Ethylene bis[1,3-dihydro-1,3-dioxoisobenzofuran-5-carboxylate]

Skin sensitisation

Administrative data

Endpoint:

skin sensitisation, other

Remarks:

(Q)SAR predictions

Type of information:

(Q)SAR

Adequacy of study:

key study

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

accepted calculation method

Justification for type of information:

According to Annex VII (Regulation (EC) No 1907/2006) and the testing strategies for the 2018 registration deadline (ECHA Guidance, Chapter R.7a, 2016), the information needed for the classification or risk assessment of a substance should obtained through non-animal methods as a first step.
Before to start any new testing an assessment of all available in vitro data, in vivo data, historical human data, data from valid (Q)SARs and data from structurally related substances (read-across approach) is requested.
Testing does not need to be conducted if the conditions specified in column 2 of Annex VII section 8.3 of the REACH Regulation and include:
a) the available information indicate the substance should be classified for skin corrosion or skin sensitisation, or
b) the substance is a strong acid (pH< 2.0) or base (pH> 11.5), or
c) the substance is (spontaneously) flammable in air at room temperature.

The test item is a mixture consisting of >80 % ethylene bis[1,3-dihydro-1,3-dioxoisobenzofuran-5-carboxylate] and between 5 and 15 % trimellitic anhydride (TMA). From the reported study results, there is evidence that TMA is at least a strong sensitiser (LLNA EC3 = 0.18 %; median of four reported values). According to ECHA Guidance on the Application of the CLP Criteria, no sub-categorisation is required when mixtures are classified as sensitizing based on the presence of a sensitizing substance at a concentration at or above the generic or specific concentration limit. With a specific concentration limit of 0.18% (sub-category 1a) TMA triggers classification of the mixture as Category 1. However, according to Annex VII, section 8.3, in addition to the assessment of whether the substance is a skin sensitiser or a non-sensitiser, the potency (Cat. 1A or not) of skin sensitising substances must be addressed.
From predictions with various (Q)SAR models there is evidence that TMEG is a skin sensitiser. With three anhydrides a category of analogues could be identified, which covers the range of possible skin sensitisation mechanisms of TMEG. The predicted EC3 of < 2% and ≥ 0.135 from the anhydrides is therefore considered reliable and indicates that TMEG is a strong sensitiser. Taking further into account the median EC3 of TMA, there is an argument for weight of evidence that the mixture of TMEG and TMA is a strong sensitiser. It is therefore proposed to classify the test item as sensitiser Category 1a.

Data source

Reference

Reference Type:

other: Expert Statement and Prediction Report

Title:

Unnamed

Year:

2017

Materials and methods

Test guideline

Qualifier:

according to guideline

Guideline:

other: REACH Guidance on QSARS R.6

Principles of method if other than guideline:

(Q)SAR predictions were performed to estimate the skin sensitation potential of ethylene bis[1,3-dihydro-1,3-dioxoisobenzofuran-5-carboxylate] (TMEG, CAS 1732-96-3).

GLP compliance:

no

Remarks:

not applicable

Justification for non-LLNA method:

(Q)SAR predictions were performed to estimate the skin sensitation potential of ethylene bis[1,3-dihydro-1,3-dioxoisobenzofuran-5-carboxylate] (TMEG, CAS 1732-96-3). With four reliable predictions there is an argument for a weight of evidence that TMEG is a sensitiser. Moreover, potency predictions with OECD Toolbox and DEREK resulted in EC3 values of < 2% and ≥ 0.135, thus indicating that TMEG is at least a strong sensitiser. Therefore further testing was not necessary.

Test material

Test material information

Constituent 1

Reference substance name:

Ethylene bis[1,3-dihydro-1,3-dioxoisobenzofuran-5-carboxylate]

EC Number:

217-062-1

EC Name:

Ethylene bis[1,3-dihydro-1,3-dioxoisobenzofuran-5-carboxylate]

Cas Number:

1732-96-3

Molecular formula:

C20H10O10

IUPAC Name:

2-(1,3-dioxo-1,3-dihydro-2-benzofuran-5-carbonyloxy)ethyl 1,3-dioxo-1,3-dihydro-2-benzofuran-5-carboxylate

impurity 1

Reference substance name:

Benzene-1,2,4-tricarboxylic acid 1,2-anhydride

EC Number:

209-008-0

EC Name:

Benzene-1,2,4-tricarboxylic acid 1,2-anhydride

Cas Number:

552-30-7

Molecular formula:

C9H4O5

IUPAC Name:

Benzene-1,2,4-tricarboxylic acid 1,2-anhydride

impurity 2

Reference substance name:

not assignable

IUPAC Name:

not assignable

Specific details on test material used for the study:

CAS number: 1732-96-3
EC number: 217-062-1
Chemical name: Ethylene bis[1,3-dihydro-1,3-dioxoisobenzofuran-5-carboxylate]
IUPAC: 2-(1,3-dioxo-2-benzofuran-5-carbonyl)oxyethyl 1,3-dioxo-2-benzofuran-5-carboxylate
SMILES: O=C(OCCOC(=O)c1ccc2C(=O)OC(=O)c2c1)c1ccc2C(=O)OC(=O)c2c1
InChI: 1S/C10H10O2/c1-2-10(11)12-8-9-6-4-3-5-7-9/h2-7H,1,8H2

