before the
House Committee on Science
Subcommittee on Energy and Environment
June 17, 1999
Karen Florini, Environmental
Defense Senior AttorneyI.
Introduction and OverviewIntroduction and Overview
Virtually everyone would agree that it's important to have basic information about whether widely used chemicals pose hazards to human health or the environment (including other species). After all, nobody can take steps to reduce hazards they don't know about. Likewise, efforts to choose safer products are thwarted by lack of data on which chemicals are less toxic, and assertions that particular chemicals are safe cannot be assessed without solid data one way or the other. And, fundamentally, the public's right-to-know about chemicals includes having access to basic data on chemicals' health and environmental effects, particularly for chemicals that are produced in high volumes.
Three recent studies – in 1997 by Environmental Defense2, and in 1998 by the U.S. Environmental Protection Agency3 and the Chemical Manufacturers Association4 – independently evaluated the public availability of basic information about hazards that widely used chemicals may pose to humans and the environment (including other species). While those reports were by no means exhaustive reviews of all studies ever published on the chemicals in question, they clearly demonstrated that existing programs for generating and distributing toxicity information have fallen short to date.
In response to this surprising and disturbing finding, Environmental Defense, the Environmental Protection Agency, and the Chemical Manufacturers Association jointly developed a framework for the High Production Volume (HPV) Chemical Challenge Initiative. Under this program, producers of about 2,800 widely used industrial chemicals can voluntarily agree to assure that basic hazard information on their chemicals is generated and available to the public within the next six years (see Attachment 1). The Chemical Manufacturers Association is working with its members and others in industry to encourage their participation, Environmental Defense is monitoring progress and will if necessary publicize any problems, and EPA will develop regulations where needed testing isn't conducted voluntarily.
In a nutshell, the framework provides that companies can sign up throughout 1999 to sponsor chemicals they produce (if there are multiple manufacturers, companies can join together to form consortia to jointly sponsor work)5. A sponsor's first obligation is to review existing data, compare them against an internationally agreed-upon template known as the Screening Information Data Set, and determine whether any gaps exist. The sponsor then develops a test plan to fill any data gaps and posts it on the Internet for public review before testing begins (any testing must follow established protocols). Members of the public can comment on whether the test plan is flawed, e.g., if the sponsor missed relevant studies. Test plans are due at the beginning of the "start year" for the chemical (2001, 2002, or 2003 for most chemicals). Sponsors are encouraged to group related chemicals into "categories" and interpolate results within the category as scientifically appropriate. Once data gathering and any necessary testing is completed, data will be made publicly available.
II. Animal Welfare Issues
Following announcement of the joint framework for the voluntary initiative, people concerned about animal welfare have raised questions about use of animal testing in the Initiative. Numerous thoughtful communications have prompted efforts to assure that every feasible step is taken to minimize the number of test animals used while still producing scientifically useful data. For example, the Environmental Protection Agency developed recommendations for ways to significantly reduce numbers of animals tested, where existing data aren't sufficient.6 In addition, Environmental Defense launched Project TestSmart, in partnership with three leading academic institutions, to promote use of non-animal test methods in the HPV Initiative and beyond (see Attachment 2)7. Most recently, on May 10, Environmental Defense and the Humane Society of the United States sent a joint letter to EPA urging the Agency to take specific additional steps to minimize use of animals in the HPV Initiative (see Attachment 3).8 Of course, there are some in the animal rights movement who believe that no animal testing is morally justifiable, regardless of the need for the information or the unavailability of alternatives. Even though we do not share it, Environmental Defense respects this position, which is often grounded in the view that animals have the same rights as humans. Unfortunately, some animal rights organizations – though by no means all – have issued materials misrepresenting the nature of the Challenge Initiative. The attached set of Questions and Answers aims to clear up some of the misunderstandings that have arisen.
III. Data-Availability Issues
Some critics of the HPV program have asserted that the three reports by Environmental Defense, the Environmental Protection Agency, and the Chemical Manufacturers Association all missed significant amounts of relevant data, particularly from more-sophisticated studies that did not directly generate screening-level data but that nonetheless provide relevant information for assessing toxicity9. Because of the methodology used in the three studies, it's entirely possible – even likely – that some additional relevant information is indeed available. That's precisely why the very first step that a sponsor takes as part of the HPV challenge is to conduct a review of all relevant data. If adequate information already exists for a particular chemical, all participants will regard this as very good news – and no additional animal tests will be conducted for that chemical. Tests on animals will be conducted only to the extent that these critics are wrong in asserting that adequate information is already available for most HPV chemicals.
