Click for next page ( 6


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 5
INTERIM FINDINGS AND RECOMMENDATIONS 5 eluded in present databases appear to be available at the National Ar- chives for abstraction and use by researchers and other interested parties in future studies. Given current knowledge and available data, the contractor has adequately demonstrated that the draft model is a valid means of assess- ing wartime herbicide exposure of Vietnam veterans. On the basis of these findings, the committee concludes that a valid exposure reconstruction model for wartime herbicide exposures of US veterans of Vietnam is feasible. It therefore recommends that the De- partment of Veterans Affairs and other government agencies facilitate additional epidemiologic studies of veterans by nongovernmental or- ganizations and independent researchers. The foundation for the findings and the conclusion and recommenda- tion are addressed below. This discussion will make reference to but not reiterate materials published or in preparation for publication by the Co- lumbia University researchers. The committee's 1997 report encouraged the dissemination of scholarly findings resulting from conduct of the re- search and encouraged dissemination of the findings through publication in peer-reviewed journals. The committee commends the Columbia Uni- versity researchers for their pursuit of that goal and for their accomplish- ments. It strongly believes that discussion and publication of the results in scholarly venues strengthens the review process and advances the wide dissemination of the materials. The committee will continue to oversee the work of the Columbia University researchers. It will produce a second report that will review the completed research effort, transmit the contractor's report and sup- port materials to VA, and offer any additional findings, conclusions, and recommendations that it deems appropriate. The data developed during the research effort will be made available to interested parties without restriction. FOUNDATION FOR FINDINGS The contractor has developed databases of wartime spraying and acci- dental dispersion of herbicides, of troop locations and movements, and of land features and soil typology. The Columbia University researchers have developed a geographic information system (GIS) into which they have placed an extended and refined database of spray missions and other documented herbicide re- leases based on the HERBS files discussed in the first National Academy of Sciences report on the effects of herbicide spraying in Vietnam (1974~.

OCR for page 5
6 EXPOSURE OF VETERANS TO AGENT ORANGE AND OTHER HERBICIDES This work has led to a substantially expanded inventory of herbicide operations and better information on flight paths of aerial spray missions, number of gallons sprayed (gallonage), and chemical agents. The GIS also includes documented spray mission targets; herbicide storage, transport, and unplanned-dispersal information; military unit identification codes; locations of military units, bases, structures, air fields, and landing zones; movements of combat troops; land features2; soil typology; and locations of civilian populations. The development and contents of the GIS are addressed at length in a 2003 Environmental Health Perspectives paper pub- lished by the Columbia University researchers (Stellman et al., 2003a). As part of this effort, they have expanded and are correcting and validating an archive of previously tracked locations of combat battalions stationed in III Corps Tactical Zone in 1966-1969; this zone was the site of extensive spraying, and the resource may be of great importance in future epide- miologic studies. The contractor has developed an effective exposure assessment tool to assign a metric the E4 Exposure Opportunity Index (EOI) for herbi- cide exposure that is based on proximity to spraying in time and space and on the amount and agent sprayed. The databases discussed above form the basis of the computation of the E4 EOI. In brief, this EOI combines the location, gallonage, and herbi- cide data in the GIS with an environmental-decay factor to generate an estimate of the potential for exposure at a particular location and time. The E4 EOI developed as part of this research has been refined from earlier formulations and now incorporates a more sophisticated method- ology and extensive additional data (Stellman et al., 2003a). The Colum- bia University researchers have made various assumptions regarding the values of variables and how they influence exposure potential, but users may modify the assumptions. A software system that implements the calculation of the E4 EOI is in an advanced stage of development and will be delivered as part of the contractor's final submission. Figure 1 illustrates the level of detail contained in the GIS. The range of calculated EOIs and information gathered to date on troop locations is sufficient to demonstrate the feasibility of future epidemio- logic studies. Additional location data for troops not currently included in present databases appear to be available at the National Archives for 2 Including coordinates of a variety of features: elevations and land contours, rivers and streams, mountains and highlands, coastal areas and mangrove forests, bays and estuaries, and man-made structures such as roadways and utilities.