Results and discussion

In vitro / in chemico

Results

Results 1

Run / experiment:

other: (Q)SAR prediction using OECD Toolbox

Parameter:

other: EC3 % value prediction

Value:

2

Vehicle controls validity:

not applicable

Negative controls validity:

not applicable

Positive controls validity:

not applicable

Results 2

Run / experiment:

other: (Q)SAR prediction using DEREK

Parameter:

other: EC 3 % value prediction

Value:

0.135

Vehicle controls validity:

not applicable

Negative controls validity:

not applicable

Positive controls validity:

not applicable

Any other information on results incl. tables

Appraisal of (Q)SAR Modelling

Except TOPKAT, all models predicted TMEG to be sensitising. The TOPKAT prediction is considered not to be reliable, while there is confidence in all positive predictions.

The mechanistic domain of the Toolbox prediction refers to protein acylation by anhydrides, while DEREK also includes azlactones and analogue structures and Toxtree identified only the domain of acyl transferring agents.

The compounds most similar to TMEG in training set of CAESAR trigger various mechanistic alerts when using the protein binding profiler in OECD Toolbox, thus representing no unique category and may therefore suggest some uncertainty. A fuzzy partition activity score O of 0.87 (of belonging to the class of sensitiser versus O = 0.13 of belonging to non-sensitiser) together with high scores in all other assessment indices indicate however confidence in the prediction.

The four analogues used for read-across in Toolbox are all sensitising anhydrides including trimellitic anhydride, thus indicating confidence in the prediction of the anhydride TMEG. The predicted EC3 of 2.59 % is based on the arithmetic mean of values from the four analogues. With an EC 3 of 9.2 % trimellitic anhydride contributed the highest value. In more recent LLNA studies on trimellitic anhydride EC3 values between 0.11 and 0.2 are however reported (for references, see software printout DEREK). The predicted EC3 of 2.59 % is therefore considered overestimated and a value of < 2% seems to more realistic.

DEREK predicted an EC3 of 0.029 % based on 12 compounds with values ranging between 0.0014% and 26%. With a median of 0.0014 % (based on five study results ranging between 0.0007 and 0.0026 %) the prediction result is notably influenced by extreme sensitising 4-Ethoxymethylene-2-phenyl-2-oxazoline-5-one (CAS 15646-46-5). This compound belongs to a specific class of azlactones, is a very strong experimental skin sensitiser in the LLNA and is considered to have a unique sensitisation potential (Roberts et al., 2007; QPRF DEREK). Its structural similarity to TMEG is 37 %, while the two most similar structures used in the calculation of EC3 belong to the same category of anhydrides and have similarities of 90 % (trimellitic anhydride) and 68 % (phthalic anhydride). Moreover, almost other compounds used in the calculation are also azlactones with similarities < 40 %. They are members of a homologues series, where the sensitisation potential increases with log K_owbetween the C₄and C₆homologues, and then beyond some point between the C₆and C₉homologues (in the log K_owrange 2.9-4.4), sensitization potential becomes a decreasing function of log K_ow(Roberts et al., 2007). The above arguments suggest that azlactones have a different sensitisation potential than anhydrides, though both belong to the acyl transfer agent domain. It is therefore considered justified to exclude all azlactones and to estimate the EC3 for TMEG from the LLNA results of the two anhydrides. Using the average of the lowest values as reported for each structure in the DEREK report results in a conservative EC3 of 0.135 %.

Table1    Prediction results

Model	Prediction result	Reliability
TOPKAT	Not sensitising	Low
CAESAR (VEGA)	Sensitising	Reliable
OECD Toolbox	Sensitising (EC3 = 2.59 %)	Reliable EC3 < 2 %
DEREK	Sensitising (EC3 = 0.029%)	Reliable EC3 = 0.135 %
Toxtree	Alert for acyl transfer agent	Not applicable
Danish (Q)SAR Database	Sensitising	Reliable^1)

¹⁾All three models reported predicted sensitisation, where TMEG was within the applicability domain of each model

Applicant's summary and conclusion

Interpretation of results:

Category 1A (indication of significant skin sensitising potential) based on GHS criteria

Conclusions:

With four reliable predictions there is an argument for a weight of evidence that TMEG is a sensitiser. Moreover, potency predictions with OECD Toolbox and DEREK resulted in EC3 values of < 2% and ≥ 0.135, thus indicating that TMEG is a strong or even extreme sensitiser.
 

Executive summary:

(Q)SAR predictions were performed to estimate the skin sensitation potential of ethylene bis[1,3-dihydro-1,3-dioxoisobenzofuran-5-carboxylate] (TMEG, CAS 1732-96-3). With four reliable predictions there is an argument for a weight of evidence that TMEG is a sensitiser. Moreover, potency predictions with OECD Toolbox and DEREK resulted in EC3 values of < 2%and ≥ 0.135,thus indicating that TMEG is at least a strong sensitiser.