united states
             Environmental Protection
             Agency
             Office of Toxic Substances
             Washington, DC 20460
EPA 560/13-79-009
July, 1979
             Toxic Substances
oEPA
Acute Toxicity
Testing Criteria for
New Chemical Substances

-------
                            DISCLAIMER
This report has been reviewed  by  the  Office  of  Toxic  Substances, EPA,
and approved for publication.  Approval  does not  signify that the
contents necessarily reflect  the  views  and  policies of  the
Environmental Protection  Agency,  nor  does mention of  trade  names or
commercial products constitute endorsement  or recommendation for use.

-------
                                         EPA-560/13-79-009
                                         July,  1979
     ACUTE TOXICITY TESTING CRITERIA
         FOR CHEMICAL SUBSTANCES
         Contract No.  W 2227-NA5X
Prepared for EPA by:  Mary Janet Normandy
                      Robert Reynolds
   EPA Project Officer:   Daphne Kamely
     ENVIRONMENTAL PROTECTION AGENCY
     PREMANUFACTURING REVIEW DIVISION
        OFFICE OF TOXIC SUBSTANCES
   U.S. ENVIRONMENTAL PROTECTION AGENCY
         WASHINGTON, D.C.  20460

-------
                                 ABSTRACT
This report addresses the rationale, considerations, and limitations of
acute toxicity testing.  General procedures are described for acute tests
including lethality studies in oral, dermal, and inhalation toxicity, and
irritation studies in dermal and eye toxicity, phototoxicity and skin
sensitization.  Recommendations are given for the acute toxicity tests
which may be used to evaluate the risks associated with the manufacture
and processing of chemical substances.  Suggested minimum protocols are
offered including the choice of test animals, dosage  levels, laboratory
practices, animal diets, and necropsy requirements.  A method is sug-
gested for interpreting the results of these tests and extrapolating to
some guidelines for a safety factor in human exposure.

This report was submitted in fulfillment of contract W2227-NA5X by Enviro
Control, Inc. under the sponsorship of the U.S. Environmental Protection
Agency, and was prepared from April to July, 1979.

-------
                                 FORWARD

This study was conducted for the Premanufacturing Review Division, Office
of Toxic Substances, Environmental Protection Agency, to collect data  and
develop procedures for the assessment of toxicity as an integral part  of
the risk assessment scheme for new chemical substances.  This  publica-
tion, which focuses on acute toxicity assessment, was directed primarily
toward the initial screening operation of  chemical  substances.

While studies  on  acute toxicity  have appeared  in the literature, this  is
the first report  useful  for the  evaluation of  the risks posed  by new
chemical  substances to human  health  and  the environment.   Although this
study focuses  on  new  chemical  substances,  it  is  important  to  note that
the procedures developed herein  may  have direct  application  to all areas
of  environmental  health  analysis.

This  document  was published  with the expert advice  of  Dr.  Peter Voytek,
Office  of Toxic Substances.   Further information can be obtained from  the
Premanufacturing  Review  Division;  telephone:  202-426-2601; address:
Office  of Toxic Substances  (TS-794), U.S.  Environmental Protection
Agency,  401  M  Street,  SW,  Washington,  D.C.  20460.
                                        Daphne Kamely,  Ph.D.
                                        Premanufacturing Review Division
                                    m

-------
                                 CONTENTS
                                                                    Page
Abstract	 ii
Foreward	 i i i
Section
  1       Introduction	   1
  2       Types of Acute Toxicity Tests
               Acute Oral Toxicity	   6
               Acute Dermal Toxicity	   7
               Acute Inhalation Toxicity	   8
               Acute Eye Irritation Toxicity	   9
               Dermal Irritation Toxicity	  14
               Dermal Sensitization Toxicity	  17
               Phototoxicity	  19
               Oral Irritation Toxicity and
                 Corrosive Effects	  21
  3       Acute Toxicity Testing Criteria
               When Should Acute Toxicity Testing
                 Be Performed?	  22
               What Acute Toxicity Studies Should
                 Be Conducted?	  23
               What are  the Minimal Protocols for
                 the Acute Studies?	  24
               How Should the  Results  be Interpreted
                 and Used to Estimate  a Tolerable  Level?	  31
               How Should the  Data be  Submitted  For
                 Easy Interpretation by Reviewers?	  34
References	   35
                                     iv

-------
                                  Tables
1.   Comparisons of Oral and Dermal L^QS within the
      Same Animal Species 	 2

2.   Grades for Ocular Lesions 	12

3.   Classification of Test Substances 	13

4.   Erythema and Edema Scoring 	15

5.   National Institute for Occupational Safety and Health
      Interpretation of Skin Test Ratings 	16

6.   NIH-7 Open Formula Rat and Mouse Ration 	27

7.   Acute Study Price Test 	29

8.   Prediction of Human TLEs from Oral LD50s Derived From Animal
      Studi es  	33

-------
                                SECTION 1

                               INTRODUCTION


    This report includes various suggestions for the collection and
utilization of acute toxicity data in the screening of chemical
substances.

    Acute toxicity studies include lethality, irritation, sensitization
and corrosion testing.  Acute lethality protocols usually involve expos-
ing an experimental animal to a single dose of a test substance for 24
hours or less.  The animal is then observed for up to 14 days after
exposure (1).  The results may be expressed as a simple mortality res-
ponse (death or no death) over a broad range of dose levels which are
reported as mg test substance per kg body weight of experimental animal.
These dose levels and associated mortality responses are then used to
extrapolate the LD50  (lethal dose for 50% of a group of test animals)
for the test substance.  Specific toxic responses can also be measured
and used to determine another useful statistic—the TD5Q.  This is the
dose at which 50% of  all test subjects develop a given toxic response
such as tremors, erythema, convulsions, loss of consciousness, or
malfunction of a specific organ system.

    Acute  irritation, sensitization, and corrosion tests are used to as-
sess a chemical's potential for injury to specific tissue, usually the
eye or skin.  Interpretation of test results is more complex than for
lethality testing and may include complex systems for grading toxic
responses.  Because death is not expected to result from these studies,
LD^Q'S are not calculated.

    It would be of obvious benefit to be able to use the results derived
from one type of acute test to predict the level of toxicity that would
be seen for that same compound using the same animal model in  another
type of acute test.   A major impediment to the feasibility of  such corre-
lations is that the toxicity data most commonly available for  any given
compound are usually  collected for two different animal  species - an oral
LD5Q in the rat and a dermal 1059 in the rabbit.  However if both
oral and dermal LDcgs are available in the same animal species, even
when the data has been collected by two different laboratories, there  are
some potentially useful generalizations to be made.  Table 1 shows the
oral and dermal 10595 where both are available for the same animal
species for 51 compounds using 4 animal species in 57 tests (2).  Dermal
LD5ps are higher than, or at least equal to, oral 10595  in about 88%
of the cases.  In every case but one, hexabutyl-distannoxane tested in
the rat, the oral and dermal LD50s are within one order  of magnitude of
each other.  Data such as these suggest that substances  which  show
toxicity upon dermal  application will probably be as toxic or  more toxic
when administered orally.  This is generally attributed  to the more rapid
systemic absorption of ingested substances.

-------
                                 TABLE 1
                    COMPARISONS OF ORAL AND DERMAL LD5QS
                       WITHIN THE SAME ANIMAL SPECIES
 CHEMICAL

 ACROLEIN
 ACRYLONITRILE
 AMMONIUM, (2-CHLOROETHYL)
  TRIMETHYL-CHLORIDE
 ANILINE
 p-CHLORO-ANILINE
 NITROBENZENE
 BENZENETHIOL
 4,4' - BIPYRIDINIUM, 1,V-
 DIMETHYL-DICHLORIDE
 2-ETHYLBUTANOL
 t-BUTYLHYDROPEROXIDE
 CARBAMIC ACID METHYL -,
 1-NAPHTHYL ESTER
 CARBON TETRACHLORIDE
m-CRESOL
 o-CRESOL
 p-CRESOL
4,6 - DINITRO-o-CRESOL
 p-DIOXANE
HEXABUTYL - DISTANNOXANE
4-METHYL -6- PYRIMIDINYL
 PHOSPHOROTHIOATE
 1,2 - DIBROMO - ETHANE

2-BUTOXY-ETHANOL
2-(2-BUTOXY-ETHOXY) ETHANOL
 ACETATE
2-CHLORO-ETHANOL
2-ETHOXY-ETHANOL
ORAL LD50 DERMAL LD5Q
SPECIES mg/kg mg/kg
RABBIT
RABBIT
RABBIT
RAT
RAT
RAT
RAT
RAT
RABBIT
RAT
RABBIT
RAT
RAT
RAT
RAT
RAT
RABBIT
RAT
RAT
RAT
RABBIT
GUINEA PIG
RABBIT
RAT
RABBIT
7
93
150
440
420
640
46
57
1200
406
710
2800
242
121
207
10
2000
87
2000
108
55
1200
2600
58
3100
562
250
232
1400
3200
2100
300
80
1260
790
2000
5070
620
1100
750
200
7600
11,700
8000
300
300
230
15,000
84
3500
DERMAL LDso
ORAL LD50 '
80.29
2.69
1.55
3.18
7.62
3.28
6.52
1.40
1.05
1.95
2.82
1.81
2.56
9.09
3.62
20.00
3.80
134.48
4.00
2.78
5.45
0.19
5.77
1.45
1.13

