7 31RT100 3.
                Technical Review Branch
                 Registration Division !
               Office of Pesticide Programs
            U.S. Environmental Protection Agency
                   September 1997



This report was developed by the Registration Division with the collaboration of the California
Department of Pesticide Regulation, Health Canada, the American Crop Protection Association
and the Chemical Manufacture Association.

This document was written by Ian Blackwell and Lucy Markarian of the Registration Division /
Office of Pesticide Programs. The following OPP staff contributed to the development of this

Tina Levine
Deborah McCall
Mark Perry
Carol Glasgow
Pam Hurley                                         ;
John Whalan                                        
Deborah McCall
Frederick Johnson
Ed Budd
A special thanks is given to Tina Levine for her contributions to the development of this



I.  General Study Deficiencies	2

n.  Acute Oral Toxicity		8

IE.  Acute Dermal Toxicity  	10

IV.  Acute Inhalation Toxicity	12

V.   Primary Eye Irritation	16

VI.  Primary Skin Irritation  	19

VII.  Dermal Sensitization	 22

    In response to the July 1993 Pesticide Reregistration Rejection Rate Analysis for Toxicology
(EPA 738-R-93-004) and realizing laboratories were often having difficulty conducting acute
toxicity studies to the satisfaction of the Agency, the Registration Division of The Office of
Pesticide Programs (OPP) performed a rejection rate analysis for the six acute toxicity studies
submitted to RD as end-use product data. The results  of the rejection analysis are listed below:
    Acute Oral Toxicity
    Acute Dermal Toxicity
    Acute Inhalation Toxicity
    Primary Eye Irritation
    Primary Skin Irritation
    Dermal Sensitization
   RD did not attempt to determine how often any particular study deficiency was found nor how
often the deficiency caused a study to be rejected. Initially RD conducted only a rejection rate
analysis, it was later decided that information on the proper conduct of studies was also needed.
RD realized there were several regularly observed deficiencies in acute toxicity studies that were
not previously addressed.  About the time RD was developing a document on the conduct of
studies, the American Crop Protection Association (ACPA) informed RD that they were also
working on a similar document hi preparation  for a self-certification program for acute toxicity
studies. On May 23,  1995, several representatives from the Environmental Protection Agency
(EPA), ACPA, the Chemical Producers and Distributors Association (CPDA), the Chemical
Manufacturers Association (CMA), Health Canada and the California Department of Pesticide
Regulation (CPDR) met to discuss acceptable methods for the conduct of acute toxicity studies.
The decisions made in the May 1995 meeting were incorporated into a preliminary RD document
that now forms this present document.

   OPP realizes that complete guidance for the proper conduct of acute toxicity studies according
to EPA policy is not found hi one source,  but must often be pieced together from several different
sources.  The goal of this document is to  compile this information into a supplement to the
November  1984 Subdivision F Guidelines so that there will be a reduction hi the number of studies
rejected or  flawed due to insufficient reporting or incorrect methodology. To insure that the
correct procedures are followed for acute toxicity studies, the laboratory  should follow the
Subdivision F Guidelines, November 1984 edition hi addition to the recommendations given hi this

   Studies  of concern are the FIFRA Subdivision F Guidelines 81-1 through 81-6  (OECD
Guidelines  870.1100,  870.1200, 870.1300,. 870.2400, 870.2500, and 870.2600): acute oral
toxicity, acute dermal toxicity, acute inhalation toxicity, primary eye irritation, primary dermal
irritation and dermal sensitization. The intent of the six acute toxicity studies (except  for the

dermal sensitization study) is to provide data sufficient to allow the placement of a product into a
specific toxicity category. The toxicity category is then used for the purposes of determining
product specific precautionary labeling.

   In addressing deficiencies observed in the past studies, the Agency hopes to inform laboratories
and registrants of errors so that these mistakes  may be avoided hi future submissions. OPP
realizes the following list does not contain all possible study deficiencies that may or may not lead
to study rejection. In addition, this document provides guidance on the conduct of studies that
have traditionally been used to support product labeling.  The Agency recognizes that decisions
concerning labeling may be made hi the absence of such data.  RD encourages the use of batching
and bridging of information from extant tested products in lieu of the conduct of new studies, the
request of data waivers for studies that may not be necessary or appropriate given the use pattern
or form of the product.  Also, RD encourages  the continued development of alternative testing
methodologies, such as in vitro testing, although such tests are not yet accepted by the OECD or
the Agency.  Such approaches reduce the use of animals, save Agency resources and speed the
review process.

   A registrant may submit a rebuttal anytime they feel the Agency has misinterpreted a study.
The rebuttal submission should include all pertinent information such as, the product's registration
number, the MRID numbers of the study, the problem(s) identified by the Agency with the study
and a clear explanation of the registrant's disagreement.  Secondary rebuttals will nol be
considered unless relevant new information is provided.  This new information should be a clarifi-
cation of procedures or data, new data, product formulations, etc.

   Study deficiencies may be separated into two broad categories, general deficiencies that may be
common to many or all studies and deficiencies specific to a particular study. Both categories may
include errors of conduct and/or errors of reporting.  Generally,  errors in the reporting  of a study
are correctable, while errors hi the conduct of a study are not. Once a study has been improperly
conducted, correcting errors without having a new study  conducted is not possible.  However,
errors of reporting, although correctable, can cause an avoidable delay hi review time.

A draft version of this document was released  to the pesticide manufacturers, registrants,
laboratories and other members of the public for comment. These comments, and answers to
them, have been incorporated into this document so that these organizations can read the direct
Agency responses to their questions.
I.  General Study Deficiencies

    A. Errors hi Study Conduct

      1.    Stadias are submitted on a test material that are too dissimilar to the registration product
           to be useful for making labeling decisions. When submitting studies to support the

           registration of a product, the studies should be conducted on the product formulation to
           be marketed.  When this is not possible, it is advisable to submit studies on a
           formulation or product that were found to be lexicologically similar either by a specific
           determination of lexicological similarity (done by the Agency) or through the batching
           process of a Registration Eligibility Decision (RED).  It is sometimes acceptable to
           "bridge" data to support one product with data conducted on another product. Data may
           be bridged from all acute toxicity studies, except the dermal sensitization, study to
           support another product  that has a similar but more dilute formulation.  Also the
           registrant may use data from a product that has toxicity category I for any of the five
           acute toxicity studies or  is a dermal sensitizer to support a category I or dermal
           sensitizer rating for a similar but more concentrated product.

     2.    The test animals were previously or concurrently used in another study, or were used to
           test more than one test material at a time.  According to the February 11, 1992, Robert
           P. Zendzian and  March 17, 1992, Penelope  Fenner-Crisp memos on this subject, OPP
           considers the reuse of test animals and the testing of multiple chemicals simultaneously
           on test animals hi acute toxicity studies or any other toxicology studies to be
           scientifically unacceptable.                        ;

     3.    The test animals were unhealthy. The 40 CFR 160.90 states that at the beginning of a
           study, test systems shall  be free of any disease or conditions that might interfere with
           the purpose or conduct of the study.                ;

     4-    The mortality exceeds more than one animal per sex for either sex in a limit test. When
           there is more than one mortality per sex at a dose level, the laboratory should test
           additional dosages to determine an LD50.

jfadustry Comment;  Additional dose levels should be tested as necessary to determine the
label category  (not necessarily an  LDSO).

