Dr Reach: Would Paracelsus have liked exposure-based waiving?

What would Paracelsus think about REACH? I expect that his reaction would be similar to that of many present-day toxicologists.

REACH is hazard-based. Compare it to the pre-manufacturing notification scheme under the US EPA, for instance, which incorporates exposure elements from the onset.

REACH focuses on collecting hazard data. In turn, these data feed in to other REACH processes: the need for exposure scenarios usually depends on the presence of a hazard: without identified hazard a substance is unlikely to need to undergo evaluation, restriction or authorisation.

Toxicologists often gasp when presented with the amount of information that REACH will generate; they recognise the massive undertaking needed to interpret hazard data.

Although toxicological information on each substance is quite ‘basic’ relative to some other regulatory dossiers, hazard data on tens of thousands of substances must ultimately be viewed in terms of hundreds of thousands of mixtures. And interpreting limited data sets for multiple exposure pathways and populations can prove particularly challenging.

Through consideration of exposure, the concept of ‘risk’ offers solid ground in the potential regulatory quagmire. This is often stressed by toxicologists following Paracelsus’s adage of ‘the dose makes the poison’.

Tonnage bands make REACH enforceable. But as a proxy for exposure, these are rough and undesirable measures.

Thankfully, the ‘small print’ of REACH offers flexibility to adapting hazard assessment through exposure consideration. The starting point is the adaptations to standard testing requirements (ie ‘Column 2’ of the Annexes).

When there is low potential for exposure, a test can be omitted. Where high exposure exists, lower tier studies can initially be substituted with higher tier testing proposals. The technical guidance lays out the necessary decision-making processes, including how intrinsic properties of a substance influence exposure (eg dermal uptake).

Proceeding through the REACH Annexes, further testing can often extend beyond tonnage band mandates. Even if there is an absence of concern in vitro there may nevertheless be a need for in vivo testing.

Exposure assessment can help draw a line under the need for testing. But compared to the intrinsic properties that describe ‘hazard’, stochastic measures of ‘exposure’ have proved unwieldy to legislate for under REACH and are likely to prove just as   difficult to implement, based on experience with other legislation.

The concept of Thresholds of Toxicological Concern (TTC) provides a potential solution. It can be used to distinguish between significant and negligible exposure, as well as to delineate ‘relevant’ exposure and can thereby limit and avoid animal testing.

TTCs relate to levels above which exposure is deemed to be relevant. This means that exposure below the level is of ‘no toxicological concern’ thereby establishing a guiding DNEL (Derived No-Effect Level). This can be based on endpoint-specific data pools from a set of appropriate substances.

Originating from polymer system assessment, TTC is used for food safety, as illustrated in the table shown – taken from the Norwegian Scientific Committee for Food Safety, 2006. Substances are sorted according to toxicological categories, such as the Cramer toxicity classes, based on available data.

Approaching fertility and developmental toxicity endpoints in a similar way has been recently proposed for REACH by a group of toxicologists at the German Federal Institute of Risk Assessment. For ecotoxicity, since a key publication in 2005, Dr Watze de Wolf at DuPont has been instrumental in developing and promoting the ‘environmental exposure threshold of no ecotoxicological concern’ (ETNC).

The concept of TTC profoundly impacts REACH, particularly on reducing testing requirements through the application of exposure-based waiving. Of course, application of TTC requires a range of experts, as does the refinement of any DNEL. Data-sets need to be selected and may include parent compound metabolites, routes of exposure must be accounted for (eg oral vs. dermal) together with differences in internal dose, etc.

Ultimately, REACH aims to put the founding principle of toxicology established by Paracelsus into mass practice, in order to establish safe levels of exposure to substances.

‘Safe levels’ must be set in an appropriate context, for instance, according to type of hazard (ie. endpoint) in question, ability to manage the risk, and the associated societal values and perceptions of risk with respect to production and use of a material. This is perhaps more of a hurdle to regulatory acceptance of TTC than understanding the science behind it.

I believe that, if alive today, Paracelsus would be fascinated to see the TTC manifestation of his key principle and the keen debates on its application under REACH.