3

Evaluation of the Evidence Base

This chapter describes the approach and methods that the committee used to identify and evaluate the scientific and medical literature on exposures to herbicides that occurred in U.S. military personnel during the Vietnam War as well as the process that the committee used to reach conclusions on the association between exposures to the chemicals of interest (COIs)—2,4-dichlorophenoxyacetic acid (2,4-D); 2,4,5-trichlorophenoxyacetic acid (2,4,5-T); picloram; dimethylarsinic acid (DMA or cacodylic acid); and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)—and a given health outcome. The committee’s process entailed the following steps: a literature search, screening of abstracts, a full text review of studies flagged in the abstract screening, and the evaluation of a final set of studies identified as relevant after the full text review. The first part of this chapter details the methodology used to identify and screen the literature. The second part of the chapter details the evaluation criteria used to review the relevant studies, including the types of studies considered, the health outcomes considered, and the categories of association used to draw conclusions about the strength of the evidence of possible health effects resulting from herbicide exposure. The committee also describes some of the issues it encountered when reviewing the literature on Vietnam War exposures and health outcomes, such as multiple exposures and individual variability. Because the current committee closely followed the approach used by prior Veterans and Agent Orange (VAO) and Update committees, much of the following information was previously described in those volumes as well, particularly Update 2014 (NASEM, 2016a).

IDENTIFICATION AND SCREENING OF THE LITERATURE

Literature Identification

Much as was the case with previous VAO committees, this committee was tasked with comprehensively reviewing, evaluating, and summarizing the scientific literature that has appeared since the previous biennial update regarding associations between TCDD and other chemicals present in the herbicides used by the U.S. military in Vietnam and health outcomes and then adding this new information to the existing compendium of evidence to draw conclusions on the strength of associations between exposure and health outcomes. To begin, the committee oversaw extensive searches of the scientific literature using a strategy adapted from prior committees’ literature search methodology (see Box 3-1). This committee’s search included additional terms to evaluate specific conditions called out in the Statement of Task: possible generational health effects, myeloproliferative neoplasms, and brain cancer, in particular, glioblastoma multiforme (see Chapter 1 for the full Statement of Task).

For this update, electronic searches of the medical and scientific literature were carried out on four databases: Web of Science, Scopus, Medline, and Embase. In the previous VAO updates, only Medline and Embase were used. The four searchable databases index biological, chemical, medical, and toxicological publications. The full texts of the articles were searched so that if any of the search terms was included in the title or abstract or indexed in the key words or text of the article (excluding the cited references section), the article would be included in the results of the search. Search terms included full and abbreviated chemical names, common and manufacturer trade names, the Chemical Abstracts Service numbers, and MeSH¹ descriptors for each of the COIs and other similar chemicals on an assumption that compounds with similar chemical structure may have analogous biologic activity. The aryl hydrocarbon receptor (AHR) and related MeSH descriptors were also included as primary search terms since this protein appears to mediate essentially all of the toxicity of TCDD.

Using the search terms in Box 3-1, the databases were searched in two phases, with the searches spanning over timeframes that were extended from those used in prior updates. In the spring of 2017, the databases were searched for articles published between January 1, 2014, and March 31, 2017. Then in early February 2018 the databases were again searched for any articles with the relevant search terms published between March 1, 2017, and December 31, 2017. The timeframes for those searches overlapped each other and also overlapped with the Update 2014 committee’s search (October 1, 2012–September 30, 2014). The overlapping dates made possible the capture of articles that might

___________________

¹ MeSH descriptors are sets of terms naming descriptors in a hierarchical structure that permits searching at various levels of specificity. For example, MeSH terms for 2,3,7,8-tetrachlorodibenzo-p-dioxin include “TCDD” and “dioxin,” without those terms having to be specified individually.

have been published but not indexed at the time of the first search. Other than dates, no limitations (such as language, populations, or species) were put on the search. In addition, potentially relevant articles were also identified by searching the reference lists of relevant review and research articles, books, and reports. Exact duplicate articles and those that had been summarized and referenced in Update 2014 were deleted. The committee became aware of a few studies that reported updated findings on relevant exposed populations (such as the Seveso, Italy, cohort and New Zealand phenoxy herbicide producers) published following the December 31, 2017, search cutoff and reviewed these studies as well.

Literature Screening

The search strategy was devised to ensure that abstracts of all potentially relevant articles were subjected to closer screening, but it also resulted in the

identification of a large number of non-relevant studies. The first search produced in excess of 12,000 “hits,” and the second search identified more than 1,600 articles of potential relevance. The articles were generally in English, but VAO committees have traditionally obtained translations for crucial ones that were not in English. Article titles and abstracts were screened for relevance by committee members and the Health and Medicine Division staff to determine which studies should be considered for full-text retrieval using the criteria in Box 3-2.

In the VAO series, neurologic deficits associated with Vietnam service are distinguished from psychiatric/psychologic conditions—such as posttraumatic stress disorder (PTSD), depression, and anxiety—and their sequelae. While, as past VAO committees have noted, the increased risks of these types of mental health conditions among veterans of all U.S. conflicts are of scientific and public health concern, military service alone, including deployment and service in Vietnam, is known to confer a range of exposures to potentially traumatic events that may be expected to increase the risk of developing PTSD and related psychologic conditions. For example, compelling evidence has established that the prevalence of PTSD is more than twice as high for operational infantry units exposed to direct combat than it is in the general population (Kok et al., 2012). Previous committees further drew the conclusion that it would be infeasible to disentangle potential adverse effects from exposure to the COIs on mental health outcomes that may occur independently of psychological effects accrued through military service. The current committee expands upon that perspective by placing it in a framework that underscores the relevance of the concepts of multifactorial causation, the literature on which has recently begun to mature and offer new insights. Both independent and interacting risk factors may play a role. An example of multiple independent effects can be seen in a recent study of Vietnam veterans in Australia, which found that both sons and daughters of veteran fathers diagnosed with PTSD had an increased risk for PTSD even after controlling for PTSD in the mother and for other diagnoses (O’Toole et al., 2017). The statistical interactions of risk factors, which can have synergist or antagonistic effects, can result in effects of combined exposures that would not have been predicted based on their independent impacts. An example of a synergistic interaction is the association with lung cancer from combined exposures to workplace arsenic and smoking: in this case, the risks from arsenic are much higher among smokers than among non-smokers (Hertz-Picciotto et al., 1992).