EPA Response: The Agency agrees.

     5-    A reduced testing scheme was used without establishing the more sensitive sex.  A
           reduced testing scheme allows testing fewer than three dose levels for each sex. To
           qualify for a reduced testing scheme in either the acute oral, acute dermal or acute
           inhalation toxicity study, the laboratory must first establish whether one sex demon-
           strates increased mortality to the test material. This demonstration is preferably done
          via the acute oral toxicity study.  Dose levels must be sufficient to clearly identify the
           appropriate toxicity category for the test material.

     6.   Dose levels were not properly chosen.  The testing scheme did not allow the Agency to
          determine the toxicity category of the study.   Ideally an acute toxicity study may be
          accepted if no more than one animal of either or both sexes dies (for a total of 2/10

          animals) during a limit test.  The Agency has received older acute inhalation toxicity
          studies that failed hi their attempt to reach the 5.0 mg/L limit test concentration.  In
          studies of this type where 2/5 animals of one sex (and 0/5 of the other sex) die in a 5.0
          mg/L limit test, the study will be classified toxicity category III for acute inhalation
          toxicity. The registrant will be given the option of retesting (at 2.0 mg/L) to achieve
          toxicity category IV for acute inhalation toxicity.

     Discussion: Bracketing. If hi an acute oral toxicity study, the lab tested only the
     concentration of 50 mg/kg and 0/10 test animals died, this would not be sufficient to
     categorize the test material. This product could conceivably have an LD50 that would place it
     into toxicity category II,  III or IV. The laboratory should then test at least one additional
     dose to decide if the product meets the limit test for Category III or Category IV.

Industry Comment:  It is only necessary to test enough doses to establish the correct label
category (not necessarily an LD50).

EPA Response:  The Agency agrees.  We have modified our discussion of bracketing to note
that deciding an IDs, is not necessary.  However, a test done at a very low cut-point is not
sufficient for labeling. Limit tests are useful for labeling only if they allow placement of the
product into Category BDE or Category IV.  Otherwise, over labeling could result.

       a. If 3-4 test animals die at the limit dose (e.g., 5000 mg/kg for acute oral or dermal
          toxicity or 2 mg/L  for acute inhalation toxicity study) and no more than one animal dies
          at the next limit dose level (e.g., 500 mg/kg for an  acute oral toxicity, 2000 mg/kg for
          acute dermal toxicity or 0.5 mg/L for an acute inhalation toxicity study). The toxicity
          category for the study would be category ffl.

       b. Another example is an acute inhalation toxicity study where two  dose levels were tested,
          with 7/10 animal deaths at 0.5  mg/L and 0/10 (or 1/10) deaths at 0.03 mg/L. As the
          upper limit for acute inhalation toxicity category I is 0.05 mg/L, this would not
          eliminate the product from falling into toxicity category I.  This  study would not be
          acceptable without a third dose level conducted between 0.03 and 0.5 mg/L.

       An acute oral, dermal  or inhalation toxicity study where one or more test concentrations are
       tested that do not allow LD50 determination or bracketing will be rejected. Examples of this
           An acute oral toxicity study is submitted with test material concentrations of 50 and 300
           mg/kg.  The study had no mortalities. This study does not allow placement into any
           toxicity category as it could only be said that the LD50 is above 300 mg/kg.  This could
           place the toxicity category into category II, III or IV; or

          An acute dermal toxicity study is conducted at 20QO mg/kg and the mortality rate is
          10/10.  No other test material concentration is tested.  This results of this study only
          show that the product could not be placed into toxicity categories III or IV. However,
          this product could fall into either toxicity category I or II for acute dermal toxicity.
          Thus, this study would not be acceptable.

Industry Comment:  A general principle that should be followed in the review of acute
toxicity studies should be that all studies that conform to the current Subdivision F or OECD
guidelines must be acceptable if they allow a label category to be assigned.

EPA Response; The Agency agrees.  For example,  fewer than 3 animals may be used for
irritation and sensitization studies if toxicity category I is demonstrated.

     7.   The laboratory fails to follow a test method approved by the Agency and the resulting
          data is questionable. Laboratories should assure their  protocols do not conflict with
          guidelines, guidance or test methods approved by the Agency. They should also ensure
          their protocols are followed.  OECD protocols for acute toxicity studies are accepted
          by the Agency.

     8.   An unacceptable test species was used. For the acute oral and inhalation toxicity
          studies, the preferred species is the rat.  For the acute dermal toxicity study, the
          preferred species is the albino rabbit, followed by the rat and the guinea pig. The
          preferred species/strain for the primary eye and skin irritation studies is the albino
          rabbit.  The preferred species for the Buehler Method  and many other dermal
          sensitization studies is the guinea pig. Although the species mentioned above are not
          the only acceptable species for these particular studies, they are the preferred species.
          If one of these species is used, no further justification  is  needed.  Any laboratory .
          conducting studies using other than the preferred species must provide justification for
          using another species. The Agency may reject studies based on the test species used

     9.   The animals were quarantined for an unacceptable length of time. The minimum
          quarantine tune is five days.  Lack of sufficient quarantine period may cause undue
          stress hi the test animals.  Such stress may contribute to the mortality rate and thus
          adversely influence the toxicity.

     10.  The lab failed to conduct sufficient observations of the test animals.   The lab should
          consult Subdivision F guidelines  for guidance on the observation periods of test animals.

     11.  The test animals used were not of an acceptable weight and/or age. Rats to be used
          hi acute toxicity studies should be 8-12 weeks at study initiation.  Rabbits should be
          at least 12 weeks of age at study  initiation.  Body weights for the test animals should
          be in the range of those for normal animals at that age. The weight range should not

          exceed  20% of the mean pre-exposure weight for that sex.  The laboratory should
          report the weights and ages of the test animals at study initiation for each of the six
          acute toxicity/irritation studies and be able to provide a growth curve from the
          supplier to show that the test animals used were of normal weight for their age. It is
          not necessary to report the ages of the animals upon receipt.  This applies to all
          acute toxicity/primary irritation studies. However, allowances will be made for
          primary skin irritation studies where animals were chosen outside of the weight
          range hi order to pick animals with skin suitable for this  study.

Industry Comment; The requirement for the growth curve should pertain only to_Eals used in
the acute toxicity studies.  The age requirement for rabbits should pertain only to acute
dermal toxicity studies, not to skin and eye irritation studies where the age and weight of the
animal are not important.

EPA Response; There is no age or size requirement for primary irritation studies.  However,
proper growth is an indication of a healthy animal.  Therefore,  the laboratory should be able
to demonstrate that the test animals used were of normal weight for then* age hi order to
demonstrate the health of the animals.

     12.  A lack of equipment calibration.  Although records on equipment calibration are not
          required hi the report, these data must be maintained by  the laboratory.

     13.  Incorrect replacement of animals other than at study initiation when a dosing accident
          has occurred. Animals on study should only be replaced when deaths have occurred as
          a result of dosing accidents or other unrelated events. Necropsies should be conducted
          to prove deaths were the result of dosing accidents or other events.

Industry Comment; Please clarify that if a dosing accident occurs that an animal may be
replaced at study initiation.

EPA Response; If a dosing accident occurs  at study initiation, an animal may be replaced.

