Report of the Independent Fact-Finding Panel on Herbicide 2,4,5-T

Executive summary

April 2013

Submitted to the Government of Ontario

Leonard Ritter PhD, Fellow, Academy of Toxicological Sciences, Panel Chair
Aaron Blair PhD, Fellow, American College of Epidemiology, Ramazzini Collegium, American Epidemiological Society
Nancy I. Kerkvliet PhD
Elliot A. Sigal HBSc
Jeanne Mager Stellman PhD

Beginning in the late 1940s to the late 1970s, the herbicide 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) was used widely for brush control in Ontario, across Canada, and throughout the world. In Ontario, the herbicide was applied by or on behalf of provincial government departments and agencies, municipalities, and private citizens. The U.S. military used 2,4,5-T as a component of herbicide mixtures designed to defoliate forests during the Vietnam War. The most widely recognized of these herbicide mixtures was Agent Orange, a 50:50 mixture of 2,4,5-T and 2, 4-D, so-called because of the identifying orange stripe on the outside of the drum.

The herbicide 2,4,5-T was contaminated with several chemical entities, the most problematic being 2, 3, 7, 8- tetrachlorodibenzo-p-dioxin (TCDD), more commonly known as dioxin. By the early 1970s, the Government of Canada had begun to express concern about the possible health effects of exposure to 2,4,5-T and, notably, its dioxin contaminant. In 1979, recognizing these emerging health concerns in Canada and internationally, and to better control the use of 2,4,5-T, the Government of Ontario instituted a permit requirement for using 2,4,5-T in the province. As the health controversy continued to build, the province chose not to issue any permits, thereby preventing further use of 2,4,5-T in Ontario.

Charge to the Ontario Independent Fact-Finding Panel on Herbicide 2,4,5-T

On March 11, 2011, the Minister of Natural Resources announced the formation of the Independent Fact-Finding Panel on Herbicide 2,4,5-T. The panel was given the following mandate by the Government of Ontario:

• Investigate and document the scope and scale of the use of 2,4,5-T herbicide in the province by Ontario government ministries and agencies, including those acting as agents or as contractors
• Determine the specific time period when 2,4,5-T herbicide was used in the province by Ontario government ministries and agencies
• Determine the geographic area where 2,4,5-T herbicide was used in the province by Ontario government ministries and agencies
• Examine whether exposure to 2,4,5-T herbicide in the affected areas may have potential health impacts
• Document the methods 2,4,5-T herbicide was deployed by employees of provincial ministries and agencies, and the interaction of those employees and the general public with 2,4,5-T herbicide application operations in affected areas
• Review the preparation, application, and storage of 2,4,5-T herbicide as well as provincial occupational health and safety, and laws, standards and workplace practices including the use of personal protection equipment and applicable training in place at that time
• Refer, where appropriate, to the Workplace Safety and Insurance Board any findings that could assist its work

The panel’s approach to its mandate

Dr. Leonard Ritter, professor (now emeritus) of toxicology in the School of Environmental Sciences at the University of Guelph, was appointed as chair of the panel. The first task for the panel chair was to identify the skills and expertise that would be required to fulfill the mandate of the panel and to recruit the most highly qualified individuals for the task. The recruitment process included direct contact with experts and announcements in scholarly media. Those who wished to serve were invited to submit expressions of interest highlighting relevant education and experience.

Following a thorough search, the panel chair advised the Minister of Natural Resources that he had appointed the following members:

• Jack Bend, distinguished university professor in the departments of pathology, physiology and pharmacology, and pediatrics at the Schulich School of Medicine and Dentistry, University of Western Ontario
• Aaron Blair, scientist emeritus at the U.S. National Cancer Institute, U.S National Institutes of Health, an expert in epidemiology.
• Elliot A. Sigal, president and senior scientist at Intrinsik Environmental Sciences, Inc., an expert in exposure assessment and risk assessment
• Jeanne Mager Stellman, professor emerita and special lecturer at the Mailman School of Public Health, Columbia University, an expert in exposure assessment

Citing personal reasons, Dr. Bend withdrew from the panel early in 2012. Dr. Nancy Kerkvliet, professor in the Department of Environmental and Molecular Toxicology and deputy director of the Environmental Health Sciences Center at Oregon State University, was then brought in as a toxicology expert.

