Risk Assessment of Manufactured Nanomaterials - Environmental and Human Health Impacts of Nanotechnology - page 396

able, for example, as regards thresholds used in some legislation (as reported by
Royal Commission as Environmental Pollution, http://www.rcep.org.uk/reports/27-
novel%20materials/27 - novelmaterials.htm ).
The implementation of legislation and the use of regulatory instruments
created by legislation remains a particular challenge. Documents that support
implementation, particularly in relation to risk assessment, adopted within the
context of current legislation will have to be reviewed in order to ensure that they
effectively address the risks associated with nanomaterials and make best use of
the information becoming available. Similarly, authorities and agencies will have to
pay special attention to risks in relation to nanomaterials where protection and
marketing are subject to pre-market control. To properly develop, modify or, in
particular, implement legislation, the scientifi c knowledge base needs to be improved
(ECC, 2008 ).
In principle, the REACH risk assessment process for chemicals has been judged
by the EU Commission SCENIHR (2005, 2006, 2007) as being appropriate to
nanomaterials, although it may need to be amended in the future as knowledge
gaps are identifi ed and resolved (ECC, 2008; Crane et al. , 2008 ; Section 10.6 ). In
principle, the major steps in risk assessment appear fi t for purpose, although the
technical details in each step require some consideration (see below).
The most over-arching concern is that chemical risk assessment assumes that the
chemical toxicology ultimately drives the adverse biological effects. This might hold
true for some nanomaterials (e.g. nanoparticulate TiO 2 as an oxidising chemical
causes oxidative stress in organisms (Handy et al. , 2008c). However, it is also pos-
sible that physical properties alone (e.g. shape, size, aspect ratio) are a key driver
of biological effects. Indeed, experience with materials such as asbestos has shown
that it is the aspect ratio (ratio of length to width) and biopersistence that are
important mediators of pathogenicity rather than the inherent chemical composi-
tion of the material. This can be resolved by ensuring that additional physical
parameters are added to the physico-chemical properties measured in the hazard
identifi cation step. These additions for nanomaterials, and the relevance of the
existing list of physico-chemical measurements for new substances, are discussed
at length elsewhere (Handy et al. , 2008a ; Crane et al. , 2008). The toxicity of differ-
ent sizes and shapes of the same nanomaterial should be measured in the hazard
identifi cation step so that risks associated with chemical reactivity can be separated
from other physical effects. If these risks change with size/shape, then it may be
necessary to risk assess individual size ranges of the same nanomaterial.
Pragmatically, this is something best avoided if possible, but at the very least it may
be necessary to consider each particle size range as a separate strand of evidence
in the risk analysis.
Hazard identifi cation requires collection of existing data on known toxic effects
(e.g. reviews by Handy et al. , 2008a, 2008b). New experimental evidence is rapidly
emerging. One fundamental concern about the hazard identifi cation step is that
novel materials may impart novel biological effects that have not yet been deter-
mined. This is a source of uncertainty, but the review process expected of all risk
assessments can reduce this as new data emerges. In principle, this is no different
to any other substance, as any new substance has the theoretical potential to show
new toxic effects. The consensus view is that the standard endpoints used in hazard