FRAMEWORK FOR ASSESSING POSSIBLE RISKS … FOR ASSESSING POSSIBLE RISKS ... and other asbestiform...

FRAMEWORK FOR ASSESSING POSSIBLE RISKS POSED BY THE PRESENCE OFASBESTOS AND RELATED STRUCTURES FROM EMISSIONS FROM

SOUTHDOWN QUARRY

8/9/00

1. EXECUTIVE SUMMARYThe New Jersey Department of Environmental Protection and Environmental Protection Agencyrequested that their staff assemble a team of experts to make recommendations regarding the bestmechanism to determine whether the presence of asbestos in marble mined at the Southdown Quarryin Sussex County, NJ poses a risk. This report presents a framework for evaluating conditions at theSouthdown Quarry to address the health-related concerns that have been raised relative to thepossible presence of asbestos. The framework is intended to define the questions that need to beanswered to resolve these concerns, and to indicate the evaluation process that will be followed foranswering the questions. The kinds of data (and the quality of the data) needed to complete theproposed evaluation are also identified along with potential tasks for developing the required data.

This framework presents a general strategy for the determination of risk which employs two differentapproaches in parallel to estimate the risk from asbestos and related structures - the "protocolstructure" approach, and the PCM-based approach. The protocol structure approach includes aworking definition of "protocol structures," which defines an exposure index for measuring asbestosthat captures characteristics which determine biological activity. Protocol structure-basedassessments to be conducted at the site will be based on measurements of protocol structureconcentrations and an appropriately matched set of dose-response factors. PCM-based assessmentswill be based on the older approach which relates risk to phase-contrast (PCM) light microscopedetectable fibers. Based on these two approaches and the reported situation at Southdown quarry,the following questions are identified for addressing the situation.

Question 1 Are current or future emissions of protocol structures and other relevant structuresfrom Southdown quarry sufficient to pose an elevated risk to downwind residents?

Question 2 Is re-entrainment of particle deposits in household dust or outdoor soil from pastemissions at Southdown quarry sufficient to release protocol structures and otherrelevant structures at a rate that poses an elevated risk to downwind residents?

Question 3 If residents are being unduly exposed to protocol structures and other relevantstructures from Southdown quarry, what actions may be necessary to adequatelycontrol/eliminate such exposure?

The framework includes a detailed evaluation of the types of investigation that might be conductedto address these questions, and recommends a 2-phase approach. The first phase, which can beconducted relatively quickly and at reduced expense, is designed as a screening study, to determinewhether exposure to protocol structures and other relevant structures being emitted from SouthdownQuarry operations pose a risk which is either clearly above or clearly below a pre-determinedthreshold for action. The results will be provided to DEP and EPA to determine what, if any, actionis necessary and whether a second phase is warranted.

INTRODUCTION

The New Jersey Department of Environmental Protection and Environmental Protection Agencyrequested that their staff assemble a team of experts to make recommendations regarding the bestmechanism to determine whether the presence of asbestos in marble mined at the Southdown Quarryin Sussex County, NJ poses a risk. Operations at the Southdown Quarry have attracted recentattention due to (1) the observed presence of tremolite in the marble that is mined at the quarry, (2)a report that tremolite asbestos structures were detected on an air conditioner filter at a residence thatis located downwind of the quarry, and (3) the accepted premise that inhalation of tremolite asbestosfibers can lead to cancer in humans.

It is important to understand that natural release of asbestos from geologic materials is pervasive. Measurable concentrations of airborne asbestos structures have been observed throughout the UnitedStates and the rest of the world. Therefore, it will be necessary to distinguish between anyenvironmental asbestos concentrations that are potentially attributable to Southdown quarry andconcentrations that are related to local ambient conditions.

This report presents a framework for evaluating conditions at the Southdown Quarry to address thehealth-related concerns that have been raised relative to the possible presence of asbestos. Theframework is intended to define the questions that need to be answered to resolve these concerns,and to indicate the evaluation process that will be followed for answering the questions. The kindsof data (and the quality of the data) needed to complete the proposed evaluation are also identifiedalong with potential tasks for developing the required data. Such tasks should help to focus aproposed site investigation (to be defined in the accompanying Workplan document), allowing datato be collected quickly and efficiently. Requirements for designing an efficient investigation are alsodiscussed.

2.1 Overview of Proposed Framework

This proposed framework is based on the well-established principles of risk assessment. Theseprinciples are applied broadly in our society for identifying and managing potential hazards. Thesteps of a risk assessment are depicted in Figure 2-1.

As indicated in Figure 2-1, the first step of a risk assessment is to break down the problem into oneor more questions to be answered. In this case, the problem is that tremolite asbestos that may bereleased from Southdown Quarry may pose elevated health risks in the nearby community. Beforethe appropriate assessment questions can be posed, however, it is necessary to clarify the types andforms of tremolite, and other asbestiform materials that are associated with potential health effects,and the terminology that will be used in this framework. This is necessary because both the popularand technical literature contain conflicting and ambiguous definitions of what constitutes dangerousforms of tremolite or asbestos. In this study, two parallel definitions of the types of asbestos andasbestiform materials which are associated with health effects will be employed. The first ispresented in detail in Appendix A to this Framework, and is summarized as follows:

All fibers and cleavage fragments derived from one of the six mineral types (serpentine andamphibole) which fall within the dimensional range that is defined as biologically activeunder this approach (i.e., longer than 5 μm, and narrower than 0.5 μm) will be referred tocollectively in this framework as “protocol structures.” and will be considered under thisdefinition to contribute to health effects as a function of their length (see Appendix A fordetails ).

The second definition is based on phase-contrast light microscope (PCM) detectable fibers and isdiscussed in the U.S.EPA’s Airborne Asbestos Health Assessment Update (U.S.EPA, 1986), and canbe summarized as follows:

All fibers which are longer than 5 μm in length, and wider than 0.25 μm in width which havean aspect ratio (i.e., length/width) greater than 3 will be considered to contribute equally tocancer potency. Although this definition is intended to apply specifically to chrysotile andamphibole fibers only, the PCM analytical method is not capable of distinguishing betweenasbestos and non-asbestos fiber types, and therefore, in practice, this definition applies to allfibers.

Based on these definitions, and the reported situation at Southdown quarry, the following questionsneed to be answered to address this problem.

Question 1 Are current or future emissions of protocol structures and other relevant structuresfrom Southdown quarry sufficient to pose an elevated risk to downwind residents?

Question 2 Is re-entrainment of particle deposits in household dust or outdoor soil from pastemissions at Southdown quarry sufficient to release protocol structures and otherrelevant structures at a rate that poses an elevated risk to downwind residents?

Question 3 If residents are being unduly exposed to protocol structures and other relevantstructures from Southdown quarry, what actions may be necessary to adequatelycontrol/eliminate such exposure?

These three questions form the basis for the rest of this framework and for any ensuing investigation.

If it is ultimately found that residents are being exposed to protocol structures and other relevantstructures, but that no link with the quarry can be established, it may be necessary to consider otherpotential sources of asbestos. However, at this point the emission of asbestos from SouthdownQuarry is the primary focus of this assessment..