The panel met several times in closed session to assign responsibilities within the panel, to review progress, and to discuss issues of relevance to its mandate. The panel also conferred frequently by email and phone, as required.

To assist the panel in its activities, a support office was established, staffed by two Ontario Public Service employees seconded from their home positions for the duration of the project. Staff in the support office reported directly and exclusively to the panel chair on all matters related to the panel. They provided administrative support to the panel and gathered additional information as requested by the panel members and the chair. The panel, with the assistance of the support office, also requested and received many other reports directly from various government departments.

Information reviewed by the panel

To ensure the panel could fulfill its mandate, all provincial ministries, the Cabinet Office, the Premier’s Office and all boards, agencies, and commissions of the Government of Ontario were instructed (subject to any privilege or other legal restrictions) to help the panel to the fullest extent possible. They were told that the panel could request any information necessary to fulfill its mandate and that (subject to any privilege or other legal restrictions), government ministries should cooperate with and assist the panel.

In addition to the support office, a single coordinating office was set up in the Ministry of Natural Resources (MNR ^{footnote 1[1]} ) to ensure that government departments were consistent in responding and in collecting, scanning, and storing information. All government departments and agencies submitted information to the MNR coordinating office, where staff scanned records and organized them into a searchable database. Approximately 4700 records were submitted to the panel, in batches as they became available, with the final submission in September 2012. The panel also relied on the expertise of the individual panelists to identify important individual research publications, reports of internationally recognized expert agencies, and reports published by various government departments at the provincial, federal, and international levels to augment the nearly 4700 records it received from Ontario government departments and agencies. All records on which the panel relied are available to the public, on request, and all resource material used by the panel is cited in the bibliography of this report.

Review of the toxicological and epidemiological evidence linking exposure to 2,4,5-T and TCDD to adverse health effects

Studies of the possible health effects from exposure to 2,4,5-T and TCDD have been reviewed by a number of respected national and international institutes and agencies. In addition, individual scientists have conducted independent reviews. Although the panel reviewed papers published by individual scientists, conclusions about potential health hazards from 2,4,5-T and TCDD were based primarily on the reviews and assessments of national and international organizations, including the Institute of Medicine (IOM), the United Nations World Health Organization, the U.S. Environmental Protection Agency (U.S. EPA), Health Canada, and the Ontario Ministry of the Environment (MOE). The panel was satisfied with the completeness and integrity of the institutional reviews and did not consider it necessary to re-assemble and re-evaluate the core studies on which the institutional reviews were based.

Similarly, the panel considered reports and reviews of experimental studies in laboratory animals that addressed the hazards of 2,4,5-T and TCDD in its overall assessment of potential adverse health effects. The panel did not attempt to derive its own tolerable daily intake values for TCDD or 2,4,5-T but was confident that the process and basis on which others had derived tolerable daily intake values for these compounds was sound.

As discussed in Chapter 6 of this report, the panel adopted a threshold model to characterize the potential for 2,4,5-T and TCDD to cause adverse health effects in humans. The threshold model is the internationally accepted basis for establishing safe exposure or dose levels for prescription drugs, residues in food, and contaminants in air and water. The approach is based on the presumption that a threshold exposure level can be identified below which adverse effects are not anticipated, even with lifetime daily exposure. The panel concluded that due to the lack of carcinogenicity associated with 2,4,5-T, the indirect carcinogenic effects of TCDD, and the general lack of mutagenicity associated with 2,4,5-T and TCDD, the threshold model was appropriate for application to 2,4,5-T and its contaminants.

