WORKBOOK FOR USER WORKSHOP
EPA Guidelines
for
Health Risk Assessment
of
Chemical Mixtures
DENVER, REGION VIII
May 3, 1988

-------
U S EPA Region 8 Libra
80C-L
999 181 h SI , Suile 500
Donvct, CO 80202-2-16:"
GUIDELINE-SPECIFIC
briefing slides

-------
RISK ASSESSMENT
GUIDELINES: Chemical Mixtures
EPA
Environmental
Protection Agency

-------
EPA CHEMICAL
MIXTURES GUIDELINES
Hazard
Identification
Dose-Response
Assessment
Exposure
Assessment
Risk
Characterization
Characterizing toxicity of
mixture, similar mixture,
or components
Evaluating dose-response
data for mixture, similar
mixture, or components
Refer to exposure
assessment guidelines
Estimating risk
using all appropriate
methods

-------
WHY ARE MIXTURES
IMPORTANT?	
Mixtures, rather than single compounds,
are often found in:
Landfill leachate
Pollutants in ambient air
Purification byproducts in
drinking water

-------
WHAT IS NEW?
Framework for working within
data constraints
Criteria for judging quality
of risk assessment data
Emphasis on flexibility, judgment,
a clear articulation of assumptions
Option not to quantify the risk

-------
ASSESSMENT OF
DATA QUALITY
Information on Interactions
Actual ^	Insufficient
Data	Data
Health Effects Information
(Dose - Response)
Full Health	^ Insufficient
Effects Data * ^ Data
Exposure Information
Monitoring ^ ^ Insufficient
Data	Data
Without adequate information, no quantitative
assessment is made!

-------
OPTIONS FOR MIXTURE ASSESSMENTS
Assess data quality
(adequate)
Qualitative risk assessment
i
(inadequate for quantification)
Data on mixture 	
(yes)	(no) (yes)
On similar mixture
On components
(no)
Risk assessment
Risk assessment
Hazard index
i
Interactions risk assessment
i
Additivity risk assessment
Develop integrated summary with uncertainties

-------
OPTIONS FOR MIXTURE ASSESSMENTS
Assess data quality	
(adequate)	(inadequate for quantification)
Qualitative risk assessment
i
Data on mixture 	
(yes)	(no) (yes)
On similar mixture —~ On components
(no)
Risk assessment
Risk assessment
Hazard index
i
Interactions risk assessment
i
Ad
ditivity risk assessment
-*¦ Develop integrated summary with uncertainties

-------
OPTIONS FOR MIXTURE ASSESSMENTS
Assess data quality	
(adequate)	(inadequate for quantification)
Qualitative risk assessment
i
Data on mixture 	
(yes)	(no) (yes)
On similar mixture
On components
(no)
Risk assessment
Risk assessment
Hazard index
i
Interactions risk assessment
i
Ad
ditivity risk assessment
Develop integrated summary with uncertainties

-------
OPTIONS FOR MIXTURE ASSESSMENTS
Assess data quality 	
(adequate)	(inadequate for quantification)
Qualitative risk assessment
i
Data on mixture 	
(yes)	(no) (yes)
On similar mixture
(no)
On components
Risk assessment
Risk assessment
Hazard index
i
Interactions risk assessment
i
Ad
ditivity risk assessment
Develop integrated summary with uncertainties

-------
UNCERTAINTIES
Evaluate and express uncertainties in:
• Composition
Interactions
Health effects
Exposure

-------
ASSUMPTIONS AND
LIMITATIONS
Discuss information such as:
Interactions
Modeling assumptions
Data limitations

-------
SUMMARY - MIXTURE
ASSESSMENT
Science and art
Judgment must be exercised
Divergence of situations requires
flexibility in approach
Assessor must pass on assumptions,
judgments, and uncertainties to the
decision maker (risk characterization)

-------
CASE STUDIES

-------
MIXTURE CASE STUDY 1

-------
PARTICIPANT'S OVERVIEW
MIXTURES ASSESSMENT
CASE STUDY 1
The following pages contain a case study that should
guide you through the mixture assessment guidelines and
give you an opportunity to do some straight forward
calculations for both systemic and cancer health risks.
As noted in the case, the goals of the exercise are to
familiarize the students with the specifics of the
mixtures guideline and also to provide the student with
hands-on experience in working through the guidelines.
As well as responses to the specific questions raised
in the problem, you should discuss the extent to which
the stated goals of the guidelines — to encourage
consistency and scientific quality in risk
assessments — have or have not been demonstrated.
Be sure to make extensive use of the Guidelines and
provide, whenever appropriate, specific page and column
numbers from the Guidelines in your responses.

