You are on page 1of 8

226 Scandinavian Cell Toxicology Congress 11

Validation ofIn Vitro Cytotoxicity Tests - Past


and Present Strategies

Bjorn Ekwall 1, Inger Bondesson", Sven Hellberg3, Johan Hogberg", Lennart


Romert'', Kjell Stenberg" and Erik Walum7

'Department of Toxicology, Biomedical Center, University ofUppsala, Box 594, S-751 24


Uppsala, Sweden;2SwedishPoisonInformation Center, KarolinskaHospital,Box60500,S-104
01 Stockholm, Sweden; 3Research Group for Chemometrics, Department ofOrganic Chemistry,
University of Umea, S-901 87 Umea, Sweden; "National Institute of Occupational Health,
Ekelundsoiigen. 16, S-171 63 Solna, Sweden; 5Department ofGenetic and Cellular Toxicology,
Wallenberg Laboratory, University ofStockholm, S-1 06 91 Stockholm, Sweden; "Department
ofMedical Radiobiology, KarolinskaHospital, S -10401 Stockholm, Sweden; 7Department of
Neurochemistry and Neurotoxicology, University ofStock holm , S -10691 Stockholm, Sweden

Summary - In recent years, conventional toxicity testing in animals has been reinforced by
in vitro methods. As a result, toxicity testing in some sectors has become more effective and at
the same time more ethical. This trend is probably only at its beginning, as many of the
newly-developed methods have not yet won general acceptance as a basis for the large-scale
replacement and reduction of animal experimentation. What limits the wider use of these
methods is validation, i.e. the evaluation of their reliability and relevance. The present paper
is a short review of the validation efforts made hitherto, including projects being planned and
under discussion. Our own MEIC approach is compared with other strategies. Finally, our
opinion on the effectiveness of one large consensus project relative to several different smaller
validation programmes is expressed - we advocate the latter strategy, because it will save time
and reduce costs.

Key words: cytotoxicity assays, evaluation, toxicity testing, validation.

Introduction actual performance of validation has been


modest. However, in the last few years, new
Progress in the in vitro testing of general strategies have been proposed and discussed
(non-genetic) toxicity depends on the (3-5) in parallel with the practical start of
development of new methodology, as well as many new programmes (6-12).
on evaluation of the relevance and reliability
of the new methods. The latter task is
commonly called validation, and seems to be Single Laboratory Validation
more difficult than method development.
Thus, several hundred different in vitro The mere description of a new in vitro test,
assays of general toxicity have been proposed even including in vitro/in vivo toxicity
in the last twenty years, but only very few of comparisons for 10-20 chemicals, cannot be
them have been validated and accepted. It is considered as validation. However, some
generally believed that validation is the key developers of tests have passed that stage and
to extending the use of in vitro methods as performed validation of their methods by
supplements or alternatives to the animal testing and comparing a larger number of
testing of the systemic, target organ and local chemicals (e.g, 50-100), including
toxicities of chemicals. Since the recognition collaboration with other laboratories in blind
of validation as a key process in the early studies for evaluation of the method's
1980s (1, 2), the speed and efficiency of the interlaboratory variability. In some
Scandinavian Cell Toxicology Congress 11 227