The toxicity of 2,4,5-T and TCDD has attracted the attention of the scientific community for decades, and many publications are available on this topic. International consensus from both laboratory studies in animals and population studies in humans is that TCDD is a potent toxicant and potential human carcinogen, while exposure to 2,4,5-T, by itself, is associated with only limited toxicity. The United Nations Joint FAO/WHO Expert Committee on Food Additives has determined “…that a tolerable intake could be established for TCDD on the basis of the assumption that there is a threshold for all effects, including cancer….”

This determination is based on the indirect mechanism of action by which TCDD promotes cancer via activation of the aryl hydrocarbon (Ah) receptor. As described in Chapter 3, activation of the Ah receptor is considered an essential primary event that underlies all of the toxic effects associated with exposure to TCDD in both humans and laboratory animals. The United Nations Expert Committee also concluded that the establishment of a tolerable intake (a “safe” level) for TCDD based on effects other than cancer would also be protective of any cancer risk. The committee established a tolerable monthly intake of 74 µg/kg body weight (BW)/month, equivalent to a tolerable daily intake of 2.3 µg/kg BW/day. Health Canada and MOE have adopted the same tolerable intake of 2.3 µg/kg BW/day. For its risk assessment, the panel adopted the same tolerable daily intake level for TCDD as accepted by the governments of Canada and Ontario and the Joint FAO/WHO Expert Committee on Food Additives.

The panel also considered the need to establish a tolerable daily intake for the parent herbicide, 2,4,5-T. Since 2,4,5- T has little chronic toxicity, the panel adopted an acute tolerable daily intake of 20 µg/kg BW/day for short-term exposures to 2,4,5-T.

Exposure assessment

The panel used a two-part process to carry out the exposure assessment. The first step was an extensive review of every document in the MNR database that referred specifically to herbicide use. Information on where the chemical was used, how it was handled, and who was involved in the operations was extracted and summarized. Historical summary tables of spraying activities were constructed to estimate the volumes and concentrations used. Maps of the spraying were plotted to give a visual representation of the areas covered by historical spray activities. Exposure scenarios were developed for MNR spray programs to promote the regeneration of tree species with commercial value in northern Ontario; for Ontario Hydro maintenance of rights-of-way clearance around power transmission lines; and for Ministry of Transportation (MTO) maintenance to control weeds and brush along highways.

The exposure scenarios were then used to model potential human exposures to the chemicals of concern. Exposure assessment models took into consideration the chemical characteristics of 2,4,5-T and TCDD that determine how the chemical reacts in the ecosystem (e.g. how long does it persist? will it decompose in sunlight? what is the route of entry into the body?). Scenarios were created for the extent of impact of exposure from consumption of wild berries, plants, and wildlife; for aerial drift of sprayed herbicides; and for persistence of the chemicals in the soils and sediment in bodies of water. Because data on usage were incomplete, the panel used low, central, and high input parameter scenarios in its calculations to provide estimates of exposure across a range.

Exposure assessment modelling also took into account human physiological factors, such as body weight and breathing rate, which influence the extent to which a toxic substance may gain entry into the body and how it will interact once it enters. Where possible, the panel relied on standard values for environmental and physiological factors that were consistent with Health Canada and MOE. For occupational exposures, the panel used the U.S. EPA Occupational Pesticide Handler Unit Exposure Surrogate Reference Table and Pesticide Handler Exposure Database in its modelling. The occupational models take into account the use of protective equipment, and such equipment was part of MNR, Ontario Hydro, and MTO operating procedures. However, the panel had no way of knowing whether or not appropriate protective equipment was actually used, so again low, central, and high input parameter estimates were used.

The re-creation of exposure events that occurred decades ago is rife with unknowns and uncertainties. As such, many decisions were made along the way that influenced the outcome of the assessment. The quantitative—or numerical— exposure assessment required the input of many assumptions that affected large amounts of data and numerical variables. Some of these input variables were obtained from the published literature, while other information had to be scenario-specific and was obtained from the various sources of information available to the panel. Each of the decisions and input variables contain some element of variability and uncertainty, which can affect the final results and conclusions. The goal of quantitative exposure assessment was to produce a conservative model so that estimates of disease risks were not underestimated. Given the tendency for the assumptions used to overestimate exposure, it is likely that this assessment errs on the side of conservatism (safety).