-------
MIXTURES ASSESSMENT TRAINING
CASE STUDY # 1
ESTIMATION OF RISK FOR POLLUTED GROUNDWATER
The Situation
Ground-water contamination is one of the nation's more
common pollution problems and one of the most difficult to
address in terms of identifying human health risk. Too often,
monitoring data are incomplete making an adequate assessment
impossible.
As the groundwater risk assessor for your office, you have
been asked to analyze the situation confronting the public health
officials of Mudville: five chemicals have been found in their
drinking water wells! In addition to examining the specifics of
the applications of the Guidelines to the risk assessment
process, this case study will emphasize the need to identify and
evaluate the uncertainties inherent in the assessment.
Concentrations of five chemicals (benzene, bromoform, carbon
tetrachloride, 1,1-dichloroethylene and toluene) have been
measured in four drinking water wells in Mudville. The
pollutants, all of which were measured above their respective
detection limits, are generally assumed to have originated in and
leached from a landfill near the wells. The landfill has been
capped with clay so that the only present exposure route of the
chemicals to the citizens of Mudville is through the drinking
water. Moreover, during the past three years, the contaminant
levels in the wells have declined about ten percent. The results
of the most recent monitoring tests are shown in Table 1.
In this exercise, you are asked to assess the human health
risk posed by the contaminant mixture in the wells, and to be
prepared, if necessary, to brief the Regional Administrator on
these risks. Using data provided in the tables at the end of
this handout and the Agency's Guidelines for the Health Risk
Assessment of Chemical Mixtures, you will be guided by a series
of tasks through the risk assessment process.
SUMMARY OF DATA TABLES (ATTACHED)
Table 1 shows the most recent monitoring data for the 5
chemicals in four Mudville drinking water wells. Notice that
"maximum" levels are also identified. In order to simplify this
exercise, we suggest you use these maximum values in your
calculations. In at least one of the exercises you will need to
discuss the implications of this simplification. Moreover, we

-------
2
suggest you consult the Exposure Assessment Guidelines to decide
how you might deal with data in the "real" world.
Tables 2 and 3 respectively show the systemic toxicity and
carcinogenicity information available for the well pollutants.
Table 2 (column 3) shows the Reference Dose (RfD) for each
compound. The RfD is an estimate, with uncertainty spanning an
order of magnitude, of the amount of a substance thought to be
without adverse effect in humans, even if exposure at this level
occurs for a lifetime. Column 6 shows the "allowable"
concentration in drinking water, estimated from the RfD, assuming
daily consumption of 2L of water by a person weighing 70 kg.
(The word "allowable" should not be understood to connote safety.
Among Agency risk assessors this concentration, since it is based
on the risk reference dose, is now more commonly referred to as a
"reference" level.)
Table 4 provides data on the cancer risk estimates for the
individual wells. Data on toxic interactions (which is very
important in mixture risk assessment) are shown in Table 5.

-------
3
Task A: Given the data in hand, and using the rating schemes
(Table 2 of the Guidelines), decide whether a quantitative risk
assessment (QRA) can be performed. If your answer is
affirmative, which of the procedures (whole mixture, similar
mixture, etc.) should be used?
Suggestions: In performing this evaluation, you will want to
consider, among other things, the following:
1.	the 'preferred' type of data that would be used in an
assessment;
2.	the nature and extent of interactions of all types
among the mixture components;
3.	the availability of health effects and exposure
information on the well water or its components;
4.	the extent to which professional judgement enters into
this evaluation

-------
4
Task B: Table 5 summarizes the available data on toxic
interactions of the chemicals found in the wellfield; benzene and
toluene, and carbon tetrachloride and toluene. The data are not
sufficient to determine whether there are long-term interactions
between the chemicals present in the wells.
What use can be made of these data in the risk assessment?
Suggestions: Consider the following:
1. the various types of interactions that can take place,
and their temporality;
2.	the effects of synergistic or antagonistic interactions
on a mixture risk assessment;
3.	the effect such interactions can have on exposure, the
hazard index, or the cancer risk assessment;
4.	the conclusions, from the data in Table 5, for the risk
assessment conducted for Mudville's drinking water.

-------
5
Task C: Using information on actual and "allowable" reference
levels of exposure to a compound(s), the guidelines recommend the
development of a "hazard index" (HI) as a rough measure of the
degree of toxicity of a mixture.
Use the data available in the tables and the information
provided in the guidelines to perform such an evaluation for non-
cancer effects. Decide whether there is cause for concern for
the non-carcinogenic effects of this mixture.
Suggestions; Among the factors you will need to consider are the
following:
1.	similarity of action of the components (are any data
provided to enable you to judge this?);
2.	the use of the formula given for the HI in the
guidelines;
3.	the extent to which additivity in response or dose can
be applied to the components of the well water;
4.	the confidence in the RfD, and the uncertainty factors
used in its estimation;
5 the extent to which the HI provides a quantitative
estimate of hazard;
6. the significance of the differences in toxic endpoints
for the mixture components.

-------
6
Task D: The excess cancer risk (R) from a lifetime exposure to
the mixture is calculated using the following formula:
R= U1 x El + U2 X E2 +	Un x En
Un = unit risk estimate (risk from 1 ug/L) of the nth toxicant
En = monitored level in the well (ug/L) of the nth toxicant
Using the above formula, and the data in Table 1, calculate
the excess cancer risk from a lifetime exposure to the mixture
for well 1. Insert your answer in column 1 of Table 4.
Please discuss the differences between the excess risk
determined using the "maximum levels" and the risks calculated
for the individual wells. Also, please discuss the reasons for
the two overall cancer risk estimates shown in Table 4.
Suggestions: In your review, you should consider the following:
1.	the applicability of the formula at various exposure
levels;
2.	the extent to which additivity can be applied to the
cancer data for the components of the well water;
3.	the similarities and differences between this formula
and that used to describe the potential systemic
toxicity;
4.	the influence of weight of evidence of carcinogenicity
in the overall risk assessment.

