72
based on empirical evidence and/or important scientific or social values identifying
a favored position—e.g., protection of children, using linear extrapolations to
identify the carcinogenic potential of a substance—or a disfavored view—e.g.,
toxicants having related features should be treated similarly (Cranor 2018).
Defaults or presumptions create provisions for departing from the status quo
ante, identifying existing or new assumptions to create defaults, and avoiding “full
ad hoc examination of data and the [full] spectrum of inferences…” (NRC 2009;
Cranor 2018). They provide starting places, clarify discussions, avoid ad hoc
choices, facilitate predictable and consistent risk assessments, and provide a basis
for the public assessment actions (NRC 2009; Cranor 2018).
For instance, the Food Quality Protection Act mandates at least a ten-fold safety
factor for toxicants in food (NRC 2009). Second, positive findings in animal cancer
studies indicate that an agent can have human carcinogenic potential (NRC 2009).
Third, “[T]he basic pathways of metabolism and the occurrence of metabolites in
tissues in … species-to-species extrapolation of cancer hazard and risk” are similar
(NRC 2009). Fourth, linear extrapolations from high doses to low dose effects are
health protective, “When the weight of … all available data are insufficient to establish the mode of action for a tumor site (NRC 2009). There are numerous others, but
the above examples suggest the idea.
Presumptions can facilitate assessing and removing toxicants from commerce.
For example, the International Agency for Research on Cancer (IARC) and others
found that vinyl chloride (VC) is highly toxic to humans based on occupational
exposures in polyvinyl chloride plants (Heath et al. 1975). The chemically similar
compounds vinyl fluoride (VF) and vinyl bromide (VB) act by similar biological
mechanisms. Thus, IARC listed VF and VB as “probable human carcinogens” without having statistically significant evidence that VF and VB causes cancer in
humans; knowledge of the toxicity of VC, the results of animal studies, and mechanistic similarities of VF and VB to VC were sufficient for the inferences (Cogliano
et al. 2008).
In addition, IARC used mechanistic data to upgrade six probable human
carcinogens to known human carcinogens and to upgrade thirty-nine substances to
probable human carcinogens (Cogliano et al. 2008).
Presumptions can facilitate quicker identification and ultimately, removal of
products from commerce because of increased scientific understanding of biochemical interactions. Of course, once there is such a body of knowledge it can also
be used in premarket contexts to better identify products as toxic before they enter
commerce and put the public at risk.
A final scientific advance meriting comment is a series of recent papers
identifying “key characteristics” of toxicants leading to disease. The idea is to
identify “upstream” indicators that portend the development of disease and combine
those with other knowledge so that one does not need to wait for a full fledged
disease to show up in a population (Smith et al. 2016).
Both facilitating quicker removal of products from commerce and having the
same scientific tools to identify toxic substances before they even enter commerce
C. F. Cranor
based on empirical evidence and/or important scientific or social values identifying
a favored position—e.g., protection of children, using linear extrapolations to
identify the carcinogenic potential of a substance—or a disfavored view—e.g.,
toxicants having related features should be treated similarly (Cranor 2018).
Defaults or presumptions create provisions for departing from the status quo
ante, identifying existing or new assumptions to create defaults, and avoiding “full
ad hoc examination of data and the [full] spectrum of inferences…” (NRC 2009;
Cranor 2018). They provide starting places, clarify discussions, avoid ad hoc
choices, facilitate predictable and consistent risk assessments, and provide a basis
for the public assessment actions (NRC 2009; Cranor 2018).
For instance, the Food Quality Protection Act mandates at least a ten-fold safety
factor for toxicants in food (NRC 2009). Second, positive findings in animal cancer
studies indicate that an agent can have human carcinogenic potential (NRC 2009).
Third, “[T]he basic pathways of metabolism and the occurrence of metabolites in
tissues in … species-to-species extrapolation of cancer hazard and risk” are similar
(NRC 2009). Fourth, linear extrapolations from high doses to low dose effects are
health protective, “When the weight of … all available data are insufficient to establish the mode of action for a tumor site (NRC 2009). There are numerous others, but
the above examples suggest the idea.
Presumptions can facilitate assessing and removing toxicants from commerce.
For example, the International Agency for Research on Cancer (IARC) and others
found that vinyl chloride (VC) is highly toxic to humans based on occupational
exposures in polyvinyl chloride plants (Heath et al. 1975). The chemically similar
compounds vinyl fluoride (VF) and vinyl bromide (VB) act by similar biological
mechanisms. Thus, IARC listed VF and VB as “probable human carcinogens” without having statistically significant evidence that VF and VB causes cancer in
humans; knowledge of the toxicity of VC, the results of animal studies, and mechanistic similarities of VF and VB to VC were sufficient for the inferences (Cogliano
et al. 2008).
In addition, IARC used mechanistic data to upgrade six probable human
carcinogens to known human carcinogens and to upgrade thirty-nine substances to
probable human carcinogens (Cogliano et al. 2008).
Presumptions can facilitate quicker identification and ultimately, removal of
products from commerce because of increased scientific understanding of biochemical interactions. Of course, once there is such a body of knowledge it can also
be used in premarket contexts to better identify products as toxic before they enter
commerce and put the public at risk.
A final scientific advance meriting comment is a series of recent papers
identifying “key characteristics” of toxicants leading to disease. The idea is to
identify “upstream” indicators that portend the development of disease and combine
those with other knowledge so that one does not need to wait for a full fledged
disease to show up in a population (Smith et al. 2016).
Both facilitating quicker removal of products from commerce and having the
same scientific tools to identify toxic substances before they even enter commerce
C. F. Cranor