The Toxicologist - Society of Toxicology

sitizers were 100% concordant among the repeat LLNA: DA tests. ICCVAM concludes
that accuracy and reproducibility of the LLNA: DA support its use to identify
potential skin sensitizers and nonsensitizers. ICCVAM recommends SI ≥ 1.8 to
identify ACD hazards since there were no false negatives relative to the LLNA. In
testing situations where dose-response information is not required, or negative results
are anticipated, ICCVAM recommends that the single-dose reduced LLNA:
DA should be considered and used, thereby reducing animal use by up to 40%. The
ICCVAM-recommended protocol formed the basis for the recently adopted
OECD Test Guideline 442A. Because the LLNA: DA does not require radioactive
reagents more institutions can take advantage of the reduction and refinement benefits
afforded by the LLNA compared to traditional guinea pig methods for ACD
testing. The LLNA: DA will also eliminate the environmental hazard associated
with use and disposal of radioactive materials used in the LLNA.
1521 USING FEWER ANIMALS TO IDENTIFY CHEMICAL
EYE HAZARDS: REVISED CLASSIFICATION CRITERIA
NECESSARY TO MAINTAIN EQUIVALENT HAZARD
LABELING.
W. Stokes 1 , J. Haseman 2 , E. Lipscomb 3 , J. Truax 3 , N. Johnson 3 , B. Jones 3 and
D. Allen 3 . 1 NICEATM, NIEHS, Research Triangle Park, NC, 2 Consultant, Raleigh,
NC and 3 ILS, Inc., Research Triangle Park, NC .
U.S. Federal Hazardous Substances Act (FHSA) regulations specify eye safety testing
procedures and hazard classification criteria for chemicals and products regulated
by CPSC and OSHA. Current regulations require up to three sequential tests
of six animals per test, with decisions on the need for subsequent tests based on the
number of positive responses observed. Testing conducted in accordance with the
OECD test guideline for eye irritation and corrosion can also be used, but current
FHSA regulations do not provide criteria to classify results from this 3-animal test.
Therefore, an analysis was conducted to determine classification criteria for results
from a 3-animal test that would provide hazard labeling equivalent to that provided
by current FHSA regulations. This analysis compared the frequency at which the
current FHSA classification criteria would identify substances as ocular irritants
with the frequency at which a classification criterion of either one or two positive
animals out of three would identify these substances. The resulting classifications
that would be assigned by each of the three criteria were also compared using four
different underlying response rates (20, 40, 50, and 75%). For these response rates,
current FHSA requirements would identify 20, 73, 88, and >99% of substances as
irritants, respectively, while using a criterion of at least one out of three positive animals
in a 3-animal test would identify 49, 78, 88, and 98% of substances as irritants,
respectively. In contrast, using a criterion of at least two out of three positive
animals in a 3-animal test would identify far fewer irritants, with detection rates of
10, 35, 50, and 84%, respectively. We conclude that using classification criteria of
one or more positive animals in a 3-animal test will provide the same or greater level
of eye hazard labeling as current FHSA requirements, while using up to 83% fewer
animals. ILS Staff supported by NIEHS contract N01-ES-35504.
1522 FROM HAZARD-BASED TO TISSUE DOSE-BASED
TESTING—REDUCTION AND REFINEMENT.
J. Bessems. National Institute for Public Health and the Environment (RIVM),
Bilthoven, Netherlands.
Research into apparently clashing paradigms in current chemical safety testing led
to unorthodox opportunities to replace, reduce and/or refine obligatory animal
testing. In order to prevent human exposure to hazardous levels of chemicals, hazardous
levels in animals have to be established. Effects assessment to establish a
point of departure comes upfront in risk assessment and is targeted at finding a hazard.
Without a LOAEL, a NOAEL can not be established. Unfortunately, replacement
of bioassays for systemic effects is a long-lasting challenge. Focusing on reduction
and refinement might be more successful. In risk assessment, human
exposure is measured, estimated or predicted completely independent from toxicity
testing. Finally, the point of departure and exposure are compared, either by taking
into account various assessment factors or via a margin of safety approach. This
contribution provides an outline how increased focus on exposure assessment including
assessment of target tissue dose by taking toxicokinetics into account can
result in reduction and refinement using tissue dose-informed testing. Predicting
tissue doses (using physiologically-based modeling) and establishing in vitro effect
concentrations should come upfront. Comparing resulting figures may lead to significant
reductions by waiving obligatory bioassays (tissue dose far below in vitro
effect concentrations). It would leave only the need for some confirmatory testing
at much lower tissue dose levels which will cause significantly less animal stress (refinement).