The second step of a risk assessment is to develop a conceptual model that describes the nature ofthe conditions and activities that need to be evaluated to answer the questions of interest. Typically,this is documented by identifying a set of exposure pathways that indicate:

(A) the sources (e.g. soils or rock) of hazardous materials of concern (in this case,asbestos);

(B) the mechanisms by which hazardous materials protocol structures and other relevantstructures may be released from each source (e.g. excavation, crushing loading, etc.);

(C) the mechanisms by which protocol structures and other relevant structures might betransported from sources to locations where exposure might occur (e.g. windtransport);

(D) the characteristics of the population(s), commonly called receptors, that mightbecome exposed; and

(E) the route of exposure (i.e. the mechanism by which a hazardous material is taken upinto the body from the environment). In the case of protocol structures and otherrelevant structures, the primary route of exposure of concern is inhalation.

Moderately elevated levels of asbestos have been reported in surface water which appears to containrunoff from the quarry. There is currently no reliable scientific evidence which clearly linksingestion of asbestos (or asbestos-like structures) to disease. Therefore, this study focuses oninhalation as the potentially important route of exposure.

At Southdown quarry, the source of the asbestos is the marble which is actively quarried. Tremolite,the parent mineral form from which tremolite asbestos may be derived, may occur as discrete veinsin the marble and may also be dispersed diffusely throughout the marble.

At Southdown quarry, emissions of protocol structures and other relevant structures potentially occurfrom multiple sources of varied types that are distributed spatially over an area of approximately 300acres (approximately 1/2 square mile). The kinds of activities that potentially contribute to suchemissions may include (but are not necessarily limited to):

• blasting;• drilling;• excavating;• hauling;• dumping;• crushing and milling;• sorting (classifying); and• wind entrainment from uncovered storage piles.

Importantly, each of the above-listed activities occur in different areas of the site. This means thatthe quarry is essentially an area, or line emissions source rather than a point source or group ofdiscrete point sources such as stacks. The potential exposure of nearby residents cannot, therefore,be comprehensively estimated from data on any of the individual sources. An illustration of the

relative locations at which several of the major activities are conducted is presented in Figure 2-2. The annotations in Figure 2-2 were developed based on a brief tour of the facility that wascompleted in November, 1999.

It is the potential releases of protocol structures and other relevant structures associated with eachof the above-listed activities, individually or in combination, along with the characteristics of localwinds (e.g. speed and direction), and the distance and direction to neighboring residents (termed“downwind receptors”) that define the specific exposure pathways that need to be evaluated toaddress the above-listed questions. Based on such considerations, a detailed description of theexposure pathways of concern for Southdown is presented in Table 2-1.

Because important contributions to exposure may occur within residences in association with thissite, the complex dynamics of retention and transport of protocol structures and other relevantstructures within a residence also need to be incorporated into the exposure pathways that will beused to characterize this site. This is represented in Table 2-1 by components of pathways involvinginfiltration from outdoor air, track-in from outdoor soils, and re-suspension of indoor dusts.

As indicated in Figure 2-1, the next two steps of a risk assessment are typically performedsimultaneously. The left side of the figure indicates that the toxicity of the hazardous material(asbestos - in the form of protocol structures and other relevant structures) is evaluated, primarilyby reviewing and evaluating literature studies that indicate the potency of the material. The goalsare to identify relevant health effects and to establish a mathematical representation of therelationship between dose (exposure) and response (induction of disease). Both the form(mathematical shape) of the relationship between dose and response (e.g. linear, a power function,etc.) and some estimate of the potency of the hazardous substance (i.e. a dose-response factor) needto be developed.

Because the toxicity of asbestos dust is a strong function of the mineralogy and the size and shapeof the fibrous structures within the dust, an appropriate “index of exposure” must be defined in orderto specify the form in which asbestos measurements should be reported in order to support a riskassessment. An index of exposure is a specified set of structure sizes, shapes, and mineralogy thatmust be quantified and combined to determine the concentration of hazardous structures. Twoindices of exposure will be used in parallel in this study. One index of exposure is the “protocolstructure” approach as set forth in equation A-1 of the Framework. Structures meeting this definitionare referred to as “protocol structures”. For protocol structures, the toxicology models and dose-response factors that are recommended for use are presented in Appendix A and in Section 3. Thesewere developed based on a detailed review of the asbestos literature and a series of recent,supplemental studies, which are described in a recent report (see Berman and Crump 1999a). Theother index of exposure is the phase contrast light microscopy (PCM)-based index as set forth in theU.S.EPA’s IRIS database file for asbestos (http://www.epa.gov/ngispgm3/iris/subst/0371.htm), andin the U.S.EPA’s 1986 Airborne Asbestos Health Assessment Update. The cancer dose-response(potency) estimates based on these two approaches will be used in parallel.

Framework Figure 2-1

FIGURE 2-2

TABLE 2-1: CONCEPTUAL MODEL - POTENTIAL HUMAN EXPOSURE PATHWAYSRELEVANT TO THE SOUTHDOWN QUARRY, NEW JERSEY

Source ReleaseMechanism

TransportMedium

TransportMechanism

Location(s) ofExposure

Route ofExposure

ExposedPopulation

Active face ofQuarry

EmissionsduringBlasting,Excavation,Loading

Outdoor Air WindDispersion

Outdoor Air inResidentialArea

Inhalation LocalResidents

Settling Local Soils inResidentialArea

(This is a secondary source, SeePage 2)

Infiltration Air InsideLocalResidences

Inhalation HouseOccupants

Settling Surface DustsInsideResidences

(This is a secondary source, SeePage 2)

Haul Roads EmissionsfromVehicularTraffic onRoads andfromTransportedMaterial

Outdoor Air The transport mechanisms, exposure points, routes of exposure, andexposed populations relevant to this source are identical to thoselisted above for emissions from the active face of the quarry.

Crushing andMillingOperations

StackEmissionsfrom Crushingand Milling


ClassificationOperations

StackEmissionsfromClassification


TABLE (PAGE 2): CONCEPTUAL MODEL (CONTINUED)

Source ReleaseMechanism

TransportMedium

TransportMechanism

ExposurePoint

Route ofExposure

Exposed Population

StoragePiles

EmissionsfromLoading andDumping

OutdoorAir

The transport mechanisms, exposure points, routes of exposure, and exposedpopulations relevant to this source are identical to those listed above foremissions from the active face of the quarry.

OutdoorSoils

WindEntrainment

OutdoorAir


OutdoorResidentialActivities

OutdoorAir


Track-in IndoorSurfaceDust

Indoor Surface Dust serves as a secondary source, see below.

IndoorSurfaceDust

Re-suspensionfrom IndoorResidentialActivities

IndoorAir

AdvectionandDispersion

Air InsideLocalResidences

Inhalation House Occupants

As indicated on the right side of Figure 2-1, exposure is characterized next. This means that eachof the potentially important exposure pathways identified in the conceptual model needs to beevaluated. Exposure can be characterized by direct measurement at points of exposure, or bymodeling based on measurements or estimates of emission and transport. Often, a combination ofthese approaches proves to be the most cost-effective for estimating the exposure.