Disentangling the separate effects of combined exposures or risk factors in relation to a particular outcome does raise serious challenges, however, and it may indeed be infeasible when the correlations among those exposures are exceedingly high, to the point of inseparability, or when sufficiently large studies cannot be conducted. For example, service-related emotional trauma, such as occurs in combat, and exposures to the COIs in Vietnam are associated (J. M. Stellman et al., 1988), but whether they are separable depends on the extent to which those experiencing service-related trauma also had high exposures to the

COIs and, conversely, the extent to which those with high exposures to the COIs invariably had a high degree of service-related trauma.

Current epidemiologic thinking recognizes that multiple risk factors may operate not only independently, but also synergistically, and that sometimes the

combined impacts may have both an independent and synergistic or antagonistic component. Thus, a nuanced and comprehensive approach to combined exposures is critical to understanding causation. Underlying susceptibility is not always genetic, but can instead be a prior or concomitant exposure, and thus the possibility of multifactorial causation requires paying attention to confounding as well as to interactions. Few studies have rigorously addressed the combination of combat duty and the COIs through an assessment of interactions. In addition, a review of the vast toxicology literature that relates to the COIs reveals that there is a dearth of reports that address the potential associations and mechanistic explanations relating to how exposure to the COIs may influence the risk of developing mental health conditions. This applies specifically to an overall absence of published evidence as to how TCDD exposure could be etiologically implicated in the development of PTSD and related psychological comorbidities. Animal models will not be informative for studying potential associations between exposure to the COIs and development of PTSD. Because the literature appears to be quite limited, it does not provide the opportunity to address effects on PTSD or other mental health conditions that may occur as a result of the combined effect of military service/combat duty and the COIs. Thus, at this time the committee was not able to determine whether the COIs might magnify the impact of traumatic events.

Although exposure to 2,4-D, 2,4,5-T, TCDD, cacodylic acid, and picloram are most germane to the committee’s charge and given the most weight in the review of the evidence on a particular health outcome, chemicals that are structurally similar to the herbicides and contaminants found in the tactical herbicides used in Vietnam are assumed to have similar biologic activity and thus were also included in the committee’s review of the literature, consistent with prior VAO Update committees. These related chemicals include other phenoxy herbicides—such as 2-(2,4,5-trichlorophenoxy) propionic acid (Silvex), 2-methyl-4-chlorophenoxyacetic acid (MCPA), 2-(2-methyl-4-chlorophenoxy) propionic acid (Mecoprop or MCPP), and 3,6-dichloro-2-methoxybenzoic acid (dicamba)—as well as dioxin-like chemicals such as those found in mono-ortho and non-ortho-substituted polychlorinated biphenyls (PCBs) and also metabolites of organic arsenic (DMA^V). Very few epidemiologic studies on exposure to picloram or cacodylic acid have been published, which is another reason for the committee to consider metabolites of these compounds.

Examining the structural representation of the COIs shown in Figure 2-1, one can readily see the basis of an assertion heard repeatedly from individual Vietnam veterans that “benzene is contained in TCDD.” Indeed, the two rings at the ends of the three-ring structure constituting the basic structure of dioxin compounds, to which chlorine molecules or other chemical radicals can be attached, do have the molecular structure of a single benzene molecule, and the “dibenzo-dioxin” in TCDD’s chemical name does mean that the molecule is a benzene-substituted dioxane. The benzene ring structure is a basic building block of a vast number of organic compounds, both industrial (such as polyaromatic

hydrocarbons, the phenoxy herbicides, picloram, and PCBs) and natural (such as estradiol, a hormone present in both men and women). However, the biologically active compound benzene does not emerge from dioxin, whose three-ring structure is extremely stable and resistant to metabolism.

An interaction or synergism among the COIs or in conjunction with other agents is another theoretical concern. The committee was not charged with attributing effects to specific COIs, and joint effects among them should be adequately identified by the committee’s approach. The combinations of the chemicals with other agents that might lead to problems are virtually infinite, and hence, not feasible for systematic and comprehensive evaluation. Real-life experience, as investigated with epidemiologic studies, effectively integrates any results of exposure to a target substance in combination with other substances that may be etiologically relevant.

Thus, in aggregate, the primary COIs evaluated by the committee with respect to potential associations with adverse health outcomes among Vietnam veterans are 2,4-D, 2,4,5-T, picloram, cacodylic acid, and TCDD. As explained, inorganic arsenic and benzene were not considered as relevant service-related exposures among Vietnam veterans and thus were not evaluated in terms of their risk for adverse health outcomes.

