Europe overreaches on chemical controls

The European Union’s REACH regulations on chemicals came in for serious criticism during the planning stages, and since the programme was launched in 2006 it has been under close scrutiny, both expert and activist. In tomorrow’s edition of Nature, toxicologist Thomas Hartung and chemist Costanza Rovida look in detail at animal testing of chemicals under the new EU regulations, and argue that the costs in terms of animal lives and money spent are likely to rise dramatically.

REACH was negotiated between 2001 and 2005, and during that time a number of attempts were made to estimate the financial and other costs of regulation, including the number of animals used in toxicity tests. Official figures suggest probable costs of €1.6 billion and 2.6 million animals, and these estimates are based on chemical production between 1991 and 1994.

The latest list of REACH-controlled chemicals contains a total of 143,835 substances. Allowing for redundancy, errors in deposition, and a 5% annual growth in the European chemical industry since 1994, the actual number is more likely to be around 68,000. This is the figure used by Hartung and Costanza in their study.

Even with the most optimistic assumptions regarding animal numbers and the necessity for additional tests, Hartung and Costanza found that REACH will require some 54 million vertebrate animals and €9.5 billion over the next 10 years. This is 20 times more animals and six times the financial cost of the official estimates, which by anyone’s standards is a considerable accounting discrepancy. The authors claim that the feasibility of REACH is therefore under threat.

Much of the projected increase in animal testing is the result of two-generation studies for reproductive toxicity in which chemical effects are monitored in the offspring of exposed rodents and their immediate descendants. In practical terms this means an average of 3,200 rats per chemical, compared with 784 for a single generation study, and a five-fold increase in financial cost. In addition, changes to the REACH regulations introduced a requirement that two-generation studies be repeated in a second species, further increasing animal use and money spent.

Better safe than sorry? Hartung and Costanza point out that multi-generational animal studies have serious problems with false positives, and provide marginal gains in safety information. With false positive rates of 40–60%, REACH may lead to the costly withdrawal of widely used chemicals, and instil fears in the minds of consumers. The reaction of animal rights campaigners is another matter, and one not explored in the Nature article.

Alternatives to in vivo toxicity tests are under development, but none are likely to be validated within the next decade. Computational modelling is limited by the complexity of reproductive toxicity when chemical mixtures are involved. The only real alternative, say Hartung and Costanza, is an extended single generation study, guidelines for which are currently being drawn up by the Organisation for Economic Co-operation and Development (OECD).

Hartung and Costanza estimate that the use of the OECD test would reduce the need for animals in reproductive toxicity tests by 40–60%, and overall animal use by 15%. The authors recommend a moratorium on reproductive toxicity testing until the OECD guidelines are completed. They go on to discuss the need for longer-term developments in toxicity testing, but we are talking here about research that could take many years to emerge from the laboratory.

Only a couple of decades ago, those of us with an interest in animal welfare were looking forward to an age in the not-too-distant future in which non-human species are no longer sacrificed on the altar of industry for the benefit of mankind. It turns out that this was a naive hope.

Further reading

Thomas Hartung & Costanza Rovida, “Chemical regulators have overreached”, Nature 460, 1080 (2009)