Page 438 - Environmental Nanotechnology Applications and Impacts of Nanomaterials

418 Potential Impacts of Nanomaterials

MR imaging of the central nervous system can be performed with
superparamagnetic iron oxide nanoparticle MRI contrast agents instead
of gadolinium-based contrast agents. These iron oxide–based agents
include the laboratory preparation MION-46 (Weissleder et al., 1990),
the clinically approved agent ferumoxides (Jung CW and Jacobs, 1995;
Ros et al., 1995), the investigational agents ferumoxtran-10 (Combidex;
Advanced Magnetics Inc., Cambridge, MA) (Anzai et al., 2003; Hoffman
et al., 2000; Nguyen et al., 1999), and the new agent ferumoxytol (Ersoy
et al., 2004; Prince et al., 2003). These iron oxide contrast agents con-
sist of an iron oxide core with a variable coating that determines cellu-
lar uptake and biological half-life. Ferumoxide, a superparamagnetic
iron oxide, is 60 to 185 nm in size, incompletely coated with dextran, and
rapidly opsonized and endocytosed. The other agents are ultrasmall
superparamagnetic iron oxides that are completely coated either with
native dextran (ferumoxtran-10) or with a semi-synthetic carbohydrate
(ferumoxytol) ([Muldoon et al., 2005). These agents have particle diam-
eters of 20 to 50 nm, show little opsonization and endocytosis, and have
long plasma half-lives of 14 to 30 hours (Jung et al., 1995). Despite the
growing use of the iron oxide–based MRI contrast agents in the CNS,
little is known about the long-term effects of these agents. The distri-
bution and toxicity of superparamagnetic iron oxide magnetic nanopar-
ticles were studied in rat brains and cerebral tumors (Muldoon et al.,
2005). Iron has been implicated in formation of free radicals, which
would be harmful to neural tissues, particularly in cases of stroke or
neurological disease (Dobson, 2004; Moss and Morgan, 2004). The cel-
lular loading experiments specifically tested for formation of reactive
oxygen species. No evidence of in vitro toxic effects was found from high
efficiency iron loading and no increase in reactive oxygen species (Arbab
et al., 2004).
Chronic exposure to iron nanoparticles over time is of concern. It has
been thought that the particles would dissolve and superparamagnetic
iron oxide signal would decrease through natural metabolic pathways
(Muldoon et al., 2005). It was shown that different iron oxide particles
may be detected in the CNS over extended periods, including strong
staining in CNS parenchymal cells. This agrees with previous ultra-
structural studies showing intracellular localization of iron particles
(Muldoon et al., 1999; Neuwelt et al., 1994). In human biopsy specimens,
iron uptake was demonstrated in cells morphologically identical to cells
in adjacent sections that stained for glial fibrillary acidic protein
(Neuwelt et al., 2004a). Additional studies are needed to demonstrate
that the iron labeling is still in the implanted cells at the end of a study,
rather than taken up secondarily by macrophages or reactive astro-
cytes (Muldoon et al., 2005).