    B.  Errors hi Reporting

     1.   Failure to properly identify the test material. Often, studies identify test materials by a
          pesticide manufacturer's internal code, an obsolete name or some form of identification
          other than the current product name or EPA registration number. Ideally, the test
          material name should be identical to the product name.

       The test material should be properly identified in the report.  If this is not possible, the
       registrant should include a statement identifying the test material hi the submission to the
       Agency. When the test material used hi a study is not the product for which registration is
       sought, the registrant must include the  name of the test material (a product name if

       possible), the CSF of the test material and the test material's EPA registration number (if
       possible), the test material's relationship to the product (if any) and clearly explain why this
       test material was submitted to support the product.

       At times, a laboratory may include a copy of the product formulation in the study report. A
       similarity determination based on chemical formulations will be conducted.   When this
       reported formula is different from the product CSF, this study may not be accepted.

     2.   The test material is not properly described. The description of the test material should
       a.  The physical state (e.g., liquid, paste, aerosol spray, granular, etc.) Other special
          properties of the test material that may affect testing;  e.g., very viscous  liquid, gel,
          encapsulated pesticide in suspension, color, etc.
       b.  pH of the test  material.
       c.  Percentage of  an active ingredient.
       d.  A manufacturer's lot or batch number of the test substance.

Industry Comment;  The pH should only be required when appropriate, for aqueous liquids
used in skin and eye irritation studies.

EPA Response;  The pH of the test material is appropriate for many additional situations.  It
is appropriate for moistened solids that will be applied to the skin and is also appropriate for
solids in solution or suspension for oral toxMty studies.  Obviously, pH would  not be
appropriate for a solid material instilled in the eye.

     3-   Noncompliant or missing OA or GLP statements may lead to study rejection or delayed
          review.  Refer to the 40 CFR part 160 and Subdivision F Guidelines for guidance.  If
          the data raises  concerns as to its validity and the study lacks appropriate  QA and GLP
          statements, it will be rejected.  If the data does not otherwise raise concerns about its
          acceptability, deficiencies in QA/GLP reporting will not influence the acceptance of the
          study; however, such  studies will be referred to the Office of Enforcement and
          Compliance Assurance (OECA) for follow-up.
     4.   Unreported data or other missing information. On occasion, pages are missing from the
          study report or the laboratory may fail to include information such as a legend
          explaining abbreviations used in reporting study results. At times reports do not include
          the complete details of a study. Perhaps this is because the laboratory assumes that
          certain aspects of the study are unimportant or the  laboratory has included a protocol
          that covers the particular type of study. Laboratories  should submit reports that address
          all details of that individual study. The report should relate how that particular study
          was conducted.   Laboratory QA units must insure the reports are complete and

     5.   The report or parts of the report are illegible. Although this usually concerns reprints
          or "blowbacks" from microfiche of reports, it is not restricted to reports submitted as a

     6.   The report contains incorrect calculations that the reviewer is not able to clarify. In
          such instances, the reviewer must consult the registrant/laboratory for clarification or
          the submission may have to be rejected if prompt resolution of the misunderstanding is
          not attainable.

     7.   Dilutions of a test material are not reported or are not adequately reported.  When a test
          material is diluted, the dilutions must be defined and the reasons for the dilution must be
          clearly explained.

     8.   The ages, weights and/or source of the test animals is (are) not reported.  This is a
          requirement for all acute toxicity studies as per Subdivision F guidelines 81-1 through
          81-6 and 80-4.
H.  Acute Oral Toxicity

    A. Errors hi Conduct

      1.   Unnecessary or improper dilution of the test material.

       -  Liquids should be tested undiluted.
       -  The highest workable test material concentration should be used for solids and viscous
       -  Justification must be provided for any dilution.

       Caustic materials should only be diluted if it is necessary for intubation. The dilution of
       caustic test materials may reduce their corrosive effects, thus giving an inaccurate
       representation of their potential threats.  If it is necessary to dilute such a test material, the
       report should give the pH of the diluted test material.  Dilution should be held to a

    Discussion: Is the recommendation for constant dose volume or constant dose
    concentration? For purposes of precautionary labeling, constant concentration is more
    important than constant volume.

    Discussion: Is analytical confirmation of dosing solutions a necessity?  Analytical
    confirmation of dosing solutions is not a requirement for any toxicity/irritation study.

Industry Comment;  The following guidance for acute oral dosing was proposed by the ACPA
group and is recommended to be included as guidance:

Acute oral dosing procedures have typically employed either constant dose volume across all
dose levels (EPA, OECD guidelines) or constant dose concentration. Systemic toxicity can
usually be determined using a constant dose volume to minimize the effects of gastric volume
on absorption. However,  EPA has expressed concern that excessive dilution of test materials
may not provide a correct assessment of the true toxicity of the test material for hazard
labeling purposes, particularly for those that are corrosive.

The following guidance was proposed:

1) Either constant volume or constant concentration administration is acceptable, provided
   the guidance below on dilutions is employed.

2) When possible, liquid test material should be dosed neat.

3) If dosing with neat material is not possible, due to high viscosity or toxicity that would
   preclude accurate low dose volumes, or if constant volume has been deemed to be the
   more appropriate method, the test material may be diluted. The highest concentration
   possible  should be administrated, although volumes less than 0.5 ml per animal would not
   be required. Lower dose volumes are acceptable if they can be accurately administered.

   [Note that the use of the 0.5 ml/ animal dose volume  may require a 50X dilution of a
   highly toxic (category I, <50 mg/kg) test material to insure accurate dosing.
   Alternatively, if a Category I result can reasonably be anticipated, waiving the study may
   be possible.]

4) If possible, the maximum dose volume should not exceed 1.0 ml/ 100 grams body weight
   for all vehicles, although volumes up to 2 ml/ lOOg for aqueous vehicles are acceptable
   with justification.

5) Solid materials should be suspended or dissolved in the nunimum amount of a vehicle and
   dosed at the highest concentration possible, following the above guidance.

EPA Response: The Agency agrees with the guidance outlined above.

     2.    The lab did not fast the animals before dosing. The presence of foodstuffs in the
          digestive tract of the test animals can have the effect of diluting the test material or
          carrying portions of it through the alimentary canal without digestion.  This action can
          thereby reduce the toxicity or corrosive effects of the test substance.

     3.   Lack of necropsy.  Although the guidelines say that gross necropsies should be
          performed on all animals under a test, lack of necropsy is not a reason for rejection.
   B.  Reporting Errors

     1.   Insufficient observation (lack of specificity).  Correct identification of symptoms in
          specific scientific terms and not generalized statements or lay expressions that could
          mean several different things are preferred.

     2.   The absence of necropsy results.  When a necropsy has been conducted, the results
          should be included hi the report.
DDL  Acute Dermal Toxicity

   A. Errors in Study Conduct

     1.   The dilution of liquid test materials or over-moistening of dry test materials. All liquid
          test materials should be undiluted for the acute dermal toxicity study.  Dry test materials
          must be moistened with water before application to ensure good contact and no loss of
          the test material.  Dry/solid materials may be moistened in a beaker or other suitable
          vessel. The laboratory should state the amount of water used to moisten dry test
          materials. The dry test materials should not be moistened beyond that point which is
          necessary to assure proper contact with the skin.