-------
                          TABLE 1 (CONTINUED)
CHEMICAL
SPECIES
ORAL LD50
  mg/kg
 DERMAL   DERMAL
 mg/kg     ORAL  LDso
2-(2-ETHOXYETHOXY) -
 ETHANOL                      RAT       6500
Z'-HYDROXY - 2,4,4' -
 TRICHLORODIPHENYL ETHER
N,N - DIMETHYL - FORMAMIDE
N '-(4-CHLORO-o-TOLYL)-N,N-
 DIMETHYL-FORMAMIDINE
FORMIC ACID ETHYL ESTER
METHYL-HEPTANETHIOL

2-ETHYL-l,3-HEXANEDIOL

4-METHYL-3-PENTEN-2-ONE
PHENETHYL ALCOHOL

PHENOL
2-sec-BUTYL-4,6-
 DINITROPHENOL                RAT         25
4,4' - ISOPROPYLIDENEDI-
 PHENOL

PENTACHLORO-PHENOL
PHOSPHONIC ACID (2,2,2-
 TRICHLORO-1-HYDROXYETHYL)-
 DIMETHYL ESTER
PHOSPHORIC ACID 1,2-DIBROMO-
 2,2-DICHLORO-ETHYL DIMETHYL
 ESTER                        RAT        250

S-((5-METHOXY-2-OXO-l,3,4-
 THIADIAZOL-3 (2H)-YL)        RAT         25
 METHYL) 0,0-METHYL PHOS-     RABBIT      63
 PHORO-DITHIOATE

PHOSPHORODITHIOIC ACID,       RAT         25
 S-((2-ETHYLTHIO)ETHYL)
 0,0-DIMETHYL ESTER
PHOSPHOROTHIOIC ACID, 0,0-    RAT         76
 DIETHYL 0-(2-ISOPROPYL-6-
 METHYL-4-PYRIMIDINYL ESTER)

PHOSPHOROTHIOIC ACID, 0,0-    RAT       1000
 DIMETHYL ESTER, 0,0-DIESTER
 WITH 4,4'-THIODIPHENOL
 6000
                          0.93
RAT
RAT
RAT
MOUSE
RABBIT
RABBIT
RAT
RABBIT
RABBIT
GUINEA
RAT
3700
2800
170
160
625
2075
85
2600
1000
PIG 400
414
9300
5000
4000
225
640
20,000
1954
2000
5990
5000
669
2.51
1.79
23.53
1.41
1.02
9.64
22.99
0.77
5.99
12.5
1.62
  80
 800

  20
 375

 179


 455


1370
                          3.20
RABBIT
RAT
RAT
RABBIT
2230
50
450
1450
3000
105
2000
5000
1.35
2.10
4.44
3.45
                          3.20

                          0.80
                          5.95

                          7.16
                          5.99
                          1.37

-------
                           TABLE 1 (CONTINUED)
CHEMICAL

PHTHALIC ACID, BIS (2-ETHYL-
 HEXYL) ESTER
POLYPROPYLENE GLYCOL MONO-
 BUTYL ETHER
1,2-DIBROMO-3-CHLORO-PROPANE
1,2-EPOXY-PROPANE
TOXAPHENE
2-CHLORO-4-ETHYLAMINO-
 6-ISOPROPYLAMINO-S-TRIAZINE
4-UNDECANOL, 7- ETHYL-2-
 METHYL-,HYDROGEN SULFATE,
 SODIUM SALT
2,6 - XYLENOL
SPECIES
RABBIT
ORAL LD5Q
  mg/kg
  DERMAL
  mg/kg
   34
RABBIT   23,900
RABBIT      180
GUINEA PIG  690
RAT          40
RABBIT
  750
GUINEA PIG  650
MOUSE
RABBIT
  980
  700
    25

21,000
  1400
  8640
   600

  7500

   650
   920
  1000
DERMAL LD5Q/
  ORALLDsn
    0.74

    0.88
    7.78
   12.52
   15.00

   10.00

    1.00
    0.94
    1.43

-------
Acute toxicity testing is valuable  in risk assessment  in  these ways:

•     as a screening procedure to help  identify substances  of such
      low toxicity that extensive further acute testing is  not
      justified;

•     as a method to indicate specific  toxic effects of a substance
      which might be associated with either a  single massive exposure
      or with frequent use  at a high level of  exposure.

t     as a range finder for studies by  producing  data  concerning
      lethal  levels, signs  of intoxication, and possible  target organ
      effects.

•     as an aid  in the design of  appropriate clinical  management  pro-
      grams for  individuals involved in the misuse  of  or  an accident
      with a  chemical substance.

•     as a means to  develop an  index of the relative hazard of acute
      exposure to various substances by comparing LDgQS and the
      slopes  of  their dose-response curves.

-------
                                SECTION 2

                      TYPES OF ACUTE TOXICITY TESTS
ACUTE ORAL TOXICITY
    Acute oral toxicity studies  are designed  to  investigate  the  qualita-
tive and quantitative nature of  the toxic  effects  from  a  single  oral  ex-
posure to a large dose of the  chemical  substance in  question.  The doses
are selected  to provide data sufficient to estimate  the LD50 and to
determine the  slope of the  dose-response curve.   The numbers of  test
animals per dosage level should  be  sufficient to make these  statistical
analyses according to a particular method  of  evaluation.   The test ani-
mals should be characterized as  to  species, strain,  and physiological and
morphological  characteristics.   There  is no standardized  animal  that  is
suitable for  all  acute oral tests.  It  is  extremely  important that the
test animals  be randomly selected for  the  dose level groups.

    The test  substance, if  it  is not  a liquid,  should be  administered in
an  appropriate carrier which facilitates absorption.

    It  is essential to note carefully  all  toxic  signs including  type,
time of onset, severity, and duration.   The time at  which deaths occur or
signs  of toxicity appear is important,  particularly  if  there is  any
tendency for  deaths to be delayed.  Observations of  the animal should
continue until signs  of toxicity are  absent  in surviving animals.  A
14-day  observation period is sufficient for most compounds.   At  that  time
survivors should  be sacrificed and  submitted  to  a gross pathologic
examination.

     Ideally,  to  assess potential health hazards  to humans, toxicity
studies should be conducted only in those  species of animals whose meta-
bolism  of the test  substance  is  similar to that  of humans.  Since com-
parative metabolism  is unknown for  most new substances, and  since studies
 in  humans  are rarely  feasible  at this  stage,  rodents make good initial
test  subjects.   While extrapolation of the results to humans may not
always  be  valid,  the  correlation is reasonably good  for single oral
doses.  Acute toxicity testing can  be  conducted in several species if
necessary.   A similar degree  of toxicity in several  species  would suggest
that  toxicity to  humans might  be comparable;  marked  variation in the res-
ponses  of  different  species necessitates the  assumption that humans  are
 at  least  as  sensitive as the most  sensitive species  studied.

     It  is  suggested  that these tests  be performed on two species of
animal-the  albino rat, and  the albino mouse.   All female animals should
be  nulliparous.   The  minimum  sample number and dosage schedule should be:

    -  5 male  and  5 female  animals  of  each  test species at each of five
    dose  levels  - preferably  to include 2  levels above and 2 below the

-------
    expected LD5n.  A suggested dosage schedule is: 5000 mg/kg, 500
    mg/kg, 50 mg/kg, and 5mg/kg.

ACUTE DERMAL TOXICITY

    The ability of  some chemicals  to penetrate either  intact  or abraded
skin and produce  systemic toxicity has been well established.  A  test  for
acute dermal toxicity should evaluate the  potential for  systemic  toxic
effects of chemicals expected to come in contact with  the  skin.   This  is
done by determining the LD5g of a  single dermal exposure to the com-
pound in question by the animal test species.  Since this  LD50 will be
used in hazard evaluation, the test conditions should  be related  to the
anticipated  human exposure, if at  all possible.

    The albino rabbit is the animal most frequently used in assessing
dermal toxicity and the most commonly used procedure is  the method of
Draize.(S)   It is also suggested that other animal species, such  as mice,
rats, guinea pigs,  or dogs, be  used in  addition to rabbits.   The  rabbit
appears to be exquisitely sensitive to  dermal  insult,  and  thus the eli-
cited reactions may not be valid for humans.   The  skin of  guinea  pigs  has
permeability characteristics more  like  those  of humans.  The  albino rat
is  somewhat  less  reactive than  the rabbit, and more reactive  than the
guinea pig or humans.  The rat  is  a preferred  species  for  dermal
lethality testing because it  is  the model  used most often  for acute oral
studies and  as well as for other types  of  toxicological  studies.

    Shortly  before  testing, fur  from the trunks of healthy animals  should
be  clipped so that  no less than  10% of  the dorsal  body surface area  is
available for application of  the  substance.   Care  should be taken to
avoid abrading the  skin,  since  this would  alter  its permeability.  How-
ever, a dermal LDcQ for  an  animal  with  abraded skin may be desired
since the human skin which may  eventually  be  exposed to the test  sub-
stance will  not necessarily be  intact.

    For some applications,  it may  be appropriate  or  necessary to  use  a
vehicle.  If such is the  case,  any effect  of  the  vehicle on  the  penetra-
tion of the  test  substance should  be established.  The test  dose  must
remain in contact with the  skin  throughout the exposure period,  usually
24  hours.

