Relevance of UV filter/sunscreen product photostability to human safety

Photodermatology, Photoimmunology & Photomedicine REVIEW ARTICLE Relevance of UV filter/sunscreen product photostability to human safety J. Frank Nash & Paul R. Tanner Central Product Safety, Global Product Stewardship, The Procter & Gamble Company, Cincinnati, OH, USA. Key words: adverse events; photoallergy; photo-degradation; phototoxicity; sunscreen products; UV filters Correspondence: Dr J. Frank Nash, PhD, Central Product Safety, Global Product Stewardship, The Procter & Gamble Company, 11530 Reed Hartman Highway, Cincinnati, OH 45241, USA. Tel: +1 513 626 3749 Fax: +1 513 626 7349 e-mail: nash.jf@pg.com Accepted for publication: 13 January 2014 Conflicts of interest: None declared. SUMMARY Photostability or photo-instability of sunscreen products is most often discussed in undesirable terms with respect to human safety. The health risks, specifically associated with sunscreens, photostable or photo-unstable, include phototoxic/photoirritation or photoallergic responses and, longer-term, an increased risk of skin cancers or photoageing. The aims of this paper are to define photostability/photo-instability and objectively assess the acute and chronic toxicological consequences from the human exposure to UV filter/ sunscreens and any probable photo-degradation products. The reported prevalence of photoirritation and photoallergic responses to sunscreens is rare compared with adverse events, for example, skin irritation or sensitization, produced by cosmetics or topically applied drugs and do not directly implicate potential photo-degradation products of UV filters. Moreover, for at least one photo-unstable combination, octyl methoxycinnamate and avobenzone, the long-term benefits to humans, i.e., reduction in skin cancers, seem to outweigh any potential adverse consequences attributed to photo-degradation. Sunscreen products are formulated to achieve maximum efficacy which, by necessity and design, incorporate measures to support and promote photostability since all organic UV filters have the potential to photo-degrade. Current performance measures, in vivo SPF and in vitro UVA, conducted under standardized conditions, in part account for photostability. The concerns expressed when considering human exposure to potential photounstable UV filters or sunscreen products may not manifest as health risks under conditions of use. Still, improvement in sunscreen product photostability continues to be a key strategic objective for manufacturers. Photo-instability is not a term that imparts a sense of confidence when discussing sunscreens. In fact, the negative connotations associated with the terms photoinstability or photostability have been used by diverse stakeholders, including nongovernment organizations, for example, Environmental Working Group; regulators, for example, US Food and Drug Administration (FDA) (1); and manufacturers to evoke uncertainty or promote product differentiation often in the inferred context of human safety. Moreover, in most cases when photostability is mentioned, it is done so with ambiguity and little technical explanation or context which further promotes doubt about sunscreen product integrity and safety. There is no disagreement that sunscreen products should be photostable. However, as the popular expression goes, the devil is in the details. For example: 88 doi:10.1111/phpp.12113

Photostability and human safety What is the definition of photostability/photo-instability? How should photostability be determined and is there a threshold separating photostable and photo-unstable products? What are the consequences, toxicologically or otherwise, of human use of a photo-unstable sunscreen product? The aim of this paper will be to address these questions and provide an objective assessment of the relevance of ultraviolet (UV) filter/sunscreen product photostability to human safety. What is the definition of photostability/photo-instability? This is not an easily answered question and requires some basic understanding of how sunscreens reduce solar UV radiation. In very simplistic terms, energy in the form of photons is absorbed by a chromophore, in the case of sunscreens, the UV filter. This results in the excitation of the molecule that absorbed the energy. The excited chromophore, that is, UV filter, dissipates the absorbed energy in most cases very rapidly, for example, 10 15 s, in the form of heat or light, for example, fluorescence, phosphorescence, returning to its native or ground state. In theory, this process could continue over and over. However, there are other pathways of dissipating absorbed energy which may be destructive to the chromophore or, at a minimum, reduce/eliminate its capacity to absorb energy, through processes such as isomerization, fragmentation, reactivity with other molecules, or production of free radicals (2). This is the context that photostability is most often being considered, that is, degradation of UV filters. Importantly, photostability is not an all-or-nothing response, but