In the last step of a risk assessment - the risk characterization, exposure estimates are combined withdose-response models, along with consideration of uncertainties and overall confidence in the riskestimate. The questions of interest can then be answered, providing input for any appropriate risk-management decisions.

Section 4 of this Framework lays out the investigative process that is proposed for this project,incorporating what is already known about the situation in Sussex County along with the riskassessment principles and appropriate dedication of resources, including time. The process will takepart in two phases. The first phase is an initial screening process to determine expeditiously andefficiently whether the risks associated with the quarry operations are clearly above or below a pre-determined action threshold. The results of the first phase will be provided to management at DEPand EPA management who will then determine what, if any, further action is warranted and whethera phase 2 study will be budgeted or necessary.

2.2 Requirements for Designing an Efficient Site Investigation

Due to the potentially large variability introduced by several factors that affect risk at this site (whicheach must either be characterized or controlled) and the severe limitations that are particularlyassociated with protocol structure measurement1, a variety of tradeoffs must be considered. This isto assure that valid conclusions can be drawn from the results of the site investigation whilesimultaneously assuring that the investigation remains cost-effective. Generally, the more samplescollected, the higher the confidence in the results, but if the samples are grouped properly, greaterconfidence can be achieved than from less careful sample planning.

Typically, conservative (i.e. intentionally biased in a health-protective sense) studies are performedin support of a risk assessment because such studies are generally less expensive than unbiasedstudies. One specific type of conservative study is a “worst case” study. Such studies apply

1 Given the constraints on sampling rates and the volumes of air required to keep analytical costs

reasonable, the minimum time required to collect a protocol structure air sample will likely runbetween 3 and 6 hours. Meteorological conditions will not frequently remain stable over suchintervals of time. Moreover, there may be a need to conduct a pre-study (especially in dusty areas)to gauge the maximum sampling time acceptable to assure that samples do not become too loadedto allow preparation using a direct transfer procedure; samples prepared using indirect transferhave been shown not to relate quantitatively to risk, although they are typically considered tosupport conservative (in a health protective sense) estimates of risk.

assumptions which are not necessarily likely, but which reflect the conditions and assumptions whichwould be the most adverse from a public health standpoint. Similarly, "reasonable worst case"studies seek to reflect conditions and assumptions which while not necessarily the most likely, arenonetheless, plausible. The risk predictions derived from such studies are intended to place an upperbound on the estimate of the true risk. Such conservative “worst case” studies are useful andappropriate as long as the general “laws of inequalities” are not violated when drawing conclusionsfrom such studies. Thus, for example:

• it is valid to use a negative result from a “worst case” study to conclude that a problem does notexist, since such a result is intended to reflect an upper bound on the true risk, and

• it is valid to use a positive result from a “best case” (minimized) case study to conclude that aproblem does exist.

Other combinations of results and conclusions from these types of biased (also called “judgmental”)studies are not generally valid.

Unbiased studies generally attempt to define the true level of risk, while biased studies generallyattempt to bound the risk in one direction or another. Unbiased studies generally need to be moresophisticated (and expensive and time-consuming) to provide adequate precision to support requireddecisions with adequate confidence. In unbiased studies, formal statistical procedures are typicallyemployed to control for uncertainty (error). It is general (and prudent) practice to “err on the sideof reasonable caution” to protect public health. Therefore, the statistical tests employed in suchstudies are typically set up to keep the chance of falsely missing a potential health threat to a pre-defined (acceptably small) level independent of sample size. Under such conditions, the chance offalsely reaching the opposite (positive) conclusion (i.e., concluding that a threat does indeed exist,when it truly does not) becomes a function of the quantity of data collected, relating, in turn, to thecost of the investigation .

By carefully considering the capabilities and limitations of the various tasks for evaluating each ofthe components of the exposure pathways of interest, an integrated approach can be developed forthe proposed investigation that should be optimally cost-effective. Such considerations areaddressed in Section 4.

Costs for the proposed investigation of this site can also be controlled by “getting the biggest bangfor each buck.” Among other things, this means that:

• all component measurement or modeling efforts should be designed to ensure that usefulinformation will be obtained whether the results of the effort are positive or negative;

• cost-savings can be realized by designing a study that simultaneously addresses all of the threequestions identified above to exploit any overlap in data needs for addressing these questions;and

• a well-designed study can be phased to allow the canceling of later stages if definitive results areobtained early.

3. ASSESS THE TOXICOLOGY OF ASBESTOS

Asbestos poses a threat to human health when dust is released and the resulting structures areinhaled. Whether asbestos also poses a threat to human health when it is ingested is less clear. There is no direct human evidence that ingested asbestos causes disease and the existing animalstudies are equivocal (U.S. EPA 1986). Therefore, the focus of this discussion is on health effectsresulting from inhalation exposure.

Importantly, there is strong evidence and general agreement within the scientific community that theincidence of disease (i.e. response) is a strong function of dose and that, in addition to concentration,both the mineralogy and sizes and shapes of the inhaled fibers determine the magnitude of theresponse to each particular exposure (see, for example, Berman and Crump 1999a or Lippmann2000).

3.1 Adapting Asbestos Dose-Response Models for Use at Southdown

As stated previously, this study will investigate, in parallel, two different definitions of carcinogenicasbestos and asbestiform structures, each of which is associated with a different cancer potency (orset of cancer potencies). The application of appropriately calculated exposure estimates to the dose-response ( cancer potencies) estimates associated with each definition, will yield estimates of thecancer risk associated with those exposures.

Measured or estimated exposure concentrations used to assess risk must be specifically matched tothe same exposure index to which the models and dose-response factors of the protocol have beennormalized. Thus, in the protocol structure approach, measured or estimated exposureconcentrations must represent concentrations of protocol structures. The structures that are longerthan 10 μm must also be distinguished from those between 5 and 10 μm in length because they areto be weighted more heavily than the shorter structures. The precise definition for protocolstructures (including the weighting scheme) is presented in Equation A-1 of the Appendix. Furthermore, all structures exhibiting the appropriate morphological dimensions are included in thecount, whether such structures are isolated (parent) structures, separately identifiable componentsof more complex clusters or matrices, or structures which did not initially form as fibers, but, whichthrough weathering or human activity, have assumed the appropriate dimensions.

The methods in the protocol recommended for assessing risks at Southdown (which are describedabove) can be used to derive target concentrations for protocol structures in air. Such targets canserve as quick references against which measured or estimated airborne protocol structuresconcentrations can be compared.

Table 3-1 provides an example of such target concentrations. Table 3-1 assumes, as an example,that the threshold for action for either lung cancer or mesothelioma is one-in-one hundred thousand(This table is provided merely as a example of the mathematical approach to such calculations. Itis not the action threshold which will be employed in assessing the results of this study). Thecorresponding maximum airborne concentration of protocol structures is given for a range ofexposure scenarios and a range of receptor populations. For the purposes of assessing exposure andrisk to residents near Southdown, the most appropriate exposure scenario is that presented in columnA, lifetime continuous exposure. The other exposure scenarios (columns B,C and D) are presentedfor comparison purposes. Similar tables could be constructed for other possible action thresholds(e.g., one-in-one-million). As illustrated in Table 3-1, the risk for cancer and mesothelioma variesdepending on whether the exposure is to chrysotile or amphibole. For the PCM-based approach, thecancer potency is assumed to be the same for all fibers within the defined range.