The committee recognizes that in real-world conditions, human exposure to TCDD virtually never occurs in isolation and that there are hundreds of similar compounds to which humans might be exposed, including other polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans (PCDFs), and PCBs which have dioxin-like biologic activity. The literature on the other compounds, particularly PCBs, has not been reviewed systematically by the committee except for those reports in which TCDD was identified as an important component of the exposure or when the risks of health effects were evaluated in relation to exposures to dioxin-like chemicals. In many cases, when dioxin-like chemicals are present the exposure is expressed in terms of toxic equivalencies (TEQs), which are the sums of toxicity equivalence factors (TEFs) for individual dioxin-like chemicals as measured by activity with AHR. Studies that report TEQs based only on mono-ortho PCBs (which are PCBs 105, 114, 118, 123, 156, 157, 167, and 189) are considered even though their TEQs are several orders of magnitude lower than those of the non-ortho PCBs (77, 81, 126, and 169), based on the revised World Health Organization (WHO) 2005 TEF values (La Rocca et al., 2008; van den Berg et al., 2006). This is because the lower TEQs of the mono-ortho PCBs may be counterbalanced by their abundance, which is generally many orders of magnitude higher than the non-ortho PCBs (H.-Y. Park et al., 2010). For example, PCB congeners 118 and 156 are present in numerous human populations (United States, Europe, and Japan), at levels that are 4 to 5 orders of magnitude greater than the levels of TCDD itself (Bake et al., 2007; Becher et al., 1995; Lorber et al., 2009, Tsuchiya et al., 2003), with the levels of mono-ortho TEQs closely approximating those of TCDD or in some cases exceeding them (H.-Y. Park et al., 2010).

Exposure to non-dioxin-like PCBs only has not been considered by VAO committees for two reasons. First, Vietnam veterans may have been exposed to substantial amounts of the other chemicals, and the amount of exposure to those chemicals relative to TCDD has not been documented. Second, the most important mechanism for TCDD toxicity involves its ability to bind to and activate the AHR. Many of the other chemicals act by different or multiple mechanisms, so it is difficult to attribute toxic effects after such exposures specifically to TCDD. Furthermore, an individual’s exposures to dioxin-like chemicals and their non-dioxin-like counterparts tend to be correlated, which means that it is difficult for epidemiologic studies to attribute any observed association to a particular chemical configuration (Longnecker and Michalek, 2000); carrying out analyses in terms of TEQs somewhat circumvents this problem. The interaction or synergism among the COIs or in combination with other agents is a concern, but the combinations of the COIs with other toxicants, or physical, biological stressors is outside of the committee’s charge.

In cases when the title or abstract mentioned only broad chemical categories or used non-specific terms such as “herbicide” or “pesticide,” the full text was reviewed for mention of the specific COIs. It was evident from most of the abstracts that the article did not address health effects in association with exposure to the specific COIs. The committee only deleted studies from further consideration if it verified that exposure information specifically did not include at least one of the COIs or if the study was not at all relevant, such as studies that evaluated the efficacy of the COIs in killing specific plants or that measured concentrations of the COIs in environmental samples, studies that merely provided a description of manufacturing processes, or studies that did not include a health outcome. All studies that discussed health effects or changes in pathophysiology or cell signaling were considered if the text indicated that any of the herbicides of interest (or any of their components) may have been investigated. The committee only included literature that had undergone peer review or government reports and invited presentations that were provided to the committee, under the assumption that they have been carefully reviewed. The process of peer review by fellow professionals increases the likelihood that high-quality studies will appear in the literature, but it does not guarantee the validity of any particular study or the ability to generalize its findings.

Because the use of the general search terms “pesticides” or “pesticide,” “defoliants,” “defoliant agent,” and “herbicide” or “herbicides” resulted in more than 57,000 “hits,” the committee focused its search on the specific COIs as well as related chemical classes. Even using the more targeted terms for the COI identified many studies that examined the relationship between exposure to “pesticides” or “herbicides” and adverse health outcomes without identifying specific chemicals. After carefully reading the full text of an identified article, if none of the COIs were specified as contributing to the exposures of the study population, the committee generally excluded such a study from further consideration and

did not summarize it. An exception to this practice was made for studies that indicated exposure to herbicides but did not characterize exposure with sufficient specificity for their results to meet the committee’s criteria for inclusion in the evidentiary database. For example, numerous case-control studies characterized exposure to pesticides or herbicides on the basis of job titles, farm residence, or longest-worked industry. Consistent with prior VAO reports, such studies were not given full evidentiary weight because their results are not regarded as fully relevant for the purpose of this review, but they are presented more briefly under the heading of “Other Identified Studies.” However, given that the exposure was not adequately described for these studies, their results are not included in supplementary summary tables.

Articles using the generic term “herbicide” were kept if a published report specified that the COIs were among the pesticides or herbicides used by the study population or the COIs are used commonly for the crops identified in the study or the COIs are used commonly for a specific purpose, such as the removal of weeds and shrubs along highways. For instance, this rubric would apply to any published articles from the Agricultural Health Study because 2,4-D was one of the most frequently used pesticides in this large prospective cohort, but some results have lumped all herbicide exposure together. In general, if the COIs were not presented separately in the results of a study, the study was considered to be of limited relevance and acknowledged as an “other identified study.” Exposures to dioxin-like chemicals, such as mono-ortho and non-ortho substituted PCBs, PCDFs, and dioxins other than 2,3,7,8-TCDD, were also considered relevant to the committee’s charge and were included as long as the results were expressed in terms of total or individual congener toxic equivalents. The most weight was given to studies that used quantitative measures (such as TEQs) and specified exposure.

Studies with original data collection and analyses were preferred over studies that were re-analyses of a population (without the incorporation of additional information), pooled analyses or meta-analyses, reviews, and so on, and the former are the type of evidence that the committee preferentially considered when assessing the strength of association between herbicide exposure and a health outcome when drawing its conclusions. While studies of the latter type may be informative and may be discussed in conjunction with primary results or in synthesis sections on a given health outcome, they are not themselves part of the evidence dataset and therefore were not considered in the final count of new literature considered in this volume.

EVALUATION PROCESS

This section details the methods used by the committee for evaluating and synthesizing the relevant studies identified by the literature search and screening and for making its conclusions on the relationships between the COIs and health

outcomes among Vietnam veterans. The quantitative and qualitative procedures underlying the committee’s literature evaluation have been made as explicit as possible, but ultimately the conclusions about associations expressed in this report are based on the committee’s collective judgment. The committee has endeavored to express its judgments as clearly and precisely as the data allow.