rather a continuum. In this regard, it might be surprising to know that even the most photostable UV filters/ sunscreen products may photo-degrade to some extent under the right conditions, that is, fluence, heat, time, application density, etc. For the sake of this paper, photostability of sunscreen products will be defined as functional rather than analytical. In general and practically speaking, the photostability of sunscreen products may be determined in vitro using absorbance/transmittance 1 measures (3), that is, functional, or by analytical methods, for example, highperformance liquid chromatography, mass spectrometry. Regarding the latter, the quantitative determination of 1 According to Beer s law, A = log 10 1/T, where A is absorbance and T is transmittance. It is common practice to use absorbance to describe the functional characteristics of UV filters/sunscreen products. Therefore, the term absorbance will be used throughout this paper. individual UV filter(s) is often methodologically complex and provides, at best, incomplete information regarding the photoprotective benefit. Therefore, measurement of UV absorption has been more commonly used by manufacturers and others to assess photostability of UV filters alone and in combination with other filters and raw materials (4). How should photostability be determined and is there a threshold separating photostable and photo-unstable products? As stated, a change in absorption of UV filters/sunscreen products following irradiation with solar-simulated UV is a common approach used to functionally assess photostability (Fig. 1a,b). To do so, a fixed dose of Absorbance Absorbance (a) 1 0.8 0.6 0.4 0.2 (b) 1 0 290 300 310 320 330 340 350 360 370 380 390 400 0.8 0.6 0.4 0.2 Wavelength (nm) 0 290 300 310 320 330 340 350 360 370 380 390 400 Wavelength (nm) Before UV Exposure After UV Exposure Before UV Exposure After UV Exposure Fig. 1. Absorbance curves for photo-unstable 7.5% OMC + 2% avobenzone oil/water sunscreen formula (a) and photostable 10% octocrylene + 2% avobenzone oil/water formula (b) before and after irradiation with 30 J/cm 2 of solar-simulated UV light from an Oriel 1000 W solar simulator. These experimental formulae were applied 2 mg/cm 2 to hydrated Vitro Skin with a finger cot and allowed to dry for 15 min. Absorbance curves were measured using Labsphere UV 1000 instrument. This was repeated following exposure to the solar simulator. UV, ultraviolet. 89

Nash & Tanner sunscreen product is applied to an optically acceptable substrate. The absorption of UV from a scanning or fixed UV/visible light source is measured. Then product is exposed to solar-simulated UV radiation after which the spectral absorbance is measured again. This spectrum provides information regarding the functional consequence of solar-simulated UV exposure on the absorption properties of the sunscreen product. Conceptually, this has been incorporated into in vitro test methods for UVA metrics (5, 6). Again, this is a gross oversimplification and the experimental details are complex where subtle changes can introduce differences in the final measured spectral absorbance. Specifically, the choice of substrate, dose of product and application technique, fluence of solar-simulated UV preirradiation, and temperature are among the experimental variables that affect the repeated measure of spectral transmittance of sunscreen products. Importantly, the evaluation of changes in absorption is a continuum. There is no single threshold separating photostable from photo-unstable UV filters/sunscreen products. It would be entirely subjective and arbitrary to apply any singular descriptor to categorize photostability. Moreover, it would be challenging to interpret what a change in absorption means with respect to human safety and/or efficacy. What are the consequences, toxicologically or otherwise, of human use of a photo-unstable sunscreen product? Perhaps, this is the most important question. If a sunscreen product is photo-unstable, will there be a greater risk of adverse health effects? The most probable human safety considerations would be in response to single or shortterm product application plus sunlight exposure, that is, during expected use scenarios. This might manifest as skin irritation or photodermatosis, including phototoxicity/ photoirritation and photoallergic reactions produced by exposure to photo-degradation products which might form in a photo-unstable product. As well, long-term health concerns might include the effect of repeated exposure to such photo-degradation products which could have toxicological consequences in and of themselves. Further, an increase in UV exposure due to diminishing UV protection while using a photo-unstable sunscreen product might, in theory, lead to increased risk of acute, for example, sunburn, and chronic skin damage, for example, skin cancers. These will be considered examining toxicological/safety data from preclinical and clinical studies with photo-unstable UV filters alone and in combination. 