instances, such single laboratory validation toxicokinetic data (18, 21, 22).
has led to the practical use of the method in These validation studies have led to the
industry. Recent interest in multicentre widespread acceptance of in vitro cytotoxicity
programmes must not make us forget the tests by industry (23), as so-called prescreens
contribution of such validation to of acute toxicity for use during product
achievements in the field, some examples of development as well as a vital part of the
which will be presented in the following detailed investigations of pharmaceuticals
paragraphs. (24, 25), to be conducted along with animal
tests, toxicokinetic studies and
Local irritancy of implants biotransformation studies (often carried out
on human hepatocyte cultures).
As early as 1965, Rosenbluth et al. (13)
demonstrated that a mouse cell line toxicity
assay resulted in a reasonably good prediction Eye irritancy
of the irritancy ofintramuscular implants and As a result of animal rights campaigns in the
injections in rabbits. In their study, early 1980s against the performance of the
qualitative (+/-) in vitro/in vivo comparisons Draize eye irritancy test in rabbits, academic
were made for 112 samples (13). Many false scientists and industry in the US and in
positive results were eliminated when an agar Europe have focused their attention on the
overlay modification of the test was used (14). development and validation of in vitro eye
A later qualitative validation study with 1013 irritancy tests.
materials confirmed the earlier findings (15). One group developed a battery of tests,
As a result, cellular toxicity assays have for including a neutral red uptake assay and a
a long time been accepted as routine protein assay for measuring cell viability and
supplements in biomaterials testing. cell proliferation, and performed intra-
laboratory and interlaboratory evaluation
Screening of acute systemic toxicity with several classes of chemicals (26). The
One of the laboratories in our group has predictive value of the tests was evaluated by
performed a series of validation studies on determining correlations between in vitro and
the prediction of acute human toxicity by a Draize test data. The choice of the Draize test
HeLa cell assay (summarised in [2]). The score as the endpoint of evaluation may be a
cytotoxic concentrations of 50 non-selected handicap, since it is somewhat arbitrarily
drugs to HeLa cells were compared composed of several measurements of injury
quantitatively with human lethal and toxic in the rabbit eye, which are of doubtful
blood concentrations, including a study of relevance to concentration- and time-
lethal and toxic symptoms. Surprisingly, it dependent quantitative human eye irritancy.
was found that the HeLa cell results Separate external studies of a total of 150
(representing basal cytotoxicity) were coded substances generally resulted in rank
relevant for the target organ toxicity of a correlations of 0.5-0.6. When all classes of
majority of the drugs. Further evaluation, chemicals were considered together, the
with the use of other cell types (16) and other correlations were not as good as those
classes of chemicals (17), supported the obtained for each separate class.
finding that simple cytotoxicity assays with Another laboratory has developed and
cell lines are very often predictive of acute validated the CAM test, utilising the chorio-
systemic toxicity, provided that allowance is allantoic membrane (CAM) of hens' eggs as
made for the toxicokinetics ofthe chemicals. the test material (27). An improved CAM test
Evaluation of the correlation between was challenged with 15 products, with a
cytotoxicity and acute lethal dosage for resulting rank correlation with Draize test
rodents has also been performed (17-22). scores of 0.6-0.8 (28). A modification of the
Typically, 25-50 chemicals have been used in test has been developed in Germany - the
quantitative in vitro/in vivo comparisons with Hens Egg Test/Chorio-Allantoic Membrane
a resulting linear correlation of about 0.75. (HET/CAM) test (29). According to
This correlation probably corresponds to an developers, a validation study of250 products
even better correlation between toxic demonstrated good relevance and reliability,
concentrations to target cells in vitro and in but raw data were not presented (30). Other
vivo, which cannot be demonstrated in the laboratories have failed to obtain such good
absence of biotransformation and other results (30, 31).
228 Scandinavian Cell Toxicology Congress 11

