Learning Center
Plans & pricing Sign in
Sign Out

General principles for classification and labelling of repeated



More Info

                             Additional guidance
  for the assessment of evidence, and criteria to apply, for the classification
        and labelling of chemicals which cause repeated dose toxicity

Introductory remark
The discussion on this subject has not been finalized in Germany so far. Therefore, this paper
is dealing mainly with general principles of data evaluation and interpretation.
Moreover, as a first step into the controversy about the objectives of classification, this paper
is directed to elucidate the criteria for classification of chemicals with locally irritating effects
after repeated exposure with regard to the application of R37 or R48 as provided in Annex VI
of Directive 67/548/EEC (below referred to as Annex VI).
Different strategies in the hazard assessment and addressing are set out below as Approach 1
or Approach 2.
It is not the intention of this paper, for the time being, to deal in particular with other effects,
e.g. neurotoxicity, endocrine disruption or test results of reproductive toxicity studies.
It is generally agreed that the assessment of the SPECIFIC TARGET ORGAN/SYSTEMIC
classification criteria for repeated dose toxicity are qualitative or semi-quantitative, expert
judgement is required to interpret the data in the light of all available toxicological
information on a substance. The integrative consideration of functional and morphological
end points (i.e. correlation between clinical findings, clinical pathology parameters as well as
gross and histopathological effects) is essential.
An expertise on the assignment of R48 should include the aspects as set out in Annex I of this
submission. The check list form has the advantage of a clear argumentation line to be
The experts also agree that classification should be disconnected from any automatism in
setting downstream regulations. Taking regulatory consequences into account puts an undue
burden on the classification process.

Aim of the paper
To achieve harmonization of the hazard assessment by the different committees responsible
for the evaluation of toxic effects, either dealing with new chemicals or with existing
chemicals, additional guidance is suggested to give reasonable interpretation of human
experience or results from repeated-dose toxicity tests in animals in order to decide as to if the
criteria on the application of R48 are fulfilled.

fbcf12c1-b85f-470b-8c07-24040360b3cd.doc                                                               1
September 00
Six major problems with the assignment of R48 have been identified:
       Intention of R48 assignment
       Validation of data
       SAR
       Dosage
       Seriousness of health effects
       Assessment of local effects
This paper reflects the discussion within Germany and should contribute to a harmonised
interpretation of Annex VI within the EU Classification and Labelling Working Group.

Intention of R48 assignment
The general introduction of Annex VI section 3.1.4 includes the statement: "When the
classification is to be established from experimental results obtained in animal tests, the
results should have validity for man in that the tests reflect, in a appropriate way, the risks to
man". Classification and labelling should “.. take(s) account of all potential hazards which are
likely to be faced”. However, with respect to a “slim classification/labelling” there is a clear
advice in the following:“Although the final choice of the most appropriate risk and safety
phrases is primarily governed by the need to give all necessary information, consideration
should also be given in the clarity and impact of the label. With clarity in mind, the necessary
information should be expressed in a minimum of phrases” (Annex VI section 7.1.2).
Effects of toxicological significance (adverse effects) at or below the dose range of guidance
levels for each of the standard administration routes (oral, inhalation, dermal) in test animals
indicate a danger of serious health effect after repeated exposure. The integrative
consideration of functional and morphological end points (e.g. correlation between laboratory
and pathological data) is essential.
According to the general principles developed within the internationally harmonised
classification system (OECD, 1998) hazard classification processes refer to hazards arising
from intrinsic properties of chemicals and mixtures thereof. R48 is intended as a warning of
danger of serious health effects after prolonged exposure.
Repeated-dose toxicity has to be evaluated, interpreted, and addressed as a separate endpoint
of its own. Quantitative consideration of long-term exposure/response-relationship (e.g.
deriving acceptable exposure limits) is part of a risk assessment and does not belong to the
objectives of classification and labelling.