A total of 14,750 abstracts were retrieved from the two literature searches. Full text was then obtained for any articles that were considered potentially relevant based on their titles and abstracts and after applying the inclusion and exclusion criteria. Full-text articles were distributed among the committee members based on their areas of expertise, with at least two committee members reviewing each paper. As with all VAO committees, the committee began its assessment of the literature by assuming neither the presence nor the absence of an association between exposure and any particular health outcome. Because of the variability in the descriptions and diagnoses of the health conditions considered in this report, the committee made no a priori assumptions about the usefulness of any article or report for a health outcome. Each study was reviewed and objectively evaluated for each health outcome it presented. If a study examined more than one health outcome, it was considered separately for each of those outcomes. After reading, if the full text revealed that the study met one of the exclusion criteria (see Box 3-2), it was excluded from further consideration.

After review of the full text of the identified articles, studies that were considered relevant (165 epidemiologic studies and nearly 100 toxicologic studies) were discussed and evaluated thoroughly, and are included in this report. New evidence on each health outcome was reviewed in detail, but the committee’s conclusions are based on the accumulated evidence of all VAO reports, not just on recently published studies.

The responsible committee members then presented the information from each new relevant study—including the methods used for selecting the study populations and conducting the research (i.e., design, measures of exposure and health outcomes, statistical analyses used, adjustment factors, etc.), the results, and an assessment of the strengths, limitations, and potential biases—to the full committee for discussion. Based on the details of exposure and the description of how exposure was measured, an epidemiologic study was classified either as a primary article, in which case it was given full evidentiary weight, or as a secondary article, in which case it was reviewed and more briefly described under the heading of “Other Identified Studies.” In a secondary study, the exposure was either not adequately described or considered not to be specific enough for its findings to contribute evidentiary weight. An epidemiologic study was also classified as secondary if the outcome was a biologic marker of effect as opposed to a recognized condition or disease. Studies that measured well-defined functional outcomes indicative of mild impairments and likely of worsening function that have not yet progressed to frank clinical diagnoses, such as cognitive function and loss of grip strength, as evaluated in Chapter 9: Neurologic Disorders, were

considered primary evidence. Mechanistic and toxicologic studies contributed to the evidence for biologic plausibility but were not considered primary studies, so that based on those studies alone, their weight would not be enough to change the level of evidence of an association.

Following the discussion of each individual study for a specific health outcome, the lead committee member for an outcome then reviewed and summarized for the full committee the epidemiologic evidence reported in prior VAO updates and the most recent conclusion from Update 2014 on that health outcome. The toxicologists on the committee provided a summary of previous and new mechanistic or toxicologic studies for that health outcome. Although both primary and secondary studies contributed to the committee’s conclusion regarding the evidence of the COIs to be associated with a particular health condition or outcome, primary studies were given more weight. Using all of the available information, the full committee then came to a consensus regarding the conclusion and assigned a category of association (discussed later in this chapter) on the strength of the evidence of exposure to at least one of the COIs and the health outcome under scrutiny. When drafting language for a conclusion, the committee considered the nature of the exposures, the nature of the specific health outcome, the populations exposed, and the quality of the evidence examined.

The draft text was reviewed and discussed in further plenary sessions until all committee members reached a consensus on the description of the studies and the conclusion for each health outcome. The committee did not use a formulaic approach to determining the number of primary or supporting studies that would be necessary to assign a specific category of association. Rather the committee’s review required a thoughtful and nuanced consideration of all the studies as well as expert judgment, and this could not be accomplished by adherence to a narrowly prescribed formula of what data would be required for each category of association or for a particular health outcome.

The supplementary tables to this report (available online at www.nap.edu/catalog/25137) were revised to include only the new primary epidemiologic studies; secondary studies, including mechanistic and toxicologic studies, were not included in the compendium evidence tables for each outcome. If no new primary studies for a health outcome were identified, the evidence table from Update 2014 was included. Effect estimates, data, and units of measure are presented as reported in the cited studies, except where otherwise noted. The committee did not collect original data, nor did it perform any secondary data analyses, such as meta-analyses.

Types of Studies Considered

The committee focused on epidemiologic studies because epidemiology deals with the determinants, frequency, and distribution of disease in human populations rather than in individuals and animal models, which have several

limitations as discussed below. Epidemiologic studies effectively integrate any results of exposure to a target substance in combination with other substances that may be etiologically relevant. Several types of epidemiologic studies were evaluated, including cohort, case-control, and cross-sectional designs. Other reports of health outcomes among veteran populations have described these major types of epidemiologic studies and chronicled the limitations inherent in conducting them and in using them to make conclusions based on associations between deployment-related factors and health conditions (IOM, 1994; NASEM, 2016a).

The committee weighed the importance of the epidemiologic studies in the following order: Vietnam veterans, occupationally exposed workers, and people who were exposed environmentally. Exposures gleaned from case-control designs were considered separately. Vietnam veterans (U.S. veterans and those from allied countries) are presumed to have been exposed to all the COIs; the limitations of such an assumption are discussed in Chapter 2. Although Vietnam veterans constitute the source population of interest, the committee has taken into account the potential for more precise quantification and evaluation of the risks of adverse health outcomes associated with the COIs in better characterized cohorts (for example, occupational and environmental cohorts). Including these more highly exposed populations had the additional advantage that epidemiologic studies of them were likely to have greater statistical power to detect any adverse effects that might occur with exposure. Within the occupational-study populations, due to the substantial differences in the nature and intensity of their exposures, workers in the production of herbicides and other industrial products contaminated with TCDD were prioritized, followed by those involved in occupational use of the herbicides of interest (such as agricultural workers). The committee for VAO and the first several updates gave more weight to results that were based on job title (for example, “farmer” with no additional information) than have the committees since Update 2006 (where such studies are now considered secondary and given less weight). Those first committees also entirely excluded findings from the Yusho and Yucheng PCDF and PCB poisonings, whereas recent committees have considered these and other environmental studies that analyzed for dioxin-like PCDF and PCB congeners and expressed the results in terms of TEQs. Toxicologic studies, particularly in animal models, are included to inform the understanding of biologic plausibility through the toxicology of the chemicals and their exposure pathways.