Arguably the best validated in vitro test multicentre programmes known to us will be
for eye irritancy at present is EYTEX™, a discussed.
simple protein-aggregation test which
involves the use of a protein matrix based on The FRAME programme
jack bean powder (32). A 90% correlation with
Draize test data was obtained for 130 The first multicentre vaidation programme
substances and materials tested externally by was initiated in 1982 by FRAME (Fund for
12 industrial companies. the Replacement of Animals in Medical
A two-laboratory double-blind study of 49 Experiments; 1). One of the first parts of the
products demonstrated the same predictive programme was to develop a reliable cell line
value and also a 93% agreement of results toxicity assay to measure fundamental
from the two laboratories (33). A database of (basal) cytotoxicity. This was done in
over 1500 tested products has been reported cooperation with four laboratories, resulting
(34). When EYTEX™ and the neutral red in the FRAME ken acid blue (KB) assay (36).
uptake assay were compared in a recent study Reliability was assured by extensive multi-
of 17 solvents, both predicted the corneal laboratory testing of coded substances. A
opacity score of an EEC modification of the recent, similar project has been to develop a
Draize test equally well (r=0.84; 35). To some sensitive eye irritancy assay, resulting in a
extent, the good correlation between neutral red release test, with a short exposure
EYTEX™ and Draize test scores can be period of one minute (37).
explained by the heavy weight of the corneal A second part of the programme extended
opacity score in the total Draize score, corneal the testing of the coded substances to
opacity probably being the result of protein international laboratories. The primary goal
denaturation by chemicals. was to evaluate the reliability of the various
The single-laboratory validation exercises methods, but relevance was also evaluated
described have already led to the practical by comparison of in vitro data with toxicity
use of all the above-mentioned methods by profiles of some of the 150 chemicals (36). The
industry. In all cases, the validation process correlation between the FRAME KB test and
has resulted in effective modifications of rat LD50 values was found to be good (20).
original methods. For example, the original Recently, one international laboratory
EYTEX™ method has been modified to a compared their Hep-2 cell results with four
membrane partition assay, to cope with British assays (38). Cytotoxicity was similar
insoluble substances (34). in all the systems, independently of the types
of cells and toxic endpoints employed,
Multi-laboratory validation studies supporting the basal cytoxicity concept.
FRAME has emphasised the many possible
Multicentre studies have several advantages pitfalls of multicentre programmes (39, 40).
over single-laboratory validation studies (1, Among the lessons learned were that good,
2). Firstly, the credibility of results will and preferably human, toxicity data should
increase if an independent group, not itself be secured for the reference substances and
involved in the cytotoxicity assays, has that general cut-off points f~r the cytotoxicity
selected the chemicals for further blind data should be avoided.
testing among the participating laboratories.
Secondly, the work ofgathering, selecting and SDA and CTFA validation programmes
standardising the in vivo toxicity data is
reduced. Thirdly, only a multicentre study Two US industry associations, the Soap and
can directly evaluate the predictivity of a Detergent Association (SDA) and the
combination of methods. Fourthly, once the Cosmetic, Toiletry and Fragrance Association
standard procedures for validation have been (CTFA), have organised evaluation
determined, the relevance of an unlimited programmes for eye irritancy tests. The SDA
number of methods can be evaluated by use programme consisted of two phases (7, 41),
of computer techniques. Thus, no potentially In phase I, invited laboratories tested 8 coded
valuable method has to be excluded from an substances in 14 cellular assays, to sort out
evaluation of relevance. Finally, inter- the best methods. In phase II, 15 coded
laboratory reproducibility studies are materials were used to challenge nine
facilitated by this approach. In the following candidate systems (two CAM modifications,
sections, all the on-going and proposed corneal plasminogen activator assay with
Scandinavian Cell Toxicology Congress 11 229

