PETITIONED PUBLIC HEALTH ASSESSMENT

BURLINGTON NORTHERN LIVINGSTON COMPLEX
(a/k/a BURLINGTON NORTHERN RAIL YARD)
LIVINGSTON, PARK COUNTY, MONTANA

APPENDIX A: FIGURES

Figure 1. Burlington Northern Rail Yard Site Location Map

Figure 2. Location of private and city wells

Figure 3. Mission Wye area

Figure 4. River sediment sampling locations

Figure 5. Ambient air monitoring sites

Figure 6. Burlington Northern Rail Yard

APPENDIX B: DATA TABLES

* Conversion factor for air:

* Conversion factor for air:

* Conversion factor for air:

a. One micron is equal to one-millionth of a meter or 0.000001 meter.

b. Contour plots are lines that run through sample collection points that show similar concentrations.

c. All values from reference (16).

d. ND = not detected.

e. All values from reference (16).

f. BDL = below detection limit.

g. All values from reference (16).

h. Detected at the same concentration in duplicate samples.

i. Not analyzed.

j. Below Detection Limit.

k. All values from reference (16).

l. All values from reference (16).

m. All values from reference (16).

APPENDIX C - COMPARISON VALUES

Comparison Values

ATSDR comparison values are media-specific concentrations that are considered tobe safe under default conditions of exposure. They are used as screening values inthe preliminary identification of site-specific "contaminants of concern". The latterterm may be misinterpreted as an implication of "hazard". As ATSDR interprets thephrase, a "contaminant of concern" is merely a site-specific chemical substancethat the health assessor has selected for further evaluation of potential healtheffects. Generally, a chemical is selected as a contaminant of concern because itsmaximum concentration in air, water, or soil at the site exceeds one of ATSDR'scomparison values.

However, it must be emphasized that comparison values are not thresholds oftoxicity. While concentrations at or below the relevant comparison value mayreasonably be considered safe, it does not automatically follow that anyenvironmental concentration that exceeds a comparison value would be expected toproduce adverse health effects. The purpose behind highly conservative, health-based standards and guidelines is to enable health professionals to recognize andresolve potential public health problems before they become actual health hazards. The probability that adverse health outcomes will actually occur depends on site-specific conditions and individual lifestyle and genetic factors that affect the routeand duration of actual exposure, and not on environmental concentrations alone.

Listed and described below are the various comparison values that ATSDR uses to select chemicals for further evaluation, along with the abbreviations for the most common units of measure.

Cancer Risk Evaluation Guides (CREGs) are estimated contaminant concentrationsexpected to cause no more than one excess cancer in a million persons exposedover a lifetime. CREGs are calculated from EPA's cancer slope factors.

Minimal Risk Levels (MRL) are estimates of daily human exposure to a chemical(i.e., doses expressed in mg/kg/day) that are unlikely to be associated with anyappreciable risk of deleterious noncancer effects over a specified duration ofexposure. MRLs are calculated using data from human and animal studies and arereported for acute (< 14 days), intermediate (15-364 days), and chronic (> 365days) exposures. MRLs are published in ATSDR Toxicological Profiles for specificchemicals.

Environmental Media Evaluation Guides (EMEGs) are concentrations that arecalculated from ATSDR minimal risk levels by factoring in default body weights andingestion rates.

Intermediate Environmental Media Evaluation Guides (IEMEG) are calculated fromATSDR minimal risk levels; they factor in body weight and ingestion rates forintermediate exposures (i.e., those occurring for more than 14 days and less than 1year).

Reference Dose Media Evaluation Guide (RMEG) is the concentration of acontaminant in air, water or soil that corresponds to EPA's RfD for thatcontaminant when default values for body weight and intake rates are taken intoaccount.

EPA's Reference Dose (RfD) is an estimate of the daily exposure to a contaminantunlikely to cause noncarcinogenic adverse health effects. Like ATSDR's MRL,EPA's RfD is a dose expressed in mg/kg/day.