Vietnam Veterans

Because Vietnam veterans are the target population of the charge to the VAO committees, studies of these veterans (whether American or otherwise) have always been accorded considerable weight in the committees’ deliberations and are presented first in the literature for each health outcome. The committee reviewed all identified studies of U.S. and international Vietnam veterans

published since Update 2014. In general, few studies include objective measures of TCDD or other chemical concentrations; those that are available were performed in small subsets of Vietnam veterans, such as those from the Air Force Health Study and Army Chemical Corps. Instead, having served in Vietnam or participating in the Agent Orange Registry is often considered a proxy of herbicide exposure. Therefore, it is difficult to quantify the risk of specific health outcomes when the exposures of the total at-risk population have not been measured or estimated. In the absence of actual measures of exposure, comparisons between deployed and non-deployed Vietnam-era veterans are considered the next most relevant comparison. Moreover, in many studies of Vietnam veterans, not all health outcomes of interest were reported (in some cases there were too few cases to report, only specific health outcomes were of interest, or the veteran population was too young for a particular manifestation). Consequently, several other groups known or thought to have potentially higher and better-characterized exposure to TCDD or phenoxy herbicides than the Vietnam veterans themselves were considered.

Human Studies Among Non-Veterans

Whereas exposure estimation and characterization of the COIs has been lacking in many (especially early) studies of Vietnam veterans, studies of occupational exposure (for example, among chemical-production, paper and pulp, railroad, agricultural, and forestry workers) and of environmental exposure (for example, residents of dioxin-contaminated sites such as Seveso, Italy, and the areas around some former U.S. military installations in Vietnam) to TCDD and the other COIs provide evidence of exposure and health outcomes to supplement the studies of Vietnam veterans. Some occupational and environmental cohorts that received exceptionally high exposures have produced many informative results and provide stronger evidence about health outcomes than studies of Vietnam veterans because the exposures were better characterized and measured sooner relative to the exposure. Moreover, in several studies of chemical-production plant workers, the magnitude and duration of exposure to the chemicals were measured and generally greater than among Vietnam veterans, so the likelihood that any possible health consequences would be manifested was greater. Some of the populations used in these types of studies were large enough to examine dose–response relationships. Other populations, such as the Agricultural Health Study, a prospective cohort study of U.S. agricultural workers with specific information on the COIs, continue to contribute new evidence of association with health outcomes.

As the information on populations that had established exposure to the COIs has grown, VAO committees have become less dependent on data from studies that had non-specific exposure information (e.g., exposure to non-specified herbicides) and have been able to focus more on the findings of studies that had refined

exposure specificity. For each health outcome, occupational and environmental studies are presented after the studies of Vietnam veterans.

Animal and Mechanistic Studies

The committee used studies of toxicology data to determine whether there is a plausible biologic mechanism or other evidence of a causal relationship between herbicide exposure and a health effect. A positive statistical association between an exposure and an outcome does not necessarily mean that the exposure is the cause of that outcome. Data from toxicology studies may support or conflict with a hypothesis that a specific chemical can contribute to the occurrence of a particular disease. Insights about biologic processes inform whether an observed pattern of statistical association might be interpreted as the product of more than error, bias, confounding, or chance. Discussions on biologic plausibility are presented after new epidemiologic evidence and before the synthesis of all the evidence. The degree of biologic plausibility itself influences whether the committee perceives positive findings to be indicative of a pattern or the product of statistical fluctuations. Ultimately, the results of the toxicology studies should be consistent with what is known about the human disease process if they are to support a conclusion that the development of the disease was influenced by an exposure.

Aryl Hydrocarbon Receptor Many of the available toxicologic and mechanistic studies involve the AHR because it has been found that essentially all of the toxic effects of TCDD involve interaction with this protein (receptor). The AHR can modulate transcription by binding TCDD and other aromatic hydrocarbons with high affinity. The formation of an active complex that involves the intracellular receptor, the ligand (the TCDD molecule), and other proteins is followed by an interaction of the activated complex with specific sites on DNA, which can alter the expression of specific genes involved in the regulation of cellular processes. The affinity of TCDD for the AHR is species- and strain-specific, and responses to the binding of the receptor vary among cell types and developmental stages.

Although studying AHR biology in transformed human cell lines minimizes the inherent error associated with species extrapolations, it is still not clear to what extent toxicity is affected by the transformation itself or by the conditions under which cell lines are cultured in vitro. Genetically based differences in the properties of the AHR are known to exist in human populations (Zhou et al., 2010), as they are in laboratory animals, so there are genetically based differences in people’s responses to TCDD, which leads to some people having an intrinsically greater risk of toxic effects from the same TCDD exposure, while others have less risk. Therefore, as humans have AHR with differing affinities for dioxin, a single transformed human cell line will not accurately reflect the spectrum of responses observed in the entire human population. As a result, the

TEQ approach should be seen as qualitatively generalizable, but cannot be definitively quantitative when discussing differences between tissues or individual humans.

Limitations of Animal Models Animal models are the basis for many of the toxicologic and mechanistic studies, although cell lines and in vitro cell cultures (human or animal) are also used. Studies that use isolated cells in culture also can elucidate how a chemical alters cellular processes. The objectives of those toxicology studies are to determine what toxic effects are observed at different exposure levels and to identify the mechanisms by which the effects are produced.