IV. Participation Issues
One concern raised by animal-protection groups is that they were not consulted during the development of this Initiative. That is true, and in retrospect a significant oversight. In part, this occurred because the voluntary Initiative was viewed essentially as an acceleration of a existing program created by the Organization for Economic Cooperation and Development – a program that had been non-controversial throughout its 10-year history10. More importantly, following the launch of the HPV initiative there have been numerous meetings about the initiative11. Representatives of animal-protection groups participated in these sessions, and their views are helping to influence the initiative's implementation. For example, EPA has promoted a strategy for combining protocols in order to reduce the number of animals used in testing. In addition, the agency recently announced that it no longer has a preference for in vivo rather than in vitro genetic toxicity studies. Environmental Defense plans to continue efforts to coordinate with responsible members of the animal advocacy community.
The HPV Initiative and Animal Welfare: Questions and Answers
Already-Tested Chemicals
Question: Is it true that Environmental Defense wants additional animal tests to be conducted on substances that have already been adequately tested for their toxicity?
Answer: Absolutely not – and under the HPV Initiative, existing data are reviewed before deciding what, if any, additional testing now needs to be done. A lot of confusion has arisen because the HPV Initiative involves 2,800 chemicals on the list of high-volume chemicals. But being on the high-volume list means only that the chemical is produced in large quantities (over one million pounds in the US annually) – not necessarily that the chemical needs any additional testing. Indeed, once a company signs up to sponsor a specific chemical, its first step is to gather and evaluate existing data on that chemical. Because testing is expensive, companies have a strong incentive to rely on existing data to the extent those data meet criteria for scientific adequacy. The extent to which existing data will suffice is not yet clear, since sponsors have not yet completed their reviews of existing information.
Previously-Unpublished Studies
Question: What about existing studies that may already exist in company files but that have never been published?Answer: Companies are free, and encouraged, to use previously-unpublished data (assuming it is of adequate scientific quality). Some critics have argued that there needs to be an "amnesty" for existing data that shows negative effects, since the Toxic Substances Control Act of 1976 imposes a legal obligation to provide to the Environmental Protection Agency any data that show a chemical "presents a substantial risk of injury to health or the environment." Those critics argue that companies will conduct new tests rather than admit they have existing data which show adverse effects. However, in the early 1990s EPA conducted a massive compliance initiative to assure industry-wide compliance with the reporting requirement, and received over 10,000 submissions which have been placed in the publicly available TSCATS database (and are available for review as existing data). Almost by definition, any company that ignored the compliance initiative is a bad actor and is most unlikely to agree voluntarily to sponsor a chemical in the HPV initiative.
Human Data
Question: Shouldn't human data be taken into account as well?
Answer: Certainly, and they will be. Assertions that relevant human data will be excluded are simply wrong. All existing data – whether from screening studies, more-sophisticated animal or non-animal tests, or human studies – are relevant in determining whether additional testing is needed. However, it is not adequate to assert "people have been using chemical X for years with no apparent ill effects" to say that testing isn't needed. Especially for developmental and longer-term effects, it takes carefully designed studies to determine whether or not health effects are associated with particular exposures. Where it's clear that ongoing human exposures won't occur – for chemicals that are only used as intermediates in a closed system, for example – only tests for short-term effects are included. (Because of the possibility of accidents, such data are needed even for closed-system intermediates.)
Number of Chemicals
Question: I've heard that testing will be performed on tens of thousands of chemicals. Is that true?
Answer: Only 2,800 high-volume chemicals are even nominally within the scope of the HPV Initiative. Some of those will turn out to have adequate information already available, as noted above, and won't have any further testing. Others will be placed into "categories" of structurally related compounds; selected members of the category will be tested (if existing data aren't sufficient) and the results extrapolated to other category members using Structure/Activity Relationship (SAR analysis). These strategies will sharply reduce the amount of additional testing. Assertions that Environmental Defense has called for testing of individual chemicals, and threatened to denounce companies that choose to test chemicals in categories, are flat wrong. To the contrary, Environmental Defense has repeatedly called for reliance on categories to the greatest extent scientifically appropriate.
Additional Health Effects of Known Toxicants
Question: If you already know a chemical is toxic, isn't it pointless to conduct additional testing for other health effects?