Industry Comment:  Can we specify that vehicles other than water or saline such as gum
arabic, ethanol + water, carboxymethyl cellulose, glycerol, propylene glycol, PEG vegetable
oil and mineral oil can be used, if water or saline cannot be used, as long as the vehicle is not
irritating and the inability to use water or saline is justified in the report.

EPA Response: Yes, the Agency agrees as long as the replacement vehicle is non-toxic, non-
irritating, and will not substantially change the properties of the test material. Also the effect
the vehicle has on the permeability of the test substance should be taken into consideration.

     2.   The size of exposure area is incorrect.  The exposure site should be approximately 10 %
          of the animal's body surface area.  The exposure area is to be from the scapula
          (shoulder) to the wing of the ileum (hipbone) and halfway down the flank on each side
          of the animal.

Industry Comment;  Less than 10% body area may be exposed if the material is highly toxic.
The second sentence in the above paragraph should read:  The prepared area for a rat or
rabbit dermal toxicity study will be defined as a shaved or clipped area starting at the scapula


 (shoulder to the wing of the ileum (hipbone) and half way down the flank on each side of the

 EPA Response;  The Agency agrees that for highly toxic substances, less than 10% of the
 surface area may be used.  This is consistent with our general policy of using a test material
 as close to the actual product as possible. If the product is highly toxic or irritating by the
 dermal route, it may also be appropriate to request a waiver of the study and label the
 product with Category I labeling.

      3-   Improper occlusion, covering and wrapping of the test site. An improperly wrapped
          test site may result in loss of test material and thus reduce the effective dose of the
          pesticide. Important factors to strive for when wrapping the test animals for the acute
          dermal toxicity study are:

       a.  Powders or other solids should be slightly moistened (not runny) to a paste before
          application to the test site.
       b.  It is preferred that the test material  be applied to the dorsum.
       c.  The gauze covering is added to act  as a reservoir for the test material and to keep it
          localized.  It is important that not too much gauze1 is used.  Too many plies or layers of
          gauze can have the effect of absorbing  the liquid test material or the water used to
          moisten the solid test material thus  minimizing the amount of test material that is
          available to the skin.
       d.  The test material is further covered with an occlusive material (such as plastic sheeting
          or rubber dam) or semi-occlusive material (such as perforated plastic) to prevent
          evaporation of liquids  from the test site and to prevent ingestion of the test material.
    Discussion: Guidance for occlusion.
    occlusive dressing is acceptable.
Although semi-occlusive dressing is recommended,
Industry Comment: We suggest the following wording be added as agreed in our May
meeting:  When possible the test substance should be applied directly to the skin, otherwise it
may be applied directly to a porous gauze dressing that is immediately placed in contact with
the animal's skin.

The test substance must be held in contact with the skin with gauze and non-irritating tape for
a 24-hour exposure period. The test site must be covered in a suitable manner to retain the
test material in contact with the skin, avoid wicking of material from the skin surface, and to
ensure that the animal cannot ingest the material.  To minimize wicking, the gauze should be
no more than 8-ply; fewer layers of gauze may be needed for small test volumes. Although a
semi-occlusive dressing is preferred, an occlusive dressing will also be acceptable.
EPA Response! The Agency agrees.

   B.  Errors in Reporting

     1.   Systemic toxicity, and dermal irritation are not reported or are not reported sufficiently.
          This information is required and is often helpful in evaluating other studies.
          Evaluations of systemic toxicity and local irritation should be made frequently on the
          day of application, and daily thereafter. Mortality checks alone are not sufficient
          evaluations.  If information from the dermal toxicity study is to be used to support
          dermal irritation labeling, i.e., if a waiver of the dermal irritation study may be
          appropriate,  observations of dermal reactions should be made with sufficient frequency
          and hi sufficient detail to obviate the need for a specific dermal irritation study.
IV.  Acute Inhalation Toxicity

   A. Errors in Study Conduct

      1.   Historically, one mam reason acute inhalation toxicity studies have been rejected was
          that the labs were often unable to achieve a small enough Mass Median Aerodynamic
          Diameter (MMAD).  However, in 1994, the Agency changed its acceptance criteria.
          Currently, the Agency accepts acute inhalation toxicity studies with MMADs of 1-4
          micrometers.  Please refer to the 2/1/94 HED Memorandum: Interim Policy for Particle
          Size and Limit Concentration Issues in Inhalation Toxicity Studies,  by John Walen and
          John Redden.

       a.  For dry products that do not readily form aerosols with MMADs at or below 4 microns,
          further attempts to reduce their particle size must be made. Granular products should
          be milled (by air mill, ball mill, hammer mill, etc.) for 24 hours if necessary.  If the
          product will still not form the proper aerosol and the product is dissolved in liquid
          before application, the laboratory must attempt to dissolve it in the vehicle to obtain an
          aerosol. The test material concentration must be calculated based on the amount of test
          material without the diluent.  The acceptable concentration must allow for the diluent.
          Water is the recommended diluent.  If the diluent is some material  other than water, a
          vehicle control study must be conducted.

Industry Comment:  The granular products should be milled for 24 hours as necessary ox
until particle size plateaus.  If the product will not form a proper aerosol and the product is
dissolved or suspended in water or another vehicle under conditions of field use, the study
should be conducted in a suspension or solution of water or the vehicle that is used under
commercial use conditions.

EPA Response; The Agency agrees with the comment.  Granular products should be milled
for a reasonable amount of time until particle size plateaus. This may be less than 24 hours.

  b.  Products that do not deliver an acceptable particle size and are not water soluble are
     excellent candidates for a waiver of the acute inhalation toxicity study.
     Microencapsulated products with a large percentage of capsules above four microns are
     also excellent candidates for waivers.  Waivers may be  granted for technical and end-
     use products that cannot conceivably be generated in sufficient concentration to pose an
     inhalation hazard. All waiver petitions must be considered on a case-by-case basis.
     The following are likely waiver candidates:
     1.  Non-volatile products that cannot be readily aerosolized (e.g., viscous liquids,
         waxes, and resins), and which are  not heated or diluted to an inhalable state during
     2.  Tree injections or thick liquids such as lotions,  waxes, etc., which contain non-
         volatile active ingredients.
     3.  Corrosive or highly irritating agents (the chemical may be designated inhalation
         Toxicity Category I by default). The Agency will categorize a study by actual acute
         inhalation toxicity data over the corrosivity of a product and may also use systemic
         toxicity data where available.
     4.  Slow release collars and ear tags (plasticized).
     5.  Products with fewer than 1 % of their particles being below 50 microns under
         conditions of use, unless that product is toxicity category I by the oral or dermal
         routes, or induces severe systemic  toxicity upon ocular dosing.
     6.  Non-friable granules.  The registrant must demonstrate that the granules do not
         produce fines when subjected to shipping and handling.
     7.  Microencapsulated products that are not readily fractured, dissolved, time released,
         leaky, or, small enough to be respirable.

If an end-use product cannot be aerosolized, but the addition of a diluent under conditions of
use yields an inhalable aerosol, an inhalation study shquld be performed using the most
concentrated label dilution.  A vehicle control group should be included if a diluent other than
water is used.