90 SPF 30 25 20 15 10 5 0 4 8 12 16 20 Cumulative Erythemal UV Irradiation (MED) OVERVIEW OF SUNSCREEN PRODUCT FORMULATION Photostable Not Photostable Fig. 2. Example of the variation in SPF protection offered by two hypothetical SPF 20 sunscreen products when exposed to UV light. The 100% photostable product maintains SPF 20 performance regardless of UV irradiation. In contrast, the photo-unstable SPF 20 sunscreen initially delivers much higher SPF protection, but that protection slowly decreases with continued UV irradiation. UV, ultraviolet; SPF, sun protection factor. A discussion of photostability of sunscreen products needs some understanding of how such products are formulated. Sunscreen products are not simple solutions with a single UV filter. They are complex mixtures designed to form a uniform layer covering the skin which provides protection against solar UV exposure (7). In the development of a sunscreen product, the initial consideration is the desired efficacy target, that is, sun protection factor (SPF)/broadspectrum (UVA), followed by attributes such as water/ wear resistance or moisturization. The efficacy targets will help define the choice of UV filters that will be used and the attributes will determine choice of vehicle/formulae raw materials. A key focus in designing sunscreen products is to select a UV filter combination that delivers the SPF/UVA targets with maximum efficiency, thus minimizing the high cost, negative skin feel, and other trade-offs of these materials (7). Importantly, to achieve maximum efficiency photostability must be considered, as it has a significant impact on the tests, in vivo SPF and in vivo/in vitro UVA testing, used to label such products. Specifically, as illustrated in Fig. 2, to achieve a given SPF target, a sunscreen that has poor photostability must initially, before solarsimulated UV exposure, provide protection that is significantly greater than the target SPF. 2 Thus, a UV filter 2 Assuming SPF is a determination of the total number of photons absorbed and that exceeding this threshold will produce defined erythema.

Photostability and human safety combination with poor photostability is not very efficient, as it must have excess concentrations of UV filters to compensate for the subsequent loss of UV absorbance that will occur upon exposure to UV in SPF testing and similarly in UVA testing. For modern sunscreens utilizing UV filters allowed by FDA, photostability is part of the formulation consideration particularly if the UVA filter avobenzone (a.k.a. Parsol 1789, 4-tert-butyl-4 methoxydibenzoylmethane) is used. In the past 15 years, the inclusion of avobenzone in sunscreen products marketed in the United States has more than doubled (8), because of its strong UVA attenuation and the increased understanding of the negative health effects of chronic UVA exposure (9, 10). To manage the photostability of avobenzone, thus ensuring maximum efficiency, a variety of photostabilizers are typically combined with it, including other UV filters such octocrylene as illustrated in Fig. 1b (11), and excited state quenchers such as diethylhexyl 2,6-naphthalate and ethylhexyl methoxycrylene (2, 12). Whereas these formulation developments reduce photo-degradation, they do not completely eliminate it. Generally speaking, part of the motivation behind photostabilization of avobenzone or any other UV filter is driven by efficiency gains and the resulting improved photoprotection rather than direct human safety concerns. Regardless, modern sunscreen formulation includes efforts designed to maximize the photostability of commercial products. PHOTOSTABILITY OF UV FILTERS: TOXICOLOGICAL CONSIDERATIONS The direct toxicological study of photo-degradation products of UV filters is rare. That said, there are many studies which have investigated the human safety of UV filters alone and following exposure to UV from a variety of sources which indirectly consists of an evaluation of the parent and any photo-degradation products. Clearly, a comprehensive review of UV filter/sunscreen product toxicological studies is beyond the scope of this paper and readers are referred to reviews of this subject (13 15). There are, however, noteworthy examples of photounstable UV filters, alone and in combination, which will be discussed. Para amino benzoic acid (PABA) One of the first UV filters widely used was p-amino benzoic acid, better known as PABA. It is no longer used to any great extent globally but serves as a case study in the context of photostability as it is clearly known that irradiation of PABA solutions results in photo-degradation (16 18). It is important to consider photo-degradation of PABA in two toxicological constructs: (1) photoallergy or photocontact hypersensitivity, and (2) photogenotoxicity and photocarcinogenic effects. PABA is a weak photoallergen first reported in clinical case reports and later confirmed in human photopatch testing (19 22). The prevalence of PABA-mediated photocontact hypersensitivity is unknown. However, it is reported to be infrequent based on diagnostic photopatch testing