Registration Dossier

Administrative data

Description of key information

Key value for chemical safety assessment

Skin sensitisation

Link to relevant study records
Reference
Endpoint:
skin sensitisation: in chemico
Remarks:
in silico
Type of information:
(Q)SAR
Adequacy of study:
weight of evidence
Study period:
3 June 2011
Reliability:
1 (reliable without restriction)
Rationale for reliability incl. deficiencies:
other: see 'Remark'
Remarks:
A DEREK prediction of sensitisation was undertaken by an experienced licensed Llasa user, using the latest version of the DEREK software, Nexus V2.0. All potential isomers of the substance are assessed. The calculation is considered appropriate for the rule base utilised, as the DEREK prediction software is fully audited by the JRC, and has an approved QMRF (see below).
Justification for type of information:
QSAR prediction
Reason / purpose:
reference to other study
Qualifier:
no guideline required
Principles of method if other than guideline:
Derek for Windows assigns various levels of non-numeric likelihood to its predictions ranging from CERTAIN to CONTRADICTED, via PROBABLE, PLAUSIBLE, EQUIVOCAL, DOUBTED, IMPROBABLE, IMPOSSIBLE and OPEN. In the Derek lexicon, PLAUSIBLE indicates that the structure has activated an alert for the endpoint in question. There is sufficient information in the database to support the prediction. EQUIVOCAL – indicates that there is an equal weight of evidence for and against the proposition. OPEN – means that there is no evidence that supports or opposes the proposition.
GLP compliance:
no
Positive control results:
Not applicable.
Key result
Parameter:
other: structural alerts
Value:
0
Remarks on result:
no indication of skin sensitisation

Result

 

TXP

No alerts triggered for skin sensitization

  

Comment and discussion

 

Derek for Windows assigns various levels of non-numeric likelihood to its predictions ranging from CERTAIN to CONTRADICTED, via PROBABLE, PLAUSIBLE, EQUIVOCAL, DOUBTED, IMPROBABLE, IMPOSSIBLE and OPEN. In the Derek lexicon, PLAUSIBLE indicates that the structure has activated an alert for the endpoint in question. There is sufficient information in the database to support the prediction. EQUIVOCAL – indicates that there is an equal weight of evidence for and against the proposition. OPEN – means that there is no evidence that supports or opposes the proposition.

 

A specific isomer of TXP was analysed for skin sensitisation using Derek Nexus in a range of mammalian species. The results indicate that TCP does not contain any alerts for skin sensitisation.

 

Derek Nexus contains over 70 separate alerts for skin sensitisation and a further 8 for photoallegenicity. In 2003, Hulzeboset alstated in a review of Derek for Windows 5.0 (the predecessor of Derek Nexus) that “the accuracy of prediction was around 60% for sensitisation” (Helzeboset al.,2003).

 

Using a later version of DEREK for Windows a review of a guinea pig archive data for compounds eliciting allergic contact dermatitis and a local lymph node assay (LLNA) data set was conducted. DEREK for windows correctly predicted 82.9% and 73% of the results from the guinea pig and LLNA data (Fedorowiczet al., 2005). New alerts are added each year to Derek software and this indicates increasing accuracy of the software with development of new versions.

  

Conclusion

 

TXP was analysed for skin sensitisation using Derek Nexus in a range of mammalian species. The results indicate that isomers of TXP do not contain any alerts for skin sensitisation and as such are unlikely to be skin sensitizers. The report is appended below for information.

Interpretation of results:
not sensitising
Conclusions:
TXP was analysed for skin sensitisation using Derek Nexus in a range of mammalian species. The results indicate that isomers of TXP do not contain any alerts for skin sensitisation and as such are unlikely to be skin sensitizers.
Executive summary:

TXP was analysed for skin sensitisation using Derek Nexus in a range of mammalian species. The results indicate that TXP does not contain any alerts for skin sensitisation.

 

Using a later version of DEREK for Windows a review of a guinea pig archive data for compounds eliciting allergic contact dermatitis and a local lymph node assay (LLNA) data set was conducted. DEREK for windows correctly predicted 82.9% and 73% of the results from the guinea pig and LLNA data (Fedorowiczet al., 2005). New alerts are added each year to Derek software and this indicates increasing accuracy of the software with development of new versions.

The data provided is considered to be substantial supporting evidence of the effects noted in the human exposure studies detailed in section 7.10 below. The substance is not considered to be a sensitiser on the basis of a weight of evidence approach.

Endpoint conclusion
Endpoint conclusion:
no adverse effect observed (not sensitising)
Additional information:

A Weight of Evidence approach is undertaken for Skin Sensitisation, based on a combination of study data, human health monitoring and QSAR results. The following is assessed:

 

Local Lymph Node Assay (Read-across):

 

The results for the Local Lymph Node Assay were read across to the analogous substance Kronitex TCP. Under the conditions of the study, Stimulation Indices (SI) of 3.7, 3.4 and 5.4 were calculated for applied concentrations of 25%, 50% and 100% Kronitex TCP, respectively.  Considering that the SI threshold value, as stated in Annex I of 1272/2008/EC and Annex VI of 67/548/EEC, is at least 3, the substance would normally be classified as a sensitizer.  However, a dose-response relationship was not observed, which in accordance with OECD guideline 429 is required of the LLNA test design.  That is, the guideline states,’ [t]he decision process with regard to a positive response includes a stimulation index =3, together with consideration of dose-response and, where appropriate, statistical significance.’