    At the end of the exposure  period,  any residual material  is  gently
removed with a gauze compress,  the exposed area  is examined,  and  any
lesions are  noted.  The  animals  should  be  examined at  least  once  daily
for 14 days  for signs of  systemic  toxicity and localized dermal  reac-
tions.  All  animals that  succumb or are moribund  are  necropsied.   At the
end of the  14-day observation  period,  all  survivors  are subjected to a
thorough  clinical examination,  including examination  of the exposed area
of  skin.  They  are  then  sacrificed and  necropsied.  The degree of skin
irritation,  signs of  intoxication, changes in body weight, mortality,  and
gross pathological  findings  as  a function  of dose and  time are noted.

-------
    The method of calculating the acute dermal LDgQ is the same as that
for the acute oral LD50.  Both of these LDcg's are convenient for
estimating acute toxic hazard.  Although there is always risk in extra-
polation from animals to humans, it is usually safe to presume that  sub
stances with lower dermal LD^gs in animals will be potentially more
toxic to humans than those with higher
    It is suggested that these tests be conducted on two species of
animal-the albino rabbit and the albino rat.  Female animals  should be
nulliparous.   The following sample numbers and dosages are recommended:

    - 2 male and 2 female animals of each test species at each of five
    dose levels - preferably to include 2 levels above and 2  below the
    expected LD50.  A suggested dosage schedle is:  5000 mg/kg, 500
    mg/kg, 50 mg/kg, 5mg/kg, and 0.5 mg/kg.

ACUTE INHALATION TOXICITY

    The respiratory tract is particularly vulnerable to many  substances
since it is generally less protected than most body systems.  Moreover,
it can be subjected to  injury not only when a toxicant enters the body
through the respiratory tissues, but also, in some instances, when a
toxicant leaves the body via the respiratory tract after having gained
entry by a different route.  Consequently, injuries to the lung and other
body tissues resulting from inhaled toxic substances can have numerous
ramifications depending on the degree of toxicity of the substance, the
concentration and duration of exposure, and the existence of  an immediate
or latent effect.  The  anatomy and physiology of the respiratory tract
can have great  influence on the toxicity of inhaled vapors, gases, and
particularly, inhaled particles.

    Single, high concentration inhalation exposures are used  to deter-
mine the approximate toxicity level of a chemical substance or mixture
for comparative purposes.  The nature of the toxic effect, if any, should
also be determined through this process; in this way the concentrations
to be used in subchronic inhalation exposure tests may be established.
These procedures are also applicable to brief and intermittent human
exposures.

    The most informative and useful technique for determining the  acute
toxic effects of inhalation exposure is the one used to determine the
LC5Q value for  rats (that atmospheric concentration statistically  esti-
mated to kill 50% of the exposed animals within a specified post-exposure
period, usually four hours.)  Although death is the measured  endpoint for
the LC5Q determination, observation of other toxic responses  should
also be recorded.  Test animals should be observed for at  least two  hours
for signs of irritation of eyes, nose, and  lung tissue.  Gross  changes in
the respiratory rate, diaphragmatic breathing, gasping,  and frothing  and
bleeding from the nares are some signs of  irritation of  lung  tissue.
Other evidence  of discomfort may be pawing  at the eyes or  nose.   In  addi-
tion to these observations, records should  include time  of death  and  gross

-------
pathological changes noted at necropsy.  An acceptable post-exposure
observation period is 14 days, with an optional extension to 21 days.

    Exposure by inhalation is perhaps the most time-consuming and expen-
sive of all toxicological dosing procedures.   If a test substance pro-
duces a systemic toxic effect by a route of administration other than
inhalation, one can assume that inhalation of  a similar dosage would pro-
duce at least as great an effect (4).  Thus, it may be expedient to post-
pone the inhalation tests or to perform them last, even though inhalation
might be the most likely route of exposure.  It may be possible to form a
negative decision on a substance likely to be  inhaled based on the re-
sults of exposure by routes other than inhalation.

    The effects on the respiratory tract itself however, can be measured
only by inhalation studies.  These effects may be transient such as tem-
porary inflammation, or  irreversible such as death.  A common direct ef-
fect is acute chemical irritation, which can affect any part of the in-
halation route or the gas-exchange surfaces of the lung.  These acute
effects are often reversible except in cases where they produce pulmonary
edema or inflammation of such severity that the lung is no longer
functional  (4).

     It is  recommended that these tests be conducted on two species of
animal-the  albino rat and the albino mouse.  Female animals should be
nulliparous.  The following  sample numbers  and dosage schedule are
recommended:

     - 5 male and 5 female or 10 male animals of each species at each of
     two dose levels 2.0 mg/1 and 0.2 mg/1, for four hours of exposure.

ACUTE EYE  IRRITATION TOXICITY

     Test procedures to assess the surface toxicity of chemical substances
to ocular  tissues of laboratory animals should show the potential for
substantial human eye injury.  Albino  rabbits  have been most commonly
used in these test procedures because  their eyes  are large and have no
pigmentation.  In addition,  the tractable nature  of this  animal facili-
tates handling and examination.  However, the  rabbit eye  differs  in
several anatomical and physiological respects  from the human eye.  The
rabbit cornea is thinner, their blink  reflex is not well  developed, and
they have  nictitating membranes and thick fur  on  their lids.

     Limited comparative  data from controlled exposures of humans  and
rabbits show responses of the rabbit eye to be much more  severe and long
lasting.   Other nonprimate laboratory  species  such as rats, guinea pigs,
dogs, and  cats are either less satisfactory than  the rabbit or have not
been thoroughly evaluated.   Among nonhuman primates, the  rhesus monkey
has  been used most frequently, but squirrel monkeys are also suitable.
The  potential use of monkey  species as human models seems obvious, as
their eyes  are structurally  and functionally similar to those  of  humans.

-------
However, the limited availability, cost, and potential hazards in the
handling of monkeys prevent their extensive use.  Therefore, the albino
rabbit is the species of choice, with the rhesus monkey as the preferred
second species when confirmatory data are necessary.

    In a given test only one eye of each animal should be used and the
animal should not be subjected to extraneous test procedures or stresses.
Most standard methods call for the instillation of a measured amount,
usually 0.1 gm or 0.1 ml, of the test substance directly to the cornea of
the test animal.

    Epidemiological evidence suggests that most eye accident victims
rinse their eyes with water within one minute of exposure.  Certainly
most physicians recommend prompt irrigation for accidental exposures to
chemical substances with the rationale that the chemical on the surface
is diluted  and  irrigated away.  Some experimental animal studies indicate
that irrigation may decrease the amount of irritation caused by a chemi-
cal, but is not likely to change an apparent irritant to a nonirritant.
With some chemicals, such as 1% sodium hydroxide, irrigation markedly di-
minishes the toxic effects.  With 5% sulfuric acid, irrigation exacer-
bates the reaction.  The variability of irrigation techniques and the
arbitrary nature of any one regimen further complicate this test.
Consequently, irrigation is not a recommended requirement in a test of
the ocular  irritancy of a substance.

    Interlaboratory and temporal variability in rabbit eye testing makes
it difficult to determine the accuracy of any given result.  Assuming
that the factors that cause variability consistently  affect all observa-
tions in a  single test, it should be possible to compensate for them.
This is done by testing control materials of established ocular irritancy
and by rating unknown substances with respect to them.   If the human res-
ponse to the control material is known, animal  response  data can be
extrapolated to potential human response.  In such cases, the more nearly
alike the test material and control are in irritancy, the more confidence
can be placed in the extrapolation.

    If healing of the cornea and conjunctiva follow chemical injury, it
is usually  completed within 14 days.  Therefore, observation for 21 days
is essential in any test for toxicity.  The recommended  times for obser-
vations are 1 hour, 1, 3, 7, 14, and 21 days, though  slight deviations
from this schedule will not seriously affect results.

    Scoring the results of ocular irritancy tests is  complex and several
scoring systems have been developed to  assess the degree of damage.  Ob-
servations  of the cornea, iris, conjunctivae should be reported as well
as serious  lesions such as pannus, phlyctena, and rupture of the eye-
ball.  The  grades of ocular reaction must be recorded at each examina-
tion.  Evaluation of these reactions can be facilitated  by use of a
binocular loupe or hand slit-lamp.
                                    10

-------
    At the 24 hour observation, the eyes of any or all test animals may
be further examined after applying fluorescein stain.  For this  optional
examination, one drop of fluorescein solution is dropped directly  on  the
cornea.  After flushing out the excess fluorescein with tap water  or
saline solution, the injured areas of the cornea will appear yellow in
ultraviolet light.

    A record of the discharge  from treated eyes in not required; however,
any exudate above normal can be recorded as additional information.

    In grading these tests, an animal has exhibited  a positive reaction
if the test substance has produced one or more of the following  signs at
any observation:

    a)  Ulceration of the cornea  beyond  a fine stippling effect

    b)  Opacity of the cornea, other than a slight dulling

    c)  Inflammation of the iris, other  than  a slight deepening  of the
        rugae or a slight hyperemia of the circumcorneal blood vessels

    d)  Swelling in the conjunctivae (excluding the  cornea and iris)  with
        partial eversion of the eyelids  and a diffuse crimson color with
        individual vessels not easily discernible.