Framework Table 3-1

Air Concentrations of Protocol Structures Longer Than 5 µm Equivalentto One in One Hundred Thousand Risk

ExposureScenario:

Acceptable Concentrations (g/cm3)

5% structures longer than 10 µm 30% structures longer than 10 µm

A B C D A B C D

CHRYSOTILE

MALENON-SMOKERS

Lung Cancer 1.42E-03 1.37E-02 4.16E-01 2.08E+00 2.49E-04 2.39E-03 7.28E-02 3.64E-01

mesothelioma 1.42E-03 7.14E-03 1.17E-01 5.85E-01 2.48E-04 1.25E-03 2.05E-02 1.02E-01

Combined 7.10E-04 4.69E-03 9.13E-02 4.56E-01 1.24E-04 8.21E-04 1.60E-02 7.99E-02

FEMALENON-SMOKERS

Lung Cancer 2.00E-03 1.91E-02 5.83E-01 2.91E+00 3.49E-04 3.35E-03 1.02E-01 5.10E-01



MALESMOKERS

Lung Cancer 1.76E-04 1.49E-03 4.50E-02 2.25E-01 3.08E-05 2.60E-04 7.88E-03 3.94E-02



FEMALESMOKERS




AMPHIBOLE

MALENON-SMOKERS




FEMALENON-SMOKERS




MALESMOKERS

file:///D|/depnet/wwwroot/webster/Area51/root/dep/dsr/sparta/$work/table3-1_insert.htm (1 of 2) [8/23/2000 4:13:12 PM]




FEMALESMOKERS




Exposure Scenarios:

A= Lifetime, Continuous Exposure

B= 30-yr "Occupational Exposure", Beginning at Age 0

C= 1-yr "Occupational Exposure", Beginning at Age 0

D= 60-day "Occupational Exposure", Beginning at Age 0

D. Wayne Berman, Aelous, Inc.

Workplan Timeline

4. RECOMMENDATIONS FOR AN INTEGRATED APPROACH TO THEINVESTIGATION OF SOUTHDOWN

Assessment of risk to residents near Southdown can be evaluated through the use of measurements(both at residential locations and at emissions sources) and computer-based modeling. Theappropriate choice of assessment method(s) can ensure that adequate data are available to addressthe intended scope of this assessment.

Direct measurements (primary data) are those gathered specifically for this study. These includemeasuring the following at various locations at the quarry and in the community:

a. concentration of protocol structures and other relevant structuresb. meteorological condition

Direct measurements are used to gauge the severity of current risks and to calibrate computer modelsfor long term exposure assessment. Direct measurements reflect the conditions at the time ofmeasurement. Therefore, estimates of risk from relatively short -term measurements of (e.g.,outdoor air concentration of protocol structures and other relevant structures) may not capture theentire range of variability which could be observed over longer periods of measurement, but arenonetheless useful for characterizing risk during representative periods of measurement.

Indirect measurements (secondary data) are those that were gathered not as part of this study but areuseful nonetheless. Indirect measurements for this study include previous analyses of tremolitecontent of quarry materials. The reference data will complement the current effort in characterizingthe quarry. Additionally, the information will be used in computer models to project future protocolstructure emission levels.

Computer modeling can be used to quantify the health risk for protocol structure emissionsexpeditiously and efficiently. However, one must always be cautious of its limitations. The area ofinterest (quarry and surrounding upwind and downwind areas) encompasses several square miles. It is not practical to deploy and maintain a network of monitoring stations at all areas of interest fordirect measurements to account for concentration variations due to meteorological condition,protocol structures and other relevant structures concentration, quarry material, and quarry processand production rate. The costs in time and resources can be prohibitive for a study based strictly ondirect measurements. However, computer modeling, when calibrated with adequate directmeasurements, can reduce the time and resources needed for an effective study. If the long termrange and variability of the conditions governing emission and dispersion of protocol structures andother relevant structures is known, then modeling can be used to generate an unbiased (i.e. realistic)estimate of long-term exposure and risk. While realistic data on the type and frequency of thevarious practices at Southdown which govern emission of protocol structures and other relevantstructures may be known or obtainable, long-term local meteorological data are not available. Suchdata are particularly important in this case because many of the quarry operations occur below grade,and because of the irregular topography of the local area. collection of local meteorological dataconcurrent with the study cannot provide a sufficient database to permit modeling of long-term

exposure and risk. The lack of such data limits the extent to which modeling can be used to obtainedrealistic estimates of exposure and risk. Modeling can be used, however, with estimates ofreasonable worst-case parameters of meteorology and emission rates (as necessary). Such anapproach can model an upper bound on the likely risk to the nearby residents, and can identify pointsof maximum exposure and risk.

The expert panel considered a number of possible options for collecting information to address thethree questions. They also investigated the possibility of combining these various options intocoherent study plans. The following list summarizes the various tasks that the Expert Panel selectedas the option most likely to produce results that would answer the questions effectively andefficiently. They are described in detail below. Based on the estimates of risk from this phase of thestudy, and the acceptability of the precision with which those risks can be characterized, the needfor further work will be determined.

Phase 1: Preliminary Investigation(to determine whether the risk associated with the Quarry emissions is either well above or well

below an action threshold)

Task 1:Complete simple ranking of emission sources at the quarry.Based upon existing emission inventories or information from Southdown.

Task 2:complete characterization of protocol structures and other relevant structures concentrationsin Southdown marble.

From existing NJGS core data with re-analysis of cores and/or from further core sampling on site asnecessary (Task 7).

Task 3:conduct a reasonable worst-case assessment of exposureModel-based, using emission estimates from Tasks 1 and 2, and reasonable worst-case assumptions for dispersion parameters.

Task 4:Conduct a study of house dust and soil.Measurement of protocol structure concentration in accumulated household dust andin soil in and around the community.

Task 5 Conduct a study of outdoor and indoor airborne concentrations of protocol structures, andother relevant structures

Short-term (3-6 weeks) sampling in conjunction with known and typical operating activities atSouthdown Quarry, and controlled typical in-home activities.

Task 6.Perform a summary risk characterization.Based on the results of Tasks 1 - 5 to determine whether the results indicate clearly that a elevated riskdoes or does not exist resulting from Quarry emissions.

Phase 2: Quantitative Characterization of Transport and Exposure(to more realistically characterize exposure and risk with a smaller margin of uncertainty) (May not be requested by DEP and EPA).

Task 7:Detailed horizontal and longitudinal sampling and characterization of Southdown marbleWith detailed measurements, providing more definitive results than Task 2, if necessary.

Task 8:Perform realistic modeling of the transport and dispersion of protocol structures and otherrelevant structures

With refined emissions data and site-specific meteorologic data.

Task 9:Conduct long-term indoor/outdoor air sampling at selected residences.Approximately 6 months of sampling to include seasonal and quarry production variability.

