Workshop Report 20 – Workshop on Guidance on Assessment Factors to Derive a DNEL

WR 20 : Workshop on Guidance on Assessment Factors to Derive a DNEL | December 2010

 Executive Summary

Under REACH, chemicals produced in or imported into the EU, in amounts of = 10 tons/year, require detailed registration dossiers including a chemical safety assessment. One key element for the chemical safety assessment is the derived no-effect level (DNEL) which in turn depends upon the point of departure of the hazard assessment and the assessment factor (AF) applied. The REACH ?Guidance on information requirements and chemical safety assessment' (REACH TGD) contains in Chapter R.8 a number of AF for extrapolation of animal data to man that are based on previous experience and convention, and are thus proposed as default values.

Previous ECETOC Task Forces have developed concepts for so-called ?informed' AF based on animal data (TR 86) and for the use of human data (TR 104) to develop AF. A recently convened ECETOC Task Force looked at further published literature to substantiate the proposed informed AF. It also worked examples, based on SCOEL data, in order to show differences in DNEL when applying default or informed AF. The draft report from the Task Force was presented to the Workshop participants that came from regulatory bodies, academia and industry. This report summarises the presentations given at the workshop and the outcome of the discussions.

In conclusion, it was realised that there are some sensitivities regarding the suggestions made and the timing of the report. For some of the issues addressed in ECETOC's draft report, e.g. on the question of the necessary conservatism, it became evident that the positions taken by the different Workshop participants were not far apart and some common ground could be gained. However, for a number of approaches in the REACH TGD and those proposed by the ECETOC Task Force on the justification for the application of informed AF, e.g. the need for a residual AF of 2.5 for remaining overall uncertainty, some of the participants raised the need for more scientific evaluation.

The valuable discussions in the breakout sessions revealed a number of helpful suggestions for improvement of the draft report. Among those were:

• Registrants should be very precise when changing default AF.
• The ECETOC report should demonstrate better where the TGD AF are too conservative.
• The wording in the ECETOC report should also respect this, i.e. not to advise against the TGD default AF. The use of informed AF need to be put appropriately into context.
• A point estimate, i.e. a single AF, does not reflect the distribution of data. Several participants criticised the ECETOC approach since in deriving AF not only scientific considerations are to be taken into account. To decide on the proportion of the population to be protected by a given DNEL a political and societal agreement is required on the necessary level of confidence, or remaining uncertainty, and level of protection.
• Human data should be ranked and weighted in the light of the relevant animal data. Also negative human data can be of relevance, e.g. in the case of irritation and sensitisation.

All of these proposals were taken up into the final report of the Task Force (meanwhile published as Technical Report No. 110).