3.   Studies that do not demonstrate that the particle concentration was consistent over the
     period of the exposure are not acceptable.  Studies  that have not conducted particle size
     analysis at least twice during the exposure are not acceptable.   Studies that conduct
     particle size analysis once during the four hour exposure have not demonstrated that .the
     particle size was consistent during the  exposure. Chamber concentration and
     size measurements should be made at least twice during the study at time points
     well apart.  Chamber concentrations may be measured using gravimetric analysis or by
     analytical measurement.  If these are "reasonably consistent" (+_ 10% for liquid aerosol,
     gas or vapor, +.20% for dry aerosol), then two measurements should be sufficient. If
     the measurements are not consistent, then further measurements (a total of 3 or 4)
     should be considered.  Pretest measurements may be helpful.

     4.   Studies with MMADs that are not within an allowable range will not be acceptable. All
          particle size analysis should show the MMAD to be within the acceptable range of 1-4
          microns.  Otherwise, the test animals will not have been exposed to an acceptable respi-
          rable particle distribution throughout the exposure.

     Discussion: Studies where the MMAD exceeds four microns.  The Agency will evaluate
     such studies on a case-by-case basis.  If the MMAD is only slightly above four microns  and
     there are no other reasons to question the acceptability of the study, the Agency will accept
     the study.

Industry Comment:  Studies should not be rejected, if reasonable  efforts to minimize
particle size are employed.

EPA Response; The Agency agrees.

     5.   The lab did not obtain a concentration sufficient to determine an appropriate toxicity.
          category or conduct a proper limit test.  At tunes, OPP has received acute inhalation
          toxicity studies that tested one or two aerosol concentrations, the concentration(s) tested
          was below the limit test concentration, no LC50 was determined and no explanation of
          the study was given.  If the laboratory can demonstrate to the satisfaction of the Agency
          that the test was actually conducted at the maximum attainable concentration, and there
          is no evidence of toxicity, the product will be assigned toxicity category  IV.  The report
          must include a description of the physical and chemical nature of the test material, and a
          justification for the acceptance of the study.   Otherwise, this study will not be

          Often tunes, laboratories had problems attaining limit test concentrations. An acute
          inhalation toxicity limit test was originally a four hour exposure to a test atmosphere at
          a concentration of 5.0 mg/L.  However, hi 1994 the Agency changed the limit test
          concentration to 2.0 mg/L.  Acceptable acute inhalation toxicity studies which show no
          mortality at concentrations above 2.0 mg/L will be placed  into toxicity category IV.

          Corrosive test materials that cannot achieve appropriate particle distributions and/or
          particle concentrations may be placed into toxicity category I if there is no other
          evidence indicating low toxicity. Other evidence would likely be the result of an acute
          oral toxicity study.

     6.   Studies that do not measure the test material concentrations from the breathing zone of
          the animals will not be accepted.  Studies not measuring the test material concentration
          from the breathing zone will not be taking representative samples.

     7.   Studies that do not provide a continuous four hour test material exposure.  The
          exceptions to this rule are studies with a 100% mortality before the end of the exposure.


       If a 100% mortality level is found and the study does not allow the placement of the
       product into toxicity category I, other concentrations must be tested. A mean or time
       weighted average should be used if more than two measurements were taken and
       significant variability exists between the measurements.

       Studies that report analytic or gravimetric concentrations that are higher than  the
       nominal  concentration.  In these situations, the laboratory appears to be measuring more
       test material than was introduced into the test chamber.

       Studies displaying mortality with the oxygen concentrations below 19%  or fewer than
       ten Chamber air changes per hour, A study may have an insufficient oxygen level due
       to the concentration of the oxygen hi the supplied air or-due to an insufficient number of
       air changes.  If such a study reported mortality, it could not be stated that the mortality
       was solely attributable to exposure to the test material.  Studies whose only deficiency is
       the lack of reporting the oxygen concentration or number of chamber air changes per
       hour may be accepted.

       Failure to remove excess test material  from animals' coats after exposure. An effort
       should be made to remove excess test material. Removal of excess material eliminates
       concerns of exposure to the test material from ingestion or dermal absorption. Delayed
       deaths  cannot be solely attributed to  inhalation of test material if this precaution is not

       Animals  (bv volume^ occupy more than 5% of the exposure chamber. This requirement
       is stated in the Subdivision F guidelines.  Animal volume is calculated assuming that
       each gram of an animal occupies one cc of volume within the chamber.

  12-  Lack of necropsy,  Although it is not a requirement, gross necropsies are

B. Reporting Errors

  1     A lack  of information of the study conditions. The  following is a list of data that must
       be included in the inhalation toxicity study report.  The mformation to be included in
       the report is not limited to the list below.  This list should be used hi conjunction with
       the Subdivision F guidelines.  Failure to report any  of the following factors may cause a
       study to be classified as unacceptable data and thus be rejected until receipt of the
       mformation by the Agency.

   a.  The length (of time) of the exposure  of the test animals to the aerosol.
   b.  The method of exposure of the test animals, e.g., whole body or nose only exposure.
   c.  The chamber rate of airflow, temperature, humidity, and how each was measured.

      d. The actual concentrations of test material in the chamber atmosphere measured from the
         breathing zone of the test animals and how they were measured.  This must be in
         tabular form.
      e. The chamber concentration must be reported in mg/L, not ppm or mg/m3.
      f. Reports should state whether concentrations were measured from the breathing zone.
      g. The rate of air flow, volume of air taken and length of time the air was measured for
         aerosol concentration measurements.
      h. The oxygen content of the chamber air during the exposure.

JndustryJ^omment;  Oxygen content should not be required as long as ten air changes are
done according to EPA guidelines.

EPA Response; The Agency agrees.

      I. The Mass Median Aerodynamic Diameter (MMAD) and Geometric Standard Deviation
      j. The equipment used to determine the MMAD and GSD, the sampling zone, the rate and
         volume of air measured, and the actual values and percentages obtained from each stage
          of the impactor.
      k.  The chamber volume.

     2.    Failure to identify  the test atmosphere generation system.

     3.    The failure to report the time taken to reach 90% or 99% of the desired particle
          concentration.  (The tune needed to reach 99% is more desirable.)
V. Primary Eye Irritation

Industry Comment: We recommend that the Agency allow for the use of relevant human eye
testing and human use data if available.  The Drake eye irritation test is known to be over-
predictive for many chemicals and formulations.  If human data are available, they should be
considered by the Agency.

EPA Response:  The Agency agrees. However, the Agency does not feel that registrants
should seek to conduct human studies instead of animal studies.

   A. Errors in the Conduct of the Study

      1.   Failure to test aerosol products by spraying test material into the eves. Aerosols should
          be sprayed into an eye held open for one second from a distance of 10 cm.  It is
          necessary for aerosol products to be sprayed so that the physical damage that may occur
          because of the  force of the spray may be evaluated. The exception to this rule is total


          release foggers.  In testing most total release fogger sprays, instillation of the liquid
          from the fogger spray will have to be carried out.  Child resistant packaging will be
          required for such products and labeling identifying the physical hazard will be required.

       The Agency will allow industry to develop a system to determine the ocular damage that
       would be caused by the physical force of various aerosol sprays.

          The use of fluorescein stain stopped before resolution of opacity.  Fluorescein aids in
          sighting lesions that otherwise may not be apparent.

         Comment:  Fluorescein does not measure opacity and therefore it may be stopped
before the resolution of opacity. Opacity is recorded as a separate observation. Lack of
fluorescein stabling is not a reason for study rejection; however, the use of this method is

EFA Response: The Agency agrees that the use of fluorescein staining is encouraged, but not
required. Although fluorescein may not measure opacity, if it is used and anomalies are
observed, the use should be continued until the anomalies are resolved. Otherwise, the results
are ambiguous and difficult to interpret.