rabbit corneal cells, SIRC/cloning efficiency Proposed programmes


of rabbit corneal cells, the uridine uptake
assay, a cell protein accumulation assay, the In 1986, the Johns Hopkins Center for
neutral red uptake assay, Tetrahymena Alternatives to Animal Testing (CAAT)
motility assay, and chromium release from arranged a conference entitled Approaches to
murine mastocytoma cells). The methods Validation. One keynote speaker preferred a
generally could discriminate between sensitive and flexible validation process
materials indicated by the Draize test to be rather than a detailed road-map leading
irritants and non-irritants. An algorithm of inevitably to the hidden prize (43). He pointed
pH and cytotoxicity was needed for out that models in some cases might make
discrimination, rather than cytotoxicity predictions on a functional basis, without the
alone. The CTFA has started their prog- necessity to be mechanistically defined. A
ramme (8). The phases of the CFTA second introductory speaker emphasised that
programme correspond to evolution of mutagenesis test validation has shown that
different classes of chemicals, in Phase I, ten a high correlation for single endpoint tests
alcohol-containing generic prototypes from will probably not be achieved, since multiple
the cosmetic industry have been tested in 25 mechanisms are often involved in the lesion
in vitro assays. of interest. Therefore, validation cannot be
based on a unifying concept (44). The latter
speaker also advocated the use of human,
Other on-going multicentre programmes clinical data for validation purposes.
A third US industrial toxicologist recently
A programme was. started. in. 1988 ~n outlined a collaborative validation of
Germany, with the aim ofvahdatmg tw.o m candidate in vitro eye irritancy tests (4): a
vitro tests for eye irritancy (a combmed committee should select two panels of a
neutral red uptakelKB cytotoxicity assay and representative variety of test compounds with
the HET/CAM assay; 9, 42). The study is known human and rabbit eye irritation (20
similar to the Amden protocol (vide infra; 5). and 30 materials, respectively). All
Firstly, the reliability of the methods will be investigators with candidate systems should
determined by an intralaboratory and be invited to test the coded panel A
interlaboratory study involving 14 contract compounds. After the data obtained had been
laboratories testing 35 chemicals. Later, an returned, the systems could be evaluated ~s
in vitro database with 200 chemicals will be "go" or "no go", based on results and economic
established to be used in the evaluation of factors. Finally, the "go" systems would test
predictions' from non-animal tests against the panel B compounds, and this could result
Draize test scores. So far, an acceptable in identification of the best single tests and
reproducibility of both tests has been found, test batteries, which later could be
as well as a similar tendency of the tests to recommended for use in industry.
produce false-positive resul~s. . . CAAT has proposed an outline of future
In France a similar validation of two m validation of in vitro toxicity tests (3), in part
vitro tests for eye irritancy has started, with built on a previous model (45). Validation is
use of 40 substances (10). In the same visualised as having four phases: phase I
country, an interesting multicent~e includes the preliminaries, such as definition
validation of in vitro tests for acute hepatic of the objective of each candidate test; phase
toxicity/hepatic metabolism has also been II is a microvalidation of 20 substances,
performed (12). Six laboratories tested 30 performed in the primary laboratory; p~ase
compounds in assays using primary cultured III is a macrovalidation of 50-100 chemicals
rat hepatocyte or a rat hepatoma cell line, at performed in a contract laboratory, followe.d
a set exposure time of 20 hours. Most of the by statistical evaluation; and phase IV IS
compounds tested were equally cytotoxic ~o battery optimisation. All the testing is
both cell types, indicating a basal cytotox~c performed with coded substances. Th~ e~~ly
effect. Some specifically hepatotoxic feasibility study is focused on reliability
compounds were relatively more toxic to ~he evaluation, while macrovalidation is focused
primary cultured rat hepatocytes, confirmmg on evaluation of the relevance of the methods
the results from a previous study of involved. This outline has been comp-
differential cytotoxicity in hepatocytes and rehensively expanded in a subsequent OECD
cell lines (16). document (46).
230 Scandinavian Cell Toxicology Congress 11

In February 1990, European and American All the reference chemicals have data on