results (23 25). Nonetheless, the clinical findings have been investigated and confirmed in guinea pig (26) and mouse (27) models of photoallergy. It is likely that excitation of the PABA molecule increases the reactivity of the primary amine group (16). This photoreactivity, coupled with its demonstrated dermal penetration (28, 29), contribute to the mechanistic likelihood of photo-hapten formation, a step in the induction of photoallergy. There is no evidence that photostabilizing PABA in a sunscreen formulation eliminates the photoallergic nature of this UV filter. However, it is quite certain that the removal of this UV filter from sunscreens was the result of reports of photoallergy to the point where claims of PABA free were present on many commercial products. It was reported that PABA and its photo-degradation products interact with DNA, which has been interpreted as photomutagenicity (17). Such photomutations have been postulated as the first step in the process leading to tumor formation. This hypothesis, that is, PABA and/or PABA photo-degradation products may be photocarcinogens, was tested in a mouse photocarcinogenicity study (30, 31). Topical treatment once daily for 30 weeks (5 days/week) of female hairless mice with PABA or a partially photodegraded solution of PABA did not have a tumorigenic effect. In this study, treatment with the photo-degraded PABA solution before repeated exposure to a UVB light source significantly reduced the appearance of skin tumors. These data are suggestive that photo-degradation of PABA does not contribute to the appearance of UV-induced skin tumors in mice. Certainly, solutions of PABA photo-degrade following irradiation. There is no evidence that photo-degradation products of PABA are the cause of the photoallergic response to this material. Treatment of mice with photodegradation products of PABA did not augment UV-induced skin tumor formation. On the contrary, it reduced such events. Thus, for this material, the question of photostability did not appear to mediate toxicological events, which seem more the result of photoactivation of the parent molecule. Importantly, photoactivation of a UV filter is not synonymous with photo-degradation even if the former may ultimately contribute to the 91

Nash & Tanner destruction of the parent molecule upon absorption of UV energy. Avobenzone or Parsol 1789 or 4-tert-butyl-4 methoxydibenzoylmethane As previously stated, one of the principle points of discussion for avobenzone is photostability. It has been known for several years that exposure to UV radiation will cause avobenzone to breakdown (32, 33). Prior to the inclusion of avobenzone as a generally recognized as safe and effective over-the-counter sunscreen active (34), there were at least two products marketed in the United States under a New Drug Application (NDA) containing avobenzone, both of which would likely be considered photo-unstable. Without the benefit of directly reviewing data supportive of the NDA, one can only speculate that characterization of any photo-degradation and safety assessment of such products would have been done prior to approval. As well, such consideration of human safety would apply to the 1997 FDA approval of avobenzone (34) prior to widespread marketing of products containing this photo-unstable material. Thus, at least for one photo-unstable UV filter, independent regulatory approval would imply no or limited human health concern following topical application. Like PABA, there are clinical reports of photoallergy attributed to avobenzone (35, 36). As well, a solution of avobenzone evaluated in the 3T3 Neutral Red Uptake Phototoxicity Test (OECD 432) has been shown to have phototoxic potential (37). The photoallergic potential of avobenzone may be the result of the photoproducts formed following exposure to UV. Karlsson et al. (38). showed that there were two major classes of dibenzoylmethane photoproducts, arylglyoxals and benzils. In this superb study, arylglyoxals were shown to be strong sensitizers in the mouse local lymph node assay. These data suggest that photo-degradation of avobenzone forms classes of photoproducts which have strong potential for sensitization. Given the photoallergic hazard potential of avobenzone and its increasing use worldwide (8), it follows that some evidence of photocontact sensitization should be evident. Studies where photopatch testing has been conducted provide conclusive evidence that avobenzone is one of several UV filters responsible for photoallergic responses (39 41). Yet what is surprising, given the mechanistic understanding and known photo-degradation of avobenzone, is that the reported prevalence and positive photopatch test remain relatively infrequent. In this regard, the work of Gaspar et al. (37). provide a possible explanation because the phototoxicity potential as detected by 3T3 NRU 92 PT assay was positive for avobenzone alone and in combination with other UV filters. However, when tested on a human skin model, the positive results were no longer observed. Thus, it is possible that