-------
7
Task 5 f Part 1): The evaluation of uncertainty and the
presentation of the uncertainties, assumptions, and limitations
of the assessment is an important part of risk assessment.
Discuss the major points of uncertainty in the assessment of
risk of the sampled drinking water wells of Mudville.
Suggestions; You should, at a minimum, discuss the uncertainties
in the interaction data, in the exposure information, and in
hazard estimates done in the assessments. In so doing you should
consider the following:
1.	data on interactions;
2.	the precision of the RfDs and the effect of this
precision on the uncertainty of the mixture assessment;
3.	the implications of the use of "maximum" exposure
levels (do the Exposure Assessment Guidelines help?);
4.	the lack of uniformity in the cancer data;
5.	the absence of data on other exposure routes or on
individuals who do not fit EPA's assumptions (e.g.,
children, individuals drinking more or less than 2L per
day, persons who may be exposed at work, with lower
body weights or other special considerations).
Task E (Part 2): Summarize the most important "overall"
conclusions of the risk assessment?
Suggestions: You should consider the following in your review:
1.	your conclusion of the cancer and non-cancer risk to a
person drinking water from the wells;
2.	your summary of the quality of the available data and
the uncertainties of the assessment;
3.	the extent to which this exercise demonstrated (or
failed to adequately bring out) the stated goals of the
guidelines - to promote quality and consistency in risk
assessment.

-------
TABLE 1. MONITOR ING DATA
CHEMICAL	CONCENTRATIONS IN WELLS (ug/l)a

Well 1
Well 2
Well 3
Well 4
Maximum
Level
Benzene
22.
17.
10.
30.
30.
Bromoform
82.
30.
34.
42.
82.
Carbon tetrachloride
20.
21.
14.
20.
21.
1,1-0 ic hioroethylene
33.
27.
41.
22.
41.
Toluene
540.
470.
600.
520.
600.
3 All concentrations are above the detection limits. These are
the latest measurements, but the concentrations seem to be
decreasing over time.

-------
TABLE 2. SYSTEMIC TOXICITY INFORMATION FO* THE EXAMPLE SITE ASSESSMENT

MAXIMUM
REFERENCE DOSE8
REFERENCE

CRITICAL
chemical
CONC1Nd
RfD
CONFIDENCE UF USED
LEVEL (RL)C
E/RL
TARGET

(ug/1)
(mg/kg-d)
IN RfD
FOR KfD
(ug/D

ORGAN
Benzene
30
•0007e
	
1000
25.
1.2
b 1 ood
Bromoform
82
• 006f
1 ow
10
210.
.39
1 i ver
Carbon tetra-
21
.0007
med i urn

25.
.86
1 i ver
chl oride







1,1-DCE
41
.009
medium

320.
.13
1 i ver
Toluene
600
.3
medi um

11000.
.057
blood
Hazard Index (liver) = 1.4
Hazard Index (blood) =1.3
¦a: Maximum monitored level (See Table 1).
b: An estimate (with uncertainty spanning perhaps an order of magnitude)
of the daily exposure to the human population, including sensitive
subgroups, that is likely to be without appreciable risk of deleterious
effects even if exposure occurs during a lifetime.
RfDs are estimated to one significant digit. RLs and the E/RL ratio
are carried with two digits since they are calculated intermediate
values. The Hazard Index is only given to one digit to reflect the
precision of the RfD.
c: RfOs are converted to RLs assuming daily consumption of 2 L of water
by a 70 kg person, and the application of a factor of 1000 to convert
mg to ug.
d: The target organ affected in the critical study, i.e., in a series of
studies, the study showing an adverse effect at the lowest dose level
(see Appendix to IRIS).
e: No reference dose has been established for non-cancer effects of
benzene. This value was derived by EPA's Office of Drinking Water,
U.S. EPA 1985.
f: No reference dose has been established fo^ bromoform. The value
reported here is fictitious, as are the uncertainty factor and level
of confidence in the fictitious RfD.

-------
TABLE 3. CARCINOGENICITY INFORMATION FOR THE EXAMPLE SITE ASl>CLIENT
CHEMICAL
MAXIMUM
CONC'N. (ug/1)a
UNIT RISKb
(/ug/L)
EXCESS RISKC
WEIGHT OF
EVIDENCE*1
Benzene
30
8. 2x10~7
2.5xl0"5
A
Ca rbon
Tetrachl oride
21
3. 7x10-6
7.8xl0"5
B2
1,1-DCE
41
1. 7xlO-5
7.0x1O"4
C
Mixture cancer risk (without DCE) = lxlO-4
Mixture cancer risk (including DCE)= 8x10"^
a Maximum monitored levels (See Table 1).
b Upper bound of the estimated excess cancer risk from lifetime exposure to
1.0 ug/1 in drinking water, assuming consumption of 2 L water per day by
a 70 kg person. The actual risks are not likely to be greater, and
could be significantly smaller than this estimate.
Note that this value (and the excess risk) are not rounded because
they are intermediate values in the risk assessment
c Upper bound of the estimated excess cancer risk from lifetime exposure at
the maximum monitored concentration.
d See the cancer risk assessment guidelines (US EPA, 1986b).