fbcf12c1-b85f-470b-8c07-24040360b3cd.doc                                                         2
September 00
Validation of data
Human data on chronic effects of chemicals are scarcely available and sometimes of poor
quality. Hazard identification for repeated dose toxicity is predominantly based on subacute,
subchronic or chronic studies conducted in experimental animals as outlined in the “General
Introduction” of Annex VI.
Standard test guidelines have been developed using the rat as the preferred test species if not
otherwise indicated (tests methods B.7-9, 26-30, 32, and 33 of Annex V 67/548/EEC).
Data have to be evaluated for relevance and validity on a case by case basis (see Annex VI
section 1.7.2). Studies conducted under GLP conditions may, as a rule, considered to be valid.
The evaluation of repeated dose data on existing chemicals are often hampered because
studies do not fulfil current standard requirements. Evaluation should be based on all
available data, including peer-reviewed publications and additional data from authorised study
reports. The adequacy of these studies is to be considered in the procedures on hazard
identification and classification. Minor differences from a standard study design may be of no
relevance, e.g., lack of a single test parameter. For major differences, it has to be proven
whether there is sufficient evidence on a particular repeat-dose toxic effect. If this effect is
identified in more than one study and/or more than one species, this can be regarded as
sufficient evidence.
Species-specific effects in test animals, that are demonstrated to be plausible, and spontaneous
alterations within a well known normal biological variation are not relevant for classification.
Data from concurrent control animals are of higher value than data from historical control
database except they are plausibly invalid.

The results of validated structure-activity relationships and expert judgement may also be
taken into account where appropriate.Generally, they are not sufficient as such for
classification with R48 on their own.

With respect to repeated dose toxicity, substances are classified on the basis of the weight of
all evidence available, including the severity of effects, the duration and the level of exposure
in comparison to the guide values as set out in Annex VI section 3.2.2 to 3.2.4.
Historically, the criteria for R 48 were developed in 1983 (83/467/EEC) assuming that toxic
effects show a time-dependent increase in severity due to cumulation of toxicity.
The guide values of relevant dose levels are intended for the standard design of a 90-day
study and can be extrapolated to other treatment schedules following the assumption that the
product of dose x time keeps constant.
In many cases the selected doses may be above or next to the guide values. If it is considered
to be likely that toxic effects may also occur at doses in the range of guide values, which have
not been tested, classification should also be taken into consideration.
Otherwise, extrapolation of dose-related effects may also be necessary when the maximum
dose applied in repeated dose studies is far below the critical range of the guide values.
Range-finding studies may have identified the maximum tolerated dose that justifies the low

fbcf12c1-b85f-470b-8c07-24040360b3cd.doc                                                            3
September 00
level tested in the main study. Those range finding studies at higher levels might provide
useful information and should be considered.
In cases where doses in the lower range of the guide values give indications of toxicity that
would not lead to classification, the knowledge of serious health effects at higher doses
observed in the range-finding studies has to be taken into account. If dose-effect relationship
and pathogenetic experience show a reasonable biological correlation, classification with R48
has to be considered.
If the effect is already covered by acute toxicity testing after single exposure, classification
with R48 should not be applied. That means an identical effect, e.g. mortality, was observed
in acute studies and repeated dose studies at the same dose/concentration range and at
comparable time frames.

Seriousness of health effects
According to Annex VI section 3.2.4. for the application of R48 serious damage to health is to
be considered to include death, clear functional disturbance or morphological changes which
are toxicologically significant. It is particularly important when these changes are irreversible.
It is also important to consider not only specific severe changes in a single organ or biological
system but also generalised changes of a less severe nature involving several organs, or severe
changes in general health status.
In general, toxic effects are irreversible in tissues/organs with no or very limited potential for
regeneration and/or functional reserve capacity.
Irreversibility is not an imperative criterion for the classification with R48. But, persistence of
toxic effects within the recovery period of repeated dose toxicity studies requires specific
consideration. However, irreversibility of lesions is not limited to these organs; also lesions in
other tissues/organs may have the potential being non-reversible or being reversible to a
limited extent. This is depending on the progress of pathogenetic processes, the severity
and/or the distribution of a lesion.
Evaluation of morphological damage and/or functional impairment of tissues/organs, which
have a substantial potential for regeneration and high functional reserve capacity, is done
under consideration of quantitative aspects.
Quantification of adverse effects includes the incidences (events per number of animals per
group), the severity grades of the lesion, and the extension of a lesion (e.g., focal, multifocal,
diffuse, unilateral or bilateral in paired organs). In addition, the dose/time-relationships of all
quantitative parameters may be of considerable importance. For lesions that may occur
spontaneously, the test substance-related increases of incidences, severity grades, and/or
distribution in comparison to those in control groups is to be considered.
Adaptive responses, e.g. hypertrophy of liver cells due to increased metabolic activity, are not
considered as adverse effects unless they are either associated with relevant functional
impairment of the tissue/organ or irreversible.
Approach 1
Adverse effects occurring in organs with no or very limited potential for regeneration and/or
functional reserve capacity are always regarded as serious health effects (e.g. toxicity to the
nervous system and sensory organs) irrespective of quantitative aspects. For example,
degeneration of olfactory epithelium in a subchronic animal study generally will be