Task 10: Perform detailed risk characterization.Based on the results of Tasks 5, 6, and 7 to determine risk levels associated with Quarry operationsand to serve as a guide to further Agency actions.

The following task options for completing an investigation of Southdown Quarry are derived basedon the above-described evaluation of issues. These task options are not mutually exclusive, and nomore interdependent than is indicated. Each is designed to provide information that would bevaluable for determining any effects of quarry emissions. Thus, some can be pursued, and othersand can be combined in various ways. The proposed Southdown study is divided into two potentialphases. Phase 1 is designed to provide a cost-effective screening of the possible risks associatedwith emissions from the Quarry. It will include emissions characterization of multiple sources inthe quarry, and indications of long term deposition of emissions from the quarry within theresidential neighborhoods, and will provide short-term measurements of exposure at residences nearto the quarry emissions (<0.5 miles) during periods of known quarry activity, and dose estimationsfor risk characterization. The Phase 1 tasks are designed to determine whether the riskassociated with the Quarry emissions is either well above or well below an action threshold. The study is not, however, intended to provide an accurate estimate of the absolute risk. That is,if the risk is found to clearly exceed the action threshold, the true level of risk may not be clear. Ifthe results from the Phase 1 risk characterization indicate that the risk is clearly above or below theaction threshold, then further characterization may be unnecessary. The results from Phase 1 willbe provided to DEP and EPA management to determine what, if any, further action is necessary andwhether a phase 2 study will be budgeted or necessary. It is important to note that a complete anddetailed Quality Assurance Project Plan (QAPP) will need to be developed and signed byrepresentatives of all participating organizations (both Agencies and contractors) prior to thecommencement of any data collection activities. The QAPP must conform to both the US EPA andNJDEP requirements. It is possible that separate QAPPs will be developed for each Phase of theproject, or certain combinations of the Phases. In such case, the QAPPs, or a separate document alsosigned by the participants, will need to clarify the connection between the activities covered by thedifferent QAPPs. It is also possible that the QAPP(s) will rely upon data collected previously by thesame or other people or groups. The QAPP(s) must account for this so-called secondary data,ensuring that the data quality and associated quality assurance and quality control procedures aresufficient for the use of the data in this project. Note, too, that detailed estimates of the number ofsamples required for the various component tasks will need to be developed based on the data qualitycriteria that are specified during QAPP development and as the evaluation of data from earlier tasksis completed.

Phase 1: Preliminary Investigation

Task 1: Complete simple ranking of sources at the quarry.

The relative emissions of dust from each of the various point and fugitive sources at the quarry canbe estimated and ranked to complete this task. The NJDEP has emissions data provided by thequarry in support of existing permits, as well as recent data on ongoing fugitive emissions. If thesedata are not sufficiently robust to support adequate ranking of sources, it will be necessary to obtaininformation from the quarry concerning the numbers and types of equipment used in their variousprocesses, the average throughputs of material in their various processes, and other supporting

information. Once such information is obtained, it should be possible to complete this task in a veryshort span of time (a couple of weeks).

Task 2:Complete characterization of protocol structures and other relevant structuresconcentrations in Southdown marble.

The concentrations of protocol structures and other relevant structures in Southdown Quarry marblecan best be determined in phases. For the initial phase of this investigation, available studies of thequarry will be reviewed. Core samples previously collected will be studied to gauge thehomogeneity of the material. These data will be evaluated before deciding whether it will benecessary to conduct further coring and analysis. Even if the coverage represented by existing coresis adequate, it will be necessary to extract samples from the cores and to analyze these samples toobtain information on protocol structures and other relevant structures, such that the information willbe useful in the subsequent modeling and in planning monitoring activities.

The samples from the existing cores will be composited to reduce analytical requirements. Detailedprocedures for conducting each of these steps are provided in Chapter 8 of the Superfund Method(Berman and Kolk, 1997)

Task 3: Conduct A Reasonable Worst-Case Assessment of Exposure

Preliminary dispersion models, employing reasonable worst case assumptions about local and site-specific meteorology, and conservatively accounting for the influence of local topography, and usingthe emissions models and data from Task 2, can be used to estimate downwind protocol structureconcentrations beyond the quarry fence line. Health risk can then be evaluated using Table 3-1 onthe assumptions of on lifetime, continuous exposure at the concentrations derived from thedispersion models.

Task 4: Conduct a Study of House Dust and SoilThis task is required to address the potential that past emissions from Southdown may haveadversely impacted surrounding areas. This work can be carried out simultaneously with Task 3.

An array of houses will be selected so as to provide for: (1) the greatest chance of detecting offsiteprotocol structures and other relevant structures emitted from the quarry (i.e. sample the closesthomes), (2) the opportunity to judge the degree of variability among adjacent homes, and (3) thegreatest possibility of observing a clear trend in concentrations with increasing distance from thequarry.

Soil samples will be collected at residential sampling locations to determine whether protocolstructures and other relevant structures deposited onto soil over time may serve as a reservoir forresuspension and track-in residences.

Concentrations of protocol structures and other relevant structures measured in house dust willprovide evidence of the integrated deposition over time, and will be evaluated for spatial trends. Inaddition, if the results of the spatial trends analysis indicate such a trend and/or if the concentrationof protocol structures and other relevant structures appears clearly elevated, the measuredconcentrations will be combined with published dust resuspension models to produce boundingestimates of risk to residents from indoor resuspension of the dust..

If the results of this task indicate that accumulated protocol structures and other relevant structuresin the homes would not be associated with elevated exposure, this should provide sufficient evidenceto suggest that past emissions from the quarry were not a problem. Coupled with a negative resultsfor the soil sample study and negative result for Task 3, it should also provide sufficient evidenceto end the study.

Positive detection of sufficient protocol structures and other relevant structures in house dust tosuggest elevated exposure, coupled with evidence that protocol structure deposition has notdecreased with increasing distance from the quarry should trigger consideration of the possibility ofsources of protocol structures and other relevant structures other than Southdown quarry.

Task 5: Conduct a Study of Outdoor and Indoor Airborne Protocol structures and otherrelevant structures Concentrations.

This task involves the simultaneous measurement of airborne concentrations of protocol structuresand other relevant structures inside and outside a limited number of homes immediately down-windof the quarry, along with measurements up-wind of the quarry (to assess background contributions). The purpose of such sampling is to validate the results of the modeling tasks, to provide anindependent estimate of the ambient concentration of protocol structures and other relevantstructures, and to assess indoor exposures during typical indoor activity.

In addition, careful simultaneous upwind and downwind sampling will provide an indication of thepresence of possible sources of airborne protocol structures and other relevant structures in thecommunity other than Southdown Quarry.

The outdoor sampling in this task will be used independently (to validate the modeling and to assessbackground contributions). The indoor sampling will assess a composite of various sources ofindoor exposure. These are expected to be dominated by re-suspension of previously depositedmaterial indoors, but will also represent a contribution from the infiltration of protocol structures andother relevant structures into the indoor environment from outside sources.