     3-   The study ended before resolution of all irritation.  For eyes showing irritation at 72
          hours,  scoring should be continued and as conducted as frequently as needed until  all
          irritation has subsided or for 21 days, whichever comes first. The exception to this rule
          is irritation so severe that the test material will obviously be placed into toxicity
          category  I whether the study is continued or not.

     4-   Incorrect volume of test material used, The correct volume is 0.1 ml for liquids, and for
          solid test substances, the weight equivalent of 0.1 ml or a weight of not more than 100
          mg  should be used.  Aerosols should be sprayed into an eye held open for one second
          from a distance of 10 cm.
          The test material was a solid consisting of large granules or other large solid particles,
          and was not ground before instillation. Where possible, test materials should be ground
          to a fine powder before instillation into the eye.  If the solid test material cannot be
          readily ground, but can still be placed into the conjunctival sac, then it should be tested
          as is.
Industry Comment: If the granules are large and not amenable to grinding then a waiver
may be obtained.

EPA Response: The Agency agrees.

     6.   The data submitted was only on animals whose eyes had been rinsed immediately after
         test material instillation. The report must include data on test animals whose eyes were
         not washed out before 24 hours after instillation.

     7.   The test material was diluted before instillation into test animals' eves.

     8.   Studies using animals that display irritation, "background opacity" or  "naturally
         occurring opacity" before test material administration. Such animals should not be
         selected for testing.

     9.   The method of observation was incorrect. Evaluations should be made using white
         light, resembling day light. The use of magnification and fluorescein stain is
         encouraged.  The slit lamp is considered the most accurate way of detecting lesions  in
         the eyes.  Absence of fluorescein staining may not  always signify healing of all corneal
         lesions, because the eye heals from the corneal epithelium inward.

EPA Comment;  There has been considerable discussion concerning acceptable methods of eye
examination in the primary eye irritation study.  Initially,  EPA objected to the use of bright
room light or a pen light. Industry disagreed. Subsequently, industry acknowledged that
room light alone was unacceptable, but that pen light was  still allowable. Following
consultation with experts in this area, EPA generally agrees with the most current industry
position. A pen light is adequate for examining changes hi and near the eye. It gives a
bright enough and focal enough light to see inflammation, engorgement of vessels,
discoloration, etc. However, a slit lamp allows the viewer to see where the injury is in depth
through the clear optical tissues and fluorescein stain is important for looking for
compromised epithelium. That is why  the use of magnification and fluorescent stain is
encouraged. If observation of the reactions is aided by the use of a binocular loupe, hand slit
lamp, biomicroscope or other suitable device, these findings should be thoroughly described,
recorded and reported.

      10.  Discussion:  How many animals are required for primary eye irritation study?
          Although the Agency prefers six test animals for this study, it will accept three in
          accordance with OECD protocol.

      11.  Discussion: Are scores of "1" considered to be positive?  Scores of 1 for conjunctival
          redness, discharge and/or swelling are not considered to be positive and do not
          influence precautionary labeling.

jTndustry Comment; Animals with only these scores do not need to be examined past 72

EPA Response;  The Agency agrees.

    B. Errors in Reporting

      1-   The ocular grading scale used is not defined in the report. The grading system should
          be the Draize Scale and must be included in the report.

     2.   The method of examination was not reported.  The use of fluorescein stain, anesthetics,
          the type of light used, the frequency of observations, etc., must be reported.

     3-   Disregarding the presence of stippling when it is stained. The Agency may count
          stained stippling as positive ocular irritation and it may affect the toxicity category of a
VI.  Primary SMn Irritation

   A.  Errors in Study Conduct

     1>   The size of the exposure area is incorrect. The appropriate size is 1 in2 (6 cm2).

     2-   An improper amount of test material was used.  The proper quantity is 0.5 milliliters
          for liquids and 0.5 grams for solids.

     3-   The dry (solid, powder) test material was not moistened for application.  The
          moistening of the test material helps it to penetrate the dermal membrane.  The test
          material must be moistened before application to assure uniformity.  The dry test
          material should be slightly moistened (not runny) to a paste using water.  The amount of
          water used to moisten the test material should be stated in the report.
          The test material was diluted rather than moistened without giving an explanation.
          Dilution (over-moistening) of the test material may reduce the irritation obtained.
          The exposure site was not properly occluded.  The exposure site should be semi-
          occluded.  In cases where the test material contains a significant amount of an alcohol or
          petroleum distillate, care should be taken to preclude a reaction between the adhesive
          and the test material.  Interaction between some test materials (containing an alcohol or
          petroleum distillate) and the adhesive could easily change the outcome of the study.  It
          is important that the laboratory not cover the test material with enough gauze to pull
          liquid test material or a moistening agent away from the skin site. Other considerations
          in dermal application are discussed under acute dermal toxicity.

Industry Comment: Please see the specific wording for semi-occlusion under the industry
comments for dermal toxicity.  At the industry meeting in May it was also agreed that use of
an occlusive dressing would not be a criterion for study rejection.


]EPA Response; The Agency agrees.

     6.   The exposure period was not four hours.  The Agency will accept studies with
          exposures of less than four hours if toxicity category I is demonstrated or if the
          exposure period was greater than four hours and minor irritation (toxicity category III
          or IV) was noted.

     7.   The exposure site was not rinsed with water or another suitable vehicle after the four
          hour exposure and severe irritation was observed. Without wiping the exposure site,
          the residual test material may continue to cause irritation and fallaciously increase the
          toxicity category of the product.

     8.   Preparation of the exposure site is incorrect. One example could be that the exposure
          site was not shaved or otherwise cleared of fur.

     9.   Abrasion of test sites.  This is thought to exacerbate the irritation. The study will be
          rejected if the abrasion may have placed the study in a higher toxicity category (I or II).

     10.  Observations were not conducted until irritation had subsided.  Observations should be
          continued until irritation has subsided (no scores other than " 1" for dermal erythema
          and/or edema) or 14 days has been reached. If there is still irritation at 14 days the
          study may be ended. The exception to this rule would be ending a study early when it
          is evident that it will be placed into toxicity category I.

     11.  Test animals were not young adult.   Healthy adults should be used. The test animals
          should be of normal weight for their age.

Industry Comment;  The requirement should read "Healthy adult animals" (not necessarily
young) should be used. As long as animals are healthy, then* weight is not important, it was
agreed in the industry/regulator's meeting in May. There will be no weight restrictions since
there is no scientific evidence that age or weight affects the results of eye or skin irritation
studies. Reference EPA  guidelines.

EPAJEtespojise; As previously noted, there are not weight restrictions.  However, the health
of animals that are not of normal weight for their age  would be questioned.

     12.  Unjustified use of a moistening agent other than water (or saline). Water is the
          preferred agent to be used for moistening dry test materials for primary skin irritation
          studies because it resembles sweating  by humans.  Other vehicles may cause irritation
          by themselves or exacerbate irritation caused by the test material. Other moistening
          agents are not acceptable. The study may be upgraded upon subsequent explanation of
          the choice in vehicle. Liquid materials should be tested undiluted.