photostability of sunscreen products containing avobenzone is acceptable or perhaps constantly improving or the risk of photoallergy is low or a combination of these. As well, this may simply illustrate the difference between hazard and risk wherein the exposure to avobenzone and photo-degradation products under conditions of use as a sunscreen may be below a threshold needed to induce/elicit photoallergy. Finally, avobenzone has been tested in a 12-month photocarcinogenicity study in hairless mice by P.D. Forbes et al., at The Center for Photobiology, Temple University (Givaudan-Roure, Listing of Animal & Human Testing, July 29, 1993). In this study, avobenzone and any potential photo-degradation products were found to protect against solar-simulated UV-induced skin tumor formation. Similarly, in a study by Harrison et al. (42), 0.75% and 2.0% avobenzone in an oil/water emulsion was tested alone and in combination with octyl methoxycinnamate (OMC). Skh-1 albino hairless mice were irradiated 8 h a day, 5 days/week (Monday Friday) for 16 32 weeks using different lighting conditions, that is, full spectrum or UVA only. The topical application of avobenzone reduced biochemical changes and connective tissue damage at 2% but not at 0.75%. Thus, under conditions of repeated exposure to UV in the presence of avobenzone and, implicitly, any of its photo-degradation products, there was no evidence of augmentation of UV damage or reversal/diminution of the photoprotective effects of this well known, photounstable compound. Like PABA, the UVA filter avobenzone photo-degrades when exposed to UV. Avobenzone has been reported to produce positive responses in photopatch testing. In contrast to PABA, there is a proposed mechanism by which avobenzone, following UV exposure, may photo-degrade forming at least one photoproduct that has strong sensitization potential in a mouse model. However, to date the reported prevalence of avobenzone photoallergic responses has been nominal despite increasing use of this filter, specifically, and sunscreen products over the past couple of decades. Finally, avobenzone in repeat exposure animal studies has been shown to be protective when tested in phototoxicological/photobiological studies. Combinations of UV filter A number of studies have evaluated commercial products and demonstrated a range of photostability/photoinstability (43 45). As illustrated in Fig. 1a, the most

Photostability and human safety notable photo-unstable combination of UV filters is OMC and avobenzone (46, 47). OMC, a UVB filter, and avobenzone provide broad-spectrum coverage against UV and as such have been a popular combination despite known photostability concerns. What may not be as well appreciated is that this combination of UV filters was used in the most significant sunscreen study of the 20th century conducted by Adele Green and colleagues (48). The product, labeled as SPF 15+ water resistant was reported to contain 8% OMC +2% avobenzone in a cream base. No information is available to definitely establish the photostability of this product. However, when the study was initiated in 1992, photostabilization of this combination was not widely discussed. If we assume even a modest level of photo-instability, the outcome of the study is even more remarkable. The Nambour Skin Cancer Prevention study was a prospective, population-based controlled clinical trial which found that this combination of UV filters reduces nonmelanoma and melanoma skin cancers (49, 50) and evidence of photoaging (51). The ability of this photounstable combination to prevent skin cancers has been shown in an animal study as well (52). Thus, it would seem that the benefits of even possible photo-unstable combinations of UV filters outweigh the possibility of adverse events. Finally, there is no questioning that the nano form of titanium dioxide (TiO 2 ) and zinc oxide (ZnO) is stable after exposure to sunlight. However, these metal oxides may produce free radicals after exposure to UV which could degrade other ingredients in the product that come into contact, so called photo-oxidation (44). It has been shown that TiO 2 surface treatments commonly used on UV filters, including common cosmetic ingredients both inorganic (e.g. silica, alumina) and organic (e.g. stearic acid, various silicones), greatly reduce this photo-oxidative behavior (53). Toxicologically, the diversity of nano TiO 2 and ZnO products, that is, possible combinations with organic filters, makes generalizations overreaching. Nonetheless, there is no evidence from clinical case reports or other literature sources of adverse health effects, for example, photoirritation/phototoxicity or photoallergy, from the hypothetical photo-degradation of organic UV filters by metal oxides. SUMMARY AND CONCLUSION Photostability of UV filters and sunscreens is an important part of the formulation of these products. Ideally, the lowest concentration and fewest number of UV filters to achieve maximum target efficacy would be incorporated into an aesthetically pleasing formula. Such a sunscreen