Given that supporting data on sensitization all indicates that aryl phosphates do not cause sensitization, and that a dose-response relationship was no observed in this test, it is considered that the results of the study are inconclusive.

Human Health Monitoring Data:

 

Medical data from the site of point manufacture has been reviewed for a period of 50 years, in order to assess sensitisation. In addition, research has determined a human patch test that assessed aryl phosphate exposure for both irritation and sensitisation. The data is considered appropriate for inclusion due to the fact that this is used as part of due diligence occupational exposure monitoring, and is an accurate representation of endpoint effects (lack of) at the point of key exposure. Both studies are presented together here as they are supportive of the occupational works conducted at the site.

 

On the basis of the available human data, both historical and study specific, it is concluded that the aryl phosphate substances manufactured by Chemtura do not cause skin sensitization in man. The evidence presented in the historical data on epidemiological effects detail that skin effects are not prevalent within the workforce involved in the production, handling and maintenance of the areas associated with the aryl phosphates. This conclusion is reinforced by the study specific data where irritation and sensitization effects in man where studied via the use of human patch testing. The results of this study where negative, with no effects seen in any of the volunteers assessed.

 

Aryl phosphates are therefore not considered to be skin sensitisers in man on the basis of the evidence available.

 

QSAR – DEREK Sensitisation Prediction

 

TXP was analysed for skin sensitisation using Derek Nexus in a range of mammalian species. The results indicate that TXP does not contain any alerts for skin sensitisation.

Using a later version of DEREK for Windows a review of a guinea pig archive data for compounds eliciting allergic contact dermatitis and a local lymph node assay (LLNA) data set was conducted. DEREK for windows correctly predicted 82.9% and 73% of the results from the guinea pig and LLNA data (Fedorowiczet al., 2005). New alerts are added each year to Derek software and this indicates increasing accuracy of the software with development of new versions.

 

The data provided by DEREK is considered to be substantial supporting evidence of the effects noted in the human exposure studies.

 

QSAR – CAESAR Sensitisation Prediction

 

A QSAR assessment using the CAESAR project system for skin sensitisation was conducted. The CAESAR model for skin sensitization is based on a data set that includes 209 compounds extracted from the paper by Gerberick et al. As the system is not yet listed on the QSAR Model Reporting Format Inventory (QMRF), and certain molecules are outside of the applicbility domain it has been assigned a reliability of 3. This is based on the fact that the system has undergone extensive validation in the CAESAR validation process and is considered viable, but has not yet been accepted as an approved QSAR model. As the substance is a reaction mixture, a selection of key isomers as detailed in the registration have been evaluated for the purposes of the assessment. On the basis of these results, it can be seen the molecule assessed is a non-sensitiser. Given that other QSAR systems predict non-sensitisation potential, plus human health assessments for sensitisation detail no evidence of sensitisation, it is deemed appropriate to consider that the results overall indicate negligible potential for sensitisation. The data provided is considered to provide supporting evidence of the effects noted in the human exposure studies detailed in section 7.10 below. The substance is not considered to be a sensitiser on the basis of a weight of evidence approach. This study is provided as supporting information only.

Overall conclusion:

 

The data provided is considered to be substantial supporting evidence of the effects noted in the human exposure studies. The substance is not considered to be a sensitiser on the basis of a weight of evidence approach.

Respiratory sensitisation

Endpoint conclusion
Endpoint conclusion:
no study available
Additional information:

No direct studies on respiratory sensitisation are available; however experience in use would indicate that this hazard is not experienced with manufacture and use of the substance.

Justification for classification or non-classification

The above studies have been ranked reliability 1 to 3 according to the Klimish et al system. Reliability 3 ranking was deemed appropriate to the main LLNA study was conducted to GLP in compliance with agreed protocols, but the results did not fulfil the definitive requirements of the OECD guideline. The DEREK QSAR prediction is considered reliability 1 as the methodology is well documented, and the results fell within the model system that was used in the assessment. The human health monitoring data was assigned reliability 1, based on the historical data that is available in support of the registration, plus the additional human patch testing that was conducted on the substance in the past. Finally, the CAESAR QSAR prediction is assigned as reliability 3, as there is no existing QMRF validated method available and the results are outside of the scope of the model, although the CAESAR program has conducted extensive validation in conjunction with a number of parties.

Justification for classification or non classification

On the basis of the weight of evidence approach, the above results triggered no classification under the Dangerous Substance Directive (67/548/EEC) and the CLP Regulation (EC No 1272/2008). No classification or labelling is applicable for this endpoint.