Exposure from proximity to spraying

The likelihood and extent of exposure to the chemicals is a function of how close to spray activities an individual was. Herbicide spraying occurred in remote areas as well as close to more populated, residential areas. Remote spraying was generally by aerial application, while residential spraying was ground based.

MNR used herbicides primarily on forests in remote areas, with most spraying done aerially. MTO used ground-based methods to apply herbicides adjacent to roadways in both urban and rural areas. For the MTO scenario, properties were assumed to be located no closer than 30 m from the roadway. Ontario Hydro applied along transmission corridors in both remote and populated areas via both aerial and ground-based methods. For the Ontario Hydro scenario, residential properties were assumed to be located no closer than 30 m from the right of way and recreational areas set back several hundred metres from the right of way.

The drift of herbicides from aerial spray was estimated using a standard model (AgDrift) and was used to estimate the concentrations in soil, food crops and wild berries, and wildlife contamination.

At-risk populations

The at-risk population (called receptor groups in the analysis) consisted of groups of people with potential exposures from occupational sources, from residences near sprayed areas or from recreational activities in or near sprayed areas (see list of receptor categories below). Given the many years over which herbicide spraying occurred, it is likely that many different individuals were occupationally involved in these activities for various periods between 1948 and 1979.

Occupational

• Mixer/loaders
• Applicators
• Flagmen
• Junior rangers
• Supervisors

Non-occupational

• Residents
• Recreational visitors
• Hunters/anglers
• First Nations people (dietary)
• Bystanders of aerial spray drift or overspray
• Spouses/other family members

Several possible exposure routes with the chemicals or with contaminated soil were considered in the assessment models. These included inhalation; ingestion of soil or contaminated game, fish, plants, or berries; and direct skin contact. Exposure to contaminated clothing was possible for several receptor groups, but because the panel thought it to be less likely and less easily quantifiable it was not included in the mathematical modelling.

Documentation of transport, storage, and disposal of the herbicides was also examined and summarized.

The development of standards and regulations for pesticide use in Ontario

As part of its mandate, the panel was asked to review the preparation, application, and storage of 2,4,5-T herbicide as well as provincial occupational health and safety laws, standards, and workplace practices, including the use of personal protective equipment and applicable training in place at that time.

Phenoxy herbicides were first used in North America in the late 1940s, and uses were largely unregulated. The first regulations focused on minimizing drift and damage to neighbouring crops rather than protecting the health and safety of applicators and bystanders. Standards and regulations were evolving from the 1950s, and language specific to protecting human health began to be considered in the 1970s, either via controlling occupational exposures or minimizing contamination of food intended for human consumption. Although the panel assessed and documented the evolution of pesticide use standards and regulations in Ontario, it could not determine the extent to which practices intended to protect the general public or to protect workers involved in pesticide handling, mixing, loading, and application were practiced, monitored, or enforced.

As early as 1954, Ontario Hydro produced the Manual of Spray Practices, which noted that “herbicides are known to be non-toxic to man and animals but operators should minimize breathing in the mist, or having it contact their faces.” By 1962, the Ontario Hydro Manual of Spray Practices (A0141556 ^{footnote 2[2]} ) was expanded to include matters of personal safety. In a 1965 circular, MTO listed neoprene gloves and blue coveralls as safety clothing available for weed spray crews. In 1976, the Ontario Ministry of Labour indicated in a memo that safety glasses and shields should always be worn; protective, impervious clothing was needed; and respiratory organic vapour canisters were required. The memo also noted that “…workmen have been observed spraying with no protective clothing or equipment.”

General concern about potential human health effects that might be associated with exposure to 2,4,5-T and its dioxin contaminant began to emerge in the early 1960s. In 1979, responding to safety concerns related to 2,4,5-T and the dioxin contaminant, the Government of Ontario instituted a permit requirement for using 2,4,5-T in the province. As the health controversy continued to build, the provincial government chose not to issue any permits, thereby preventing further use of 2,4,5-T in Ontario, six years prior to the de-registration of the herbicide by the federal statutory authority.

