Titanium dioxide (TiO2) and zinc oxide (ZnO) in large-particle form have long been used in various sunscreens to protect the skin by reflecting or physically blocking ultraviolet (UV) radiation. In recent years, TiO2 as well as ZnO nanoparticles have been incorporated into sunscreens and cosmetics to act as a UV shield. They have been shown to be effective barriers against UV-induced damage, and yield stronger protection against UV insult, while leaving less white residue, than previous generations of the physical sunblocks.
However, some data suggest that in nanoparticle form, TiO2 and ZnO absorb UV radiation, leading to photocatalysis and the release of reactive oxygen species (Australas. J. Dermatol. 2011;52:1-6). This column will focus primarily on the safety of TiO2 in nanoparticle form.
While numerous studies examine both TiO2 and ZnO, the primary inorganic sunscreens, the sheer number of separate investigations warrants individual articles, and ZnO was addressed in previous columns. Briefly, though, TiO2 is more photoactive and exhibits a higher refractive index in visible light than ZnO (J. Am. Acad. Dermatol. 1999;40:85-90); therefore, TiO2 appears whiter and is more difficult to incorporate into transparent products.
A 2011 study by Kang et al. showed that TiO2 nanoparticles, but not normal-sized TiO2, and UVA synergistically foster rapid production of reactive oxygen species and breakdown of mitochondrial membrane potential, leading to apoptosis, and that TiO2 nanoparticles are more phototoxic than larger ones (Drug Chem. Toxicol. 2011;34:277-84).
However, also in 2011, Tyner et al. investigated the effects of nanoscale TiO2 use on UV attenuation in simple to complex sunscreen products. They found that barrier function was diminished by none of the formulations, and that optimal UV attenuation resulted when TiO2 particles were stabilized with a coating and evenly dispersed. The researchers concluded that nanoscale TiO2 is nontoxic and may impart greater efficacy (Int. J. Cosmet. Sci. 2011;33:234-44).
In vitro and in vivo studies
In 2010, Tiano et al. evaluated five modified TiO2 particles, developed and marketed for sunscreens. They used different in vitro models, including cultured human skin fibroblasts, to determine potential photocatalytic effects after UVA exposure. The investigators found that the kind of modification to and crystal form of the TiO2 nanoparticle influences its ability to augment or reduce DNA damage, increase or decrease intracellular reactive oxygen species, diminish cell viability, and promote other effects of photocatalysis. In particular, they noted that the anatase crystal form of TiO2 retained photocatalytic activity. The authors suggested that while the debate continues over the penetration of nanosized TiO2 into the viable epidermis, their results help elucidate the potential effects of TiO2 particles at the cellular level (Free Radic. Biol. Med. 2010;49:408-15).
A 2010 study by Senzui et al. using in vitro intact, stripped, and hair-removed skin of Yucatan micropigs to test the skin penetration of four different types of rutile (the most natural form of) TiO2 (two coated, two uncoated) revealed no penetration of TiO2 type in intact and stripped skin. The concentration of titanium in skin was significantly higher when one of the coated forms was applied on hair-removed skin, with titanium penetrating into vacant hair follicles (greater than 1 mm below the skin surface), but not into dermis or viable epidermis (J. Toxicol. Sci. 2010;35:107-13).
Animal studies
In 2009, the Food and Drug Administration Center for Drug Evaluation and Research worked with the National Center for Toxicology Research using minipigs and four sunscreen formulations to determine whether nanoscale TiO2 can penetrate intact skin. Their use of scanning electron microscopy and x-ray diffraction revealed that TiO2 particles were the same size as that observed for the raw materials, implying that the formulation process influenced neither the size nor the shape of TiO2 particles (Drug Dev. Ind. Pharm. 2009;35:1180-9).
In 2010, Sadrieh et al. performed a study of the dermal penetration of three TiO2 particles: uncoated submicrometer-sized, uncoated nano-sized, and dimethicone/methicone copolymer-coated nanosized. The investigators applied 5% by weight of each of the types of particles in a sunscreen on minipigs and found no significant penetration into intact normal epidermis (Toxicol. Sci. 2010;115(1):156-66).
In 2011, Furukawa et al. studied the postinitiation carcinogenic potential of coated and uncoated TiO2 nanoparticles in a two-stage skin carcinogenesis model using 7-week-old CD1 (ICR) female mice. They found that application of coated and uncoated nanoparticles after initiation and promotion with 7,12-dimethylbenz[a]anthracene and 12-O-tetradecanoylphorbol 13-acetate at doses of up to 20 mg/mouse failed to augment nodule development. The investigators concluded that TiO2 nanoparticles do not exhibit postinitiation potential for mouse skin carcinogenesis (Food Chem. Toxicol. 2011;49(4):744-9).
Human data
Given the persistent concerns about possible side effects of coated TiO2 and ZnO nanoparticles used in physical sun blockers, Filipe et al., in 2009, assessed the localization and potential skin penetration of TiO2 and ZnO nanoparticles dispersed in three sunscreen formulations, under realistic in vivo conditions in normal and altered skin. The investigators examined a test hydrophobic formulation containing coated 20-nm TiO2 nanoparticles and two commercially available sunscreen formulations containing TiO2 alone or in combination with ZnO, with respect to how consumers actually used sunscreens compared with the recommended standard condition for the sun protection factor test. They found that traces of the physical blockers could be detected only at the skin surface and uppermost area of the stratum corneum in normal human skin after a 2-hour exposure. After 48 hours of exposure, layers deeper than the stratum corneum contained no detectable TiO2 or ZnO nanoparticles. While preferential deposition of the nanoparticles in the openings of pilosebaceous follicles was noted, no penetration into viable skin tissue was observed. The investigators concluded that significant penetration of TiO2 or ZnO nanoparticles into keratinocytes is improbable (Skin Pharmacol. Physiol. 2009;22:266-75).