Task 6: Perform Summary Risk Characterization

Perform risk characterization following standard risk characterization guidance from the US EPAand the National Research Council. Risk characterization will include estimates of the uncertaintysurrounding risk estimates. The risk will be computed based on the results from tasks 3, 4, and 5,with the resulting risk estimates used together to form a quantitative, integrated, but screening-levelpicture of the risks associated with exposure to protocol structures and other relevant structures inthe Southdown quarry area. This will include upper bound and screening-level assessments, suitablefor use in determining whether the risks are far above or far below the level associated with elevatedrisk.

Since separate dose-response models for protocol structures and PCM-based fibers will be appliedin parallel, the risk characterization step may result in different estimates of risk from each approach. The results from the application of both models will be provided to DEP and EPA management assoon as they are completed. DEP and EPA will then determine what, if any, further action iswarranted and whether a phase 2 study will be budgeted or necessary.

Phase 2: Quantitative Characterization of Transport and Exposure

Details of a Phase 2 assessment, including possible application of Tasks 6 - 10, will be developedif requested by DEP and EPA management based on their review of the results from the Phase 1study.

5.0 REFERENCES

Berman, D.W. and Crump, K.S. (1999a). Methodology for Conducting RiskAssessments at Asbestos Superfund Sites. Part 2: Technical Background Document. Prepared for: Kent Kitchingman, U.S. Environmental Protection Agency, Region 9, 75Hawthorne, San Francisco, California 94105. Under EPA Review.

Berman, D.W. and Crump, K.S. (1999b). Methodology for Conducting Risk Assessmentsat Asbestos Superfund Sites. Part 1: Protocol. Prepared for: Kent Kitchingman, U.S.Environmental Protection Agency, Region 9, 75 Hawthorne, San Francisco, California94105. Under EPA Review.

Berman, D.W. and Kolk, A.J. (1997). Superfund Method for the Determination of Asbestosin Soils and Bulk Materials. Office of Solid Waste and Emergency Response, U.S.Environmental Protection Agency, Washington, D.C., EPA 540-R-97-028.

International Organization for Standardization (1995), Ambient Air-Determination ofasbestos fibres - Direct-transfer transmission electron microscopy method. ISO 10312.

Lippmann (2000) Asbestos and other mineral fibers. In: Environmental Toxicants,Human Exposures and Their Health Effects, 2nd Ed. (M. Lippmann, Ed.),Wiley-Interscience, New York, NY, 2000, pp. 65-119.

U.S. EPA (1986). Airborne Health Assessment Update. Office of Health and EnvironmentalAssessment, U.S. EPA, Washington, D.C., EPA 600/8-84-003F.

Appendix A

Toxicology of Asbestos Exposure Under the Protocol Structure Approach(Including A Working Definition for the Airborne Particles of Interest)

Asbestos is a collective term that refers to the fibrous forms of six different mineral types. Asbestosis of interest here because inhalation exposure to certain asbestos fibrous structures has been linkedto the incidence of serious health effects in humans. For the purpose of a health effects investigation,it is important to clearly understand the specific sizes, shapes, and mineral types of asbestos fibrousstructures that contribute to disease induction and thus need to be measured and evaluated to assessrisk. The information presented in this Appendix includes a working definition of such biologically-relevant structures and indicates the manner in which such structures will be measured, the attendantexposures estimated, and risks assessed in support of the investigation of the Southdown Quarry.

The discussion begins in Section A.1. with a concise definition of asbestos. Section A.2. thenprovides a delineation of the health effects associated with exposure to asbestos fibrous structureswith the attendant dose-response models presented in Section A.3. Sections A.4. and A.5. thendescribe the effects of mineralogy, size, and shape on the ability of an asbestos fibrous structure toinduce disease. Finally, Section A.6. presents a working definition of the specific asbestos fibrousstructures that contribute to the induction of disease (i.e. the biologically-relevant structures) underthe protocol structure approach.

A.2 The Definition of Asbestos

Asbestos is a term that represents the fibrous habits (crystalline forms) of six different mineral types(IARC 1977). The most common type of asbestos is chrysotile, which is the fibrous habit of themineral serpentine (a partially hydrolyzed magnesium silicate). The other five asbestos-relatedminerals are all amphiboles. Amphiboles are also partially hydrolyzed magnesium silicates but theircrystal structures differ from that of serpentine. The amphiboles whose fibrous habits are includedin the definition of asbestos are: fibrous reibeckite (crocidolite), fibrous grunerite (amosite),anthophyllite asbestos, tremolite asbestos, and actinolite asbestos.

All of the minerals whose fibrous habits are termed asbestos occur most commonly in non-fibrous,massive forms. While unique names have been assigned to the fibrous varieties of three of the sixminerals (noted parenthetically in the last paragraph) to distinguish them from their massive habits,such nomenclature has not been developed for anthophyllite, tremolite, or actinolite. Therefore,when discussing these minerals, it is important to specify whether the common (massive) habit orthe fibrous habit (asbestos) is intended.

It is important to note that the above definition of asbestos is broadly accepted but does notnecessarily reflect the specific characteristics of asbestos fibrous structures that contribute to adversehealth effects. Therefore, to avoid confusion, a new terminology is introduced in the next few

sections of this Appendix which, under the protocol structure approach, defines the specific rangesof sizes and shapes of the asbestos fibrous structures that contribute to the induction of disease. Moreover, the dose-response models and dose-response factors that will be used to assess asbestos-related risks potentially associated with Southdown under this approach are matched for use withexposures specifically expressed in terms of the concentration of protocol structures.

A.2. Health Effects Associated with Asbestos

It is generally accepted that exposure to long and thin asbestos fibrous structures can lead to a rangeof adverse health-effects including, primarily: asbestosis, lung cancer, and mesothelioma (see, forexample, Berman and Crump 1999a, and Lippmann 1988 and 2000). Asbestosis, a chronic,degenerative lung disease, has been documented among asbestos workers from a wide variety ofindustries. However, the disease is expected to be associated only with the higher levels of exposurecommonly found in workplace settings and does not typically result from environmental exposureto asbestos.

The type of lung cancers that have been attributed to asbestos exposure are the same as thoseattributed to smoking. In fact, simultaneous exposure to asbestos fibrous structures and cigarettesmoke tends to have a multiplicative effect on the risk of developing lung cancer.

Mesothelioma is a rare cancer of the membranes that line the pleural cavity (which surrounds theheart and lungs) and the peritoneal cavity (i.e. the gut). Although there is some evidence of a lowbackground incidence of spontaneous mesotheliomas in the general population, this cancer has beenassociated almost exclusively with exposure to fibrous substances (HEI-AR 1991). In most cases,this means exposure to asbestos. In rare cases, however, exposure to other fibrous substances hasalso been linked to the induction of mesothelioma. For example, erionite (a fibrous zeolite mineralthat occurs in some volcanic tuffs) has been established as the causative agent for the high rate ofmesothelioma observed in some villages in Turkey (Baris 1987).

Gastrointestinal cancers and cancers of other organs (e.g. larynx, kidney, and ovaries) have also beenlinked with asbestos exposure in some studies. However, such associations are not as compellingas those for the primary health effects listed above and the potential risks from asbestos exposureassociated with these other cancers are much lower (see, for example, HEI-AR 1991, Berman andCrump 1999a, and Lippmann 2000). Consequently, by addressing the more substantial asbestos-related risks associated with lung cancer and mesothelioma, the much more moderate riskspotentially associated with cancers at other sites are also addressed by default. Therefore, thediscussion in this document is focused on lung cancer and mesothelioma.

