases. However, the FDA continued to require 12 month non-rodent studies based on their analysis, which identified five/27 compounds with new findings in the 12 month dog studies 12, 16.

A subsequent re-evaluation, by FDA scientists, of the data in the FDA, CMR International and JPMA databases suggested that nine months may have been sufficient to capture most of the new toxicity findings that emerged after six months. It was therefore proposed that nine month studies in non-rodents be considered as a possible basis for harmonisation by all ICH regions. In order to progress this topic, a panel of regulators from the EU, Japan and the USA undertook a detailed evaluation of the 16 cases where new toxicities had been identified after six months. As a result, a proposal that nine month studies in non-rodents should be acceptable for submission in the three regions has been agreed at Step two of the ICH process, and has entered the Step three consultation period.

The agreement to reduce the duration of repeat-dose studies has had a significant impact on the number of animals required for chronic testing, as both six and 12 month studies are no longer necessary.

Reproductive and developmental toxicity testing

The realisation that exogenous agents (including pharmaceuticals) which were harmless to the mother could impair development of a human embryo led to the recognition of the need for a testing procedure to detect this potential toxicity. The three-segment design proposed by the FDA in 1966 was adopted by most leading authorities throughout the world, including Europe and Japan 17. Although the resulting national guidelines had a number of common requirements, there were major discrepancies involving the number of species and animals, duration of treatment and types of examination required 18. These differences caused difficulty in amalgamating the segments for different agencies and it was often necessary to repeat studies in order to comply with all guidelines 19. This resulted in a certain amount of redundancy, particularly in relation to fetal examinations and postnatal, physical and functional development of the F1 generation. Because of the large numbers of litters produced in reproduction and development studies, a compromise practice which was acceptable internationally could involve the use of as many as 6000 animals (including dams and offspring) provided that no problems were encountered. However, as pointed out by Tanimura (1990), each guideline had advantages and disadvantages and it was difficult to decide generally and scientifically which were the best protocols.

Following ICH1, mutual acceptance of regional differences in protocols was agreed. At ICH2 a new, flexible and harmonised guideline was adopted (for the full text of the guideline see ICH Harmonised Tripartite Guidelines: Detection of Toxicity to Reproduction for Medicinal Products 20). As pointed out by Bass 21, this guideline has succeeded in providing an understandable, reasonable, defendable and acceptable scheme for testing, without restricting the knowledgeable expert by rigid requirements. In order to be successful this approach requires flexibility, the application of scientific concepts, and the willingness of all to discuss and improve constantly the methodologies used for testing 22.This strategy is now fully accepted in the three regions. The use of this guideline means that in defined circumstances, the extent of testing can be much reduced without compromising safety. It has been estimated that it will reduce the number of animals required for reproductive and developmental toxicity testing by more that 27% (Manson, personal communication), whilst promoting the use of scientifically designed studies rather than routine regulatory protocols 23.

Carcinogenicity testing

The basic scientific criteria for the design and conduct of carcinogenicity studies for pharmaceuticals has been a major topic of discussion within the ICH procedure, as this is an area with considerable potential for reducing the number of animals required for testing, and for refining the experimental design to take into account the scientific knowledge on mechanisms of carcinogenesis which has accumulated since the concept of the lifetime study was first developed. The two issues that have been addressed are the conditions under which carcinogenicity studies should be required, and the design of the studies,particularly the selection of the high dose and the need for using two rodent species.

Conditions under which carcinogenicity studies are required
The main conditions which trigger the need for carcinogenicity studies of new pharmaceuticals are the proposed duration of clinical use and any cause for concern regarding potential carcinogenicity that may be related to the chemical structure or biological action of the compound, or to the findings of previous toxicity or genotoxicity studies. Prior to the ICH discussions, in the USA, any compound to be used in the clinic continuously for three months or longer required a carcinogenicity study, whilst in Europe and Japan the clinical duration triggering the need for these studies was six months24. In addition, the Japanese MHW required new carcinogenicity studies for new salts or esters of existing chemicals, or when the clinical route of administration was changed even if carcinogenicity studies had been completed using the original route of administration.

These differences have been eliminated in the ICH Step 4 guideline which was signed off at ICH3, "Guideline on the Need for Carcinogenicity Studies of Pharmaceuticals". It was agreed that carcinogenicity studies are required for compounds intended to be either administered clinically for periods of six months or longer, or used repeatedly in an intermittent manner; for compounds to be delivered in a way that results in prolonged exposure; and for compounds for which there is concern about carcinogenic potential. Importantly, the guideline also clarified the types of compound or conditions under which carcin-ogenicity studies will not normally be required (Table 2). The application of this guideline should avoid the unnecessary use of animals in testing and provide consistency in the worldwide regulatory assessments of applications for marketing authorisation.

