As citizens in a world full of technological marvels, where new products appear at an ever-increasing pace, we enjoy abundant clean air and water, healthy and predictable food supplies, and lightning advances in medicine; perhaps nearly forgetting that, not so long ago, such things were luxuries, at best. As scientists in this world, Toxicologists (and related disciplines) share in the responsibility to not only maintain or accelerate the pace of scientific understanding and its practical application, but also ensure that we do so without accumulating unnecessary risks to our health and longevity along the way.
Whether arising spontaneously, or induced following some interaction with agents in our environmental, or likely following some interaction of both endogenous and exogenous factors, human cancer is one risk always considered during the evaluation of proposed new drugs, pesticides, and other products to which people may be exposed. In the United States alone, over 1.5 million people will be diagnosed with cancer this year, and over 550,000 will succumb (ACS Facts and Figures, 2015), making it second only to heart disease in terms disease-related mortality for northern Americans (http://www.cdc.gov/nchs/fastats/deaths.htm).
For decades, rats and mice have been the model organisms used for identifying potential hazards to human health, including possible cancer induction, with the two-year rodent cancer bioassay serving as the gold standard test for human hazards following chronic exposure, in the absence of sufficient human data. The National Toxicity Program (NTP), part of the National Institute of Environmental Health Sciences (NIEHS), was formed in 1978 and charged with supplying information on potential human health risks in order to support U.S. regulatory agency decision-making processes. In the initial development and optimization of the rodent cancer bioassay, NTP has found that a two-year exposure duration in adult rodents was optimal for the detection of tumors, including rare or late-appearing varieties, and that with 50 animals per treatment group, they have an approximately 80% chance to detect a 15% increase in tumor induction, if the background incidence (i.e., that which occurs naturally, without intervention) was relatively infrequent. Over its nearly 40 year history, entailing the evaluation of over 580 different agents including everything from individual chemicals to mold and mixtures (e.g., sunscreen), research at the NTP has evolved beyond just cancer bioassay testing, and is currently relied on by a wide range of users: international health agencies such as the International Agency for Research on Cancer (IARC), as well as federal and state health-protection agencies, when characterizing potential human health hazards; public media organizations like Consumer Reports and the New Scientist, when describing the scientific support for product-related health concerns; and scientists worldwide when evaluating the state of knowledge databases for research planning or product development needs.
In terms of evaluating human cancer hazards, the NTP is also under a mandate to produce a “Report on Carcinogens” (RoC), in which they attempt to determine how likely exposure to numerous compounds is to cause human cancer, based largely upon (but by no means limited to) the results of two-year bioassays in rats and mice. Recent initiatives at the NTP and elsewhere have led to the evaluation of non-mammalian models in the effort to try and predict human cancer hazards, including zebrafish as an in vivo model organism, and high-throughput in vitro batteries of cellular function and molecular signaling assays as part of the Tox21 and ToxCAST initiatives. In the most recent phase of the Tox21 testing, efforts are now focused on trying to identify 1,500 or so “sentinel” genes, and to facilitate the quantitative in vitro to in vivo extrapolation (QVIVE) of this information, which United States health protective agencies (e.g., EPA, FDA, CDC, etc) could use in their hazard estimation efforts. This research direction is consistent with the objectives typically described as “the 3 R’s:” the reduction in use of, refinement in the application of, and eventual replacement of animal test models.
In addition to the laudable goal of reducing the use of rodents as test models when they can be replaced with other suitable in vitro and/or non-mammalian systems, the sheer cost of a standard two-year rat bioassay (coming in around $3.2 million, each), is also driving innovation in the refinement of methods used to predict potential human cancer hazards. This has led to the hypothesis proposed by some researchers at pharmaceutical companies that a two-year duration may not be necessary in certain situations, e.g., when specific information is known regarding proposed compounds which would lead reasonable scientists to predict lack of a positive response in the rodents. This prospective investigation, currently underway, proposes that a two-year rat bioassay would not positively inform hazard estimation for chemicals with no evidence of histologic risk, genotoxicity, or hormone system effects in shorter duration rodent bio-assays. In a retrospective analysis, chemicals which were “clean” of those three properties were predicted to have no cancer outcomes reported in two-year rat bioassays with a 91% sensitivity: this accounts for approximately 40% of all rodent bioassays performed. However, if a chemical or agent elicited some measure of activity in the three areas noted above, then a two-year rat bioassay would likely be informative, and would still be performed. A similar initiative in the US EPA program responsible for evaluating pesticide safety would use the specific molecular characteristics and predicted mechanisms of toxicological action of proposed compounds to determine if and when further testing would be required, and would supply testing waivers when new data would be unlikely to further inform the potential for particular hazards to human health, including possible cancer hazards.
While these approaches, including the increasing application of systems biology approaches to estimate human health hazards, may soon facilitate a faster, cheaper and more humane information-gathering process, there are still some hurdles to overcome. One of these is metabolism: parent compounds may not have inherent activity, but could be metabolized to any one of potentially numerous toxicologically active metabolites in rodents and humans. Unfortunately, metabolism is not well handled by current in vitro models, which could lead to false negatives, and possibly unacceptable health outcomes. The prospective development of a metabolically competent mammalian cell line could be one way forward, but regardless of the approaches eventually employed, work to overcome this and other obstacles to the wider adoption of alternatives to the two-year rodent bioassay is well underway.
This blog discusses highlights from the SOT Annual Meeting and ToxExpo Roundtable Session The Future of Carcinogenicity Testing (Safety Assessment Approaches for Product Development)