Answer: Just knowing that a chemical causes one type of toxicity does not automatically eliminate the need for additional testing. There are many different types of toxic effects, or endpoints, and chemicals that cause one endpoint may or may not cause others as well. Existing data may not reflect the most-sensitive endpoint. To take one well-known example, lead causes cancer at fairly high doses, but causes serious neurological damage at much lower doses. It would have been a terrible mistake if lead's cancer connection had obscured its devastating neurological effects.
Scheduling
Question: Some have claimed that the "hurried schedule" for the HPV program won't allow for proper review of existing data and development of a scientifically sound testing strategy, and that testing could begin as early as March 1999. Is that the case?
Answer: The data required under the Initiative – which may be supplied using existing data where they are scientifically adequate – are a subset of those called for in the Screening Information Data Set (SIDS) program of the Organization for Economic Cooperation and Development (OECD). OECD is an intergovernmental forum of the developed countries, including the United States, Canada, Japan, European nations, and others. SIDS was originally developed in an extensive series of discussions beginning in the late 1980s. Ever since, it has been used (as modified over time) as a scientifically sound method of setting priorities for further work on chemicals in the US and throughout the developed world. In terms of reviewing existing data, sponsors have until the beginning of their "start year" to gather and review existing data. There are staggered dates for "start years," with most activity occurring in 2001, 2002, or 2003. Prior to conducting any additional testing, sponsors will post their draft test plans on the Internet (at www.hpvchallenge.com, though no plans are due as yet) for public comment. This will allow interested members of the public, including animal-protection advocates, to see if existing data or non-animal test methods have been overlooked. EPA has stated that it will examine many of the test plans, particularly those using categories, as well.
Tests of Known Toxicants
Question: Is it true that tests using animals could actually clear chemicals already known to be hazardous?
Answer: No. As noted above, where there are adequate data for a SIDS endpoint, they will be used and no additional testing will be conducted for that endpoint. More generally, animal tests cannot trump data from human studies, and screening-level tests such as the SIDS tests used in this program cannot trump data from more-sophisticated tests. Likewise, "negative" screening data cannot undercut existing "positive" screening data – they just reinforce the need for more-sophisticated testing, which the positive screening data already indicated. And, as noted in question #5 above, additional testing of known toxicants can be essential in understanding their full range of health and environmental impacts.
Relevance of Animal Studies
Question: People aren't rats, so why are the kinds of animal tests involved here relevant to human health anyway?
Answer: Every regulatory agency in the world relies on animal tests to set standards to protect both humans and animals from harm by environmental pollutants. It is true that animal effects are not perfect predictors of human effects; understandably, some chemical companies like to stress the occasional findings of differences between human and animal outcomes. But they are only occasional. Indeed, at least 90% of the chemicals that cause cancer in rats also cause cancer in people, for the limited number of compounds where there are enough data to make the comparison. And of course, tests on animals are directly relevant in assessing potential impacts on animals (thus the requirement to study effects on fish, for example).
Timing of Regulations
Question: Could animal tests cause delays in government regulation of dangerous chemicals?
Answer: There is no basis at all to believe that the HPV program will delay regulations that would otherwise come into effect, since those regulations have to be based on information already in hand. "Innocent until proven guilty" may be a questionable standard to apply to chemicals (as opposed to people), but U.S. law requires evidence of risk to justify chemical regulations. Ultimately, the HPV Challenge Initiative will help enable the government to take steps to establish protections for human health and the environment where needed.
Other Priorities
Question: Given that thousands of known toxicants remain in use, won't getting additional toxicity data just detract from efforts to regulate known problems?
Answer: This question assumes that we have already identified the highest-priority problems, which is by no means necessarily true. Even more importantly, it assumes that regulations are the only way to reduce exposure to toxic chemicals. But it's clear that both voluntary efforts by industry and pressure by the public can lead to exposure reductions even without regulatory action. For example, the highly successful Toxics Release Inventory program has focused public attention and often public pressure that helped propel dramatic reductions in releases of the 650 chemicals that it now covers, well beyond reductions required by regulations. But chemicals cannot be listed on TRI unless data exist that show potential toxic effects. Likewise, voluntary initiatives by industry to develop safer products depend on having information about which chemicals are less toxic. While the tests in the HPV program don't provide definitive data, they are an important step toward such information. In addition, toxicity isn't an on/off switch. Regulatory standards and public health both require focusing on which chemicals are "more toxic" and "less toxic." In a world that isn't about to abandon chemicals, this sort of triage is essential.