    In Table 2 a grading system for these parameters is presented.  Fre-
quency counts are then made for each ocular parameter and Table  3  can be
used  to aid in  labelling the test substance.

    The test should be considered positive  if four or more animals ex-
hibit a positive reaction.  If only one  animal exhibits a positive reac-
tion,  the  test  should be regarded as negative.  If two or three  animals
exhibit a  positive reaction, the  substance  is considered to be an
irritant.

    A clear distinction must be made between  those substances  that pro-
duce  transient  irritation and  those that produce  substantial  injury.
Substances that produce conjunctivitis which  clears  within 2-3 days with
no further reactions generally require no further  investigation.  Materi-
als that produce  serious cornea!  injury  or  internal  injury to  the  eye are
serious hazards and appropriate controls should be placed on  their use.
It is also important to note  any  signs  of  systemic toxicity  in  the course
of these tests.

    The recommended test species  are the albino  rabbit  and  albino  rat.
    The suggested sample numbers  and dosage  are:

        6  animals of each test species  at  one dose of 0.1 gm  (or O.lml)
        in a single eye of each animal.
                                     11

-------
                                 TABLE 2


                        Grades for Ocular Lesions

                                  CORNEA
No ulceration of opacity	0
Scattered or diffuse areas of opacity	1
Translucent areas, details of iris slightly obscured	2
Nacreous areas, no details of iris visible, size of pupil
  barely discernible	3
Opaque cornea, iris not discernible	4

                                   IRIS

Normal	0
Markedly deepened rugae, congestion, swelling or circumcorneal
  hyperemia, but iris still reacting to  light	1
Hemorrhage, gross destruction, or no reaction to light	2

                               CONJUNCTIVAE

Blood vessels normal	0
Some blood vessels hyperemic	1
Diffuse, crimson color, individual vessels not easily discernible	2
Diffuse, beefy red color	3

                     LIDS AND/OR NICITATING MEMBRANES

No swelling	0
Slight  swelling	1
Moderate swelling with partial eversion  of lids	2
Swelling with  lids about half closed	3
Swelling with  lids more than half closed	4
                                     12

-------
                         TABLE 3
            Classification of Test Substances
Classification

Non-irritant
Irritant
Severe  Irritant
Ocular Reaction

No positive reaction in any
category in more than 1 out
of 6 test animals at 1-3
days and all eyes normal at
7th day.

Corneal opacity grades of
1.0 to 2.0 at any observa-
tion up to 7 days, but all
corneas cleared by 14th day.

Iritis grades of 1.0 at 1-7
days, but all iritis
cleared by 14th day.

Conjunctivitis grade of 2.0
or more at 1-7 days.

Lid or nictitating membrane
swelling grades of 2.0 or
more at 1-7 days.

A positive reaction in any
category which has not
cleared by 14 days.
                             13

-------
DERMAL IRRITATION TOXICITY

    A reliable test for skin irritation should provide  a means  for  dif-
ferentiating among substances that will produce different  degrees of  ir-
ritation or corrosion of the skin.   In this context,  irritation is  the
local inflammatory response of normal living  skin to  direct  injury  by
single, repeated, or prolonged contact with a chemical  agent without  the
involvement of an immunologic mechanism.  The macroscopic  manifestations
are erythema and edema.  Corrosion  is direct  chemical  action on normal
living skin that results in its disintegration and  irreversible altera-
tion  at the site of contact.  Its  important manifestations are  ulcera-
tion, necrosis, and with time, the formation  of scar  tissue.   It is es-
pecially important to be able to  distinguish  between  materials  that will
produce minor or inconsequential  degrees  of skin  irritation  from materi-
als that can produce substantial  irritation or corrosive  injury as  a  re-
sult  of either customary use or accidental exposure.

    The voluminous literature on  primary  irritation test  methods lacks
consensus on the animal model or  procedure most likely to  give  accurate
and dependable results.  Test procedures  for  human  subjects  are as  nume-
rous  as those for animals,  suggesting that the problem does  not lie sole-
ly  in selection of the test species.  The most standardized  procedure is
a 24-hour,  semiocclusive patch test  of a  full-strength substance on the
skin  of albino rabbits.  A  common test procedure  is to demarcate about
100 cm? of  the skin  into quadrants  - two  abraded  and two  intact. The
test  substance can then be  applied  to one abraded  and one  intact quad-
rant, leaving  another quadrant of each type to serve as controls.   Most
of  the conventional  laboratory animals and some of  the more  exotic
species have been tried in  skin  irritancy testing.   None  provide perfect
models for  human skin.  The albino  guinea pig and  albino  rabbit, though
commonly used, lack  the broad spectrum of human responses  to skin irri-
tants.  They show only degrees of erythema and edema.   Both  species react
more  strongly  than humans  to mild-to-moderate irritants.   In fact,  some
materials that appear unsafe when tested  on rabbits may be nonirritating
to  human skin.  The  response to  guinea pig skin  is  more like that of
human skin  over  a wide range of materials; thus  it  is preferable to the
rabbit.

     The chemical  substance  can  be applied to  intact and abraded test
sites on clipped  animals for periods of  exposure  that range up to 24
hours.  A statistically significant number of test  sites  (usually 2-4
sites on each  of  six  animals) should be  available for evaluating each
substance.

    After removal of  the  test substance,  30-60 minutes should  lapse be-
fore  the patch sites  are  read to  allow sufficient time for pressure and
hydration effects to  subside.  Additional readings  should be made 24  and
72  hours after the patch  application.  Some  persistent effects such as
corrosion are  better  determined  at 7 days.
                                     14

-------
    Evaluation of skin effects involves using a scoring system for the
degree of redness and the degree of edema at the  site of  application  of
the test substance.  The scoring system usesd is  commonly that which  has
been published in the Federal Register in Section 191.11  of  the Federal
Hazardous Substances Act of the United States.  Their scoring system, as
seen in Table 4, involves assignment of numbers for the relative  degree
of erythema and the degree of edema formation.
                   TABLE 4.  ERYTHEMA AND EDEMA SCORING
        Erythema
          0  =
          1  =
          2  =
          3  =
          4  =
        Edema
          0  =
          1  =
          2  =
          3  =
          4  =
no erythema
slight, barely perceptible erythema
well defined erythema
moderate to severe erythema
severe, beet red erythema with injuries in depth

no edema
very slight, barely perceptible edema
slight edema with raised edges
moderate edema with surface raised approximately 1 mm
severe edema with the area raised more than 1 mm and
extending beyond the area of exposure
    The scores obtained for  both  erythema  and  edema  at  each  scoring
period and for both  the abraded  and  intact skin  are  listed  and  the mean
for each group and for each  type  of  effect is  calculated.   All  8  mean
values are then  added  together  and  divided by  4  since  there  are four mean
values for each  effect (erythema  and edema)  thereby  giving  a final
numerical figure which is  the primary irritation score.(5)

    In some  cases it may be  useful  to retain animals for  2  weeks  after
application  but  because such delayed readings  usually  only  confirm ef-
fects seen at 7  days,  the  value  of  further observation  should be  measured
against the  cost of  maintaining  the  animals.

    Table 5  reflects the kinds  of warnings which can be developed from
the numerical results  of these  tests.

    The suggested test species  are  the albino  rabbit and  the albino  rat.
Recommended  sample numbers and  dosage schedule are:

        6 animals of each  test  species at  100%,  30%, 10%,  1%, and 0.3% of
        the  test compound.   If  the  lowest  of these produces positive re-
        sults, a dose  below  0.1% is  used to find a no-effects level.
                                     15

-------
                                 TABLE 5
          NATIONAL INSTITUTE FOR OCCUPATIONAL SAFETY AND HEALTH
                 INTERPRETATION OF SKIN TEST RATINGS (6)
                   Rating
                              Interpretation
Intact skin
Abraded skin
0-0.9


1-1.9



2-4


0-0.9



1-1.9



2-4
Nonirritant; probably safe for intact
human skin contact

Mild irritant; may be safe for use, but
appropriate protective measures are
recommended during contact

Too irritant for human skin contact;
avoid contact

Nontoxic to cellular components of
abraded skin; probably safe for human
skin contact

Mild cellular toxins; may be safe for
abraded skin contact provided protec-
tive measures are employed

Cellular toxins too irritant for
abraded skin contact; avoidance of con-
tact is advised
 1-1.9
 2-4
                             Mixed reactions
Intact
skin
0-0.9

Abraded
skin
0-0.9
1-1.9

Safe
Safe

for
for

human skin



contact
intact human
skin contact;
may
 2-4


 1-1.9


 2-4




 2-4
be  safe for  abraded  skin  contact  when
protection is maintained

Safe for  intact  human  skin;  contact
with abraded skin  should  be  avoided

May be safe  for  intact and  abraded  skin
contact when protection  is maintained

May be safe  for  intact human skin con-
tact when protectio  is maintained,  but
contact with abraded skin is to be
avoided

Unsafe for intact  and  abraded human
skin contact;  avoid  contact
                                     16

-------
DERMAL SENSITIZATION TOXICITY

    Dermal contact sensitization refers to a delayed,  immunologically-
mediated, allergic reaction to a chemical.  With few exceptions, contact
sensitization develops as  a result of one or more contacts with  a chemi-
cal which initiate the sensitization process.  The  latent sensitized
condition generally develops  no sooner than 1-2 weeks  after  the  effective
exposure.  Subsequent exposure of the skin of the sensitized  individual
to a lower concentration of the sensitizer or related  substance  (cross-
sensitizer) can elicit a more intense response than to the initial expo-
sure to the chemical.  This response may take hours or even  days to
develop, hence it is "delayed."  Responses may be characterized  by pruri-
tis, erythema, edema, or induration, papules, vesicles, bullae,  or combi-
nations of these.  Reactions  generally subside over a  period  of  days  if
there  is no further contact with the sensitizer, but the state of sensi-
tization may be permanent.