-------
TABLE -i. CANCE- SISK ESTIMATES rnp i^n IVI [1UAL WELLS3
EXCESS CANCER RISK
CHEMICAL WELL 1
WELL 2
WELL 3
WELL 4
Benzene
1.4x10-5
8.2x10-6
2.5x10-5
Carbon tetrachl.
7.8x10"^
5.2x10-5
7.4x10-5
1,1-DCE
4. 6x10~4
7.0x10-4
3.7x10*4
Mixture risk



without DCE
9xl0"4
6x10-5
1x10-4
including DCE
6x10-4
8x10-4
5x10-4
a. For assumptions and interpretations see footnotes Table 3.
TABLE 5
DATA ON TOXIC INTERACTIONS FOR WELL WATER CHEMICALS
COMPOUNDS
ROUTE3
UURATION
SPECIES
EFFECT/ORGAN
INTERACTION^ #
STUDIES
toluene/
i n ha 1
acute
htnan
excretion/lung
inhibition 1
benzene
i nhal
acute
hunan
elimin./blood
none 1

inhal
acute
hunan
metabolism/body
none 1

i .p.
acute
rat
excretion/body
inhibition 3

i .p.
acute
rat
el imin./blood
inhibition 1

i .p.
acute
rat
metabolism/body
inhibition 3

i .p.
acute
rat
metabol ism/liver
inhibition 1

s.c.
acute
rat
excreti on/body
inhibition 1

s.c.
acute
mouse
function/marrow
inhibition 1
toluene/
i.p.
acute
mouse
mortality/body
potentiation 1
carbon tet.
i .p.
acute
mouse
depression/ens
potentiation 1
a: inhal = inhalation, i.p.= intraperitoneal injection, s.c.= subcutaneous
injection.
b: listed process is lower rate or less severity than expected (if
inhibition); or is higher rate or greater severity (if potentiation).

-------
MIXTURE CASE STUDY 2

-------
8
MIXTURE ASSESSMENT TRAINING
CASE STUDY # 2
The Situation:
You are the Senior Science Advisor to the Regional
Administrator. On your way to work one morning, you are greeted
with the following headline:
DEADLY DIOXIN DELUGES DOWNTOWN
EPA Officials Hold Breath	And Are Silent
"Sources close to the Regional Administrator of EPA (RA)
have informed the Gazette that the Agency has obtained 'high tech
data' showing that the city's new $100 M recycling energy-from-
waste municipal waste combustor (MWC) is emitting dioxin, which—
according to EPA — is 'the most toxic man-made chemical'. In
addition, EPA documents state in reference to dioxin that 'there
is no safe level for exposure to such a compound. The
recommended level of exposure for humans is zero.
"These air emissions daily form a plume which casts a deadly
pall over the center of the city, with a maximum impact on the
grounds of our beloved Wilma Wilder's Shelter for Widows and
Waifs. (Ms. Wilder was recently recognized by her admiring
fellow citizens when the Mayor's mother presented her with the
coveted 'Octagenarian of the '80s' award.) When interviewed in
connection with this story Ms. Wilder confessed to 'not feeling
as good as I used to.'
"Contacted at his home in the suburbs, 30 miles upwind from
the MWC facility, the RA pleaded ignorance of details of the
problem, but said he remembers not discussing the matter with his
aides — particulary his Senior Science Adviser. However, he
promised this reporter that the latter would report to him by 2
pm this afternoon and that the RA would be available to the media
in the Press Room (inexplicably called the Lion's Den) at a press
conference later in the afternoon. At that time, the RA intends
to discuss the MWC emission data and its implications."
. Your eagerness to greet the new day having been blunted, you
arrive at work and set about gathering data from recent MWC tests
as well as other background information on dioxins in general, on
the materials being emitted from the MWC, and on the proper
conduct of a risk assessment on a mixture.
After reviewing this material, you should proceed through
the tasks that are set forth in this handout. Please make
extensive use, where appropriate, of the guidelines.

-------
9
BACKGROUND INFORMATION
A. General Information
The "dioxin" that was referred to in the newspaper stories,
as being emitted from the MWC is not a single chemical. Rather,
the emissions are a mixture of chemically related chlorinated
dibenzo-p-dioxins and -dibenzofurans (CDDs and CDFs), as well as
many other components (for which no data are shown). The 75 CDDs
and 125 CDFs comprise a "family" of structurally related
compounds:
Each of the CDDs or CDFs is described as a congener. The
CDDs having the same number of chlorine atoms belong to the same
homologous group (the same is true for the CDFs). Therefore
there are eight CDD homologues (i.e., each having one to eight
chlorine atoms). Chemically distinct members of a homologous
group are called isomers. For example, there are 22 isomers in
the tetra-homologous group of CDDs (the TCDDs). One of these 22
isomers is 2,3,7,8-TCDD.
Most toxicity information is available for 2,3,7,8-TCDD and
two HxCDDs. Some information is available for 2,3,7,8-TCDF.
While much less is known about the toxic potential of the other
congeners, this information is sufficient to assess their
relative toxicity. In general, congeners with chlorine
substituents at positions 2,3,1, and 8 are significantly more
toxic than isomers not so substituted (see Appendix B).
B. Data Available on MWC emissions
Data pertaining to the MWC and its emissions are provided on
several attachments. Table 1 shows the relative toxicity (the
Toxic Equivalence Factor or TEF; see below) of the 15 most toxic
congeners. We will shortly review how these TEFs are used in the
risk assessment.
Table 2 shows the results from recent sampling of the MWC
stack and the average concentrations of the various homologues
and selected congeners (the most toxic ones, those having
chlorine atoms at the 2,3,7, and 8 positions) found in the
emissions.
i
DIBENZODIOXIN
DIBENZOFURAN