Reference Dose Concentrations (RfD-C) is a concentration derived from an EPAReference Dose with assumed body and ingestion rates factored into thecalculation.

Environmental Protection Agency Region III (EPA III) values are risk-basedconcentrations which take into account factors such as body weight, toxicity, andexposure duration and frequency for non-carcinogens and carcinogens, whenapplicable.

Drinking Water Equivalent Levels (DWEL) are based on EPA's oral RfD andrepresent corresponding concentrations of a substance in drinking water that areestimated to have negligible deleterious effects in humans at an intake rate of 2L/day, assuming that drinking water is the sole source of exposure.

Lifetime Health Advisories (LTHA) are calculated from the DWEL and represents theconcentration of a substance in drinking water estimated to have negligibledeleterious effects in humans over a lifetime of 70 years, assuming 2 L/day waterconsumption for a 70-kg adult, and taking into account other sources of exposure. In the absence of chemical-specific data, the assumed fraction of total intake fromdrinking water is 20%. Lifetime health advisories are not derived for compoundswhich are potentially carcinogenic for humans.

Maximum Contaminant Levels (MCLs) represent contaminant concentrations indrinking water that EPA deems protective of public health (considering theavailability and economics of water treatment technology) over a lifetime (70 years)at an exposure rate of 2 liters of water per day (for an adult).

Permissible Remediation Goal (PRG) are levels set by EPA under Superfund thattrigger a response or action when the contaminant concentration exceeds thisvalue. Also generically known as action levels.

Occupational Safety and Health Administration's Permissible Exposure Limit (PEL)for air is an 8-hour, time-weighted average developed for the workplace. The levelof exposure may be exceeded (for brief periods), but the sum of the exposure levelsaveraged over 8 hours must not exceed the limit.

Threshold Limit Value (TLV), according to the American Conference ofGovernmental Industrial Hygienists (ACGIH), is "the time-weighted averageconcentrations for a normal 8-hour workday and a 40-hour workweek, to whichnearly all workers may be repeatedly exposed, day after day, without adverseeffect". Many of ACGIH's TLVs were adopted by OSHA for use as PELs. TLVsand PELs, which were designed to protect healthy workers, are usually much higherthan the health-based values of ATSDR and EPA, which were designed to protectthe health of the general population, including the very young and the elderly. Although the ATSDR does not base any of its community health decisions on TLVsor PELs, agency health assessors and toxicologists may sometimes mention suchvalues in Public Health Assessments or consultations as a means of putting site-specific concentrations of contaminants into some kind of meaningful perspective for the reader.

References

Agency for Toxic Substances and Disease Registry. Health Assessment Guidance Manual. Atlanta: ATSDR, March, 1992.

APPENDIX D - PCE

PCE Carcinogenicity

PCE is a nongenotoxic animal carcinogen. In chronic bioassays (1.5-2.0 yrs), massive doses of PCE administered orally (up to 1,072 mg/kg/day) or by inhalation (100-200 ppm), have produced liver cancer in mice, but not in rats; administered by inhalation (200-400 ppm), it has also caused a statistically insignificant increase in kidney tumors in male, but not female rats (22,25,33,34). However, recent re-evaluations of these studies by various government agencies and independent scientists indicate that the tumors observed in animals were probably due to species-specific mechanisms that exhibit thresholds at near-toxic levels (33,34,36,37). That is to say that the induction of cancers in mice and rats by PCE required doses in excess of anything humans might reasonably be expected to encounter, and involved certain elements of rodent biology that are not likely to be shared by humans (peroxisome proliferation, alpha-2µ-globulin accumulation, glutathione-PCE conjugate formation). The implication is that the cancers observed in laboratory animals at very high doses of PCE have little or no relevance for human risk evaluation at environmental levels of exposure that are orders of magnitude lower. In fact, a number of epidemiological studies of men and women exposed occupationally to PCE have not identified an increased risk of cancer (27).

