Registration Dossier

Administrative data

Description of key information

Key value for chemical safety assessment

Skin sensitisation

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

A total of 6 skin sensitization assays were available within the SIEF. After thorough review by RSS/CSR author, incl. if necessary change of the original conclusions, four tests were of appropriate reliability, two otherswere K3, one of which was not included in IUCLID because it was not owned by a SIEF member ; all 6 tests are presented below.

Summary of key reliability and sensitivity features of the 6 skin sensitisation assays available to the 2-ethylhexylnitrate SIEF:

Reference Batch (purity) Sensitivity of the study* Comments on sensitivity and reliability check Original conclusion Revised conclusion Revision based on (results for undiluted product)**
J. Clouzeau (1988) op 3/85 (NR) Maximal (MK; 0.1 mL, 50%, 100%, 100%) GLP, induction phase insufficiently described, no positive control: K2- but gives useful histopathology data to understand nature of skin reactions Negative Negative No histopathological signs of sensitisation, overrides the 12/20 persistent macroscopic reactions at challenge
 S.R. Kynoch, B.I. Parcell (1989)- report 89652D B (NR) Maximal (MK; 0.1 mL, 30%, 100%, 100+50%) non-GLP, induction phase insufficiently described, 2/10 controls are positive** at challenge: K3*** NR NA 7/20 persistent macroscopic reactions at challenge, biased by invalid controls
 S.R. Kynoch, B.I. Parcell (1989)- report 89651D A (>99%) Medium (MK; 0.1 mL, 5%, 100%, 100+50%) non-GLP, induction phase insufficiently described but no impact as turns positive: K2 NR Positive 9/20 persistent macroscopic reactions at challenge
V M Davison (1988) CI-0801 (NR) Medium (MK; 0.05-0.1 mL, 10%, 100%, 100+30%) non-GLP, purity unknown; induction phase insufficiently described but no impact as turns positive: K2 Positive Positive 15/18**** persistent macroscopic reactions at challenge 
X. Manciaux (1998) 0020/98 (99.84%) Low (Buehler 3; NA; NA; 100%; 100%) GLP, adequate detail level and protocol: K1- but less sensitive method (Buehler 3 applications) Negative Negative 0/20 macroscopic reactions at challenge 
S.M. Glaza (1988) NR (>99%) Medium (MK; 0.05mL, 5%, 100%, 50%) GLP, induction not maximized, leaving doubt about negative responses: K3 Negative NA 0/20 macroscopic reactions at challenge, biased by non-maximized protocol

NR: not reported;       NA: not assessable or applicable

underlined: limited exposure inappropriately justified and/or deviating from the guideline

*: indicated in parentheses: method (MK or Buehler 3), volume for intradermal injection, concentrations for intradermal injection, topical induction and challenge

**: local or extensive reaction, at least until 48h after end of challenge, was re-counted as positive by RSS/CSR author

***: study not present in IUCLID dossier

****: animals with bandage slipped were excluded

Out of the four Klimisch 1-2 tests, three MK assays clearly involved macroscopic persistent reactions in >30% of challenged animals and a fourth test, a Buehler test, showed no reaction in any animal. Three main hypotheses could explain this apparent inconsistency:

- Inconsistencies could be related to an impact of sensitizing impurities and/or inter-lab variability: it was impossible to localize, decades afterwards, data on impurities in the tested batches; however, it is notable that studies by Clouzeau and Manciaux were done in the same lab for the same Sponsor, limiting this source of variability; furthermore, the synthesis process is mostly similar for all producers and unlikely to introduce a sensitizing impurity.

- Reactions could be irritation due to a bad protocol (too high concentrations): this is not likely if the test is correctly carried out (choice of a non-irritant challenge dose-level), and has been taken into account in Klimisch rating and review of conclusions.

- Reactions could be non-related to sensitization: this hypothesis is supported by four different points:

1) Study by Clouzeau showed macroscopic skin reactions sufficient for classification as sensitizer, but as opposed to all 5 other tests, this one included an histopathology confirmation that evidenced lesions evocating irritation (acanthosis and hyperkeratosis) and an absence of inflammatory cells, so it was concluded to be negative.

2) The substance, being an organic nitrate, has vasodilation properties (as confirmed by observation of headaches and dizziness in some workers) which could account for reddening of the skin that does not reflect the mechanism of standard irritation or sensitisation, but can be mistaken for it by macroscopic examination.

3) Skin cracking has been noted in rabbits after repeated dermal exposure (see 7.3) and may be related to interaction with skin lipids.

4) The IUCLID4 file (see attached file) of the synthesis precursor 2EH, which may be present as an impurity in some studies (no data), is attached and shows that this substance is clearly a skin irritant, so a possible hypothesis could be delayed formation of a skin irritant compound (in the tested batch or during the application time) leading to delayed skin irritation, misinterpreted as being sensitization.

A LLNA assay was not carried out to clarify the issue because it was considered unethical in the existence of 6 sensitisation assays (incl. 4 reliable ones), and one test would be unable to answer on the possible impact of impurities.

It may also be noted that the OECD QSAR Toolbox predicts an absence of protein binding, also in favour of an absence of sensitizing properties. See the attached document.

Migrated from Short description of key information:

A total of 6 skin sensitization assays were available within the SIEF. After thorough review by RSS/CSR author, four tests were of appropriate reliability, out of which 2 were concluded to be negative and 2 to be positive. One of the negative tests would have been considered positive due to macroscopic skin reactions, but was considered not to represent sensitisation but irritation (possibly delayed) due to an absence of typical inflammatory response at histopathology. This could also account for the other "positive" results in tests not including histopathological confirmation.

Respiratory sensitisation

Endpoint conclusion
Additional information:

Migrated from Short description of key information:

No data. The substance is not a skin sensitiser and does not include structural alerts for respiratory sensitisation. It is therefore unlikely to be a respiratory sensitizer.

Justification for classification or non-classification

As the test item induced delayed skin reactions which:

- are not representative of sensitisation (inflammatory cell infiltration) after histopathological examination,

- may be related to vasodilation (the substance being an organic nitrate), interaction with skin lipids, or delayed formation of a skin irritant impurity/degradation product,

And in the absence of predicted protein binding according to OECD QSAR Toolbox,

the substance is considered as non-sensitizing as a weight-of-evidence.