    Occasionally  an individual who has been sensitized through the skin
will exhibit a systemic reaction, anaphylactic shock,  whose  symptoms  in-
clude  irritability, dyspnea,  cyanosis, convulsions, unconsciousness  and
death.

    Dermal sensitization tests should detect materials that  are  capable
of inducing either a substantial incidence of or degree of sensitization
responses among individuals exposed  during use or accidental  misuse.

    Laboratory animal  species are generally much  less  responsive to  con-
tact sensitizers  than  are  humans.  The most responsive is the guinea pig,
particularly albino varieties.  Animals from 1-3 months of age are pre-
ferred as they are more  easily sensitized  than very young or older
animals.  There  is no  appreciable difference  in the proclivity of male
and female guinea pigs  to  develop sensitization,  but  pregnant females
should be avoided since  pregnancy may  alter the  allergic  response.

    Topical  application  techniques  for  the  determination  of  skin sensiti-
zation have  been  in common use for  the  last twenty  years.  A method  which
involves  intradermal  injection  of  the  test  material in guinea pigs  is
also used, although this technique  does  not represent  true topical  sensi-
tization.

    This  testing  involves  both  an  induction phase  and  a  challenge  phase.
The concentrations of  test compound  to  be  used  in  the induction  phase as
suggested by Klecak  (7)  are 100%,  30%,  10%, 3%,  1%  and 0.3%.  On the
first  day of testing,  a  O.lml solution  or  suspension  of  each of  these
                                     17

-------
concentrations of the test substance is applied to an  area measuring 8
cm2 on the clipped flank skin of six to eight  animals  per concentration
group.  Applications are repeated daily for 3  weeks, or  five  times  weekly
for 4 weeks, always using the same skin site.  The application  site re-
mains uncovered and the reactions (if continuous  daily applications are
performed) are read 24 hours after each application  or at the end of each
week.  When strong local irritation reactions  are elicited, the applica-
tion site is changed.  A control group of  six  to  eight animals  is not
induced.

    The concentrations to be used in the challenge phase will have  been
determined  largely from the results of the dermal irritation  studies.
Each  test animal  and each control will be  challenged by  each  of four
concentrations of the test substance:

     •   the minimal  irritating  concentration  (the lowest dosage at  which
        25% of the test animals  develop a  mild erythema, an  erythema
        score of  1,  but no edema.),
     •   the maximum  nonirritating concentration  (the highest  dosage which
        elicits no skin reaction), and
     •   two scaled doses below  the maximum nonirritating concentration

     On  days 21 and 35,  applications of 0.025  ml  of each  of  these four
concentrations are applied to  the contralateral  flank  of each animal  on  a
skin  site measuring  2 cm2.  Reactions  are  read at 24,  48,  and 72 hours.
This  procedure determination of  the minimal  sensitizing  concentration
necessary for  inducing  allergic  contact hypersensitivity and  the minimal
eliciting concentration necessary to cause a  positive  reaction.  A  con-
centration  is  considered  allergenic when  at least one  of the  animals  in
that concentration group concerned shows  a positive  reaction  with non-
 irritant  concentrations.

     Because cutaneous responses  are  visible,  they can  be readily evalu-
 ated by a trained observer.  The delayed  reactions of  contact sensitiza-
 tion are  best  evaluated by making sequential  observations  of test sites
on the  skin.   The area  is  scored for the  degree  of erythema and edema as
 described  above  in the  dermal  irritation  section. Numerical  scores are
then averaged  for the  animals  in each  group.   The initial  observation
 should  be made 48 hours  after  the first  application  of the test mixture
 and these  scores  are used  as  an index  of  the irritant properties of the
 substance.   A second observation should  be made 24 hours after the chal-
 lenging injection in order  to  allow  primary irritation to subside,  but
 subsequent  scorings  should  be  made  48  hours later after the  challenging
 dose.  The  scores that  are  taken following the challenging dose are com-
 pared with  the irritant scores.   If  the  challenge scores are two to four
 times the  irritant  scores  the  compound is considered to be a mild  sensi-
 tizer.   Compounds that  have  high sensitizing activity will show sensitiz-
 ing scores  that  are  four  to  seven times  the irritant scores. (8)
                                     18

-------
    Reactions to the test substance at challenge that are stronger than
reactions to negative controls or to those seen during induction should
be suspected as results of sensitization.  Responses that are marginally
more intense than control response or that occur in very few animals
should be confirmed by a second challenge after 1 or 2 weeks.  Rechal-
lenge after a longer delay can produce unreliable results, because
sensitization in guinea pigs  is short-lived compared to that in humans.
Whether or not a rechallenge  is performed, a judgement confirming the
presence or absence of sensitization should be recorded for each animal.
Mean scores or indices, which are customarily calculated for each experi-
mental group, are useful only for showing relative  intensity of res-
ponse.  Two or more unequivocally positive responses in a group of 10-20
animals should be considered  significant.  A negative, equivocal, or
single response probably assures that a  substance is not a strong sensi-
tizer, though this is best confirmed by  further testing with human
subjects.

    The albino guinea pig and the albino rat are suggested as  the test
species and the following sample numbers and dosage is recommended:

        5 male and 5 female  animals per  concentration  group each induced
        with 0.1 ml of the test substance at 100%,  30%, 10%, 3%, 1%, 0.3%
        or 0% (controls).  Each animal  is then challenged with 0.025 ml
        applications of the  test substance at four  concentrations as
        determined from the  dermal  irritation studies: 1) the  minimum
        irritating concentration, the maximum non-irritating concentra-
        tion, 3 & 4) two scaled does below the maximum non-irritating
        concentration.

PHOTOTOXICITY

    A phototoxic response refers to  irritation that is not immunologi-
cally-mediated and which depends on  light exposure  for its presence.
This response is not to be confused with photosensitization or the photo-
allergy response in which  light energy  results in the  promotion of hapten
formation and the consequent appearance  of antigen.  In this case it  is
the resultant antigen-antibody response  that  is the cause of cell
damage.  At this point neither the  need  nor  the exact  protocol require-
ments have been established  for photosensitization  testing, thus it  is
not presently recommended  at this  level  of acute testing.

    Responses to phototoxic  chemicals are elicited  by  ultraviolet  ligh-
wavelengths, those between 280 and  430  nm.   The agent  plus light energy
results in the excitation of the agent;  its  energy  state  is raised to  a
higher level.  In the process of returning to  its original ground state
of energy, the compound can  react with  constituents of the cell.
Although the mechanism of  action of phototoxicity  is not  clear,  the
chemicals which elicit this  reaction  are thought to be free  radical  ini-
tiators.  These compounds  can react with molecular  oxygen  in  the  presence
of light energy to form peroxides that  cause damage to cell membranes.(8)
                                     19

-------
    The chemical to be tested is administered  in a solvent which will  not
alter the test situation by reacting with the  chemical  or otherwise  ab-
sorbing energy upon exposure to ultraviolet  light.   Dosing should  be on  a
microgram or milligram per square centimeter basis,  simplifying  the
extrapolation to dosing in humans.  The skin site can  be conveniently
demarcated with a marking pen and the chemical can be  delivered  to the
skin with a micropipette.  Following application, the  animals  are  exposed
to ultraviolet light from a high output source.

    It is customary to administer one high dose of the  test  substance.
This may be up to 10% of the oral LD50.  No  situation  has yet  been
found in which a compound has been negative  at a high  dose and positive
at a low dose.  If the high dose elicits a positive  response,  the  least
effective dose  is then determined.  Each animal may  be  used  as its own
control.  Control sites  include 1) negative  (the vehicle), 2)  positive (a
known phototoxic chemical), and 3) unirradiated, chemical-treated  sites.
Groups of 4-10  animals,  are sufficient for this testing.

    The phototoxic response is usually elicited quickly.  For  maximum ef-
fect the site should be  irradiated within 30 minutes to 2 hours  after the
chemical application.  Scoring as described  for dermal  irritation  tests
is performed 12-24 hours later.  The result  most commonly found  in photo-
toxic responses is a visible and palpable dermatitis consisting  of ery-
thema, induration, and at times, frank necrosis.  The  phototoxic response
is dramatic; there are few tests easier to read.  Certain phototoxic
chemicals and solvents may irritate the skin without irradiation.  When
this occurs, attempts should be made to decrease the effective tissue
dose so that the irritancy is not seen at the  site where the chemical has
been applied but has not been exposed to light.