A.3 Dose-Response Models

Mathematical dose-response models that describe the observed associations between asbestosexposure and the induction of lung cancer and mesothelioma, respectively, were developed morethan 10 years ago (U.S. EPA 1986). These models were supported by a broad body of evidence from

numerous human and animal studies. There has been no firm evidence since then that theirfunctional form is inadequate, at least in terms of being protective of human health (HEI-AR 1991,Berman and Crump 1999a).

Based on these models, the nature of the respective relationships between exposure and the inductionof lung cancer and mesothelioma differ in at least two important respects (Berman and Crump1999a). While the incidence of both diseases increase linearly with increasing exposureconcentration (all else held constant), lung cancer incidence also increases linearly with duration ofexposure, independent of the time since first exposure.

In contrast, mesothelioma incidence increases as a power function of the time since first exposure. Because of this power relationship, exposure to asbestos that occurs earlier in life contributes moreto mesothelioma induction than later exposures and the overall incidence of mesothelioma increasesmore steeply with increasing (level and duration of) exposure than lung cancer.

The incidence of asbestos-related lung cancer also increases with increases in exposure to othercausative agents for lung cancer (e.g., smoking). The incidence of asbestos-related mesotheliomais not affected by simultaneous exposure to other factors that induce lung cancer.

Such details are noteworthy because, among other things, they help to explain why relatively shorterexposures appear to have induced mesotheliomas in some studies, even when other protocolstructures-related effects are not observed. At the same time, the models employed for evaluatingasbestos dose-response account for such observations and they reinforce the critical points that: (1)the risk associated with asbestos exposure is always a function of the level of exposure; and (2) suchrisks can be characterized sufficiently to answer questions of interest and to support reasonable risk-management decisions.

A.4 The Effects of Asbestos Mineralogy

There is abundant evidence and it is now generally accepted that the different asbestos types exhibitdiffering potencies, particularly toward the induction of mesothelioma (see, for example, Bermanand Crump 1999a, Lippmann 1988, 1994, 2000). Generally, the amphibole asbestos types (including tremolite asbestos, actinolite asbestos, amosite, and crocidolite) are now considered tobe more potent inducers of mesothelioma than chrysotile. The problem had been that suchdifferences have been difficult to quantify heretofore, primarily due to the need to separate the effectsof size and shape from the effects of mineralogy.

By applying an exposure model similar to that developed by Berman et al (1995) from animalinhalation data to the human epidemiology studies, the effects of mineralogy can be separated fromthe effects of size and shape (Berman and Crump 1999a). Results of this study indicate that, indeed,amphiboles are more potent than chrysotile for the induction of disease. More specifically, fiber-for-fiber, this model indicates that chrysotile asbestos is only about one tenth as potent as amphiboleasbestos toward the induction of lung cancer and only about one hundredth as potent toward theinduction of mesothelioma. The Berman and Crump model is described more fully in Section A.5.

Berman and Crump (1999a) further indicate that, once size and shape are adequately addressed, allamphiboles are equally potent toward the induction of each disease. The evidence supporting thisfinding is that, when the size-dependent model is applied to the published amphibole studies, boththe disparate lung-cancer potency estimates and, separately, the disparate mesothelioma potencyestimates from the various literature studies for amphiboles are each reconciled, converging to asingle value for each disease, respectively.

This last conclusion from this work contrasts with earlier suggestions in the literature thatamphiboles might vary in potency with crocidolite being the most potent. However, as notedpreviously such suggestions tended to be based on arguments not relating to quantitative dose-response comparisons and generally failed to address the confounding effects of size and shape. Importantly, none of the results of the earlier studies are inconsistent with the conclusions describedhere, they were just base on incomplete consideration of the relevant factors.

It has also been suggested in the past that tremolite asbestos is the most potent of the amphibolesbased on the observation that a certain Korean tremolite sample was shown to induce the greatesttumor response in animal injection and inhalation studies (Davis et al. 1985). However, this too hasbeen taken out of context.

When the effects of fiber size and shape are properly considered among the animal data, tremoliteis shown to exhibit essentially the same potency as that of the other amphibole asbestos types towardthe induction of animal tumors (Berman et al. 1995). As can be seen in Figure A-1 (which is agraph depicting the fit of a model to the animal inhalation data that is reproduced from Figure 3 ofBerman et al. 1995), the observed tumor response following inhalation of tremolite falls very closeto the dose-response curve predicted in this model. This in fact is the same Korean tremolitediscussed above. Therefore, the tremolite potency observed in the animal studies is adequatelypredicted by this animal model, which further indicates that all amphiboles exhibit essentially thesame potency.

Importantly, there is no undisputed evidence that risk associated with exposure to any of the asbestostypes cannot be adequately described by the existing models for lung cancer and mesothelioma (asmodified to incorporate size effects in the manner we recommend).

A.5 The Effects of Fiber Size and Shape

There is abundant evidence and it is generally accepted that the potency of the asbestos dustassociated with any particular exposure is a strong function of the distribution of the sizes and shapesof the fibrous structures in that dust (see, for example, Berman and Crump 1999, or Lippmann 2000). In addition the abundant evidence in the literature, two earlier statistical re-evaluations of theanimal data warrant particular attention.

Lippmann (1994) showed that lung cancer incidence in rats exposed to asbestos fibers by inhalationis significantly associated with the number concentration of long fibers (i.e. fibers longer than 5, 10or 20 μm) for both chrysotile and the various amphiboles. The best model fit is found for fiberslonger than 20 μm (R = 0.855), but the fit for fibers longer than 10 μm is almost as good (R=0.841),and both are considerably better than the fit for fibers longer than 5 μm (R=0.763). Thus this studyindicates that, while fibrous aerosols shorter than 5 μm have produced few, if any, lung cancers inanimal studies, fibers between 5 and 10 μm in length can produce lung cancers in rats and the fiberslonger than 10 μm are more carcinogenic than those between 5 and 10 μm in length.

Berman et al. (1995) indicated that a specific index of exposure (a weighted combination of theconcentrations of protocol structures in multiple size categories) could be identified that was capableof accurately predicting response (the observed tumor incidence in the set of animal inhalationstudies evaluated). Figure A-1 is reproduced here because it illustrates the goodness of fit of thedeveloped model to the data from a number of animal inhalation studies.

The best fit solid line in Figure A-1 represents the relationship between dose (exposureconcentration) and response (tumor incidence) that is predicted when exposure is expressed by theoptimum exposure index recommended in the 1995 study. The points in the Figure represent themeasured response in each of the inhalation experiments evaluated. Each point is also accompaniedby a vertical line representing a 95% confidence interval for the observed response from each study. It is clear from the Figure that most of the points lie very close to the model line and that, at aminimum, the model line passes through the 95% confidence interval of virtually all of the points. The overall appearance also indicates a strong monotonic relationship (i.e. a single “y” value for asingle “x” value) in which response increases regularly with exposure.