Genetic Toxicity protocols
Question: Could animal tests cause delays in government regulation of dangerous chemicals?
Answer: Until recently, critics justifiably complained that EPA was calling for use of animal tests for one type of genetic toxicity even though there is an approved protocol for a non-animal test. Thanks to efforts of animal protection activists, EPA announced at an April 26 public conference sponsored by TestSmart – a joint project by the Johns Hopkins Center for Alternatives to Animal Testing, Environmental Defense, and two other universities – that EPA would no longer do so. Subsequently, in a joint letter dated May 10, Environmental Defense and the Humane Society of the United States urged EPA to go further, and to expressly recommend use of non-animal tests for genetic toxicity. (EPA originally believed that the animal methods were more scientifically reliable, but now believes that the non-animal methods are acceptable for screening purposes, such as those covered by the Initiative.) Because the HPV Initiative is a voluntary program, EPA cannot require sponsors to use non-animal methods, but industry is clearly influenced by EPA's technical recommendations – especially since the non-animal approach is less expensive as well as less controversial and more humane. The May 10 letter urges EPA to distribute a statement on this and related issues directly to all chemical sponsors.
Test Species
Question: Is it true that the Initiative will involve testing on birds, rabbits, and guinea pigs?
Answer: Although OECD protocols theoretically allow testing on rabbits and guinea pigs, rats are less expensive and easier to use. Therefore, as a practical matter, any new health tests will be conducted on laboratory rats, which are bred for the sole purpose of toxicity testing. (Though of course, even laboratory rats should not — and will not — be treated inhumanely or needlessly sacrificed.) Ecological testing may also involve testing on fish, but that's testing that designed to assess hazards to fish themselves, and other aquatic species. At the April 26-27 meeting (see above), EPA clarified that birds will not be used in the Initiative. (Use of particular species in testing is one of the issues addressed in the May 10 joint letter.)
Skin Irritation Tests
Question: I've seen disturbing pictures that suggest that the Initiative involves burning animals' skin. Is that the case?
Answer: Most chemicals will be tested by ingestion; for those that occur as gases, testing will be by inhalation. There is no reason to believe that any of the testing under the Initiative will involve skin testing or burning of any kind - not skin irritation, and not dermal exposure. (This issue is also addressed in the May 10 joint letter.)
LD50 Tests
Question: Will the Initiative will involve use of the traditional LD50 test, which involves finding the dose that kills half of the animals, despite the existence of better and less cruel alternatives?
Answer: Existing SIDS protocols do not require use of the LD50, although they allow it, so that existing data from LD50 tests that have already been completed can be used, in lieu of additional testing. Where existing acute-toxicity data are not adequate, EPA has already recommended use of other approved tests for acute toxicity, which involve many fewer test animals (and thus are cheaper as well as more humane). Assertions that EPA requires use of the LD50 protocol are simply incorrect. (This issue is also addressed in the May 10 joint letter.)
Development of Non-animal Alternatives
Question: Wouldn't delaying the HPV program would allow validation of non-animal alternatives for lethal dose, skin abrasion, and acute fish toxicity?
Answer: The HPV program offers an opportunity to perform validation of non-animal alternatives for acute effects (and others), on a more rapid schedule than would otherwise occur. In any event, skin abrasion testing is not part of the HPV program.
Other Testing Initiatives
Question: Isn't the HPV program is just the tip of the iceberg, and won't the endocrine disruptor testing program means testing 85,000 chemicals on animals?
Answer: The HPV initiative is quite separate from the endocrine disruptor program. The latter would screen at most 15,000 chemicals using non-animal tests known as high-throughput screening tests. Chemicals that show potential effects on endocrine systems or that are otherwise high priority (e.g., because of significant human exposure) would then be tested using more sophisticated protocols, some of which include use of animals. Until the results of the high-throughput tests are available, there is no way to tell how many chemicals will be tested in these more sophisticated systems. Given the serious risks on reproduction and development potentially posed by endocrine disruptors to humans and animals alike, such effects warrant investigation. DDT caused the near-extinction in the United States of many species of birds by disrupting endocrine systems that control eggshell thickness – indeed, Environmental Defense was founded in 1967 by a group of scientists concerned about the effects of DDT on osprey and other wildlife.
Endnotes