cell toxicologists held a workshop in Amden, acute human toxicity, including toxico-
Switzerland, with the aim of preparing a kinetics, plus rodent oral LD50 values. Many
consensus document defining the proper of them also have other human or animal
performance of large-scale validation of in data, e.g. on irritancy in the eye and skin.
vitro cytotoxicity tests (5). The outlined Another 50-100 chemicals will be added to
scheme resembles earlier FRAME (1,36) and the list. Incoming results will be evaluated
CAAT (3, 41) proposals, and the major steps on their relevance for various types of human
of validation are defined as: toxicity (acute and chronic systemic toxicity,
target organ toxicity, local irritancy). The
1. Intralaboratory assessment;
2. Interlaboratory assessment; evaluation will be quantitative and linear. It
is presumed that results from the cytotoxicity
3. Test database development; and
tests, including in vitro tests with
4. Evaluation.
extracellular targets, must be inserted in
Validation is performed with four sets of toxicokinetic models to be fully predictive for
coded reference chemicals from a database both systemic toxicity and local irritancy. To
set consisting of 200-250 chemicals. The first be able to judge the effectiveness of in vitro
steps are mainly concerned with evaluation tests, prediction of human toxicity by
of test reliability, except for some relevance conventional animal tests for the same
evaluation in the first step with 20-50 chemicals will be evaluated in parallel. Blind
chemicals. Then the ultimate relevance of a testing will be achieved by the majority of
method is determined by an evaluation participants, although we have not used
following the test database development. One coded substances in the first, major phase of
typical feature of the proposal is that the programme. In a second, final phase, the
developing laboratories have a rather minor tests with the best prediction of human
role in the validation process. The programme toxicity will be evaluated for reliability by
is built on coded substances, contract contract laboratories and with coded
laboratories and decision-making by several substances (if they were not known to be
expert panels. Thus, decisions by a panel are reliable before). The aim is to be able to
crucial to the acceptance of the methods to recommend tests and batteries of tests as
be validated. Another typical feature is that supplements or alternatives to animal testing
relatively trivial aspects of future validation, for various types of general toxicity. The
such as evaluation of the variability and selection of test batteries will be based on soft
reproducibility of tests, integration of the multivariate modelling (47).
validation process, battery selection and At present, 62 laboratories have joined the
implementation, are discussed extensively. project. Twenty laboratories have returned
The likely tricky part, i.e, how to properly results from more than 50 assays for the first
evaluate the relevance of non-genotoxic in 10 chemicals. Methods for evaluating acute
vitro tests, is based on either previous lethal toxicity (48) and acute sublethal
qualitative (+/-) genotoxicity validation toxicity (49) have been worked out with the
studies or the previous, debatable, semi- first incoming results as training sets. Models
quantitative validation of in vitro eye for evaluating other types of toxicity are in
irritation tests against Draize test scores. progress. To facilitate evaluation of eye and
Thus, there is no consensus in the protocol skin irritancy, we will try to obtain data on
on how to evaluate methods with the use of quantitative minimal irritancy in man.
human data, toxicokinetic algorithms, and Another optional part of the programme is to
other probably more relevant approaches. encourage participants to develop key
methods which are missing at present, such
as long-term exposure cytotoxicity assays.
The MEIC programme MEIC is essentially an economical and
The Scandinavian Society of Cell Toxicology practical way for a laboratory to extend its
started a multicentre evaluation of in vitro own "single-laboratory" validation of a
cytotoxicity (the MEIC. programme) a few method on its own terms. The laboratory has
years ago (6). In the spring of 1989, to finance its own contribution, but the profit,
international laboratories willing to take part in the form of data from other methods and
were invited to test a list of 50 reference the evaluation against human data, is large
chemicals using their various in vitro assays. compared with the investment. Thus, the
Scandinavian Cell Toxicology Congress 11 231