Industry Comment:  Although water or saline is the preferred moistening agent, other agents
may be used providing the use is justified.  Acceptable alternatives are: gum arable, ethanol
+ water, carboxymethyl cellulose, glycerol, vegetable oil and mineral oil. These can be used
if water or saline cannot be used as long as the vehicle is not irritating and the inability to use
water or saline is justified hi the report.

EPA Response:  The Agency agrees.

      13.  Too few animals ..were used.  The Agency prefers that six animals be used in the
          primary skin irritation study; however, in accordance with OECD criteria, the Agency
          will accept prunary skin irritation studies conducted with only three animals. Studies
          where the laboratory can demonstrate toxicity category I using fewer than three animals
          will be accepted by the Agency.
      14.  Discussion: Waivers.  The primary skin irritation study may be waived for products
          with pHs of 2 or less, or, 11.5 or above. These products will be placed into toxicity
          category I by default. If the registrant wishes to have such a product placed into a
          toxicity category other than  I, he must have a primary skin irritation study conducted to
          prove that a toxicity category I is  inappropriate for such a product.

     The primary skin irritation study may also be waived if the material was no more than
     slightly irritating after the 24-hour acute dermal toxicity test exposure.  The sites  will have to
     have been scored using the Draize Scale. Such products will be placed into toxicity category

   B. Errors in Reporting

      1.   The irritation scoring scale used by the laboratory is not reported. Dermal irritation is
          usually reported numerically. When the scale used for grading is not identified or
          included hi the report, the Agency cannot be sure of the irritation observed.

     2.   The method of preparation of the  exposure  site is not reported. The report should state
          how the hah- was removed from the exposure site, etc.

     3.   The location of the  exposure site is not reported. The report should state  where on the
          animal's body the testing site was located.

     4.   Insufficient detail or imprecise reporting of observations.  For example, stating that the
          exposure site is exfoliated, and giving no further information is vague.  Exfoliation can
          be desquamation or sloughing. Quite often, reports fail to go into detail or fail to
          attempt to give a visual representation of the irritation observed when such irritation
          goes beyond erythema or edema.

Industry Comment;  For consistency the Agency should provide a definition of the descriptive
terms for exfoliation, desquamation and sloughing.

EPA Response; The laboratory should define the terms used in the reporting of observations.

     5.   Failure to report the condition of the skin after sloughing.  Sloughing very often leaves
          the skin thickened and denuded (hyperplasia or  hyperkeratosis and dead skin follicles).

     Observations should clearly show whether the  skin has returned to "normal"  or whether the
     skin has been irreversibly changed, i.e., scarred, blanched or denuded. Reversibility of skin
     reactions is as important as the primary skin irritation index.

    . Dermal Sensitization

    Discussion: When is the study required? The dermal sensitization study is required when
    there will be opportunity for repeated exposure to the product.  EPA will waive the dermal
    sensitization study for formulations classified as toxicity category I for dermal toxicity or
    dermal irritation.  Products placed  into toxicity category I for dermal toxicity or dermal
    irritation with use dilutions that do  not require the use of protective clothing will also require a
    dermal sensitization study.  The sensitization study is always required for the technical
    materials, regardless of the toxicity or irritation  properties of that technical.  In Canada, the
    sensitization study is always required unless the  use dilution is also irritating.   The California
    Department of Pesticide Regulation does not routinely  require skin sensitization studies to
    register pesticides. Products may also be categorized as sensitizers  if components of the
    formulation are known sensitizers.

    The dermal sensitization study is the only one of the six acute  toxicity/irritation studies that
    may be conducted using one of several different techniques. Subdivision F Guidelines  offer a
    choice of methodologies.   The seven testing methods accepted by the Office of Pesticide
    Programs for dermal sensitization studies are:

      1.    The Buehler Method/the Modified Buehler Method.  This is by far the method
           submitted most  often to OPP for dermal  sensitization.  Recommended references for  this
           study are:

       a.  Ritz, Harry L. and Buehler, Edwin V. "Planning, Conduct, and Interpretation of
           Guinea Pig Sensitization Patch Tests," in Current Concepts  in Cutaneous Toxicitv.
           V.A. Drill and  P. Lazar (eds.), Academic Press, New York, 1980, P. 25-40.

       b.  Robinson, et al. "A Review of the Buehler Guinea Pig Skin Sensitization Test and
           Its Use in a Risk Assessment Process for Human Skin Sensitization," Toxicology.
           vol. 61, 1990, P.  91-107.

  c.  "Experimental Skin Sensitization in the Guinea Pig and Man," Animal Models in
      Dermatology, H.I. Maibach (ed.), Churchill, Livingstone, Edinburgh, 1975  P

  d.  "A Rationale for the Selection of Occlusion to Induce and Elicit Delayed contact
      Hypersensitivity in the Guinea Pig," in Current Problems in Dermatology.  E.V.
      Buehler, vol. 14, Karger, Basel, 1985,  P. 39-58.
  e.  Buehler, E.V. "Occlusive Patch Method for Skin Sensitization in Guinea Pigs: The
      Buehler Method," Food and Chemical Toxicology vol.  32, no. 2,  1994, P.97-101.

 It is recommended that the laboratory personnel familiarize themselves with each of the five
 citations listed above before conducting Buehler Method studies.

 2.    The Guinea Pig Maximization Test is another method accepted by OPP for the
      determination of dermal Sensitization.  A recommended reference for this study is:  "The
      Identification of Contact Allergens by Animals Assay. The Guinea Pig Maximization
      Test," The Journal of Investigative Dermatology  vol. 52, no. 3, 1969, P. 268-276.

 3.    The Split Adjuvant Technique.  A recommended reference for this study is  "The
      Bioassay of Contact Allergens in the Guinea Pig," J.  Sac. CosmetChem.: vol. 24
      March 1973, P. 151-162.

 4.    The Open Epicutaneous Test. A recommended reference for this method is "Screening
      of Fragrance Materials for Allergenicity in the Guinea Pig," Journal of the  Society of
      Cosmetic Chemists, vol. 28, February 1977, P. 53-64.

 5.   The Maurer Optimization Test.  A recommended reference is Maurer, T., et al. "The
     Optimization Test in The Guinea Pig: Method for the Predictive Evaluation of the
     Contact Allergenicity of Chemicals," International Congress Series. Excerpta Medica:
     #376, 1975.

6.   Freund's Complete Adjuvant Technique

7.   The Footpad Technique in the Guinea Pig.

Freund's Complete Adjuvant Test, the Guinea Pig Maximization Test, The Split Adjuvant
Technique, the Buehler Test and the Open Cutaneous Test are discussed in "Identification of
Contact Allergens: Predictive Tests in Animals," Dermatotoxicology and Pharmacology, F.
Marzulli and H.I. Maibach, eds.

The Buehler Method and the Guinea Pig Maximization Test are the two skin sensitization
methods that are preferred by the Agency. If another testing method is used, the tester should
provide the reasoning for their choice of technique.

Below are study deficiencies that have been found hi dermal sensitization studies conducted by
the Buehler Method.  The Agency chose to focus on the Buehler Method because it is the
dermal sensitization technique most often submitted to the Agency.  The following Buehler
Method study deficiencies are separated into errors in the conduct of the study and reporting

A.  Errors hi Study Conduct

  1.   Failure to select the proper  induction and/or challenge concentrations from the primary
       irritation screening.  An improper induction concentration (too low) may fail to evoke
       sensitization with a substance that may have otherwise brought about sensitization.