fbcf12c1-b85f-470b-8c07-24040360b3cd.doc                                                              4
September 00
considered as a serious health effect unless there are additional data showing that the effect is
Each toxic effect has to be considered in view of its probable progression or outcome after
prolongation of the exposure. A sequence of events that are expected to lead to serious health
effects are to be considered for R48 assignment.
In cases where functional changes occur without morphological alteration (or vice versa) each
effect may represent an adverse effect by its own. Health impact and progression of the lesion
have to be carefully considered. If there is indication for a serious damage to health R 48
might be warranted.
Approach 2
The criteria for R48 assignment were derived from acute toxicity criteria assuming that
systemic effects show a time dependent increase of severity due to cumulation of toxicity and
taking also adaptive and detoxification processes into account. The effect considered in this
context was lethality. This indicates that classification was intended for the presence of severe
health damage, only.
Irreversibility on its own should not lead to classification without considering extent and
severity of the respective lesion and its relevance for man.
If morphological changes are not accompanied by severe functional deficits, classification
might not be warranted. For example focal degeneration of olfactory epithelium in subchronic
animal studies is not considered to represent a severe health effect, because this kind of lesion
is reversible (Gaskell 1990, Eldridge et al. 1995) and even pronounced lesions are usually not
accompanied by the loss of the sense of smell (Evans et al. 1995, Dorman et al. 1997)

Assessment of local effects
Approach 1
In general, toxic effects in repeated dose studies are not covered by other toxic endpoints such
as acute toxicity, sensitization or carcinogenicity. A common mode of action may sometimes
be anticipated to be responsible for effects on several toxic endpoints. Irrespective of this,
classification should reflect all possible hazards for human health.
Effects on superficial epithelia along the exposure routes are not covered by local studies on
skin or eye. Results from these studies are not predictive for lesions after repeated dose
exposure but may indicate concern. It might be assumed that substances with marked irritant
or corrosive properties will produce local effects on tissues lining the portal of entrance.
However, critical concentration ranges may be extremely different for different substances.
Moreover there is no scientific basis to predict whether these effects will increase by
extension of exposure duration. Thus local effects need to be evaluated under the aspect of
chronic toxicity.
Conventional animal studies on acute toxicity do not give indication on the nature of lesions
as long as they do not include appropriate diagnostic methods (e.g. histomorphology).
Toxic effects on surface epithelia may reflect direct effects on the site of contact. They also
may reflect indirect effects as consequence to systemic toxicity (e.g. uremic gastritis) or
secondary to systemic distribution of the test substance or its active metabolite(s). Commonly
the underlying mode of action is not clarified by routine toxicity studies.

