The U.S. Environmental Protection Agency is embarking on a multiyear journey to prioritize and assess the risks of tens of thousands of chemicals in the U.S. market—and much of that evaluation will be done without the use of vertebrate animals.

Most of the chemicals found in products sold in the U.S., except for prescription drugs and pesticides, have not been subjected to extensive toxicity testing in lab animals. Such testing is expensive and time-consuming and sacrifices numerous animals to yield results that aren’t easily extrapolated to humans. Under changes enacted last year to the Toxic Substances Control Act (TSCA), EPA has new authority to collect toxicity information from manufacturers to evaluate the health risks of chemicals in everyday products. That information, combined with the specifics of where and how those chemicals are used, will help EPA decide how to manage chemical risks.

The amended TSCA requires EPA to promote alternative testing methods that reduce or replace the use of vertebrate animals. Some of those methods include high-throughput assays involving human cells and tests that examine how chemicals affect biological pathways known to lead to adverse health outcomes. The tools also include data analysis software that compares information about a chemical with that of other substances that have similar properties to predict how new chemicals will affect human health and the environment. The agency is under a statutory deadline to develop a strategic plan by June 22, 2018, for promoting such alternative methods.

EPA sees alternatives to animal testing as playing a key role in “risk-based decision-making for both new and existing chemicals,” said Louis (Gino) Scarano, a toxicologist in EPA’s Office of Pollution Prevention & Toxics. Scarano was speaking at a meeting last month to gather input from industry, international government agencies, animal welfare groups, and other stakeholders on how to incorporate new methods into chemical risk assessments.

For the 80,000 or so existing chemicals regulated under TSCA, alternative technologies could help the agency “determine which ones to focus on for risk evaluation,” he noted. For new chemicals, there are “huge opportunities” for using new approaches to mine existing information to help regulators figure out what information to ask for, Scarano said.

The chemical industry echoed EPA’s enthusiasm for these new approaches. “The most promising applications of these new technologies” are predicting exposure and bioactivity, said Rick Becker of the American Chemistry Council, the primary trade group for chemical manufacturers in the U.S. The revised TSCA has led to a “grand experiment in the U.S.” to prioritize chemicals with new technologies in a way that focuses efforts on those chemicals that really deserve the attention, he added.

But there are significant hurdles to get tests that do not involve animals in place for regulatory decision-making. To start, EPA does not know about many of the technologies being created and used in-house by industry. Such technologies are not standardized, nor are they—or the data they generate—typically shared.

Much of the discussion at the TSCA meeting focused on a road map under development by the National Toxicology Program Interagency Center for the Evaluation of Alternative Toxicological Methods (NICEATM). That road map is being coordinated by NICEATM’s Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM), a group of representatives from 16 federal agencies and multiple interagency work groups, to help federal agencies incorporate alternative methods into the safety evaluation of new chemicals and medical products. The committee released a draft of the road map in August and expects to finalize it by mid-January 2018.

The road map reflects a shift in thinking about method development: Rather than try to replace each existing animal test with an alternative method, the road map encourages government agencies to think about what information they need to make a decision. Then they can seek methods to create that information.

For EPA in particular, if the agency can identify specific areas that need to be addressed, such as inhalation toxicity or endocrine disruption, “we will have a better likelihood of developing new methods to meet those needs,” said Warren Casey, director of NICEATM.

EPA agrees that the way it evaluates acute inhalation toxicity is ripe for change. EPA spends a lot of time considering inhalation exposures in the workplace for chemicals before they are put in commerce, Scarano said. The agency has started a pilot project to consider different kinds of tests for lung toxicity. As a first step, EPA is evaluating physicochemical characteristics, such as surfactancy and particle-size distribution, for their ability to predict lung toxicity. The surfactancy of a chemical, for example, provides clues as to whether it will disrupt the protective surfactant film of the lungs. Particle-size distribution determines how deep into the lungs a substance will travel.

Meanwhile, NICEATM and the National Institute of Environmental Health Sciences are supporting a grant for the development of an in vitro, human lung-cell-based inhalation assay. Representatives from EPA, the Department of Defense, and the Consumer Product Safety Commission are on the steering group helping the developer create the assay system, according to Anna Lowit, cochair of ICCVAM and a senior science adviser in EPA’s Office of Pesticide Programs.

Lowit spoke about the inhalation assay at a workshop in late November hosted by the National Academies of Sciences, Engineering & Medicine. The goal of that workshop was to bring together stakeholders to explore how to build confidence in the use of toxicity information from methods that don’t rely on animal testing. Participants presented a few case studies highlighting how regulators are already using such methods.

One of the case studies involves EPA’s pesticides office, which gets applications for about 20 to 30 new pesticide products annually. Those applications involve 200 to 300 animal-based toxicity studies. The office is working closely with ICCVAM to replace six acute toxicity tests that rely on rats, rabbits, and guinea pigs, with alternative tests that follow Organisation for Economic Co-operation & Development (OECD) guidelines. The six tests are oral, dermal, and inhalation acute toxicity; skin irritation; eye irritation; and skin sensitization.

EPA has also made progress in adopting high-throughput computational toxicology models in its program to screen for endocrine disruptors, chemicals that perturb hormonal pathways. The agency announced in 2015 that it would allow use of a high-throughput estrogen receptor model as a replacement for three screening assays, two of which test chemicals for estrogenic activity in rats.

EPA is now evaluating alternative high-throughput models for the other eight screening assays required under its Endocrine Disruptor Screening Program, including tests for disruption of the androgen receptor, steroid biosynthesis, and thyroid pathways. The androgen receptor model, which EPA presented to a panel of external advisers at a meeting in late November, is close to being adopted by EPA as an alternative method. High-throughput models for steroidogenesis and thyroid disruption are still a few years away.

It took EPA more than 20 years to get to this point of using high-throughput models for the screening-level assays. And those chemicals that show potential for endocrine disruption will need to be further tested using traditional animal assays. Nonetheless, the fact that EPA now accepts a high-throughput estrogen receptor model in place of three previously required lower-throughput in vitro and in vivo estrogen assays is critical for establishing confidence in alternative methods, Catherine Willett, director of regulatory toxicology, risk assessment, and alternatives at the Humane Society of the U.S., said at the TSCA meeting in early November.

Willett and others are encouraging EPA to go back and evaluate the appropriateness and reliability of animal testing methods to get a sense of how well alternative methods stack up against them. EPA’s move to adopt alternative technologies “is not just about reducing our animal use; it is fundamentally about making better public health decisions,” EPA’s Lowit said.

Regulators in other countries are also keeping a close eye on EPA’s activities related to reducing the use of vertebrate animals in chemical safety testing. Representatives from Canada’s health and environment agencies attended the TSCA meeting, as they too are grappling with the use of alternative methods for prioritization and screening-level risk assessment of chemicals. Collaboration and data sharing are “absolutely fundamental,” said Christine Norman, director of existing substances at Health Canada. Data sharing, however, comes with the additional challenge of protecting intellectual property.

EPA expects to have a draft of its plan for reducing vertebrate animals in chemical safety testing by the end of February, with the goal of finalizing it by the June 22 deadline.