OCR for page 5
INTERIM FINDINGS AND RECOMMENDATIONS abstraction and use by researchers and other interested parties in future studies. Preliminary experience with the calculation of E4 EOIs suggests that they vary over several orders of magnitude (Stellman et al., 2003a). Figure 2, which illustrates a representative calculation of indices,3 shows areas of the country where (for the time period shown) there were no known exposures and areas where estimates vary between 104 and over 9 x 106. The metric thus appears to have a sufficient range to allow for informative epidemiologic studies of Vietnam veterans. The Columbia University researchers have developed an approach to classifying military units so that they can be separated by the degree to which their missions required frequent changes in location. That approach has permitted them to develop a database of locations and changes in location over time for about 80 percent of all Army troops, most Air Force personnel, and Navy personnel assigned to construction battalions or permanent installations. Their experience indicates that additional lo- cation information may well be available in the National Archives and possibly other data repositories (albeit not in an easily accessible form). A how-to manual developed by the Columbia University researchers en- capsulates the experience they gained in generating available location information to facilitate future research (Stellman et al., 2002~. It must be noted, however, that proper implementation of the model will require effort and expertise on the part of the user. Given current knowledge and available data, the contractor has ad- equately demonstrated that the draft model is a valid means of assess- ing wartime herbicide exposure of Vietnam veterans. The 1997 IOM report indicated that validation of the exposure assess- ment approach developed by the contractor was important but that the best method of validation was not obvious and was left to the researchers. The Columbia University researchers have centered their validation ef- forts on the development of a comprehensive database to which quality control and internal consistency checks have been applied. Those efforts are described in the progress reports submitted to the committee and are addressed in published papers (Stellman et al., 2003a; Stellman et al., 2003b). The contractor's motivation for taking this approach is that it repre- sents the best means available given the scientific and practical limita- tions of seemingly more direct validation techniques like biomarker stud- 3 The E4 EOI is quantitative on a ratio scale: an EOI of 1,000 represents twice as much exposure as an EOI of 500.

OCR for page 5
8 EXPOSURE OF VETERANS TO AGENT ORANGE AND OTHER HERBICIDES ies or environmental measurements.4 Since the methodology they devel- oped bases exposure opportunity scores on the proximity of herbicide releases in space and time, the usefulness of the model depends in part on the precision and completeness of the underlying data. In brief, their validation work includes Comparing and combining disparate versions of the existing her- bicide spraying databases to make a single corrected composite record and supplementing it with new mission data from National Archives and other historical records. Those data include leaks, crashes, dumps, and other non-mission-related herbicide releases. Information has also been gathered on bases of operation, airfields, storage depots, and the like to facilitate factoring in of possible undocumented releases. Systematically identifying and reviewing possible typographic and reporting errors in the databases and correcting them as appropriate by using primary data and the expert judgment of a consensus panel. Creating a database of location histories for military and combat support units by using data from the US Armed Services Center for Re- search of Unit Records and other documentary sources. Combining the more fragmentary records of mobile combat units with expert judgment to develop a way to estimate the location of these troops. Evaluating the robustness of the E4 EOI by applying various as- sumptions to the calculation in a sensitivity analysis. Because the GIS is based on geographic partitions (grids), different breakpoints been have applied to confirm that small shifts in grid parameters do not result in large shifts in estimated exposures. Using double-keying and other accepted data-management tech- niques to minimize the possibility of introducing typographic errors in the entry of new information. Developing a quality assurance plan to test all software compo- nents of the database and exposure assessment calculation. This plan will be implemented when the model is finalized. This work provides validation of the inputs to the exposure assess- ment model. Limited qualitative and quantitative validation information for the model itself is also available. The Columbia University researchers have noted, in presentations to the committee, a correspondence between their database's records of spraying and an independent set of measure- 4 The limitations involved in more-widespread application of such techniques are dis- cussed below.

OCR for page 5
PREPUBLICATION COPYUNCORRECTED PROOF ~ jgO~ \\ ~ ~ ~ - Hi r ~ ~ ' ~ INS be, ~ }A ~ J it' ~ ---I ~ I, ~.- j,/ \ \,,,,) ~ __0 i' ~ ~ _': ,.':'?, ~ ~ .i'~ ?, ,;,^-^CS<.,~3p,~=S~ i. ~ ~ ~ >.:; ,. _~esi:~ ~~ ~ ,. ease "` ~ ~ ., ~ '. 'an ~ .t _- ~ :' ~ ~~. aims I;::::: _ of; :~ ~~ ~ . - '1 i: i. ~ . . ~ _ ;= ~ rnerous La-L'~n-s ~+' ... ~5 ,~ Hi, _ ~m ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Ye,,' Figure I. Two Agent Orange herbicide missions (numbers 383 and 384, shown as orange lines) flown January I, 1966, in the Rung Sat Special Zone showing all grid points whose centers fell within ~ km (large blue dots) and 5 km (small black dots) of the flight paths. +: elevation point; hi: local airfield. Source: contractor presentation made at 42-month update meeting, January ~ 8, 2002, Washington, DC. Characterizing Exposure Interim Findings and Recommendations Page 9 of 20

OCR for page 5
10 EXPOSURE OF VETERANS TO AGENT ORANGE AND OTHER HERBICIDES FIGURE 2. Choropleth plot of Vietnam for June 1969 showing exposure potential calculated with the E4 EOI model. Source: adapted from contractor presentation made at 48-month update meeting, July 24, 2002, Washington, DC.