To be considered an acceptable surrogate for the study of a human disease, an animal model must reproduce, with some degree of fidelity, the manifestations of the disease in humans. However, a given effect of an exposure in an animal species does not necessarily establish its occurrence in humans, nor does the apparent absence of a particular effect in animals mean that the effect could not occur in humans. In addition to possible species differences, many factors affect the ability to extrapolate results from animal studies to health effects in humans. For example, animals used in experimental studies are most often exposed to purified chemicals, not to mixtures. Even if herbicide formulations or mixtures are used, the conditions of exposure might not realistically reproduce the human exposures that occur in the field. Other variables, including the amount and duration of exposure, can be controlled precisely in laboratory settings.

TCDD is thought to be responsible for many of the toxic effects of the herbicides used in Vietnam. Attempts to establish correlations between the effects of TCDD on experimental systems and their effects on humans are particularly difficult because there are well-known species-, sex-, and outcome-specific differences in susceptibility to TCDD toxicity. Even in humans, the data on TCDD’s toxic effects are not consistent (DeVito and Birnbaum, 1995; Ema et al., 1994; Moriguchi et al., 2003). Differences in vulnerability may also be affected by variations in metabolism and by the rate at which TCDD is eliminated from the body. Although the degree of susceptibility is generally thought to be an inherent biological response, it can be influenced by life stage, past history, and co-exposures.

Many factors may contribute to differences between the results of controlled animal studies and the effects observed in humans. The following, which are elaborated on in Chapter 4, are among the most important:

• Physiologic differences. Laboratory animals are not miniature humans. Depending on the biologic process under investigation, a particular test species may match the human system more closely and so be a better experimental model than others.

• Magnitude of exposure. As is often the case for toxicologic studies of any chemical, the TCDD exposure used for animal studies has been many orders of magnitude higher than Vietnam veterans are likely to have received during military service, although the ultimate body burdens may not be as different.
• Duration of exposure. Although TCDD is a persistent organic pollutant, animal studies seldom examine the chronic low-level exposure that occurs over a period of years or even many months.
• Timing of exposure. It is well known that many organ systems are highly susceptible to xenobiotic exposure during critical stages of development, such as gestation; the response of some systems (such as the immune system) may also depend on the timing of the exposure to antigens relative to the timing of the exposure to xenobiotics such as TCDD.
• The route of exposure. The route of exposure by which an exogenous agent enters an organism may influence the nature of any toxic response elicited. The outcomes of animal studies may be perturbed by the delivery of treatment doses by “unnatural” routes of exposure, such as a bolus by gavage or intraperitoneal injection, but the route of exposure does not seem to be a major reason that the results of epidemiology studies may not agree with the findings of controlled studies for the COIs considered in the VAO series.
• Genetic constitution and expression. The etiologies of most diseases in humans and in animals are likely to be influenced by numerous genes and to involve complex gene–environment interactions, and preliminary evidence suggests that TCDD can induce epigenetic modifications of an organism’s DNA and alter expression of its genes.
• Sex differences. There are well-known differences between male and female animals (including humans) in susceptibility to xenobiotic exposures, some of which are modified by sex steroids.
• Prior and recurring exposures to multiple sources. Humans are exposed to xenobiotics from multiple sources throughout their lifetimes.
• Complex mixtures. Most xenobiotic exposures occur in complex mixtures; the makeup of these mixtures can heavily influence the ultimate toxic effects. In addition to the dietary modulation of responses to other exposures of both humans and animals, including dietary supplements in humans, prescription and over-the-counter pharmaceuticals, and other factors (such as cigarette smoking and ambient pollution) may have effects.
• Stress. Stress—of known or unknown origin—is a well-known modifier of human disease responses (such as immune responses); stress is an ever-present variable that is difficult to assess or control for in epidemiologic studies because there is substantial individual variation in response to it.

Selection of Health Outcomes

The committee was charged with summarizing the strength of the scientific evidence concerning associations between exposure to various herbicides and contaminants during service in the Vietnam War and individual diseases or other health outcomes. The list of outcomes was developed on the basis of diseases and conditions identified from searches of the scientific literature, as first done by the original VAO committee and amended and expanded as needed by VAO Update committees. For example, the current committee included chronic skin conditions, which had not specifically been addressed by prior committees. Some health outcomes in the series have been added in response to requests from the Department of Veterans Affairs (VA) and various Veterans Service Organizations and in response to concerns of Vietnam veterans and their families. Comments received at public hearings and in written submissions from veterans and other interested persons have been valuable in identifying issues to be pursued to greater depth in the scientific literature. The current committee defined a health outcome as any recognized diagnosis (based on ICD-9 or ICD-10²). Studies of exposure to the COIs that examined changes in pathophysiology, cell signaling, or other biologic markers of effect, such as hormone levels and blood counts, are more briefly mentioned as “Other Identified Studies” in applicable chapters, but because of the uncertainty of their relevance on the actual health outcome of interest, they are not considered further.

In aggregate, the health outcomes that the committee has focused on include cancers of all types, cardiovascular and metabolic outcomes such as diabetes, immune system disorders, and neurologic disorders. Other chronic health outcomes have also been considered, including respiratory disorders, gastrointestinal disorders, endocrine disorders, and bone conditions. Generational effects were specifically included in the committee’s charge and are addressed along with reproductive outcomes in a combined chapter. Although for most health outcomes the primary focus of the evaluation was on adverse outcomes in the veterans themselves, to examine potential effects, the children of Vietnam veterans and also later generations were included in the evaluation of the literature.