Finally, reductions by focused testing can be achieved in more tradi-
tional paradigms by upfront in silico, in vitro and/or in vivo (screening) studies
with respect to absorption, distribution, metabolism and excretion. Taking apparent
specificities across species, across doses/exposures, across age and sex and across
routes of exposure into account in planning effects bioassays will prevent testing in
the wrong species, irrelevantly high doses, wrong age, sex, or exposure route.
1523 PRACTICAL IMPLEMENTATION OF THE THRESHOLD
OF CONCERN AND NOAEL-TO-LC50 RATIO FACTOR
APPROACH TO DETERMINE ACUTE EFFECTS
SCREENING LEVELS.
T. D. Phillips, S. E. Ethridge and R. L. Grant. Toxicology Division, Texas
Commission on Environmental Quality, Austin, TX.
Effects Screening Levels (ESLs) are chemical-specific air concentrations set to protect
human health and welfare and are used in the Texas air permitting process to
assess the protectiveness of emission rate limits. Short-term ESLs are developed to
evaluate acute intermittent exposures of 1 hour (h). The Texas Health and Safety
Code requires the development of ESLs for as many air contaminants as possible,
even chemicals with limited toxicity data (LTD). Grant et al. 2007 used a statistical
analysis of a compiled database of LC50 and NOAEL inhalation data to establish
two different methods for determining acute generic ESLs for LTD chemicals:
Threshold of Concern (TOC) and NOAEL-to-LC50 (N-L) Ratio. In the TOC approach,
acute inhalation lethality data are used to classify chemicals in different
acute toxicity categories based on the GHS of Classification and Labeling of
Chemicals proposed by the UN. Different health-protective concentrations are
used for chemicals assigned to different GHS categories. For the N-L ratio approach,
a 4h duration LC50 value is multiplied by a factor of 8.3E-5 to estimate
health-protective air concentrations. These generic ESLs are used until adequate
toxicity data are available. The Texas Commission on Environmental Quality
(TCEQ) incorporated these approaches into the peer-reviewed Guidelines to
Develop Effects Screening Levels, Reference Values, and Unit Risk Factors in 2006.
TCEQ has implemented these approaches to determine generic ESLs for pentene
isomers and hexane. These approaches have also been used to ground truth acute
ESLs for the following chemicals derived when multiple-day studies were used (i.e.,
acceptable toxicity studies less than 24h were not available): 1,3-butadiene, 1butene,
2-butene, isobutene, TexanolTM, toluene, tetrachloroethylene, ethylbenzene,
and benzene. Generally, the TOC method was slightly more conservative than
the N-L ratio approach. A weight-of-evidence approach is used to determine which
method is more scientifically defensible.
1524 REPEATED-DOSE PRECLINICAL TOXICITY STUDIES:
HIGHEST DOSE TESTED=NOAEL—WHAT TO DO?
R. Garg, W. Bracken, D. Bury and K. Marsh. Abbott Laboratories, Abbott Park, IL.
Absence of Dose-Limiting Toxicity (DLT) in a repeat-dose toxicity study may raise
questions regarding validity of the study, e.g. a) evidence if the highest dose was the
maximum feasible dose or one that produces maximum feasible exposure, b) availability
of data on alternative formulations or dosing regimens attempted to improve
drug exposure and c) relation of achieved drug exposure at NOAEL to the predicted
clinically efficacious exposure. A pre-IND stage Abbott compound (ABT-
000) was generally well tolerated in animal toxicity studies. The highest dose in
both the 4-week rat and dog toxicity studies was concluded as the NOAEL (180
and 50 mg/kg/day for rat and dog, respectively). In anticipation of questions from
the regulatory agency and given the limited solubility and bioavailability of the
drug substance, several preclinical formulations were evaluated to optimize the systemic
exposure to ABT-000. From the test formulations, a vehicle containing
Cremophore EL and vinylpyrrolidone dimer (VP-dimer) at 80:20 ratio was selected
since it produced the highest systemic exposure to the drug. In both, rat and dog,
the systemic exposure for ABT-000 was not dose proportional and declined at the
highest dose tested. In view of this observation, the dose with highest exposure was
selected for the rat and dog toxicity studies. As VP-dimer is not a common vehicle,
a 4-week toxicity study on this vehicle component was also performed and included
in the submission. As adverse effects related to VP-dimer could not be excluded, a
saline control group was added to both the rat and dog toxicity studies. Given that
the systemic exposure at the NOAEL was 83 to 87-fold higher than the predicted
human efficacious exposure and supporting formulation data was available, the
studies were accepted for the IND. In conclusion, even when the highest dose in a
preclinical toxicity study reveals to be a NOAEL, this can be acceptable to a regulatory
agency, when appropriate supporting data are provided and a good safety margin
is evidenced
SOT 2011 ANNUAL MEETING 327