Concern-Driven Safety Assessment of Nanomaterials - Safety of Nanomaterials along Their Lifecycle - page 363

• The uptake, biodistribution, and biopersistence (biokinetics) of a NM in

an organism and the physicochemical characteristics of a NM inside the
organism

• The early and apical biological effects

Grouping of NMs, however, can not only be performed by different parameters but
also for different purposes. A broad spectrum of incentives to perform grouping are
conceivable, for example, the need to address different exposure scenarios, to prioritize
potential risks, to apply read-across, to justify waiving, or to define criteria for testing
strategies. Ranking of NMs can be performed within a group (e.g., high/low exposure,
high/low hazard; comparable to the toxic equivalency factor approaches for dioxins or
polychlorinated biphenyls; Safe 1998) and also different groups can be ranked (e.g.,
based on effect potency, amount of NMs released from a product, and internal doses).
In this concept, a given NM may and will belong to more than one group, depend-
ing on its assignment to the respective groups (e.g., physicochemical characteristics,
release, uptake and distribution—timing of—biological effects, lifecycle stages, biop-
ersistence, susceptibility of specific population groups, and species sensitivity), all of
which can govern NM toxicity. These examples show that different concerns (be they
scientific or societal, see the following) may also drive the need for grouping.
Criteria need to be developed for grouping to enable justification of the appli-
cation of read-across for regulatory purposes (Patlewicz et al. 2013). Clearly, the
rationale of and reasons for grouping define such criteria (e.g., exposure scenarios,
hazardous effects).
In principle, a grouping strategy resembles a testing strategy insofar as both may
use the same criteria and identify the same concerns. Tools for grouping using physi-
cochemical properties can be quantitative structure activity relationship models.
Determination of pristine physicochemical characteristics is relatively inexpensive
and reliable, too. Grouping based on exposure may sufficiently be served by use
of patterns whereas in other cases actual release measurements may be necessary.
Grouping based on biokinetics as well as early biological effects usually requires
in vitro or in vivo testing; in the future in silico models may assist this (e.g., the
transport across biological barriers). Grouping based on apical biological effects will
usually require in vivo testing with the advantage of integrating biokinetics and thus
providing accurate potency information.
In the beginning, grouping will be crude and incomplete (not all NMs will be
assignable to groups and not all steps of the source-to-adverse-outcome pathway will
be used) but may be refined with increasing knowledge. Nevertheless, one should
start out with easily achievable forms of grouping. As knowledge increases, more
specific grouping will become possible. Also in the context of a tiered approach,
grouping becomes more detailed and complex with increasingly higher tiers, always
with the main incentive to avoid having to test all NMs.

16.3.2 g rouPing of nm: o ngoing d eveloPments
Recognition of mutually common properties of different NMs is an intricate mul-
tidisciplinary task. Recently, mode-of-action integrated approaches for the testing