-------
10
Tables 3 and 4 respectively show exposure information and
hazard information on 2,3,7,8-TCDD which will be used to estimate
the risk posed by the emissions.
C. The TEF Procedure
Most ambient exposures to these compounds involve mixtures
of CDDs and CDFs, and in almost all cases there is no information
on the toxicity of the mixture in question. Therefore, in an
attempt to deal with the uncertainty presented by the absence of
data on the mixture, EPA has adopted an interim procedure (the
TEF procedure) based on dioxin toxicity equivalence factors
(TEFs) for estimating the risks from exposure to CDD/CDF
mixtures. 1
The following is a brief description of the TEF procedure.
In the TEF approach, the exposure level of each CDD and CDF
congener or homologous class is replaced by the concentration of
2,3,7,8-TCDD that is estimated to potentially cause the same
health effect as the CDD\CDF in question. These exposure levels
(now in terms of TCDD equivalents) are used in the risk
assessment. The TEF procedure involves the following steps:
1.	Analytically determine the CDDs and CDFs in the sample,
preferably determining both the total and
2,3,7,8-substituted congeneric concentration.
2.	Determine the appropriate values for the TEFs. These
are shown in Table 2.
3.	Multiply the congener concentrations in the sample by
the TEFs in Table 2. This expresses the measured
concentrations in common terms, i.e., in terms of
2378-TCDD equivalents. For example, the average
concentration of 2378-PeCDD of 650 ng/dscm (see Table
3) multiplied by its TEF of 0.5 (Table 2), gives the
2378-TCDD equivalent value (TEQ) for this congener as
325 ng/dscm.
4.	Sum the TEQs to obtain the total concentration of 2378-
TCDD equivalents in the mixture.
Thus, in cases in which the concentrations of the 15
2378-substituted congeners listed in Table 2 are known:
1Interim procedures for estimating risks associated with
exposures to mixtures of chlorinated dibenzo-p-dioxins and
-dibenzoforans (CDDs and CDFs) EPA/625/3-87/012, March 1987.

-------
11
2378-TCDD Equivalents = (TEF of each 2,3,7,8-CDD or CDF
congener X the concentration
of that congener)
5. This latter value, in combination with exposure and
toxicity information (tables 4 and 5), allows the
assessor to estimate the risks associated with the
mixture.
D. Risk Calculations
Although this is not a course on how to perform a risk
assessment, you need to know how to estimate exposure and risk in
order to work through the following material.
1.	Estimation of exposure
Table 2 contains data on measured stack emissions from the
MWC. The emissions contain an average concentration of 120 ng
237 8 HxCDDs/dscm. The exposure information presented in Table 4
indicates that a reasonable estimate for ground level
concentration at 1 3cm downwind is approximately 10^ to 10® fold
dilution of the stack emissions (for purposes of illustration we
will use only the latter value in these calculations). The
estimate for ground level,concentration at 1 km is therefore
approximately:
120 x 10"® ng 2378-HxCDDs/m^ air.
Table 1 shows that EPA has assigned a TEF value of 0.04 to
the 2378-HxCDDs. Therefore, the estimated exposures is:
0.04 x 120 x 10~® = 4.8 x 10"® ng TCDD/m^ air.
Information available in Table 3 indicates that
approximately 75% of an inhaled dose is absorbed in the lung.
Accordingly, the estimate of the absorbed dose is:
3.6 x 10"® ng TCDD/m^, or 3.6 x 10*" ^ pg TCDD/m^ air.
2.	Estimation of cancer risk
The estimate of the upper limit of the cancer risk resulting
from a lifetime exposure = dose x unit risk for carcinogenicity
(see Table 3) = (3.6 x 10"^ pq/m^) x (3.3 x l0~^/pg/m^) =
= 10 [B2].
Now let's get on with the exercise!

-------
12
Task A: Focusing on Figure 1 of the Mixture Guidelines, and
using the data provided in the Tables for this exercise, decide
whether the preferred approach can be used, and, if so, why. If
the preferred approach is not possible, determine which approach
can be used.
Task B: Use Table 2 to evaluate the quality of the data on
interaction, health effects and exposure. Does this evaluation
change the determination you made in Task A?
Suggestion; Consider the quality of the data on interactions,
health effects and exposure.