  Too high a challenge concentration is a problem that is  frequently seen hi dermal sensitization
  studies.  Laboratories sometimes choose a challenge concentration that elicits irritation in
  unsensitized animals. When the  laboratory chooses an irritating concentration of test material
  for challenge, it may not be possible to determine whether the irritation noted was a result of
  a sensitization reaction or simply dermal irritation.

  Another mistake that is frequently seen by the Agency is the use of induction concentrations
  that are too low. The goal is to evoke a dermal response hi each of the test subjects in the
  induction phase of the study.  A reaction causing mild to moderate irritation in the is
  preferred. When a test material concentration fails to evoke mild to moderate irritation, the
  lab should increase the concentration of test material (if possible)  during the subsequent
  induction treatments, definitely before the induction phase is over. Dermal sensitization
  studies  that use a percent of test material that fails to evoke irritation in the induction phase,
  when that percent is not the most concentrated available, will be rejected.

  The Agency encourages the use of multiple challenge and induction screens,  if necessary, to
  determine the appropriate induction and challenge concentrations.  In addition, the use of a
  depilation step during the challenge screen is appropriate.

  2.    Failure to report the results of the primary irritation screening. Without the results of
        the primary irritation screening, it is not possible  (or at least difficult) to assess the
        choice of induction and challenge concentrations.

  3.    The laboratory used the wrong vehicle for the  test material.  Although water or saline is
        the preferred moistening agent, other agents may  be used providing the use is justified.
        Acceptable alternatives include: gum arabic, ethanol + water, carboxymethyl cellulose,
        polyethylene glycol, glycerol, vegetable oils and ethanol for induction with acetone for


      challenge.  These can be used if water or saline cannot be used as long as the vehicle is
      not sensitizing and the inability to use water or saline is justified in the report.  Mineral
      oils and petrolatum should be avoided when possible because they may induce irritation
      independent of any test materials.

 4-    Use of ethanol as a vehicle for both induction and challenge.  Buehler has stated that
      ethanol alone can induce sensitization.  When sensitization reactions occur with the use
      of ethanol as the vehicle for both induction and challenge, it is not possible to decide
      whether the reactions were a result of the test material or the vehicle.  Please refer to
      "A Review of the Buehler Guinea Pig Skin Sensitization Test and Its Use in a Risk
      Assessment Process for Human Skin Sensitization," Toxicology. Robinson et al  vol
      61, 1990,  P. 91-107.

 5-    The study  does not contain appropriate  controls:

  a.  A positive control conducted within 6 months of the study on the test material must be
      included.  The positive control study must demonstrate sensitization.  The entire study,
      not just summary data, should be submitted. The positive control study is used to prove
      a laboratory's ability to properly conduct a dermal sensitization study. The positive
      control study should be conducted hi the same manner as the mam sensitization study.
      For example, the Buehler Method should be followed if that was the method of the main
      sensitization study.

  b.  All studies should include a vehicle control group.  This group would be exposed only
      to the vehicle during induction,  but would be challenged with the test material in the
      same manner as the test group.

  c.  A vehicle control group (where  the animals were not previously exposed to any  vehicle
      or test material) should be used  if it is suspected that the vehicle is an irritant, the
      laboratory  has little information on the irritant/sensitization potential of the vehicle, or
      when neat  (undiluted) test material has been used for induction. A vehicle control will
      generally not be necessary when using one of the vehicles detailed hi #3 above.
      However, when a unique vehicle is used, the laboratory will have to use a vehicle
      control where the animals were  induced using the vehicle and challenged with the same
      vehicle/ test material mixture as those animals in the main study.

  d.  A second set of naive and/or vehicle controls is needed if a rechallenge is conducted.

6-    Induction and challenge performed on the same exposure site.  Skin fatigue may cause
      irritation hi this situation that would falsely appear to be sensitization.

7-    Use of the  same concentration (other than 100% for nonirritating test materials^ for both
      induction and challenge.  The laboratory must conduct the induction at an at least


     minimally irritating concentration. Challenge should be conducted at the highest
     nonirritating concentration.  The only time the same concentration may be used for both
     induction and challenge is if testing a nonirritant and a 100% concentration is selected
     for both induction and challenge, or if the lab is testing a moistened solid at its least
     diluted concentration.

     A preliminary irritation screening study should be conducted to determine a mildly to
     moderately irritating concentration for induction as well the maximum non-irritating
     concentration to be used at challenge. Non-irritants should be tested undiluted.  Solid
     non-irritants should be tested at the most concentrated dilution possible. Although it is
     possible to induce sensitization at a non-irritating concentration, a dermal reaction
     causing mild to moderate irritation (scores of 1 to 2) in the induction phase is preferred.
     Moderate or severe irritation hi the induction phase is acceptable and will not cause a
     study to be rejected.  Concentrations of 100 %, 75 %, 50 % and 25 % are acceptable for
     the irritation screen.  Significant irritation at  the lowest concentration would necessitate
     a second range-finding study.

     The challenge concentration should be lower  than the induction concentration, unless
     the undiluted test material is non-irritating. Ideally, this concentration should cause no
     more than 50%  +/- scores (if the Buehler method is used).

8.   The chosen protocol was not followed.  The test deviated from the procedures as
     recommended.  Refer to VII (1) above.

9.   Rechallenge was not conducted when appropriate. Rechallenge is needed when the
     challenge results are equivocal. For example, 5/10 +/- or .5 scores hi the test group
     without reactions hi the naive control group,  or 9/10 +/- or .5 scores and 1/10 grade 1
     scores in the test group  with 5/10 +/- or .5 scores hi the naive control group.

10.  A different exposure time was used for  induction and challenge and a naive control
     group was not included. The preferred exposure tune  is six hours.  Preliminary
     screening exposure must also be six hours. Some laboratories tend to use exposure
     concentrations of 24 hours. While this  is generally acceptable, it may lead to problems
     if there was a 24 hours prescreen with too little irritation during a six-hour exposure

11.  Using different scales to grade different parts of a study. It is recommended that
     laboratories use the Buehler Scale (for the Buehler Method). It is also acceptable to  use
     the 0, 1, 2, 3 scale from the Magnusson & Kligman testing method.   As there is more
     than one version scale, reference to which scale was used should  be provided. This  is
     acceptable as the EPA has  agreed to accept OECD protocols.  The rating scale should
     be described hi the report.

     12.  The solid test material was not moistened or was not properly moistened.  Solids must
          be slightly moistened (not runny) to a paste, preferably with water or saline.  Other
          vehicles, including corn oil, may be used if justified and if they do not cause irritation.
          Suspensions and emulsions may be diluted with water provided that an acceptable
          irritation dose-response relationship can be demonstrated and provided that the induction
          and challenge concentrations fulfill the  appropriate criteria.  If not, then the test material
          should be dissolved, preferably in acetone or alcohol, although other non-irritating
          vehicles may be used if justified.

     13.  The volume of test material was too small.  The recommended dosage  is 0.4 ml or 0.5
          ml  for liquids and 0.4 g or 0.5 g for solids. A volume of 0.3 ml may be appropriate if
          a Hilltop Chamber is used.

Industry Comment: The references listed above in #13 should be followed. Exact  volume
may vary, since the references specify that the patch should be saturated so that there is a
maximum concentration at the interface of the patch and the skin.