expenses of the central operations can be of the programme not yet discussed (e.g. how
minimised. Further, open entry to the to perform an adequate evaluation of the
programme will ensure that few potentially relevance of tests), and it will ultimately cost
valuable methods will be missed, for instance, a lot of money, if the basic strategy of the
by too early a judgement of their supposed protocol is still adhered to at that time. A
qualities. This is achieved by beginning the better policy would be to fund on-going single-
programme with the relevance phase of laboratory and multicentre projects of various
validation. To end the study with the kinds directly, as well as initiating less
reliability phase also minimises costs - this ambitious, but more practical, projects, such
costly part of the study (contract laboratories, as the eye irritation validation process
coded substances) can be cut down by the proposed by Gad (4).
previous exclusion of clearly irrelevant
methods.
References
Comments on the Best Way to Validate 1 Balls, N. & Bridges, J.W. (1984). The FRAME
research programme on in vitro cytotoxicology. In
Alternative Methods in Toxicology, Vol. 4, Acute
It is clear that more effective and Toxicity Testing: Alternative Approaches (ed, A.M.
comprehensive efforts for validating in vitro Goldberg), pp. 62-79. New York: Mary Ann
toxicity tests are needed. These efforts should Liebert.
2. Ekwall, B. (1983). Screening of toxic compounds
probably be of the multicentre type, using in mammalian cell cultures. Annals of the New
refined and relevant evaluation methods. York Academy of Sciences 407, 64-77.
Otherwise, relatively ineffective in vitro 3. Frazier, J.M. (1987). The validation approach of
methods may be accepted by industry and the Johns Hopkins Center for Alternatives to
Animal Testing. In Alternative Methods in
regulatory agencies on the merits of vigorous Toxicology, Vol. 5, In Vitro Toxicology:
single-laboratory validation only, at the Approaches to Validation (ed. A.M. Goldberg), pp.
expense of potentially more valuable methods 87-90. New York: Mary Ann Liebert.
which have not yet entered any validation 4. Gad, S.C. (1989). Acute ocular irritation
evaluation: in vivo and in vitro alternatives and
contest. making them "standard" for testing. In
However, we do not believe that it is Benchmarks: Alternative Methods in Toxicology
necessary to reach a worldwide consensus of (ed. M. Mehlman), pp. 137-193. Princeton:
Princeton University Press.
the outline of an ultimate validation process. 5. Balls, M., Blaauboer, B., Brusick, D., Frazier, J.,
Probably such a guideline will have flaws and Lamb, D., Pemberton, M., Reinhardt, C.,
biases, like every project looked at in Roberfroid, M., Rosenkranz, H., Schmid, B.,
retrospect (compare the FRAME self- Spielmann, H., Stammati, A. & Walum, E. (1990).
criticism [39]). Instead, there is an advantage Report and recommendations of the CAAT/
ERGATT workshop on the validation of toxicity
with a series of different and smaller "local" test procedures. ATLA 18,313-337.
programmes. After consideration of the 6. Bondesson, I., Ekwall, B., Hellberg, S., Romert,
different angles (aims, methods, flaws, etc.) L., Stenberg, K. & Walum, E. (1989). MEIC - a
of a series of programmes, the common new international multicentre project to evaluate
the relevance to human toxicity of in vitro toxicity
features of the results of the various studies tests. Cell Biology and Toxicology 5, 331--347.
will be accepted as valid by the scientific 7. Booman, K.A., Cascieri, T.M., Demetrulias, J.,
community, industry and regulatory Driedger, A., Griffith, J.F., Grochoski, G.T., Kong,
agencies. Moreover, a traditionally split B., McCormick, W.C. 3rd, North-Root, H., Rozen,
M.G. & Sedlak, R.I. (1988). In vitro methods for
scientific "market", i.e. several different estimating eye irritancy of cleaning products.
multicentre validation studies, will allow the Phase I: preliminary assessment. Journal of
developers of tests a choice of validation Toxicology - Cutaneous and Ocular Toxicology 7,
schemes according to their specific needs - 173-185.
8. Gettings, S.D. & McEwen, G.N. (1990).
a method could, for instance, be evaluated for Development of potential alternatives to the
relevance in a MEIC-type study, and at the Draize eye test: the CTFA evaluation of
Harne time be evaluated for reliability in an alternatives program. ATLA 17,317-324.
Amden-type study. 9. Kalweit, S., Gerner, I. & Spielmann, H. (1989).
Validation project of alternatives for the Draize
If industry and governmental organ- eye test. Journal of Molecular Toxicology 1, 597-
isutions, such as the EC and the OEeD, 603.
favour the idea of a universal consensus 10. Blein, O. & Adolphe, M. (1991). French project of
strategy, like the Amden protocol, it will take correlation and validation of alternatives to the
Draize eye irritation test. Toxicology in Vitro, in
time to reach agreement of the critical parts press.
232 Scandinavian Cell Toxicology Congress 11