After reviewing the updated literature, the committee agreed that some reorganization of the health conditions was warranted for this volume. For example, it was more appropriate to group monoclonal gammopathy of undetermined significance, multiple myeloma, and amyloid lightchain amyloidosis under the heading “Plasma Cell Dyscrasias.” Whereas reproductive outcomes and effects on descendants were presented in a single chapter from VAO through Update 2010, the committees for Update 2012 and Update 2014 divided these outcomes into two chapters. The current committee believed that it made more sense to

___________________

² ICD refers to the International Classification of Diseases, which is maintained by the World Health Organization. Its current version, ICD-10, is the 10th revision.

present the reproductive outcomes and effects on descendants as a continuum in a single chapter.

Because any effect of the herbicides or contaminants in individuals or groups of veterans is evaluated in terms of disease or medical outcome, the committee paid particular attention to disease diagnosis and classification as it assembled pertinent data from various investigations related to a particular outcome in preparation for integrating the information. Pathologists, clinicians, and epidemiologists use several classification systems, including the International Classification of Diseases (ICD), which has been updated over time and with regard to topic. ICD codes are a hierarchic system for indicating the type of disease and site. For example, whereas many medical insurance and research investigators preferentially use the ninth revision of ICD (ICD-9) or its clinical modification (ICD-9-CM) to classify diagnoses or clinical outcomes, the tenth revision (ICD-10) is used to classify mortality information. Although outcome misclassification is still a possibility when recording a diagnosis with a specific ICD code into a record, the ICD system has been refined over many decades and is virtually universally used and understood, in addition to being exhaustive and explicit. Therefore, this and previous VAO committees have opted to use the ICD system as an organizing tool.

Many of the epidemiologic studies reviewed by VAO committees have not used the ICD approach to classification of disease and have relied instead on clinical impression alone. Self-reported diagnoses obtained from survey questionnaires often have inaccuracies due to recall bias and misinterpretations of the questions being asked. For example, a patient may report having been treated for stomach cancer when the correct diagnosis was gastric adenocarcinoma, gastric lymphoma, or peritoneal cancer. Furthermore, many epidemiologic studies report disease outcome by organ system. Sometimes this is done because there are few cases of a specific outcome and the study is lacking the necessary statistical power to make valid statistical associations using such specific diagnoses. However, such grouping into broader outcome categories can be problematic (and the same is true when categorizing potential exposure). For instance, the term “digestive system” may be used for conditions that are benign or malignant and that affect the esophagus, stomach, liver, pancreas, small intestine, large intestine, or rectum. Therefore, if a report indicated that a cohort has an increased incidence of digestive system cancers, then it would be unclear whether the association was attributable to excess cases of any single organ or type or to some combination thereof. Additionally, such generalization is complicated by the fact that the cause of cancer may differ among anatomic sites. For instance, there are strong associations between smoking and squamous cell carcinoma of the esophagus and between chronic hepatitis B infection and hepatic cancer. Furthermore, a single site may experience a carcinogenic response to multiple agents, while the same agent may cause cancer at multiple sites.

Several of the identified studies examined the association of exposure to at least one of the COIs (or a mixture containing it) with an individual health

outcome. However, some of the primary studies (for example, Collins et al., 2016; Cox et al., 2015) reported multiple health outcomes, and individuals might have been counted in more than one category. This can also be an issue in mortality studies when more than the primary cause of death is used. Designing studies to analyze concurrent health outcomes is much more difficult, and valid methods that can be applied with confidence to identify patterns among multiple health outcomes associated with a single exposure have not yet been developed. If such methods were developed, studies of multiple health outcomes could potentially provide more insight into whether the COIs cause multiple health effects, into competing risks among various health outcomes, and into the interactive effects of health outcomes.

Defining Statistical Association

Box 3-3 provides brief definitions of some of the most common terms used in the epidemiologic studies considered by this committee. The strength of an association between exposure and a condition is generally estimated quantitatively by using relative risks, odds ratios, correlation coefficients, or hazard ratios, depending on the epidemiologic design used. A ratio greater than 1.0 indicates that the outcome variable has occurred more frequently in the exposed group, and a ratio less than 1.0 indicates that it has occurred less frequently. Ratios are typically reported with a confidence interval (CI) to quantify random error. Statistical significance may be represented by a CI or a p-value. If the 95% CI for a risk estimate (such as a risk ratio or odds ratio) includes 1.0, the association is not considered to be statistically significant; however, if the interval does not include 1.0, the association is said to be statistically significant with a type I alpha error (likelihood that the association is due to chance) of less than 5% (that is, p < 0.05).

Determining whether an estimated association between an exposure and an outcome represents a real relationship requires careful scrutiny because there can be more than one explanation for an estimate. Bias is a general term for a distortion of the measure of association. There are several types of biases, and each type may affect the estimate differently. For example, misclassification bias may result in exaggerated or underestimated estimates, whereas self-selection bias affects the representativeness of the study population and can limit the applicability of the results to the larger population of interest. Another type of bias that may potentially affect studies of Vietnam veterans is detection bias, in which veterans who are encouraged to and who choose to participate in screening programs or registries, such as the Agent Orange Registry (discussed in Chapter 5) may have additional tests or follow-up exams that could potentially detect disease or a condition earlier or because more thorough assessments were conducted. This type of bias may occur disproportionately for Vietnam veterans with known or who report herbicide-exposure compared with Vietnam-era veterans who were not

known to be exposed to herbicides or who were deployed elsewhere. Detection bias may lead to an overestimate or underestimate of the true effect size.

Confounding is a common type of bias in epidemiologic studies that occurs when a risk factor for the disease is also related to the exposure and creates a spurious exposure–disease association. Potentially, if a confounder is known, there are methods that can be used to adjust for its effects; however, not all confounders are always known or identified, and unknown confounders may affect the estimate of association. Effect modifiers differ from confounders in that the

former are associated with the outcome but not the exposure. Effect modification occurs when an exposure has a different effect among different subgroups or strata. Chance is the degree to which an estimated association might vary randomly among different samples of the population studied.

