In the context of the 6th Amendment of the European Directive on Cosmetics,
several cosmetic companies concentrate their basic research on the develop
ment of the best adapted battery of in vitro tests able to be incorporated
in the ocular risk assessment process. Consequently, the European Cosmetic
Toiletry and Perfumery Association (COLIPA) has initiated an international
multicentric study with the main purpose to validate available alternatives
in vitro methods for assessing the eye irritation potential of cosmetic ra
w materials and formulations. The alternative methods assessed in this vali
dation study were chosen since all of these tests had already been used and
continue to be conducted in the risk assessment process. The different end
points of these assays are mainly biological parameters except for the bioc
hemical assay named EYTEX(TM). In this article, the defined prediction mode
ls and the different protocols used in the COLIPA study are described. Then
, the EYTEX assay results are presented and discussed in details in order t
o understand the failure of this assay during this validation study. The re
levance and the reliability of the EYTEX assay were particularly low in two
laboratories, whereas one laboratory presented acceptable data with a low
compatibility with tested samples. These results underline the problem of t
he complex qualification process of this assay, since sometimes the same sa
mple has been qualified with different protocols in the three laboratories.
This validation study also demonstrates that, in the case of EYTEX assay,
the criteria used to establish a prediction model have not been rigorous en
ough. For instance, the mixture of all the EYTEX protocols is not suitable
for the establishment of a well-adapted prediction model. Furthermore, a cl
earer definition of limitations of the EYTEX assay seems to be necessary to
better harmonize the qualification procedure in the three laboratories. Th
e COLIPA validation process clearly demonstrated that the EYTEX assay was f
irst, not suitable for the assessment of the eye irritation potential of su
rfactants and formulations based on surfactants, and secondly not ready for
a validation study requiring the establishment of adequate and well define
d mathematical prediction models. However, internal comparative studies wit
h specific benchmarks on emulsions containing a low percentage of surfactan
ts may be more adaptable to this type of assay. (C) 1999 Elsevier Science L
td. All rights reserved.