Table 2. Conditions under which carcinogenicity studies will not be required


Design of carcinogenicity studies

High dose selection
Traditionally, the criteria for high dose selection for carcinogenicity studies of human pharmaceuticals have not been uniform among international regulatory agencies. In Europe and Japan, dose selection based on toxicity endpoints or the use of an arbitrary upper limit set at a multiple of 100 times the administered therapeutic dose has been accepted. In the United States, dose selection based on the MTD has traditionally been the only acceptable practice 25,26. In October 1994, a new guideline on dose selection in carcinogenicity studies was agreed by Japan, the USA and Europe 27, suggesting that any one of several criteria may be appropriate and acceptable for dose selection (Table 3). This guideline is a significant advance as it means that drugs of low toxicity will not have to be tested at the MTD. Scales 25 estimates that the AUC criteria will be applicable for 15% of drugs. This will also avoid the suffering induced in animals by using extremely high doses, thus leading to a refinement in testing. Furthermore, the application of the guideline should reduce the number of studies which have to be repeated because of regulatory questions concerning the selection of dose levels, which is one of the most common reasons for repetition of carcinogenicity studies 28. However, this proposal is regarded as only the first step in an evolving approach to high-dose selection for carcinogenicity studies of new pharmaceuticals 29. The importance of continuing to examine the best method for dose selection, and updating the criteria as new information becomes available, is emphasised in the guideline.

Table 3. ICH Guideline: Criteria for High Dose Selection for Carcinogenicity Studies

Number of species
Historically, the regulatory authorities in all regions required that pharmaceuticals intended for chronic use in humans be assessed for oncogenic potential by tests in two rodent species, usually the rat and the mouse, for the lifetime of the rodent. However, the relevance of findings from the conventional long-term rodent carcinogenicity study to the assessment of human risk has come under critical evaluation in recent years. In particular, the rationale behind the routine use of two rodent species for these studies has been questioned since it has been shown that little new, clinically relevant, information is obtained using two rodent species compared with rats alone 30,31,32,33,34. Discussion of this issue under the auspices of ICH has led to the guideline "Testing for Carcinogenicity of Pharmaceuticals" which was signed off at Step 4 of the ICH process in July 1997. The guideline advocates a more flexible approach to carcinogenicity testing and proposes that in place of a second long-term carcinogenicity study, a short or medium term study in a rodent model may be appropriate. It is intended that the alleviation of the routine regulatory requirement for two long-term carcinogenicity studies will allow resources to be diverted into investigating other experimental approaches for the assessment of carcinogenic potential.

Impact of carcinogenicity guidelines

Implementation of the ICH carcinogenicity guidelines has resulted in refinement of testing strategies and reduction of animal numbers, as well as the replacement of one life-time study with a transgenic model (Table 4).

Table 4. Impact of ICH guidelines for carcinogenicity testing in rodents

Before ICH After ICH4 Impact
conditions for need defined for conditions for need and no need defined reduction
heterogeneous guidance on selection consensus on dose refinement
two lifetime studies: rats + mice one lifetime study + one short term model eg in transgenic mice): replacement
six or 12 months duration refinement
no full tumours refinement
no ageing refinement
half the number of animals reduction

Conclusion

Toxicity studies of new medicines are costly in terms of the total number of animals required. The ICH agreements to date have had a positive impact on refining the animal toxicity tests conducted to support the clinical use of new medicines, by ensuring that the most appropriate study designs are used based on scientific principles. There has also been a reduction in the numbers of animals required for single-dose, repeat-dose, reproductive toxicity and carcinogenicity testing. This achievement is illustrated by comparing a set of studies required for a non-clinical development project intended for worldwide regulatory submission prior to ICH1, with the package based on the situation post-ICH4 (Table 5).

In addition to reducing animal numbers, the ICH process has produced intangible benefits which have accumulated via direct discussion of important scientific concepts 35. This has led to new relationships among the regulators 36 and has resulted in the establishment of a network for interaction between regulators and industry at all levels which will outlive the formal ICH process 37. This means that a mechanism will continue to exist whereby the type, number and design of animal toxicity tests required for new pharmaceuticals can continue to be refined based on the state of current scientific knowledge.

Table 5. Animal use in a core testing battery of toxicology studies 38

Before ICH1 After ICH4
Single dose toxicity (2 routes) 2 Rodent
1 Non-rodent
200-300
16-32
50-100
0
Repeated-dose sub-chronic 1st duration (eg 1 month)
Rodent
Non-rodent
2nd duration (eg 3 months)
Rodent
Non-rodent
Recovery
Rodent
Non-rodent

80
24

160
32

200
40

*160
32

0
0

0
40
Repeated-dose chronic 1st duration (eg 6 or 9 months)
Rodent
Non-rodent
2nd duration (eg 12 months)
Rodent
Non-rodent

160
32

160
32

160
32

0
0

Reproduction Segment I
eg Japanese style
eg US/EU style
Segment II rat
eg Japanese style
eg US/EU style
Segment II rabbit
Segment III

192
96

96 + 48
96
60
96

192
0

0
96
60
96
Carcinogenicity 1st species (eg rat)
2nd species (eg mouse) 500 500
Medium or short-term study 0 160
400-500
500
0
400-500
500
160

TOTAL

2720-2936

1478-1583

* Excludes offspring and dose-range finding studies