A modified version of the exposure index derived from this size-dependent, animal model has morerecently been applied to the published human epidemiology studies (Berman and Crump 1999a). Results from this analysis form the basis for a new protocol that was proposed for conductingasbestos risk assessment (Berman and Crump 1999b).

In the protocol, asbestos concentrations are measured as the concentration of all fibrous structures(including both the true asbestos fibers and any appropriately sized cleavage fragments) that arelonger than 5 μm and thinner than 0.5 μm. The structures that are longer than 10 μm are alsoweighted more heavily than the structures that are between 5 and 10 μm in length. The weightingscheme employed in this “index of exposure” is:

Copt = 0.003CS + 0.997CL (Equation A-1)

where:“CS” is the concentration of asbestos structures between 5 and 10 μm in length that

are also thinner than 0.5 μm; and

“CL” is the concentration of asbestos structures longer than 10 μm that are alsothinner than 0.5 μm.

That this proposed protocol is reasonable is strongly suggested by the fact that the disparate potencyestimates derived from the published human epidemiology studies of amphiboles are fullyreconciled, and the disparate potency estimates derived from the published studies of chrysotile arepartially reconciled by normalizing these potency estimates to reflect the exposure indexrecommended in the protocol (Berman and Crump 1999a).

Note that, due to the limitations of the published literature database that was available for adaptingthe animal model to the human epidemiology studies, it was not possible to fully optimize the model. Specifically, the minimum length of the structures that are weighted more heavily is shorter in themodel applied to the epidemiology studies than in the “optimum” model originally reported fromthe animal data. This change was necessary because the published size-distribution data would notsupport formal analysis of longer structures size categories. However, because shorter structures arealways more abundant in protocol structures dusts than longer structures, if anything, this change canbe considered conservative in a health-protective sense; we are weighting a larger number ofstructures more heavily than would otherwise be the case. A discussion of the details of theadaptations employed to derive the human dose-response model estimates is provided in Berman andCrump (1999a). However, because shorter structures are always more abundant in asbestos duststhan longer structures such details are not discussed further in this document.

A.6 Working Definition of a Biologically-Relevant Exposure Index for Asbestos: ProtocolStructures

Exposure to the massive habits of the asbestos-related minerals has not been associated with adversehealth effects per say. When crushed, however, the massive habits of these minerals can generateelongated particles (cleavage fragments) that resemble fibers. Because the properties of suchcleavage fragments differ from true asbestos fibers, it is not clear whether such fragments contributeto adverse health effects. However, there do not appear to be any studies that have definitivelyseparated the effects of mineral habit from the effects of the dimensions of the individual structuresthat are released when either the massive or fibrous habits of the asbestos minerals are crushed.

In contrast, there is sufficient evidence available to define a limited range of structure dimensionsthat represent the set of asbestos structures that contribute to biological activity (see below). Moreover, the limits of this range of structures is sufficiently narrow to exclude the vast majority ofcleavage fragments that tend to be produced when the massive habits of asbestos minerals arecrushed.

Given the current limits to knowledge concerning the health distinctions between cleavage fragmentsand true fibers, prudence dictates that we continue to include in the count of biologically-relevantstructures, the limited number of cleavage fragments that nonetheless exhibit sufficiently extremedimensions to fall within the definition of biologically-active structures (longer than 5 μm andthinner than 0.5 μm). Thus, under the protocol structure approach, all true asbestos fibers and thoseasbestos cleavage fragments that fall within the dimensional range that is defined as biologicallyactive (longer than 5 μm and thinner than 0.5 μm) will be referred to collectively in this frameworkas “protocol structures” and will be considered to contribute similarly to health effects. The precisedefinition of the exposure index referred to as protocol structures throughout main body of thisdocument is provided in Equation A-1 above. Note, such an approach also avoids the potentialcontroversies surrounding situations in which the massive and fibrous habits of asbestos mineral co-occur within the same deposit.

A.7 ReferencesBaris, I; Simonato, L.; Artvinli, M.; Pooley, F.; Saracci, R.; Skidmore, J.; and Wagner, C.

(1987). “Epidemiological and Environmental Evidence of the Health Effects of Exposureto Erionite Fibres: A Four-Year Study in The Cappadocian Region of Turkey.” Int. J.Cancer: 39, 10-17.

Berman, D.W. and Crump, K.S. (1999a). Methodology for Conducting Risk Assessmentsat Asbestos Superfund Sites. Part 2: Technical Background Document. Prepared for: KentKitchingman, U.S. Environmental Protection Agency, Region 9, 75 Hawthorne, SanFrancisco, California 94105. Under EPA Review.

Berman, D.W. and Crump, K.S. (1999b). Methodology for Conducting Risk Assessmentsat Asbestos Superfund Sites. Part 1: Protocol. Prepared for: Kent Kitchingman, U.S.Environmental Protection Agency, Region 9, 75 Hawthorne, San Francisco, California94105. Under EPA Review.

Berman, D.W.; Crump, K.S.; Chatfield, E.J.; Davis, J.M.G.; and Jones, A.D. (1995). “TheSizes, Shapes, and Mineralogy of Asbestos Structures that Induce Lung Tumors orMesothelioma in AF/HAN Rats Following Inhalation.” Risk Analysis 15(2): 181-195. Note,an errata page was also published in association with this manuscript (Risk Analysis 15(4):541, 1995) indicating that Figures 2 and 3 were inadvertently switched during publication.

Davis, J.M.G.; Addison, J.; Bolton, R.E.; Donaldson, K.; Jones, A.D.; and Miller, B.G.(1985). “Inhalation Studies on the Effects of Tremolite and Brucite Dusts in Rats.” Carcinogenesis, 6(5): 667-674.

Health Effects Institute - Asbestos Research (1991). Asbestos in Public and CommercialBuildings: A Literature Review and Synthesis of Current Knowledge. HEI-AR, 141 PortlandSt., Suite 7100, Cambridge, MA.

International Agency for Research on Cancer (1977) Monographs on the Evaluationof Carcinogenic Risks to Man. Vol 14. IARC, Lyon, France.

Lippmann, M. (1988) Asbestos exposure indices. Environ. Res. 46:86-106 (1988)

Lippmann, M. (1993) Biophysical factors affecting fiber toxicity. In: Fiber Toxicology (D.Warheit, Ed.), Academic Press, Orlando, 1993, pp. 259-303.

Lippmann, M. (2000) Asbestos and other mineral fibers. In: Environmental ToxicantsHuman Exposures and Their Health Effects, 2nd Ed. (M. Lippmann, Ed.),Wiley-Interscience, New York, NY, 2000, pp. 65-119.

U.S. EPA (1986). Airborne Health Assessment Update. Office of Health andEnvironmental Assessment, U.S. EPA, Washington, D.C., EPA 600/8-84-003F.

FRAMEWORK FOR ASSESSING POSSIBLE RISKS … FOR ASSESSING POSSIBLE RISKS ... and other asbestiform...

Documents

Transcript of FRAMEWORK FOR ASSESSING POSSIBLE RISKS … FOR ASSESSING POSSIBLE RISKS ... and other asbestiform...