    Occasionally, extrapolation of results obtained  from animal  photo-
toxicity tests  to humans may be questionable.   In such cases,  tests  with
humans may  be necessary  if the basic systemic  toxicololgic data  are
available.  The experimental procedure resembles that  used with  animals;
however, it is  usually necessary to make human skin  more permeable to the
test substance  by removing most of the stratum corneum by repeated cel-
lophane tape stripping.  A stripped skin site  control  is also  used.   The
dose should be  administered in one small application.
                                     20

-------
ORAL IRRITATION TOXICITY AND CORROSIVE EFFECTS

    It is particularly important to be able to detect materials which,  if
ingested, can produce corrosive injury to the mucosal surfaces of the
oral cavity, pharynx, esophagus, and stomach.  Severe corrosive injury  to
these tissues can be fatal or can result in strictures or other perman-
ently disabling injuries.  Chemical properties are reliable predictors  of
corrosive potential.  Strong alkalis are likely to injure the esophagus
and strong  acids to  injure the  stomach and  duodenum.  Either may  injure
the tongue  and pharynx.

    There is no standardized procedure for  predicting corrosive potential
to the alimentary tract, though several techniques have  been described.
Materials have been  admininstered  by intraoral, intraesophageal,  and  in-
tragastric  gavage and by timed  application  to  specific tissues by solu-
tions or  impregnated tampons.   Rats, rabbits,  cats,  dogs,  and  swine have
been tested, but a preferred animal model  has  not  been identified.  The
experts  agree that more  research  is needed  before  an animal model  is  se-
lected and  a reliable procedure is established.

    The  need for a special test for esophageal corrosivity  has been ques-
tioned on the grounds that information on  the  chemical and  physical pro-
perties  of  a substance,  as well as the results of  dermal irritation
studies,  can provide reasonable presumptive evidence of  severe  irritation
or  corrosive hazard  upon ingestion even  in  the absence of specific cor-
rosivity testing.
                                     21

-------
                                SECTION 3

                     ACUTE TOXICITY TESTING CRITERIA


WHEN SHOULD ACUTE TOXICITY TESTING BE PERFORMED?

    It is recommended that all chemicals substances be screened for  acute
toxicity and their LDgQS be determined.  It has been suggested that
certain substances should not be required to undergo testing for toxi-
cological effects including those whose physical and chemical properties
and structure-activity relationships indicate that no adverse health
effects are probable.

Examples of this situation include:

    •   nonreactive  insoluble polymers which pose no potential human  ex-
        posure through inhalation of particles in air

    0   nonreactive  insoluble polymers which contain no monomeric con-
        taminants or have them at levels safely below those known to
        cause human  toxicity

    •   new chemical substances which differ only slightly in structure
        from known substances which have already been shown to be inno-
        cuous by existing toxicity data.

    These may be reasonable exemptions in theory but in practice, biolo-
gical nonreactivity  should be demonstrable and not merely assumed.   It is
difficult to assure  "safe" levels of monomeric contaminants in polymers
without first measuring those levels and evaluating their toxicity.   It
is difficult to define a "slight" modification in structure.  This is a
very ambigious exemption and may ultimately be abused.  Structural-func-
tional relationships may be useful predictors but it is also true that
small differences in chemical structures can significantly influence  the
biological effects of chemicals.

    A common example is found in the compounds benzene and toluene.   Ben-
zene is associated with severe acute toxicity including irritation of the
mucous membranes, convulsions, and respiratory failure.  Chronic effects
include bone marrow  depression, aplasia, and occasionally leukemia.
Toluene, on the other hand, which differs structurally from benzene  only
by the addition of a methyl group, is considerably less toxic than ben-
zene.  Toluene at high concentrations may cause mild macrocytic anemia
but it is not associated with the broad spectrum of hematologic toxicity
which is seen with benzene exposure.

    Another example  is found in the drug amphetamine (racemic
 -phenylisopropylamine) whose effects include central nervous system
stimulation and activation of receptors which are normally innervated by
the sympathetic nervous system.  The d-isomer is three to four times more


                                    22

-------
potent than the 1-isomer in its effects on the central nervous system
whereas the 1-isomer is twice as effective as the d-isomer  in its effect
on the heart.  An effect of valence number on the toxicity  of arsenic can
also be seen.  The trivalent arsenites are much more  lethal than are the
pentavalent arsenates for  lower plants and animals including protozoa,
bacteria, and yeast(8).  In essence, most predictions about toxicity and
structural-functional relationships should be verifiable and not simply
left as assumptions.

    Testing exemptions may be considered for minor alterations to sites
other than the reactive group on such compounds as epoxides, nitro-
samines, and nitrosoureas.  But here too there must be some consensus as
to what changes would have no effect on the biological reactivity of the
molecule.  In these  cases  it will  be easier to implicate a  compound as
toxic than to exonerate it as safe.

    Another group of compounds that has been suggested as possible exemp-
tions from toxicological testing includes those in which adverse health
effects are anticipated but human  exposure will be strictly controlled.
This includes:

    •   substances which are extremely acidic  (pH <2) or basic
        (pH >13), which are not volatile, which are used solely for
        industrial purposes, and for which human contact would be
        strictly prevented by protective measures.

    Compounds of extreme acidity or basicity in their industrial stages
may remain as trace  residues in or on some resultant  product intended for
human use.  At this  point  corrosivity may no longer be a factor but other
toxic properties might possibly become evident.  Exemptions to substances
in this category do  not seem to be justified unless it can  be shown un-
equivocally that detectable levels never remain in the products to which
humans may ultimately be exposed.

WHAT ACUTE TOXICITY  STUDIES SHOULD BE CONDUCTED:

    The first  level  of testing should provide  enough  information  to be
suggestive of the need for further testing  in  any  area of  toxicity  and
should also be as simple,  fast, and  inexpensive as possible.   It  is
apparent that complete testing for long-term,  irreversible  effects  will
not be required for  all substances,  and thus should not  be  included  in
the first  level of tests.   It seems  prudent  to require  a complete  battery
of acute toxicity tests in the initial testing stage  in  order  to  obtain
useful data  and possibly give some indication  of  the  need  for  subchronic
testing as a second  series of tests.

    Acute  toxicity tests recommended  are  as  follows:

    •    acute oral toxicity - for  all  chemical  substances  except gases
                                     23

-------
    •   acute dermal  toxicity - for all chemical substances except gases

    •   acute inhalation toxicity tests -  for those substances whose
        physical  nature suggests that they would be airborne and thus
        where inhalation is expected to be a possible route of human
        exposure  including:

           - all  gases

           - liquids with a vapor pressure higher than 1 torr

           - solids with a particle size below 5 micra in diameter

    •   Irritation and sensitization tests - for all chemical substances
        except gases.

WHAT ARE THE MINIMAL PROTOCOLS FOR THE ACUTE STUDIES?

    Several studies have examined the variability of acute toxicity test
results (specifically the LDcQ calculations)  obtained on substances
examined in different laboratories.  Even when the same testing procedure
was followed in each of eight laboratories, LDgg's differed by up to
5-fold; averaging a 2-3-fold difference between highest and lowest
values.  This indicates that general guidelines for protocols would be
more appropriate  than rigid procedures for acute toxicity studies.

    Acute toxic effects in animals can range from no observed effect to
sedation, loss of motor function, loss of consciousness, respiratory
depression, and death.  The intensity of effect is related to time, and
also to blood concentration.  Thus, there are two major factors in acute
toxicity tests:   the rate of absorption and the rate of elimination of
the test substance.  These two rates are in turn influenced by several
factors, including:

    •   physical/chemical properties of the substance

    •   concentration

    •   volume

    •   route of  administration

    0   biotransformation

    •   transport

    •   excretion routes
                                    24

-------
    t   species (age, sex)

    t   pre-existing pathology.

These various influencing factors will not be identical between
laboratories, or even within the same  laboratory at different times.

    In general, it would help to check on the testing  laboratory's  ac-
creditation as at least a general indication that the  test was conducted
in accordance with good laboratory practices.

    In toxicity testing a primary consideration is the purity of the sub-
stance to be tested.  The purity of the test substance should be known
and any contaminating components should be identified  and quantitated as
thoroughly as possible to account for  100% of the test substance.   All
materials should be  stored  appropriately to maintain their purity and
chemical integrity.  Whenever possible, all studies should be done  using
the same lot of test sample.

    It is suggested  that for each type of acute test to be performed, two
species of test animal should be chosen, and one of these should be the
albino rat.  This will allow comparison of the effects of a  given chemi-
cal substance administered  to the same animal model by all test routes.
Correlations can then be attempted between oral, dermal, and inhalation
LD5Q's.  Rats are preferred also because there is abundant reference
toxicologic data available  for  this species.

    The second species should be the  albino rabbit  in  tests  of dermal
toxicity, dermal irritation and eye irritation; the albino guinea pig
should be used in tests of  dermal sensitization; the albino  mouse would
probably suffice as  the second  species in tests of  oral and  inhalation
toxicity.

    In order to standardize results within and among laboratories,  all
protocol requirements should be as specific as possible.  Therefore,  if
the general test animal is  to be the  albino mouse,  an  example of  the
specific requirement would  be:

        C57B1/C3H mice (F-l) from Caesarean-derived animals, maintained
        free from Mycoplasmosis (CRD)  Salmonella, external parasites,
        internal parasites, and other  disease are to be used.  The  sup-
        plier should provide information about the  type of diet  the ani-
        mals have been fed  and  their  bedding.  Arrange animal  delivery in
        time to be ready for assignment to treatments  at  about 42  days  of
        age.
                                     25

-------
    Dietary requirements should also be carefully delineated.  Any  diet
which meets the general nutritional requirements of the  species  in  ques-
tion can be used as a standard diet in toxicity studies  so  long  as  the
exact formulation and processing of the diet  are known.   For  standardiza-
tion among testing agencies, the NIH Interagency Toxicology Support Com-
mittee has recently recommended use of the NIH-7 Open Formula Rat and
Mouse Ration (Table 6) as a standard diet.  This formula is available
from several manufacturers and is priced  competitively with other commer-
cial feeds.  It is suggested that this be the required diet in all
studies using rodents  as the test animal.