OCR for page 5
INTERIM FINDINGS AND RECOMMENDATIONS 11 meets made in Vietnam. Dwernychuk et al. (2002) found above-back- ground soil TCDD concentrations in an area of the Aluoi Valley that the Columbia University researchers' database indicates was sprayed with Agent Purple an early formulation believed to have been highly con- taminated with dioxin. Nearby areas that their database indicates were sprayed later with a herbicide formulation thought to have relatively lower dioxin contamination were at background concentrations. The Co- lumbia University researchers also note a 1994 paper (Verger et al.) that found a correlation between dioxin levels in adipose tissue samples of Vietnam residents and an early formulation of their EOI. The paper re- ported a Pearson correlation coefficient of 0.50 between the log of serum dioxin and the log of the EOI (p = 0.02) and a Spearman correlation coef- ficient of 0.44 for the 22 above-background measurements (p = 0.04~. When all 27 measurements were factored in, the Pearson correlation coefficient was 0.36 (p = 0.07) and the Spearman correlation coefficient 0.32 (p =0.10~. Overall, the committee believes that the Columbia University re- searchers' work provides an adequate demonstration that the model rep- resents a valid means of assessing wartime herbicide exposure of Viet- nam veterans. The Columbia University researchers have chosen not to conduct a formal analysis of their model with serum dioxin concentrations,5 other biomarkers, or environmental sampling before submitting it for commit- tee review. The committee supports that decision. While such tests have information value, they are not compulsory for determining whether the model is feasible. The use of biomarkers to validate an exposure assessment model pre- sents both methodologic and practical challenges. The primary limitation in using present-day serum dioxin concentrations in Vietnam veterans is that too much time has passed since exposure to make them a generally applicable, reliable indicator of exposure. Serum dioxin levels decline over time in the absence of new exposures. For Vietnam veterans, a con- servative estimate would predict a decline to less than 10 percent of war- time concentrations.6 Thus, although the presence of an above-back- 5 Specifically, serum concentrations of 2,3,7,8-tetracholorodibenzo-p-dioxin the dioxin that contaminated the wartime formulations of some herbicides used in Vietnam. Measure- ments of other dioxin congeners can provide useful input to epidemiologic studies (Kang et al., 2001), but their presence in serum does not indicate wartime exposure. 6 Assuming a simple, one-compartment model, a mean TCDD half-life of 7.5 years (Michalek et al., 2002), and an exposure ending on the last day of US wartime involvement in Vietnam (April 20, 1975), a veteran's serum concentration in April 2003 would be 7.5 percent of that when he left the country. If he left 5 years earlier, the figure would be 4.7 percent.

OCR for page 5
12 EXPOSURE OF VETERANS TO AGENT ORANGE AND OTHER HERBICIDES ground concentration in a veteran who has no other known dioxin expo- sures would strongly suggest wartime exposure, the absence of an above- background concentration does not necessarily indicate that the veteran was not exposed during Vietnam service. Lack of knowledge about variations in elimination rates poses addi- tional challenges. The pharmacokinetics of dioxin in humans is complex. Current research suggests that the elimination rate depends on several factors, including initial dose, time since exposure, body weight, and sex (IOM, 2003~. Therefore, individuals with identical exposures and body burdens during service might not have equal serum concentrations today. The use of present-day serum dioxin concentrations of Vietnam resi- dents is even more problematic. Vietnam residents experienced different wartime exposures from foreign combatants. Notably, their sole source of food was likely to have been local, and contamination of the food chain is a major source of dioxin exposure (Lorber, 2001~. Residents of the south may have also experienced postwar exposure through continued contact with a contaminated environment (Schecter et al., 2001~. Dioxin sources other than wartime spraying may be present in various parts of the coun- try, further complicating analysis. Serum measurements taken in Viet- nam might provide useful information, but caution would be necessary in evaluating the extent to which this information would validate a model of wartime exposure of veterans.7 Examination of dioxin concentrations in soil or other ecologic measures might also serve as a check on the accu- racy and completeness of spraying records and unintentional releases of some herbicides. However, dioxin fate and transport would need to be accounted for, and the interpretation of current concentrations is not straightforward. Finally, it must be remembered that not all herbicides used in Viet- nam were contaminated with dioxin and that the ones that were had varied and largely uncharacterized levels of contamination. TCDD is a contaminant of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T), a constituent of Agents Orange, Orange II, Purple, Pink, and Green. Agents Whites and Blue, which together account for about one-third of the gallonage sprayed, were composed of other herbicides.9 The TCDD concentration in stocks 7 The databases and model created by the Columbia University researchers could be adapted for use in studies of Vietnam residents and may have great utility in this regard. 8 The National Toxicology Program profile of 2,3,7,8-TCDD noted that it "may be present" in 2,4-D (NTP, 2002), an herbicide used in Agent White. The extent to which dioxin mea- surements may indicate Agent White exposure is therefore not clear. 9 Information on herbicide formulations and gallonage sprayed was derived from IOM, 1994 (Table 3-4, p. 89).