The International Agency for Research in Cancer (IARC) classifies PCE as "possiblycarcinogenic to humans" based on "sufficient" evidence of carcinogenicity inanimals and "inadequate" evidence of carcinogenicity in humans, and the NationalToxicology Program (NTP) classifies PCE as Reasonably Anticipated to be aCarcinogen (RAC) in humans (27). However, both IARC and NTP use a "strengthof evidence" basis of classification which, in contrast to the "weight of evidence"scheme used by the EPA, largely ignores negative studies and mechanisms ofaction. EPA's carcinogen classification scheme was developed at a time when littleor no data on mechanism of action were available for consideration, with the resultthat the carcinogen category that would best accommodate such data often doesnot exist. This is the case with PCE.

EPA currently classifies PCE as a B2-C carcinogen, which is to say, somewherebetween group B2 and group C (27). This is an unorthodox classification currentlyshared by only two other chemicals, trichloroethylene and styrene. Group B2 isreserved for chemicals defined as "probable" human carcinogens based on"sufficient" animal evidence. Group C is reserved for chemicals defined as"possible" human carcinogens based on "limited" animal evidence. The only otherchoices available (i.e., Groups A, B1, D, and E) require either reliable evidence ofcarcinogenicity in humans, which does not exist for PCE, or the absence of suchevidence in animals, which does exist for PCE. However, there is no question that,at high enough doses and administered by the right route to the right species andsex, PCE can cause an elevated incidence of certain cancers by species-specificmechanisms in laboratory animals. Thus, EPA classified PCE as a B2-C carcinogennot because it could not decide whether the evidence for carcinogenicity in animalswas "sufficient" or "limited", but rather because it's classification scheme does notinclude a more appropriate category for this type of carcinogen. The AmericanConference of Governmental Industrial Hygienists (ACGIH), which does have such acategory, recently reclassified PCE as an A3 animal carcinogen, signifying that "theagent is carcinogenic in animals at a relatively high dose, by route(s) ofadministration, at site(s), of histological type(s), or by mechanism(s) that are notconsidered relevant to worker exposure" (31).

It is important to realize that formal cancer risk assessment has its limitations (23-25). Based on a withdrawn EPA risk assessment that is currently being re-evaluated (32), it has been estimated that the drinking water standard of 5 µgPCE/L corresponds to a theoretical lifetime cancer risk of 14 in a million. However,as is the case with all numerical estimates of low-dose cancer risk based on animaldata, the true risk is unknown, and may be as low as zero (35). It is, in fact, apractical impossibility to determine experimentally the carcinogenicity of chemicalsat very low doses, due to the enormous number of animals that would be requiredto achieve statistically reliable results. Instead, the assessment of cancer risk atlow doses has traditionally involved linear extrapolation from the results of high-dose animal experiments with the aid of statistical models that incorporate highlyconservative default assumptions. These include the assumptions that (1)carcinogens have no threshold, and (2) all animal carcinogens are likely to behuman carcinogens. These assumptions are clearly defensible as prudent publichealth policy when used to provide a maximal margin of safety in the face ofoverwhelming uncertainties. However, they were never intended to take priorityover new scientific data that could otherwise reduce those uncertainties. Nor werethey ever intended to be used for the prediction of actual disease incidence.

The ATSDR recognizes that no single mathematical model is appropriate in all casesand that existing mathematical models for low-dose extrapolation may not beappropriate for nongenotoxic agents. Information on the biological mechanism isneeded to determine if there are threshold exposure levels for nongenotoxic agents. The ATSDR, therefore, evaluates the relevance of the animal data to humans on acase-by-case basis. In the case of PCE, the available animal cancer data appear toreflect species-specific mechanisms that exhibit thresholds at near-toxic levels and,therefore, have little or no relevance for the evaluation of human risk atenvironmental levels of exposure. However, in the absence of compelling data tothe contrary, the ATSDR policy considers it prudent to presume that a substancewhich causes cancer in animals may also pose a potential carcinogenic risk tohumans, and that exposure to the substance should be minimized (30). (For aperspective on cancer risk assessment, see reference 41.)

Next Section          Table of Contents