-------
13
Task C: Discuss the appropriateness of using the TEF procedure
with the data available in this case.
Suggestions: You will need to consider the following:
1.	the assumptions that must be made about the components
of the mixture to make this approach defensible;
2.	whether there are differences in justification for use
of this approach with respect to cancer or systemic
toxicity;
3.	the role of judgment and professional opinion in deter-
mining the applicability of this approach.

-------
14
Task D: Using the data presented in the Tables and the examples
shown at the beginning of the case, estimate the cancer and the
teratogenic risk to the exposed population posed by this mixture
for isomer-specific data.
1.	Estimate the average exposure and the absorbed dose (in
pg TCDD eqts./m3 of air) for a person living 1 Jem down-
wind from the MWC (you may wish to consult the sample
calculation provided above).
2.	Estimate the upper limit of excess cancer risk from
this exposure to an individual (use the material in D1
as an example).
3.	Estimate the risk for teratogenic effects for a woman
living 1 km downwind from the MWC.
4.
What additional exposure sites might be considered?

-------
15
Task E: In our review of Case Study #1, we discussed the
importance of the evaluation and presentation of uncertainties
and limitations of a risk assessment. For this task, please
discuss the major uncertainties in the risk estimation of this
mixture.
Suggestions: You should consider the following points:
1.	the uncertainties associated with the analytical data
and the data on interactions, health effects, and
exposure;
2.	uncertainty associated with the assumptions necessary
to use of the TEF procedure;
3.	why it is important to discuss uncertainties;
4.	how representative are the stack emission data? Can
you quantify the uncertainty in the use of the average
concentration and isomer distribution data;
5.
the uncertainties and limitations resulting from the
fact that only CDDs and CDFs in the emissions were
used for the risk assessment.

-------
16
Task F: Prepare a short briefing on the risks posed by the MWC
emissions for the RA. Be sure to include a description of the
way the mixture (and other) guidelines were used to develop the
risk assessment.
Be sure to include information on the following:
1.	characterization of the cancer risk; of the teratogenic
risk;
2.	assumptions and uncertainties limiting those
assessments;
3.	ideas on what could be done to limit some of the
uncertainty; (e.g., how could a better assessment be
developed?)
4.	the extent to which the guidelines bolster or limit the
evaluation.

-------
TABLE 1
CDD/CDF ISOMERS OF MOST TOXIC CONCERN3
Dioxi n

Di benzofuran

Isomer
TEFb
Isomer
TEFb
2,3,7,8-TCDD
1
2,3,7,8-TCDF
0.1
1,2,3,7,8-PeCDD
0.5
1,2,3,7,8-PeCDF
0.1


2,3,4,7,8-PeCDF
0.1
1,2,3,4,7,8-HxCDD
0.04
1,2,3,4,7,8-HxCDF
0.01
1, 2^3,7,8,9-HxCOD
0.04
1,2,3,7,8,9-HxCDF
0.01
1,2,3,6,7,8-HxCDD
0.04
1,2,3,6,7,8-HxCOF
0.01


2,3,4,6,7,8-HxCOF
0.01
1,2,3,4,6,7,8-HpCDD
0.001
1,2,3,4,6,7,8-HpCDF
0.001


1,2,3,4,7,8,9-HpCDF
0.001
a In each homologous group, the relative toxicity factor for the isomers not
listed above is 1/100 of the value listed above.
b TEF = Toxicity Equivalence Factor = relative toxicity assigned.
These factors are utilized in the "lEF Procedure": the concentration of
each CuD/F isomer (or homologue, if 1somer-specific data is lacking) is
multiplied by the appropriate TEF factor listed above, resulting in an
estimate of the TCOO equivalents for that isomer.

-------
TABLE 2
STACK EMISSION DATA*
(ng/dscm)
Compound	| Average Concentration	Range


HonoCDO
7.4
NO
-
13**
DC DDs
39
ND
-
130
TrCDDS
45
ND
-
140
TCDDs : total
230
42

450
: 2378
100
21
-
200
PeCDDs: total
1200
270

2800
: 2378
650
510
-
940
HxCDDs: total
510
140

1500
: 2378
120
100
-
570
HpCDDs: total
160
120

390
2378
110
90
-
240
OCDD
41
33
-
110
MonoCDF
19
8
-
55
DICDFs
66
48
-
98
TrCDFs
80
34
-
120
TCOFs : total
75
49

87
: 2378
30
22
-
76
PeCDFs: total
250
130

540
: 2378
130
100
-
280
HxCDFs: total
900
640

12U0
: 2378
620
440
-
820
HpCDFs: total
200
160

260
: 2378
120
90
-
180
OCOF
6
NO
-
20
* Averages and ranges derived from monitoring measurements made on
five successive days.
** NO: <0.5 ng/dscm.

-------
TABLE 3
EXPOSURE INFORMATION DATA
EPA-S USUAL EXPOSURE ASSUMPTIONS
I. ASSUMPTIONS ON THE EXPOSED INDIVIDUAL
L i f e s pa n :
70 years
Bodyweight:
Adult:
70 kg
Child:
10 kg
Breathing rate
Adult:
20 m^/day
11-.-ASSUMPTIONS ON BIOAVAILABILITY FOLLOWING INHALATION:
0 2,3,7,8-TCDD (and other CDDs and CDFs) in the emissions are
adsorbed onto particulate matter
0 about 75% of inhaled particulates are retained in the lung.
° all the 2,3,7,8-TCDD on the particulates is biologically available .
III. AIR DISPERSION MOOEL NEEDS:
Stack	exit temperature
Flow	rate
Stack	diameter
Stack height
Ambient temperature
Data assumptions about local climate
Residential pattern
IV. RULE OF THUMB AIR DISPtRSION RESULT:
"ballpark" estimate is about 10^ to 10® fold
dilution of stack emissions at the point of
maximum annual concentration.