Conclusions of the panel

Exposure assessment is a key component of the risk assessment paradigm and was a critical aspect of the panel’s undertaking. The use of 2,4,5-T in Ontario spanned a period of approximately 30 years. The panel carefully considered the issue of exposure and developed an exposure assessment effort to re-create an exposure profile for provincial workers and residents that could be used in an exposure-response evaluation, a basic premise of the toxicological risk assessment paradigm.

The panel adopted a well-documented and widely practiced risk assessment protocol to assess the potential health hazards of 2,4,5-T and TCDD, to evaluate the potential for exposure in various Ontario populations that could have arisen as a result of use by Ontario government departments and agencies, and to relate these exposure levels to exposure conditions under which adverse effects have been observed in studies of laboratory animals and from human populations. The panel could not locate any specific exposure monitoring data for humans related to uses of the herbicide by Ontario government departments and agencies. Lacking such direct information on exposure from herbicide applications by the provincial government, the panel carried out a comprehensive assessment of use records submitted by various departments and agencies to determine the locations and amounts of 2,4,5-T applied and to use this information to estimate likely human exposure to the herbicide and its contaminant during the period of use by the Ontario government. These application records indicated that while many departments used the herbicide to some extent, most uses by Ontario government departments and agencies were carried out by MNR, MTO, and Ontario Hydro. The panel assessment, therefore, focused on the uses by these three departments. The panel notes that most use of the herbicide in Ontario was not by the Government of Ontario but rather by private and municipal users. The assessment of these non-government uses was beyond the panel’s scope and not included in the evaluations.

Because of the complexity of the exposure estimation process and the uncertainties associated with it, the panel provided a range of exposure estimates (a low, central, and high estimate) for each exposure scenario. These levels were based on assumptions included about each input parameter in the exposure assessment algorithm. The high-case estimate employed a series of individual worst-case assumptions for each component of the algorithm, applied one after another, introducing a repetitive input parameter selection bias. The high-case estimates represented the likely worst-case exposure situation. The low case estimates always took the low-end exposure parameters to generate the lowest likely overall exposure based on the information at hand. The central exposure estimate used intermediate (central or average) parameters to create a mid-level overall exposure estimate. Parameters used to develop exposure estimates included the amount of the dioxin contaminant in 2,4,5-T, the use of personal protective equipment, the fraction of applied herbicide penetrating the forest canopy, and the half-lives of TCDD and 2,4,5-T in soil. A range of potential exposure values (facilitated through the use of selected low, central, and high input parameters) were developed to provide a general appreciation for the level of uncertainty and variability present within the quantitative exposure estimates. Overall, the panel was of the view that the central estimate likely represented the most likely estimate of exposure.

The panel adopted a margin of safety approach to characterize the risk to various population subsets that may have been affected by the use of 2,4,5-T by Ontario government departments and agencies. With this approach, exposure levels of 2,4,5-T and TCDD that did not result in adverse health effects were identified from laboratory animal and human population studies. These exposure levels were adjusted by applying an uncertainty factor intended to compensate for the inherent uncertainty in the approach (for example, when studies in laboratory animals were used to predict effects in humans and inherent exposure assessment uncertainties in population studies in humans). This adjusted dose level was then directly compared with the level of estimated exposure the panel developed for different scenarios for various Ontario populations. Generally speaking, when this ratio was greater than 1, exposures may have exceeded a safe threshold; if less than 1, it was within the range of anticipated “no hazardous effect” based on the scientific literature.

The panel found situations where the benchmark reference exposure was exceeded for certain subgroups of the Ontario population. These exceedances were almost entirely restricted to chronic occupational exposures to the mid and highest levels of TCDD and from the central and high estimates for Ontario Hydro, MNR, and MTO. Where exceedances were observed at mid exposures to TCDD, they were typically marginal.