    Caging requirements and animal care procedures should also be very
specific  since these features can influence not only reproducibility of
experimental results but can also affect  worker safety.   A  suggested
requirement would be:

        Test animals should be housed  as  a single animal per  cage  in
        hanging stainless steel wire cages.   The use of  wire  mesh bottoms
        without bedding is preferred since this will prevent  coprophagy
        and induction  of mixed function oxidases.  However, closed  hous-
        ing with false wire cage bottoms  may  be used if  precautions are
        taken to prevent coprophagy.

     It  is recommended  that animal care during these  studies include the
following standard requirements:

        • Temperatures will be kept at 27-29°C with a 12 hour dark-12
           hour light  cycle.

        • Clean drinking water will be provided and changed  twice  weekly.

        • Food consumption and body weights  will be recorded at weekly
           intervals.

        • Cages and racks will be cleaned and sanitized once every two
           weeks.

        • Floors will be sanitized  daily.

        • Animals will be observed  daily by  trained animal care
           personnel.

        • Sick or dead animals will be removed for  complete  necropsy,
           after notification of the investigator or his representative.

        • Proper records will be kept daily.

        • Facilities  and equipment  will  be monitored  for sanitation on  a
           monthly schedule.


                                     26

-------
           TABLE 6, NIH-7 OPEN FORMULA RAT AND MOUSE RATIONl
        Ingredient
Dried skim milk
Fish meal
Soybean meal
Alfalfa meal
Corn gulten meal
Ground No. 2 yellow shelled corn
Ground hard winter wheat
Wheat middlings
Brewers dried yeast
Dry molasses
Soybean oil
Sodium Chloride
Dicalcium phosphate
Ground limestone
Vitamin and Mineral Pre-mixes2
  5.00
 10.00
 12.00
  4.00
  3.00
 24.50
 23.00
 10.00
    00
    50
    50
  0.50
  1.25
  0.50
  0.25

100.00
    ^Calculated  approximate  composition = crude protein, 23.5%; crude
fat, 5.0%; crude fiber, 4.5%;  ash, 7.0%.

    ^Vitamin  and mineral  pre-mixes shall provide per kg diet: vitamin A
(stabilized), 6,050  IU; vitamin  D^,  5,060 IU; vitamin K, 3.1 mg;
a-tocopheryl  acetate,  22  IU;  choline 0.6 g; folic  acid, 2.4 mg; niacin,
33 mg; d-pantothenic acid, 20 mg; riboflavin, 3.7  mg; thiamin, 11 mg;
vitamin B-12, 4.4 yg;  pyridoxine, 1.9 mg; biotin,  0.15 mg; cobalt, 0.44
mg; copper, 4.4 mg;  iron, 132 mg; manganese, 66 mg; zinc 18 mg; iodine,
1.5 mg.
                                     27

-------
    Another important consideration is the gross pathological examina-
tions that are required.  It is suggested that the following requirements
be considered for adoption:

        Any animal which dies in the course of the study must be re-
        frigerated at temperatures low enough to minimize autolysis if
        necropsy cannot be performed immediately.  Necropsy must be per-
        formed on all animals who die or who are sacrificed within 24
        hours and preferably within 16 hours of death.  The gross dissec-
        tion and evaluation shall be performed by or under the
        supervision of the pathologist.

        The necropsy is defined as an external examination, including
        body orifices, and examination and fixation of the following
        tissues:

        Gross lesions                                 Lungs & bronchi
        Tissue masses or suspect tumors               Heart
          & regional lymph nodes                      Thyroids
        Skin                                          Parathyroids
        Mandibular lymph node                         Esophagus
        Mammary gland                                 Stomach
        Salivary gland                                Duodenum
        Larynx                                        Jejunum
        Trachea                                       Ileum
        Cecum                                         Spleen
        Colon                                         Kidneys
        Rectum                                        Adrenals
        Mesenteric lymph node                         Bladder
        Liver                                         Seminal vesicles
        Thigh muscle                                  Prostate
        Sciatic nerve                                 Testes
        Sternebrae, vertebrae,                        Ovaries
          or femur (plus marrow)                      Uterus
        Costochondral junction,                       Nasal cavity
          rib                                         Brain
        Thymus                                        Pituitary
        Gallbladder                                   Eyes
        Pancreas                                      Spinal cord


        Specific requirements for gross necropsy procedures, and patho-
        logy reports are described in NCI publication CG-TR-1 which might
        be distributed to the test laboratories.  Histopathological
        examinations are considered to be too costly for  this stage of
        testing and are not expected to reveal significantly more  infor-
        mation than gross pathology reports.

     Estimated costs for each test will vary with time and the location of
the  laboratory.  An example of  the costs of various acute toxicity tests
is shown  in Table 7.
                                    28

-------
                     TABLE 7, ACUTE STUDY PRICE LIST
                    	(July, 1979)	
EYE IRRITATION 	$390.00
    6 rabbits
    72 hour evaluation
EYE IRRITATION 	$510.00
    9 rabbits
    6 unwashed
    3 washed
    1 week evaluation
PRIMARY SKIN IRRITATION 	$360.00
    6 rabbits
    14-day evaluation
ACUTE ORAL TOXICITY 	$900.00
    Five levels of ten rats/level
    72 hour evaluation
ACUTE ORAL TOXICITY 	$325.00
    Six levels of two rats/level
    72 hour evaluation
ACUTE DERMAL TOXICITY 	$850.00
    Two levels of ten rabbits/level
    14 day evaluation
ACUTE DERMAL TOXICITY 	$1050.00
    Four levels of four rabbits/level
    14 day evaluation
ACUTE DERMAL TOXICITY 	$3400.00
    Five levels of eight rabbits/level
    14 day evaluation
ACUTE ORAL TOXICITY 	$900.00
    Five levels of ten mice/level
    14 day evaluations
                                    29

-------
Table 7 continued
SENSITIZATION 	$1200.00

    One level of ten guinea pigs/level
    Five week evaluation

SENSITIZATION 	$2300.00

    Two levels of ten guinea pigs/level
    Five week evaluation
                                    30

-------
HOW SHOULD THE RESULTS BE INTERPRETED AND USED TO ESTIMATE A TOLERABLE
LEVEL?

    In order to extrapolate the results of acute tests performed on
laboratory animals to "safe" levels of chronic human exposure, certain
obvious considerations are necessary.  One such consideration is the
relationship of animal toxic responses to human toxic responses and the
other is the relationship of short-term high-dose toxicity to long-term
low-dose toxicity.

    A convenient  approach to the question of the animal-human sensitivity
differential is to use the hundredfold safety margin.  This is based on a
convention which  assumes that humans may be up to tenfold more sensitive
than the experimental animals used and that additonally there may be a
tenfold variation in sensitivity among individuals.

    The second consideration, extrapolation of data from short-term high-
dose studies to predict the outcome of long-term low  levels of exposure,
is a more complex problem.  Several noted toxicologists have stated that
any acute toxic sign that will be produced at a given dose will occur
within 90 daysof  the initial administration of the test substance.
(9-17)  Also, the results from two major studies have indicated that
lifetime no-effect doses can be predicted from short-term studies.

    The Weil and  McCollister study (18) determined the 3 month and 2 year
no-effect doses of 33 compounds of diverse chemical structure, pharamco-
logic type,  and toxicity in rats and dogs.  These compounds included
pesticides,  veterinary chemicals, food additives, thickeners, stabili-
zers, antimycotics, and water treatment substances.

    The animals were observed for 36 criteria of toxicity including mor-
tality, food intake, weight, pathology, blood chemistry, hematology,
cholinesterase, fertility, and neoplasia.  The most sensitive criteria
were body weight, ratios of liver and kidney weight to body weight, and
kidney pathology.  It was the feeling of these authors that only these
sensitive parameters need to be followed.

    There were various numerical relationships by which the long-term
no-effect dose for each compound could be derived from its short-term
no-effect doses.  But in 97% of the compounds tested  (or 32 out of 33)  it
was seen that 1/10 of the 3 month no-effect dose could be given repeat-
edly throughout a lifetime without producing toxic effects; 100% of these
compounds were non-toxic given at 1/12 of the 3 month no-effect dose.

    This approach is further supported by a recent literature survey by
McNamara (19) in  which data was collected on the short-term and long-term
no-effect doses of 122 compounds of diverse chemical  and pharmacological
                                    31

-------
types.  These studies were conducted in various  laboratories using dif-
ferent methods and animal species.  The resulting frequency distributions
of the 1050 levels and short term and long term  no-effect  doses  demon-
strated three basic relationships:

        •  any compound administered to an animal at one hundredth of  its
           LD^Q will produce no effects in a short-term (3 month) study

        •  any compound administered daily at one tenth of its 3 month
           no-effect dose will produce no effects in the lifetime of that
           animal

        •  any compound administered daily at one thousandth of  its
           LDtjQ will produce no effects in the lifetime of the test
           animal.

