United States
Environmental Protection
Agency
Environmental Criteria and
Assessment Office
Cincinnati OH 45268
Research and Development
EPA/600/S8-86/033 Mar. 1987
&EPA Project Summary
Human Variability in
Susceptibility to
Toxic Chemicals -
I. Noncarcinogens
This report summarizes an initial effort
to assemble human data of the variability
in key pharmacokinetic parameters in a
form that can be applied to the quantita-
tive assessment of human risk from toxic
chemicals. From the standpoint of the risk
assessor, differences among individuals
mean that population risk vs. dose will be
systematically different from the risk that
would be faced at different doses by any
individual within the population.
This Project Summary was developed
by EPA's Environmental Criteria and As-
sessment Office, Cincinnati, OH, to an-
nounce key findings of the research pro-
ject that is fully documented in a separate
report (see Project Report ordering infor-
mation at back}.
Discussion
Papers from the recent literature (most
from 1979-1985) were selected for pos-
sible study if they contained individually
distinguishable human data for at least
five people on parameters related to
susceptibility to toxicants. The selected
papers were first sorted according to the
kind of parameters measured. Papers con-
taining data on pharmacokinetic
parameters in normal healthy adult popula-
tions were used for the present analysis.
This restriction implicitly has limited the
type of variability captured in two ways:
(1) "exposure" and "response" type varia-
bility are excluded: and (2) most variability
that comes from age, pathology, or other
special situations (such as pregnancy, in-
teracting exposures) is also excluded.
However, the selection criteria used do
allow the data to be readily compared with
observations of interindividual variability
in rats as observed in acute oral LD50 ex-
periments described by Weil (1972), and
with lethal toxicity probit risk equations as
developed for the U.S. Coast Guard's
"Vulnerability Model" (Eisenberg and
Lynch, 1975). The Weil data have recently
been used as support for the traditional
10-fold safety factor for interindividual
variability that is used in the computation
of "Acceptable Daily Intake" values
(Dourson and Stara, 1983).
As presumed by both the "Vulnerability
Model" probit risk equations and the
Dourson-Stara/Weil analysis, the basic
forms of the observed distributions of
pharmacokinetic parameters related to in-
dividual susceptibility appear to be better
described as log-normal than as normal
distributions. Fully 72 of the 101 data sets
examined here were found to have a posi-
tive value for skewness when the data
were not logarithmically transformed.
However analysis of kurtosis statistics for
the distributions after logarithmic trans-
formation suggested that the data sets
may often be flatter in the middle and
heavier in the tails than would be expected
for truly log-normal distributions. One pos-
sible explanation for this is that there may
often be some degree of bi- or multimod-
ality in the data.
Elimination half-lives and maximal blood
concentrations both were found to have
median values of Iog10 geometric stand-
ard deviations somewhat over 0.11*, not
"In the antilog form that is often used to express
geometric standard deviations, a log,0 geometric
standard deviation of 0.11 would correspond to about
1.29 (because 100'1 = 1.29).
-------
very different from the total geometric
standard deviation of susceptibility of
0.0934 - 0.114 assumed for chlorine and
ammonia in the Coast Guard's "Vulnera-
bility Model" (Eisenberg and Lynch, 1975).
There seems to be some tendency toward
greater amounts of interindividual varia-
bility in data sets where AUC (Area Under
a Curve of blood concentration by time)
or steady-state blood concentrations were
the measured parameters. In this case the
median Iog10 geometric standard devia-
tion is about 0.145 and three of the 20
data sets are over 0.3 (compared with
none of 61 data sets over 0.3 in the case
of T1/2). Of those parameters studied,
AUC has the greatest relevance for repre-
senting interindividual pharmacokinetic
differences for effects with linear non-
threshold dose response relationships.
Peak blood concentrations are more di-
rectly relevant for acute toxicity.
A similar result was obtained when
comparisons were made between the
human pharmacokinetic data and the
results in rats reported by Weil (1982). The
interindividual variability in specific phar-
macodynamic parameters for the 49
agents studied in groups of normal healthy
adult humans does not depart markedly
from the distribution of total variability in
susceptibility to acute lethal effects of
chemicals in groups of rats (Weil, 1972).
Where Weil (1972) found that about 14%
of the rat data sets had Iog10 geometric
standard deviations of over 0.25, the pro-
portion of chemicals showing this degree
of interindividual variability in specific
pharmacokinetic parameters was approxi-
mately 5% for elimination half-lives
(2/44), 20% for AUC measurements
(3/15), and 8% for peak blood concentra-
tions (1/12). One of four observations in a
miscellaneous set of "other" parameters
was also above this level of indicated in-
terindividual variability.
It should be stressed that the param-
eters studies in humans, however, are only
components of overall susceptibility to
toxic agents and do not include contribu-
tions from variability in exposure- and
response-determining parameters. The
data also include no contribution to
variability from diseased or other special
subpopulations, and it is these types of
sensitive subpopulations that have often
been of major concern for standard set-
ting under environmental laws such as the
Clean Air Act.
The numbers of chemicals and human
parameters summarized to date are clearly
small, both relative to the available human
literature and to the animal data set com-
piled by Weil (1972). Based on the limited
survey of the literature undertaken in
Phase I, it seems likely that the overall data
base assembled here could be expanded
3-fold to 6-fold with a relatively straight-
forward extension of the present work.
Future work should also clearly include
studies of "exposure" and "response"
parameters, and studies in patient and
various special subpopulations likely to
depart significantly from average suscep-
tibility among normal healthy adults. Fur-
ther, it is important to document whether
instances of relatively greater apparent
variability are associated with different
structural categories of chemical agents,
different modes of administration, dif-
ferent metabolic pathways, age-restricted
vs. not age-restricted population samples,
and other study characteristics. Finally, the
lessons from these different analyses
should be integrated into predictive mod-
els of human interindividual variability as
an aid for quantitative human risk assess-
ment for toxic chemical exposures.
References
Dourson, M.L. and J.F. Stara. 1983.
Regulatory history and experimental sup-
port of uncertainty (safety) factors. Regu-
latory Toxicol. Pharmacol. 3: 224-238.
Eisenberg, N.A. and C.J. Lynch. 1975.
Vulnerability model. A simulation system
for assessing damage resulting from ma-
rine spills. NTIS AD-A015 245.
Weil, C.S. 1972. Statistics vs. safety
'factors and scientific judgement in the
evaluation of safety for man. Toxicol. Appl.
Pharmacol. 21: 454-463.
The Project Summary was prepared by staff of the Environmental Criteria and
Assessment Office, Cincinnati, OH 45268.
Linda Erdreich is the EPA Project Officer (see below).
The complete report, entitled "Human Variability in Susceptibility to Toxic
Chemicals—I. Noncarcinogens," (OrderNo. PB87-101 242/AS; Cost: $11.95.
subject to change} will be available only from:
National Tbfprnical Information Service
5285 PortlPoyal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Criteria and Assessment Office
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
r-r>,. U.8,GFF;CiAUMk,H
Official Business
Penalty for Private 'Jse $300
EPA/600/S8-86/033
*'!*•. ^ *&>* i«. ^
0000329 PS
------- |