Bystander exposures, defined by the panel as being of brief nature and not from personal direct use of the herbicide, were found to exceed the benchmark level only in the case of Ontario Hydro and MNR and only for the highest TCDD exposure scenario. Mid and low bystander exposures to TCDD and all MTO bystander exposures were not found to exceed the benchmark margin. None of the bystander exposures to 2,4,5-T for Ontario Hydro, MNR, and MTO exceeded the benchmark level.

The panel also constructed maps of areas treated with the herbicide in Ontario to better visualize where potential opportunities for human exposure may have existed. These maps were developed from geographic information from spraying records provided by the ministries. The maps show that most spraying was remote from residences, and when standard models were applied, the risks for both exposure and for health effects for most situations in these areas were found to be well within the acceptable benchmark margins described in the report.

Given the results of the exposure assessment and margin of safety comparisons carried out by the panel, it was the panel’s opinion that:

• The central estimate of exposure was the best estimate.
• Low and high estimates provided a useful approximation of the range of exposure to be considered because of the uncertainty in the estimation process.
• MTO: Exposures exceeded the benchmark for a relatively small number of individuals involved in ground mixing/loading and application by MTO employees.
• Ontario Hydro: Exposures exceeded the benchmark for some individuals involved in ground mixing/loading and application by Ontario Hydro employees.
• MNR: Exposures exceeded the benchmark for some individuals involved in backpack mixing/loading and application by MNR employees and junior rangers.
• MNR: Exposures exceeded the benchmark for some individuals involved in aerial mixing/loading and flagging by MNR employees.

Some bystander exposures exceeded the benchmark margin of 1, but the highest estimated exposures corresponded to margins of safety of less than 2. Given the assumptions and uncertainties in the bystander estimates, the acute nature of bystander exposures, and the safety factors inherent in the toxicological reference values (TRVs) used in the evaluation, the panel was of the view that the bystander margins of safety were not necessarily reflective of adverse health outcomes.

The panel considered an array of possible adverse health outcomes that have been reported to be associated with exposure to TCDD and/or 2,4,5-T. Although individuals exposed to these chemicals may have experienced some of these outcomes, it cannot be concluded that the outcome was due to the exposure. The risk assessment process used by the panel was of necessity directed at the population level and not intended to describe risks to any specific individual. The assessment revealed that several occupationally exposed populations may have experienced exposures above the margin of safety. However, such exposures were restricted to occupational populations (Ontario Hydro, MNR, and MTO workers) who were chronically exposed. As noted above, those whose exposures greatly exceeded the benchmark margin were almost all occupationally exposed populations who had the highest levels of TCDD exposure. The TRVs used to assess exposure acceptability erred on the side of safety, as regulatory policy for assessing risk requires, but populations with the highest exposure levels will not necessarily experience adverse health effects. The assessment merely indicated that acceptable margins of safety were exceeded, and people’s health could be affected.

The panel’s approach was a population level assessment and not intended to describe risks to any specific individual. A population level assessment provides the probability of occurrence of an adverse outcome in a population group but cannot deliver an absolute determination of occurrence of a disease in any individual. Individuals vary widely in their responses to exposures and in the exposures themselves. Having said this, the panel noted that the United States Institute of Medicine (IOM) has developed a list of diseases for which individuals exposed to 2,4,5-T and its contaminants may be at increased risk. The panel agreed that the IOM’s findings reasonably reflect the current state of hazard assessment for these exposures. In Ontario, individuals with occupational exposures and a few bystander populations with the highest estimates exceeded the margin of safety benchmark. The associated risk of developing diseases as a result of 2,4,5-T or TCDD exposure for an individual in these categories would likely be very low. It is also important to note that an adverse effect will not necessarily occur, even in those cases where the margins the panel estimated have exceeded the benchmark.

This summary forms part of a larger report. Download the complete report. Single printed copies are available by contacting Service Ontario online or calling 1-800-668-9938 or TTY/Teletypewriter (for the hearing impaired): TTY: 1-800-268-7095 . (La version française peut être téléchargée).

© 2013, Queen’s Printer for Ontario
Printed in Ontario, Canada
(0.5k P.R., 13 05 07)
ISBN 978-1-4606-1564-5