-------
table 4
HAZARD INFORMATION DATA
FOR
2,3,7,8-TCDD
° Slope factor for carcinogenic response = 1.6 x 10$ per mg/kg/day,
derived from feeding study in rats. B2 carcinogen.
° Unit risk number (inhalation) (upper limit estimate of incremental
cancer risk for continuous lifetime exposure to 1 pg/m^ of
2,3,7,8-TCDD in ambient air) = 3.3x 10~5 . This estimate takes into
account the fact that 25% of the inhaled material is exhaled, and
75% is retained and absorbed.
° RfD (based on teratogenic effects) = 1 pg/kg/day.

-------
TABLE A
MIXTURE GUIDELINE EXAMPLE
CDD/CDF INCINERATOR EMISSIONS
ESTIMATION OF RISK
1. HOMOLOGUE-SPECIFIC ESTIMATION OF TCDD EQUIVALENTS
Compound	Concentration	TEF	TCDD Equivalents
(ng/dscm)	(ng/dscm)
Mono CDD
7.4
0
DCDDs
39
0
TriCDDs
45
0
TCDDs
230
1
PeCDDs
1200
0.5
HxCDDs
510
0.04
HpCDDs
150
0.001
OCDD
41
0
Mono CDF
19
0
DCDFs
66
0
TrCDFs
80
0
TCDFs
75
0.1
PeCDFs
250
0.1
HxCDFs
900
0.01
HpCDFs
200
0.001
OCDF
6
0
TOTAL
* Rounded to 1 significant figure.

-------
TABLE B
MIXTURE GUIDELINE	EXAMPLE
CDD/CDF INCINERATOR	EMISSIONS
ESTIMATION OF	RISK
2. ISOMER-SPECIFIC ESTIMATION	OF TCDO EQUIVALENTS
Compound	Concentration	TEF	. TCOD Equivalents
(ng/dscm)	(ng/dscm)
237 8-TCDD	100	1	100
other TCDDs
2378-PeCDDs
other PeCDDs
2378-HxCDDs
other HxCDDs
2378-HpCDDs
other HpCDDs
"OCDD
2378-TCDF
other TCDFs
2378-PeCDFs
other PeCDFs
2378-HxCDF s
other HxCDFs
2378-HpCDFs
other HpCDFs
OCDF
TOTAL
Rounded to 1 sign. fig.

-------
TABLE C
MIXTURE GUIDELINE EXAMPLE
CDD/CDF INCINERATOR EMISSIONS
ESTIMATION OF RISKS
ESTIMATE OF EXPOSURE AND RISK

-------
APPENDIX A
EXPLANATION OF DOSE AND RESPONSE ADDITION
MODELS FOR USE IN CHEMICAL MIXTURE RISK ASSESSMENT
DOSE ADDITION
Dose addition is one method for estimating the potential
toxic effects of a mixture of chemicals. This procedure involves
the addition of the effective dose of each component, i.e., the
ratio of the exposure dose and the RfD for that chemical. This
is illustrated in the following table.
RfD	Potency Exposure	Effective
(1/RfD) Dose	Dose*
Chemical
1
10
0.1
100
1
Chemical
2
2
0.5
20
10




Hazard Index:
11
* Exposure dose / RfD
This example illustrates that the effective dose of the
mixture, in effect, accounts for the relative toxic potencies of
the individual chemicals of the mixture.
The best justification for dose addition is knowledge that
the mixture components act by the same mechanism, on the same
target organ, in the same species. In practice these conditions
rarely obtain.
1.	Data on mechanism of action is seldom available.
2.	Target organ specificity of the RfD must be evaluated.
RfDs are based on the "critical effect", i.e., in a series of
studies, the effect observed at the lowest dose. The organ
system affected by the critical effect is the "critical" target
organ. At higher doses other effects may be observed, or
different target organs may be affected.
In general, only RfDs based on the same critical target
organ should be combined in the Hazard Index calculation. Dose
addition combining RfDs that are based on different critical
target organs, may overestimate the true mixture toxicity.
3.	The species basis of RfDs varies. A Hazard Index using
RfDs based from different species introduces errors of unknown
magnitude.

