(Q)SAR: A Way Forward Commentary
Published: December 6, 2007
Dr. Gilman D. Veith
International QSAR Foundation to Reduce Animal Testing
1501 West Knife River Road
Two Harbors, MN 55616 USA
Given the current overwhelming dependence on animal testing, a societal goal of eliminating all such testing in safety assessments for chemicals is often dismissed as an impossible goal to attain. The business community and regulatory agencies are unlikely to embrace such a goal, after decades of reliance on animal testing, without evidence of a solid scientific foundation for a new kind of safety assessment. Though hundreds of alternative methods exist to make testing faster and cheaper have been developed, validation of such methods advances at a glacial pace. Fortunately, significant efforts are underway to create the scientific underpinning for safety assessments that do not begin with animal testing.
In 2007, proposing to eliminate animal testing is a goal not unlike the proposal of a societal goal to put a human on the moon or mars before modern space science had been developed. Despite enormous skepticism in the 50’s, many early space scientists shared a vision for how different technologies could work together to make space travel not only possible…but a reality. Those scientists knew that it was only a matter of time before a human stepped on the moon. They also knew that technologies unheard of by most scientists at the time would be needed to overcome the more intractable problems.
A public forum to discuss “a way forward” toward the goal of eliminating animal testing is long overdue. For many years we have accepted the goal of merely reducing animal testing…a goal comparable to only flying part-way to the moon. With reduction of testing as a goal, animal testing has increased rather than decreased. More importantly, we as a society have lost countless other advantages that a new vision for toxicology and safety assessment might bring. The goal of eliminating animal testing is possible. It will, however, require a new paradigm for chemical safety assessments which does not begin and end with data from animal toxicity tests.
As with most major paradigm shifts, early efforts to overcome the limitations of outdated visions and practices often become a source of skepticism and resistance for the new vision. In risk assessment, the movement to greatly expand in vitro test methods as alternatives to in vivo animal testing has been filled with unmet expectations and promises. More recently, the era of “omics” and high through-put screening technology has breathed new life into the fascination with alternative test method development. Many of these alternatives make the familiar promise of reducing animal tests. Computerized models based on quantitative structure-activity relationships (QSAR) have used the gimmick phrase, in silico methods, to herald a new modeling age to replace animal testing. Through it all, we have reduced neither the costs nor the number of animals used to test chemicals.
Despite this steady stream of promises from “more rapid alternative methods,” the basic presumption that chemical assessments must begin with additional animal tests is the gravity that holds toxicology in the old paradigm. Escaping from the old paradigm will require new technology, parts of which are already being developed in scientific disciplines outside the traditional scope of toxicology. Setting a goal to eliminate animal testing is the important first step in forging the alliance of new technologies needed for a moon shot in toxicology.
There is a way forward.
We can begin by extending elements of today’s familiar assessment practices. Toxicology, like chemistry, is based on the premise that similar chemicals behave similarly and produce similar biological effects. When similar chemicals do not behave similarly, it is logical to reexamine our concepts of similarity, rather than reject the basic premise that keeps chemistry and toxicology from becoming completely phenomenological practices. For example, optical isomers of a chemical may “look alike” to the non-chemist, yet they can have much different biological activities. In this example, it is only a simplistic view of chemical structure that would consider such isomers to be similar. If we can group chemicals into true analogues of structure, we will realize that testing all members of a group for all hazards is wasteful and unnecessary.
The practice of grouping analogous chemicals into categories is widespread and well founded in chemistry, biochemistry, chemical engineering and toxicology. In physical chemistry, the quantification of chemical structures for grouping analogues with the same intrinsic chemical interactions is known as the field of QSAR. Although QSARs are often trivialized by being portraying as simple statistical models of empirical data, the primary focus of QSAR is to quantify chemical structures so that all chemical analogues, when grouped together, exhibit the same chemical behavior. When this QSAR approach is integrated with the extensive body of knowledge of metabolism and toxicity testing, entire categories of chemicals can be reliably assessed for hazards even though many of them have never been tested.
An important legislative experiment is beginning in the European Union with the (Registration, Evaluation and Authorisation of Chemicals) REACH legislation. REACH legislation requires the registration and thorough hazard identification of tens of thousands of chemicals, with the provision that QSAR estimates can be used to justify a conclusion that no further testing is needed. This provision would fail if QSAR is assumed to be a library of models for various hazard endpoints, because the needed QSAR models would not be available for another decade. However, if QSAR-based categories of 50-200 similar chemicals are formed and a few of them have already been tested or assessed, it is likely that entire categories can be exempted from any further testing or adequately classified by regulatory authorities using the Globally Harmonized System.
The cost of REACH implementation in terms of animal lives is a subject of considerable speculation. Were the European Commission to focus on the systematic grouping of registered chemicals into categories and then merge those categories with critiqued mammalian, plant and aquatic databases in ecotoxicology, QSAR methods have the potential to virtually eliminate additional animal testing under REACH, or at least reduce testing to a small fraction of today’s estimates.
The Organization for Economic Cooperation and Development (OECD) has already issued a guidance document for the formation of chemical categories based on the intrinsic mechanisms and toxicological behavior of the chemicals. To facilitate the mechanistic formation of chemical categories, OECD has also created a QSAR Application Toolbox which is free to all stakeholders. Over the next year, the use of this QSAR approach should provide many examples of eliminating further animal testing for entire categories of chemicals because we already have adequate data for analogues.
Predicting the toxicological behavior of a chemical from first principles may still be a long term goal, but predicting behavior of close analogues of chemicals which have already been tested can be achieved quickly if the necessary scientific disciplines work with a common vision.
©2007 Gilman Veith