11. Stammati, A & Pirovano, R (1988). In vitro 27. Leighton, J., Tchao, R.• Verdone. J. & Nassauser,
cytotoxicity test on Hep-2 cells. In Collaborative J. (1985). Macroscopic assay offocal injury in the
Study on Relationship Between In Vivo Primary chorio-allantoic membrane. In Alternative
Irritation and In Vitro Experimental Models. EEC Methods in Toxicology, Vol. 3, A Progress Report
Project 86E3-063/86-460. CECIV/E/3/Lux/-157/99. from the Johns Hopkins Center for Alternatives to
12. Rahmani, R., De Sousa, G., Placidi, M., Fautrel, Animal Testing (ed, A.M. Goldberg), pp. 355-370.
A., Chesne, C., Guillouzo, A., Braut, F., Pichon, New York: Mary Ann Liebert.
J., Hoellinger, H., Vintezou, P., Diarte, I., 28. Bagley, D.M., Rizvi, P.Y., Kong, B.M. & DeSalva,
Melchion, C., Cordier, A., Lorenzon, G., Penicourt, S.J. (1988). An improved CAM assay for predicting
M., Vannier, B., Fournex, R, Peloux, A.F., Bichet, ocular irritation potential. In Alternative Methods
N., Gouy, D., Cono, J.P. & Lounes, R (1991). A in Toxicology, Vol. 6, Progress in In Vitro
multicenter study of acute in vitro hepatic Toxicology (ed. A.M. Goldberg), pp. 131-138. New
cytotoxicity. Toxicology in Vitro, in press. York: Mary Ann Liebert.
13. Rosenbluth, S.A., Weddington, G.R, Guess, W.L. 29. Luepke, N-P. & Wallat, S. (1987). HET/CAM:
& Autian, J. (1965). Tissue culture method for reproducibility studies. In Alternative Methods in
screening toxicity of plastic materials. Journal of Toxicology, Vol. 5, In Vitro Toxicology:
Pharmaceutical Science 54, 156-159. Approaches to Validation (ed. A.M. Goldberg), pp.
14. Autian, J. (1973). The new field of plastics 353-363. New York: Mary Ann Liebert.
toxicology: methods and results. In CRC Critical 30. Rowan, AN. ed. (1989) The Alternatives Report
Reviews in Toxicology, Vol. 2 (ed. 1. Golberg), 1- 1, 3. North Grafton, MD: Center for Animals and
40. Cleveland: CRC Press. Public Policy, Tufts University School of
15. Wilsnack, RE. (1976). Quantitative cell culture Veterinary Medicine.
biocompatibility testing of medical devices and 31. Reinhardt, C.A., Aeschbacher, M., Bracher, M. &
correlation to animal tests. Biomaterials, Medical Spengler, J. (1987). Validation of three cell toxicity
Devices and Artificial Organs 4, 235-261. tests and the hen's egg test with guinea-pig eye
16. Ekwall, B. & Acosta, D. (1982). In vitro and human skin irritation data. In Alternative
comparative toxicity of selected drugs and Methods in Toxicology, Vol. 5,1n Vitro Toxicology:
chemicals in HeLa cells, Chang liver cells and rat Approaches to Validation (ed. A.M. Goldberg), pp.
hepatocytes. Drug and Chemical Toxicology 5, 463-469. New York: Mary Ann Liebert.
219-231. 32. Gordon, V.C. & Bergman... H.C. (1988). External
17. Ekwall, B., Nordensten, C. & Albanus, 1. (1982). validation of the EYTEXr M and EYTEX-MPATM
Toxicity of29 plasticizers to HeLa cells in the MIT- systems in 12 laboratories to determine the
24 system. Toxicology 24, 199-210. correlation of EYTEX™ results with in vivo
18. Walum, E. & Paterson, A. (1984). On the results. In Alternative Methods in Toxicology, Vol.
application of cultured neuroblastoma cells in 6, Progress in In Vitro Toxicology (ed. A.M.
chemical toxicity screening. Journal ofToxicology Goldberg), pp. 325-331. Mew York: Mary Ann
and Environmental Health 13, 511-520. Liebert.
19. Halle, W. & Gores, E. (1987). Vorhersage von 33. Soto, RJ., Servi, M.J. & Gordon, V.C. (1989).
LD50-werten mit der Zellkultur (Quantitiative Evaluation of an alternative method for ocular
prediction of LD50 estimations by the use of cell irritation in in vitro toxicology. In Alternative
cultures) (German). Pharmazie 42,245-248. Methods in Toxicology, Vol. 7,1n Vitro Toxicology:
20. Clothier, RH. & Hulme, L.M. (1989). Comparison New Directions (ed, A.M. Goldberg). New York:
of the in vitro cytotoxicities and acute in vivo Mary Ann Liebert.
toxicities of 59 chemicals. Molecular Toxicology I, 34. Gordon, V.C., Kelly, C.P. & Bergman, H.C. (1990).
571-577. Applications of the EYTEX™ method. Toxicology
21. Ekwall, B. (1983). Correlation between in Vitro 4, 314-317.
cytotoxicity in vitro and LD50 values. 35. Dierick~ P.J. & Gordon, V.C. (1990). The
Pharmacology and Toxicology 52, Suppl. II, 80-98. EYTEX system and the neutral red uptake
22. Fry, J.R., Garle, M.J., Hammond, A.H. & Hatfield, inhibition assay in cultured Hep G2 cells as
A (1990). Correlation of acute lethal potency with alternative methods for in vivo eye irritation
in vitro cytotoxicity. Toxicology in Vitro 4, 175- following the EEC protocol. ATLA 17,325-333.
178. 36. Balls, M., Riddell, R.J., Horner, S.A. & Clothier,
23. Mehlman, M.A., Pfitzer, E.A. & Scala, R.A. (1989). R.H. (1987). The FRAME approach to the
A report on methods to reduce, refine and replace development, validation and evaluation of in vitro
animal testing in industrial toxicology alternative methods. In Alternative Methods in
laboratories. Cell Biology and Toxicology 5, 349- Toxicology, Vol. 5, In Vitro Toxicology:
358. Approaches to Validation (ed, A.M. Goldberg), pp.
14. Linesman, D.A., Raczniak, T.J., Aaron, C.S. & 45-58. New York: Mary Ann Liebert.
Bacon, J.A. (1990). Comparative cytotoxicity 37. Reader, S.J., Blackwell, V., O'Hara, R., Clothier,
rankings of four aminoglycoside antibodies in the R.H., Griffin, G. & Balls, M. (1989). A vital dye
Chang, SIRC and LLC-PK1 cell lines. ATLA 18, release method for assessing the short-term
283-290. cytotoxic effects of chemicals and formulations.
25. Bradley, M.O., Noble, C. & Sina, J.F. (1989). An ATLA 17,28-37.
in vitro nephrotoxicity assay useful in drug 38. Mazziotti, I., Stammati, A. & Zucco, F. (1990). In
development. In Vitro Toxicology 2,171-182. vitro cytotoxicity of 26 coded chemicals to Hep-2
26. Shopsis, C., Borenfreund, E. & Stark, D.M. (1987). cells: a validation study. ATLA 17,401-406.
Validation studies on a battery of potential in vitro 39. Balls, M. & Clothier, R (1989). Validation of
alternatives to the Draize test. In Alternative alternative toxicity test systems: lessons learned
Methods in Toxicology, Vol. 5,1n Vitro Toxicology: and to be learned. Molecular Toxicology I, 547-
Approaches to Validation (ed, A.M. Goldberg), pp. 559.
31-44 New York: Mary Ann Liebert. 40. Clothier, R.H. & Balls, M. (1990). Validation of
Scandinavian Cell Toxicology Congress 11 233