-------
Information on the critical target organ (and some
indication of target organs affected at higher doses) may be
obtained from EPA's IRIS data base, or from toxicology references
Judicious consideration of such dose-response information may
enable the application of dose addition even when the data
providing the best justification (data in the same species and
target organ, toxicants acting by the same mechanism) are not
available.
The following examples illustrate the problems discussed
above, and their pragmatic resolution. Example 1 illustrates a
Hazard Index that is fairly well justified by the available
toxicity data. Example 2 illustrates Hazard Index estimates that
might be judged too uncertain to be used.
EXAMPLE I
GOOD JUSTIFICATION FOR SIMILAR TOXIC ACTION ASSUMPTION
Component	Crit. Target Species RfD Exposure Exposure
Organ	Dose	RfD
Chemical 1
Chemical 2
Chemical 3
blood
blood
blood
rat
rat
rat
5 90
1 35
0.1 10
18
35
100
Hazard Index:153 (=200)
EXAMPLE 2
POOR JUSTIFICATION FOR SIMILAR TOXIC ACTION ASSUMPTION
Component	Crit. Target Species
Organ
Chemical 1	liver	human
Chemical 2 blood*	rat
Chemical 3 blood	monkey
* The liver is affected at slightly higher doses. Therefore
one might be justified in estimating a hazard index for chemicals
1 and 2. However, such an index would be an uncertain estimate,
since it involves two disparate species.
RESPONSE ADDITION
In response addition, the component risks (response rates)
are summed, and the probabilities of simultaneous risks are

-------
subtracted from this summation. This is illustrated in the
following example, using cancer as the response:
Chemical
Risk
1
2
Mixture risk =
0.001
0.003
[Probability of
[Probability of
[Probability of
0.001
0.004
+ 0.003 -
cancer due
cancer due
cancer due
(0 - 001)*(0.
to chemical 1] +
to chemical 2] -
to both chemicals]
003) = 0.003997
This example shows that if the true case is independence of
toxic action, response addition is sufficiently accurate even at
the fairly high cancer risks of 0.001 and 0.003. There are
reasonable theoretical argumentsa supporting the judgment that
even if synergism is observed in a bioassay (i.e., at relatively
high doses), the response addition risk estimate for the (much
lower) ambient exposures would not be significantly increased by
interaction terms. Although these arguments do not consider
synergism in biological factors such as pharmacokinetics or
physiological transport, they support the use of simple response
addition at low doses.
In the case of systemic toxicants, considerations arguing
against detectable synergism at low doses may not be justified
since there are data^'c showing synergism at component doses
which individually are "no-effect" levels.
aThorslund, T.W. and G. Chamley. 1986 Use of the multistage
model to predict the carcinogenic response associated with time-
dependent exposures to multiple agents. ASA/EPA Conference on
interpretation of environmental data: Current assessment of
combined toxicant effects. Washington, DC., May 5-6.
^Charbonneau, M., et al. 1986. Acetone potentiation of rat
liver injury induced by trichloroethylene-carbon tetrachloride
mixtures. Fund. Appl. Toxicol. 6 654-661.
cEastmond, D.A., et al. 1987. An interaction of benzene
metabolites reproduces the myelotoxicity observed with benzene
exposure. Toxicol. Appl. Pharmacol. 91:85-96.

-------
APPENDIX B
Potencies of Dioxins Relative to 2,3.7,8-TCDD
Enjyma Induction
Chemical
Guinea	Reproductive'
pig Carcino• teratogenic Receptor
LDyt geniciry effects	binding
AHH
£R00
Animal Human
cells cells
Flat
IXBI	Immuno-
Cell cell toxicity
keratin. assay in vitro
CCDs:
Mcno thru tri
2378-TCDO
TCDOs
2378-PeCDO
PeCDOs
2378-HxCOOs
HxCDDs
2378-HpCDDs
HpCDOs
OCDD
CDFs:
Mono thru tri
2378-TCDF
< io~*
1
<.001
.67
.002
.03
.004
.002
.28: .5
.001-.01 <.001
1	1	1
<.001	<.01-. 16	<.001-.02
1	.02-.2
—	<.001
.04	.01	.OS	.001-.1
<.001
—	—	.002-004
<.001
<.00001	— <.001
£.001.02 <.001
.03-. 13 .3 ; J4 ; .4 .01-.4
<.001
.4
.01
1
<.001-.01
.5
.005 —
.»
.001 —
.05 .1
.005
1
.1 . 1
rcofs
2378-PeCDF
12467-PeCDF
PeCDFs
2378-HxCDFs
HxCOFs
2378-HpCDFs
HpCDFs
.017
.001-.05 3.001 ;.04
.13:.7:.6 <.3:.4
.IS
.001-. 1
.04-.S
.001
—	<.001
.002
s.001-2
.0S-.Z
.001 :.002
.004
<.001
.4
.8
.6
.9
s.OOS
.1
<.001
S.001
.1-.S
.006

-------
Memo to Facilitator Trainees Attending the Mixtures Session
During	the discussion of applications of the Mixture
Guidelines,	you may wish to use examples from your own
experiences	in risk assessment. Please consider the following
during your	preparation for this discussion:
1. Describe a situation where you had to evaluate
(qualitatively or quantitatively) the health risk from an
existing mixture. How did you judge the adequacy of the
toxicology and exposure data? How did you present your findings?
2. If you were to perform a risk assessment of a
mixture now, what would you change from your previous procedures?
Which of the changes would make more use of the Agency Mixture
Guidelines? Do the Guidelines assist you in describing the
uncertainties in the mixture risk assessment?

-------
CASE STUDY
Scenari o

-------
Data Tables and Summary (or both)

-------
Tasks for Trainees

-------
Guidance for Facilitators

-------