    Using the third observation, we can use the  LDgg derived from short
term  acute testing to predict the lifetime NOEL  (no observable effects
level) within the same animal species.

    The lifetime NOEL of the test animal may be  further modified, as
described above, to accomodate the human-animal  sensitivity differential.
The lifetime NOEL for the test animal divided by 100 becomes the proposed
TLE (tolerable level of exposure) for humans.

        The working hypothesis therefore becomes:

           The LDcQ derived from short term animal studies divided by
           100,000 is the TLE of chronic human exposure.

    This safety factor of 100,000 is very conservative and its use is
most  justified when the results of acute testing produce a flat  dose-
response curve or a delay in the onset of and recovery from toxic signs.
These data may be suggestive of a potential for  cumulative toxicity and
thus  the need for longer term testing (1).  Physical/chemical properties
of the toxicant may also indicate the need to apply a greater safety
factor when determining a NOEL in humans.

    In other situations, where a  less conservative approach seems to be
justified, one might reduce the human-animal sensitivity differential  to
10 and use the LD™ from animal studies divided  by 10,000  to predict
the tolerable level of chronic human exposure.

An example of how one might use each of these factors is demonstrated  in
Table 8 using some oral LD50S from the literature(21):
                                    32

-------
                 TABLE 8.  PREDICTIONS OF HUMAN TLEs FROM
                 ORAL LD50s DERVIED FROM ANIMAL STUDIES
    Compound
Species
                            Conservative
               Predicted     Predicted
OralLD50        Human        Human TLE&
 (mg/kg)      TLEa( q/kg day)  ( g/kg/day)
Methyl Sal icy late
Thiabendazole
Propylene glycol
Pimaricin
dog
rat
rat
rat
2100
3100
27000
3700
210
310
2700
370
21
31
270
370
a.  figures given are oral LD50s divided by 10,000.
b.  figures given are oral LD50s divided by 100,000.
                                         33

-------
HOW SHOULD THE DATA BE SUBMITTED FOR EASY INTERPRETATION BY REVIEWERS?

    All dose and response data should be presented in tabular form.
Doses should be calculated in mg/kg body weight.  Responses should be
listed as number of animals killed per number of animals exposed  in  each
dosage group.  Other toxic effects should also be noted and recorded for
each dosage level group.

    It is suggested that manufacturers perform their own calculations of
dose-response relationships, slopes, standard deviations,  and LD50s,
the use of a second party to observe and monitor these calculations  is
also recommended.  Simple computer programs can be used to facilitate
operations and minimize errors.

    Experimental protocols should also be fully described  by each  labora-
tory in order to check for any procedural errors which could influence
the accuracy of the data.
                                    34

-------
                                References
1.  Loomis, T.A.  1976.  Essentials of Toxicology, Vol. 2, pp 180-184,
    Philadelphia: Lea and Febiger.

2.  Registry of Toxic Effects of Chemical Substances, NIOSH, July 1979.

3.  Draize, J. H., G. Woodard, and H.O. Calvery.  1944.  Methods of the
    study of irritation and toxicity of substances applied topically to
    the skin and mucous membranes.  J. Pharmacol. Exp. Ther. 82:377-390.

4.  Nelson, N., chairman.  1975.  Acute and  subchronic toxicity.
    Principles for Evaluating Chemicals in the  Environment,  pp 97-114.
    Washington:National Academy of Sciences Printing and Publishing
    Office.

5.  Loomis, T.A., 1976.  pp 207-215.

6.  Campbell, K. I., E. L. George, L.  L.  Hall and J.  S. Stara.  1975.
    Dermal  irritancy of metal compounds.  Arch.  Environ. Health
    30:168-170.

7.  Klecak, G., 1977.  Identification  of  contact allergens,  in
    Dermatology and Toxicology , Vol.  2,  eds.   Marzulli, F.  N. and H.  I.
    Maibach, pp 321-328, New York: Halsted Press.

8.  Loomis, T.A., 1976.  pp 163-165.

9.  Weil, C. S., M. D. Woodside, J. R. Bernard  and C. P. Carpenter.
    1969.   Relationship between  single-peroral  one-week and  ninety-day
    rat feeding studies.  Toxicol. Appl.  fharmacol.   14:426-31.

10. Paget,  G.E.  1963.  Standards for  the laboratory  evaluation of the
    toxicity of a drug - Viewpoint of  the Expert Committee on Drug
    Toxicity of the Association  of British Pharmaceutical Industry.
    Proc. Eur. Soc. Study Drug Toxic.  11: 7-12.

11. Peck, H. M.  1968.  An appraisal of drug safety evaluation in animals
    and the extrapolation of results to man.  In Importance  of
    Fundamental Principles in Drug Evaluation,  eds.   D.H. Tedeschi and
    R.E. Tedeschi, pp. 449-471.  New York:   Raven Press.

12. Barnes, J. M. and F. A. Denz.  1954.  Experimental methods used in
    determining chronic toxicity.  Pharmacol Rev.  6:191-242.

13. McCollister, D. C. chairman.  October 1974.  Society of  Toxicology Ad
    Hoc Committee Comments on Health Research Group Petitions.  Soc.
    Toxicol. Newslett., p 2.
                                    35

-------
14. Bein, H.  J.  1963.  Time to appearance of toxic signs in Ciba studies
    of 46 compounds at Basel, Switzerland and Summit, New Jersey.  Proc.
    Eur. Soc. Study Drug Tox.  11:15-24.

15. Hayes, W., Jr.  1972.  Essays on Toxicology, vol. 3, pp. 65-67.  New
    York, Academic Press.

16. Frawley,  J. P.  1967-  Scientific evidence and common sense  as a
    basis for food-packaging regulations.  Food Cosmet. Toxicol.
    5:293-309.

17. Kurlyandskiy, B. A.  1975.  Premorbid conditions of chemical
    etiology.  U.S. Joint Publications Research Service (JPRS 64252).
    (Translation of an article that appeared in Moscow Farmakologiya
    Khimioterapevticheskiye Sredstva Toksikologiya Problemy Toksikologii,
    1972, vol. 4, pp. 8-42.)

18. Weil, C.  S., and D. D. McCollister.  1963.  Safety evaluation of
    chemicals.  Relationship between short-and long-term feeding studies
    in designing an oral toxicity test.  J. Agri. Food Chem. 11:486-91.

19. McNamara, B. P.  1976.  Concepts in Health Evaluation of Commercial
    and  Industrial Chemicals.  In Advances in Modern Toxicology  vol 1,  pt
    1: New Concepts in Safety Evaluation, ed. M. A. Mehlman, pp. 61-140.
    New  York, Halsted Press.
                                    36

-------
 50272-101  	
  REPORT DOCUMENTATION
         PAGE
1. REPORT NO.
 EPA 560/13-79-009
                                                3. Recipient's Accession No.
 4. Title and Subtitle

   Acute  Toxicity Testing Criteria for New  Chemical  Substances
                                                5. Report Date
                                                  July.  1979
                                                                          6.
 7. Author(s)
   M.J.  Normandy,  R.6.  Reynolds
                                                8. Performing Organization Rept. No.
 9. Performing Organization Name and Address

   Enviro  Control,  Incorporated
   11300 Rockville  Pike
   Rockville, Maryland   '20852
                                                10. Project/Task/Work Unit No.
                                                 EPA 560/13-79-009
                                                11. Contract(C) or Grant(G) No.

                                                (o W2227-NA5X

                                                (G)
  12. Sponsoring Organization Name and Address

   Premanufacturing  Review Division
   Office  of Toxic Substances
   Environmental Protection Agency
   Washington, D.C.   **"AX
                                                 13. Type of Report & Period Covered
                                                 Technical  Report
                                                 April-July, 1979
                                                14.
  15. Supplementary Notes
   Project Officer  -  Dr. Daphne   Kamely
   Technical  Assistant - Felix Santos
  16. Abstract (Limit: 200 words)

          This  report addresses  the rationale, considerations, and limitations  of acute
    toxicity testing.   General  procedures  are described for acute tests including letha-
    lity  studies  in  oral, dermal, and  inhalation toxicity, and  irritation studies in dermal
    and eye toxicity,  phototoxicity and  skin sensitization.  Recommendations are given  for
    the acute  toxicity tests which may be  used to evaluate the  risks associated with the
    manufacture and  processing  of chemical  substances.   Suggested minimum protocols are
    offered including the choice of test animals, dosage levels,  laboratory practices,
    animal  diets,  and necropsy  requirements.  A method  is suggested for interpreting the
    results of these tests and  extrapolating to some  guidelines  for a safety factor in
    human exposure.
 17. Document Analysis a. Descriptors
    Toxicity, Chemical  Compounds, Test Methods, Laboratory Animals, Evaluation Criteria
    b. Identifiers/Open-Ended Terms
   c. COSATI Field/Group   06T,14B
 18. Availability Statement

    Release unlimited
                                 19. Security Class (This Report)
                                 Unclassified
                                 20. Security Class (This Page)
                                 Unclassified
21. No. of Pages
     36
                                                           22. Price
(See ANSI-Z39.18)
                                          See Instructions on Reverse
                                                          OPTIONAL FORM 272 (4-77)
                                                          (Formerly NTIS-35)
                                                          Department of Commerce

-------