Integrating New Information into the Evidence Base

Because relatively few studies of exposure to herbicides or TCDD (especially with objective measures) and health effects have been conducted among cohorts of Vietnam veterans, much of the evidence that VAO committees have considered has been from studies of well-documented populations that received occupational or environmental exposures to TCDD or specific herbicides. In many other studies, TCDD exposure was combined with exposures to an array of “dioxin-like” chemicals, and the herbicides were often analyzed as members of a functional class. Such studies are less informative for the committee’s purposes than individual results on a specific chemical but are still considered as part of the evidence base. In the case of epidemiologic studies, exposure to multiple, possibly toxic, chemicals is common in some industries, such as agriculture, and those exposures cannot be controlled for in the same way that laboratory experiments can. In its examination of these epidemiologic studies, the committee looked for evidence of health effects that are associated with the specific compounds in the herbicides used in Vietnam and sought consideration of and adjustment for other possibly confounding exposures. When all the available epidemiologic evidence has been evaluated, Vietnam veterans are presumed to be at increased risk for a specific health outcome if there is evidence of a positive association between one or more of the COIs and the outcome.

The quality of exposure information in the scientific literature reviewed by this and previous VAO committees varies widely. Some studies relied on interviews or questionnaires to determine the extent and frequency of exposure. Such self-reported information, which has the potential for recall bias, generally carries less weight than do more objective measures of exposure, such as levels of a contaminant as measured in serum or other biospecimens. The strength of questionnaire-based information as evidence of exposure is enhanced to the extent that the information can be corroborated or validated by other sources. Similarly, greater weight is given for studies that use more objective measures of health outcome assessments (such as clinical diagnosis).

In drawing conclusions, the committee examined the most thoroughly adjusted quantitative estimates of association, judged whether an adjustment for any crucial confounders was lacking, and evaluated the potential influences of bias and chance. In integrating the findings of various studies, the committee considered the degree of statistical significance associated with every estimated risk (a reflection of the magnitude of the observed effect and the power of the study designs) and took note of whether dose–response relationships were evident with increasing

exposure rather than simply tallying the “significant” and “non-significant” outcomes as dichotomous items of evidence. The committee also considered whether controlled laboratory investigations provide information consistent with the COIs being associated with a given effect and perhaps causally linked to it.

Categories of Association

As was done in previous volumes, the current committee used four categories of association to rate health outcomes based on the strength of the scientific evidence. The criteria for each category express a degree of confidence based on the extent to which bias and other sources of error could be reduced. The coherence of the full body of epidemiologic information, including biologic plausibility, is considered when the committee reaches a judgment about association for a given outcome. As was the case with the past three update committees, this committee did not use the Bradford Hill criteria for causality (Hill, 1965) as a checklist for its strength-of-association assessments. The committee discussed the evidence and reached consensus on the categorization of the evidence for each health outcome, which appears in the Conclusion section for each health outcome. Those categories of association have gained wide acceptance by Congress, VA, researchers, and veterans groups.

Sufficient Evidence of an Association

For effects in this category, a positive association between herbicides and the outcome must be observed in studies in which chance, bias, and confounding can be ruled out with reasonable confidence. For example, the committee might regard evidence from several small studies that are free of bias and confounding and that show an association that is consistent in magnitude and direction to be sufficient evidence of an association. Experimental data supporting biologic plausibility strengthen the evidence of an association but are not a prerequisite and are not enough to establish an association without corresponding epidemiologic findings.

Limited or Suggestive Evidence of an Association

In the category of “limited or suggestive evidence of an association,” the evidence must suggest an association between exposure to herbicide compounds or COIs and the outcome in studies of humans, but the evidence can be limited by an inability to confidently rule out chance, bias, or confounding. Typically, at least one high-quality study indicates a positive association, but the results of other studies could be inconsistent. Because the VAO series has a number of agents of concern whose toxicity profiles are not expected to be uniform—specifically, four herbicides and TCDD—apparent inconsistencies can be expected among

study populations that have experienced different exposures. Even for a single exposure, a spectrum of results would be expected, depending on the power of the studies, inherent biological relationships, and other study design factors.

Inadequate or Insufficient Evidence to Determine an Association

By default, any health outcome is placed in the category of “inadequate or insufficient evidence to determine an association” before enough reliable scientific data have accumulated to promote it to the category of sufficient evidence or limited or suggestive evidence of an association or to move it to the category of limited or suggestive evidence of no association. In this category, the available human studies may have inconsistent findings or be of insufficient quality, validity, consistency, or statistical power to support a conclusion regarding the presence of an association. Such studies might have failed to control for confounding factors or might have had inadequate assessment of exposure. Several health effects have been moved into or out of this category since the original VAO committee reviewed the evidence then available.

For nonmalignant conditions, the diversity of disease processes involved makes the use of broad ICD ranges less useful, but because VAO committees could not possibly address every rare nonmalignant disease, they do not draw explicit conclusions about diseases that are not discussed, and thus, this category is the default or starting point for any health outcome. If a condition or outcome is not addressed specifically, then it will be in this category.

Limited or Suggestive Evidence of No Association

The original VAO committee defined the category “limited or suggestive evidence of no association” for health outcomes for which several adequate studies covering the “full range of human exposure” were consistent in showing no association with exposure to herbicides at any concentration and had relatively narrow confidence intervals. A conclusion of “no association” is inevitably limited to the conditions, exposures, and observation periods covered by the available studies, and the possibility of a small increase in risk related to the magnitude of exposure studied can never be excluded. However, a change in classification from inadequate or insufficient evidence of an association to limited or suggestive evidence of no association would require new studies that correct for the methodologic problems of previous studies and that have samples large enough to limit the possible study results attributable to chance.