Rabbit Rest: Can Lab-grown Human Skin Replace Animals in Toxicity Testing?

New experimental models based on three-dimensional reconstructions of human skin are helping to reduce chemical testing on live animals, but cannot yet replace animals altogether


It likely comes as no surprise that many common household chemicals and medical products as well as industrial and agricultural chemicals, may irritate human skin temporarily or, worse, cause permanent, corrosive burns. In order to prevent undue harm regulators in the U.S. and beyond require safety testing of many substances to identify their potential hazards and to ensure that the appropriate warning label appears on a product. Traditionally, such skin tests have been done on live animals—although in recent decades efforts to develop humane approaches, along with ones that are more relevant to people have resulted in new models based on laboratory-grown human skin.

The most recent chapter of this ongoing effort was written on July 22 when the Organization for Economic Cooperation and Development (OECD)—an international group that, among other things, provides guidelines to its 32-member countries on methods to assess chemical safety—officially approved three commercially available in vitro models of human skin for use in chemical testing. Specifically, the new guideline (OECD Test No. 439) stipulates that the models can serve as an alternative to animals in tests for skin irritation, one of several human health endpoints for which chemicals are tested. Similar 3-D models were approved for corrosion tests in 2004, leaving many hopeful that soon it may be possible to the assess the full spectrum of a chemical's effects on human skin—from irritation to corrosion—without using live animals.

Recent legislation in the European Union (E.U.) has made the need for nonanimal test methods an urgent matter. Since 2009 the E.U.'s Cosmetics Directive has banned animal testing of cosmetics ingredients and finished products (with a few exceptions for certain health endpoints). What's more, the ban applies to cosmetics products marketed in the E.U. As a result, manufacturers in countries outside the union must comply with the regulation to sell their products in E.U. markets. Importantly, however, E.U. law prohibits experiments on animals if alternative methods are "reasonably and practicably available."

At the same time, manufacturers are slated to use millions of animals in the coming decade to comply with a 2007 European Community program called Registration, Evaluation, Authorization and Restriction of Chemicals (REACH), which has requested more extensive safety evaluations for approximately 30,000 chemicals.

Although the new in vitro skin models will certainly reduce the need for, and suffering of, animals in some chemical tests, these methods are not yet ready to completely replace animal-based skin tests. To understand why, it is necessary to consider how toxicity testing is conducted and some of the regulatory constraints in place.

Skin Right
Since the mid-1940s researchers have tested the skin-irritancy potential of chemicals primarily on albino rabbits. In a procedure called the Draize rabbit skin test, a patch of the animal's fur is shaved and the test substance is applied to the bare skin for up to four hours. A trained technician then monitors the skin for as many as 14 days for signs of an adverse reaction and subjectively scores the severity of the reaction. The Globally Harmonized System of Classification and Labeling of Chemicals (GHS) classifies a substance as an irritant if it causes reversible damage to the skin or a corrosive if it causes burns or permanent scarring (irreversible damage).

Opponents have long criticized animal-based skin tests as too variable because the responses observed can differ among individual animals. Results also often vary among labs, partly because the measures of the reactions are qualitative.

Others are concerned that adverse reactions identified in animals do not always reflect how humans may respond to the chemical exposure. "The rabbit is not a particularly good model for human irritation or corrosion, largely because the barrier properties in rabbit [skin] are far less robust that in the human," says John Sheasgreen, president of MatTek Corp., which manufactures one of the approved models. He explains that when his company compared results from chemical testing on their in vitro skin with available human and rabbit data, the "in vitro model correlates much better with the human experience than the rabbit experience."

The artificial skin models come from normal human skin cells, which are cultured in specialized media to form a three-dimensional reconstruction of the real thing. Sheasgreen explains that the in vitro skin closely resembles intact human skin both structurally and biochemically—it consists of multiple layers of cells and has a stratum corneum, the dead layer of cells on the surface that provides a protective barrier. These properties make it amenable for use in toxicity testing.

Proponents of the in vitro alternatives claim that they yield more reproducible and more quantitative results, because the response to irritancy is based on measures of cell viability. Furthermore, because the new tests are based on human cells, the data they provide may better represent how people would react.

In formally testing the alternatives, the European Center for the Validation of Alternative Methods (ECVAM) concluded that the models can reliably and accurately predict whether a chemical is an irritant.

Saving skin
According to the new OECD guideline, the in vitro skin can serve as a "stand-alone replacement test for in vivo skin irritation testing," depending on a country's hazard classification requirements. The regulatory system in E.U. countries, for example, classifies substances as corrosive, irritant or neither, which in vitro tests can adequately identify. By contrast, some regulatory authorities in some countries, such as the U.S. Environmental Protection Agency, require identification and labeling of an additional category of so-called "mild" irritants, which the in vitro tests cannot discern. Additional testing, such as animal tests or ethical trials on human volunteers, may be necessary to identify such low-level irritants.

The approval of the human skin models would not eliminate animal testing, explains Laurence Musset, principal administrator of the OECD Environment Directorate's Environment, Health and Safety Division. "The in vitro tests have a limited scope and don't work for all chemicals," Musset says, adding that each country's regulatory agencies will decide whether the in vitro or the animal test should be used for skin irritation studies.

OECD guidelines recommend using the human skin tests as part of a sequential testing strategy, in which in vitro methods are used before animals in combination with a weight-of-the-evidence (WoE) analysis of existing data on a substance or related chemical. In this testing strategy if a corrosive or irritant is identified in vitro, follow-up testing in animals is not necessary; if a substance tests negative in vitro, however, a WoE analysis may be used to back up that finding. Otherwise, in vivo testing may be necessary to rule out potential false negatives.

William Stokes, the executive director of the U.S. Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM), says his organization is currently evaluating the applicability of the in vitro irritation tests for chemical testing in this country. Stokes points out that the in vitro corrosion tests miss nearly 20 percent of known corrosives tested, which is worrisome because there is no clear strategy to follow when a substance tests negative.

"The guideline says is that if [a substance tests negative], you may be able to conclude that it's not a corrosive using a WoE approach," Stokes says. Information that might be considered in a WoE analysis includes results from in vitro irritation tests for that substance.

But the false negatives from the corrosion tests were never examined when ECVAM validated the skin irritation models. ICCVAM is currently doing that evaluation, and preliminary results show that the irritation tests fail to classify at least two known corrosives even as irritants. Without follow-up in vivo testing, these corrosives could enter consumer products without any hazard labeling.

"We want to make sure that if something is corrosive, it's not slipped through these tests," Stokes says. "The data we have show it's a possibility. That…indicates that we need to better understand the usefulness of these assays. Our goal is to prevent injury and disease. It's important to us to have accurate safety testing methods."

Another issue is that the false positive rate for the in vitro skin irritation approaches 30 percent—a substantial number of substances are identified as irritants when they really are not.

"We don't want to excessively over-label products as hazardous or people would probably start ignoring safety labels," Stokes says.

The bottom line is that in vitro methods help to reduce animal use, but "full replacement is not something science supports at this point," Stokes says. "We're trying…to find ways and approaches that can further reduce uncertainty in those decisions, to reduce the number of circumstances where you might have to use [an animal]. As we take advantage of new advances in science and technology we'll continue to make progress in this area. It's a win–win—we're protecting both public health and animals."

Rights & Permissions
Share this Article:


You must sign in or register as a member to submit a comment.
Scientific American MIND iPad

Give a Gift & Get a Gift - Free!

Give a 1 year subscription as low as $14.99

Subscribe Now >>


Email this Article