alternative toxicity tests: principles, practices and McCulley, J.P. (1986). Validation. In Alternative
cases. Toxicology in Vitro 4, 692-693. Methods in Toxicology, Vol. 4, A Critical
41. Booman, K.A., DeProspo, J. Demetrulias, J., Evaluation of Alternatives to Acute Ocular
Driedger, A., Griffith, J.F., Grochoski, G., Kong, Irritation Testing (ed, A.M. Goldberg), pp. 113-
B., McCormick, W.C. 3rd, North-Root, H., Rozen, 117. New York: Mary Ann Liebert.
M.G. & Sedlak, R.I. (1989). The SDA alternatives 46. Frazier, J.M. (1990). Scientific Criteria for
programme: comparison of in vitro data with Validation ofIn Vitro Toxicity Tests. Paris: DECD.
Draize test data. Journal of Toxicology 47. Hellberg, S., Bondesson, I., Ekwall, B., GOmez-
Cutaneous and Ocular Toxicology 8, 35-49. Lech6n, M.J., Jover, R., Hogberg, J., Ponsoda, X.,
42. Spielman, H., Gerner, I., Kalweit, S. & Romert, L., Stenberg, K. & Walum, E. (1990).
Wirnsberger, T. (1991). Validation project of Multivariate validation of cell toxicity data: the
alternatives to the Draize test in West Germany. first ten MEIC chemicals. ATLA 17,237-239.
Toxicology in Vitro, in press. 48. Ekwall, B., Bondesson, I., Castell, J.V., GOmez-
43. Scala, R.A. (1987). Theoretical approaches to Lech6n, M.J., Hellberg, S., Hogberg, J., Jover, R.,
validation. In Alternative Methods in Toxicology, Ponsoda, X., Romert, L., Stenberg, K. & Walum,
Vol. 5, In Vitro Toxicology: Approaches to E. (1989). Cytotoxicity evaluation of the first ten
Validation (ed, A.M. Goldberg), pp. 1-9. New MEIC chemicals: acute lethal toxicity in man
York: Mary Ann Liebert. predicted by cytotoxicity in five cellular assays and
44. Rundell, J.D. (1987). Validation as applied to in by oral LD50 in rodents. ATLA 17,83-100.
vitro toxicology. In Alternative Methods in 49. Ekwall, B., GOmez-Lech6n, M.J., Hellberg, S. &
Toxicology, Vol. 5, In Vitro Toxicology: Walum, E. (1991). Good prediction of acute toxic
Approaches to Validation (ed, A.M. Goldberg), pp. blood concentrations in man in 24·hour cyto-
11-15. New York: Mary Ann Liebert. toxicity to Hep-G2 and HeLa cells for the first 10
45. Frazier, J.M., Gad, S.C., Goldberg, A.M. & MEIC chemicals. The Toxicologist 11, 227.

You might also like