fbcf12c1-b85f-470b-8c07-24040360b3cd.doc                                                            5
September 00
Local effects at superficial epithelia along the exposure routes (oral, inhalatory) or at the
exposure location (dermal) observed in repeated dose studies are considered to represent acute
effects if the nature of the lesion is comparable, and the lesion occurs at comparable
dose/concentration ranges as after single exposure and at comparable time scales after first
exposure. For these effects labelling with R 37 should be discussed.
All other local effects at the target epithelia mentioned above at doses/concentration below
that relevant for acute toxicity or local effects on skin and eye showing delayed occurrence
(non acute) are considered to be specifically related to the repeated or prolonged exposure.
Identical criteria for classification are applied for repeated-dose toxic effects on all
tissues/organs including superficial epithelia along the exposure routes. A separation into
different evaluation procedures with respect to a certain organ is not justified.
Approach 2
Generally, repeated dose toxicity has to be evaluated and interpreted on its own, unless the
overall evaluation of toxicity indicates common mechanisms of action in different endpoints.
Thus, toxic effects already covered by another R-phrase being based on the primary mode of
action should not warrant R48 additionally.
Information and data concerning possible mode of action should be considered for the final
assessment. A considerable database, human experience included, exists for local irritant
effects of a number of chemicals.
Repeated inhalation of corrosive or strongly irritating substances in irritating concentrations
cannot be expected due to the warning effect and thereby not be regarded as a condition of
“normal handling and use”.
It is an expected finding that materials with marked primary irritant or corrosive properties
produce irritation in tissues at the portal of entry depending on their pattern and magnitude of
deposition. For example, changes in the upper respiratory tract caused by vapors of
irritant/corrosive materials are frequently observed at concentrations below the classification
limit for R 48. A considerable number of materials (acids, amines, aldehydes and other
reactive compounds) would be candidates for labeling with R48. As the test guidelines on
toxicity studies require the elicitation of toxic effects at least at the high concentration, it is
obvious that the primary effects of irritant/corrosive materials are likely to occur at the site of
Isolated local effects on the skin induced by repeated administration of irritant or corrosive
substances should not lead to classification with R 48 (Annex VI section 3.2.4, 2(d)). This
exemption should be extended to other portal of entry effects (e.g. irritant gastritis after
gavage, epithelial lesions in the respiratory tract after inhalation).
Classification of local effects can be based on single or repeated exposure data which indicate
clear signs or symptoms of irritation. Marked irritant or corrosive materials should be labelled
with R phrase 37 if there is a relevant exposure potential due to their physical properties
(volatility, particle size) and/or if relevant data on respiratory irritation in real world scenarios
exist. If substances are already labelled concerning their irritancy/corrosivity as the
underlying mechanism for effects on the upper respiratory tract, an additional labelling with
R48 in the absence of independent systemic toxicity would be misleading. The same applies if
lung damage after inhalation of vapors or more frequently aerosols of irritant/corrosive
materials is present and the damage can be clearly attributed to the irritant/corrosive
fbcf12c1-b85f-470b-8c07-24040360b3cd.doc                                                              6
September 00
Local effects are to be clearly distinguished from systemic or target organ response, which
may have wider systemic toxicological consequences. For severe lung effects, based on other
modes of action, e.g. chronic inflammation and fibrosis caused by intrinsic properties of the
substance (e.g. silicosis), the above mentioned exemptions are not applicable.

67/548/EEC: Council Directive 67/548/EEC of 27 June 1967 on the approximation of the
laws, regulations and administrative provisions relating to the classification, packaging and
labelling of dangerous substances.
OECD 1998: Harmonized integrated hazard classification system for human health and
environmental effects of chemical substances. 28th Joint Meeting of the Chemicals Committee
and the Working Party on Chemicals in November 1998.
OECD 2000: Revised step 2 Proposal for the classification of chemical substances that cause
specific target organs/systemic toxicity following repeated exposure. 3rd Meeting of the Ad
hoc Expert Group on Target Organ Systemic Toxicity of the Task Force on Harmonisation of
Classification and Labelling, 7th-8th September 2000. ENV/JM/HCL (2000)13/REV2
Gaskell B.A., Nonneoplastic changes in the olfactory epithelium – experimental studies,
Environ. Health Perspect. 85, 275-289, 1990
Eldridge S.R. et al., Effects of propylene oxide on nasal epithelial cell proliferation in F344
rats, Fundam. Appl. Toxicol 27, 25-32, 1995
Evans J.E. et al., Behavioral, histologic and neurochemical effects of Nickel (II) on the rat
olfactory system, Toxicol. Appl. Pharmacol. 130, 209-220, 1995
Dorman D.C. et al., Chloroform-induced olfactory mucosal degeneration and osseous ethmoid
hyperplasia are not associated with olfactory deficits in Fischer 344 rats, Toxicology 122, 39-
50, 1997

fbcf12c1-b85f-470b-8c07-24040360b3cd.doc                                                        7
September 00
                                                                                Annex I

    Questions which should be answered if reasons are given for classification
                                  with R 48

Which study design has been used to obtain the data relevant for classification?
    animal species
    route of administration
    study duration
    scheme of dosing
    validity of the method used / sensitivity of the study

Which of the observed effects give reason for classification?
    nature of the lesion (type of effect, tissue, organ)
    severity grades
    extension of the lesion
    significance of the effect (comparison with control group)
    dose- and time-response relationship
    lowest effect level at which the lesion is judged to be serious

What biological relevance has to be assumed for the observed effects?
    toxicological relevance of the morphological damage
    functional impairment of tissues/organs
    probable progression or outcome
    reversibility
    effects on life quality and expectancy
What are the reasons to believe in a chronic character of the observed health effects?
    comparison with acute toxicity
    time dependency

Comparison with classification criteria
Which aspects have to be considered in applying the classification criteria on the study
   relevant guidance dose level on the background of the available data
   sensitivity differences between men and experimental animals
   substance-specific modifying factors (e.g. study duration, severity or seriousness of the
    effect, slope of dose response curve, weight of evidence).


To top