product would be applied, the filters would absorb solar UVB/UVA, and this cycle would be repeated over and over again without any loss of the chromophores. This, of course, would represent the ultimate photostable sunscreen. Remarkably, this theoretical product would still have the potential to produce adverse events even if it were 100% photostable and would need to undergo a full product safety analysis. Unfortunately, photostability and photo-instability are portrayed as two sides of a coin rather than opposite ends of a continuum. This all-or-nothing framework carries with it a similar implied human safety profile wherein the potential formation of photo-degradation products in the case of a photo-unstable product would be hazardous and likely to produce unacceptable human health effects. Such a reductionist view is in practice inaccurate and is not supported by historical and current understanding of adverse events associated with exposure to photostable or photo-unstable sunscreen products. Photo-instability of UV filters/sunscreen products is not a new discovery. It is critical to examine historical work done in support of the human safety of UV filters/ sunscreen products as many of these studies have been conducted with photo-unstable UV filters, alone and in combination, by design or default. For example, the work of Flindt and colleagues (18, 30, 31), in which the direct effect of photo-degraded PABA was evaluated in a photocarcinogenicity study, was investigated. As well, there are numerous studies examining avobenzone in combination with OMC (42, 52), a known photounstable combination, which show protection and no evidence indicative of augmentation or novel adverse events. This includes the work of Green and colleagues (48 51). It is in these cases that the benefit of photoprotection outweighs the risk of effects that may otherwise be difficult to observe. What is important, however, is that existing in vivo preclinical studies do not identify a hazard that could be inferentially assigned to photo-degradation of UV filter or combinations. Most certainly one of the points of discussion is avobenzone photostability. It has been known for several years that exposure to UV will cause avobenzone to photodegrade. As well, in vitro studies (37), clinical case reports (35, 36), and an extremely important mechanistic investigation by Karlsson et al. (38). show that the photodegradants of avobenzone are likely responsible for the photoallergic potential. While some studies have used high exposures (54), the photo-degradation of avobenzone alone and in combination with other UV filters has been demonstrated using a variety of experimental conditions (33, 43). Through the work of many, the photo-degradation 93

Nash & Tanner of this critical UVA filter can be reduced with the inclusion of stabilizers such as other UV filters (2, 55, 56). When combining the photo-degradation and photoallergenic potential, it is clear that photostabilization will greatly reduce the likelihood of such an event. However, the risk even in photo-unstable formulations must be low. This assertion is based on the cumulative evidence, that is, photopatch testing, clinical reports, and perhaps the bias of the findings from the Nambour Skin Cancer prevention studies using the combination of OMC and avobenzone. In any case, modern sunscreens which contain avobenzone are formulated with the photostability in mind. Sunscreen products are unique products in that efficacy as measured by in vivo SPF is assured provided proper use. By design, the SPF test is also a partial measure of functional photostability, albeit weighted for shorter more energetic wavelengths of UV. The inclusion of in vitro measures of UVA including ISO 24443 Determination of sunscreen UVA photoprotection in vitro have contributed to the standardized testing which further provide some measure of photostability. The combination of in vivo SPF and in vitro UVA provides complimentary assurance of functional photostability of final sunscreen product formulae. It is imaginable that photo-instability is not relevant when considering human safety or those adverse events are poorly or underreported, thereby leading to a false sense of security. There is most likely a combination of low risk when considering human exposure to photodegradation products that could be present in a photounstable sunscreen product. As well, reported prevalence of sunscreen product adverse events is probably underreported, although this may have less to do with photo-stability than with the subjective nature of such events, that is, some people are more tolerant than others. In summary, the awareness of photostability/photoinstability of sunscreen products has arguably improved modern sunscreens. Photostability is important to achieve maximum efficiency of UV filter performance and further reduce the risk of adverse events associated with human use of such products. It is quite probable that as more and more people use sunscreens either on a daily basis or in a recreational context, the reported prevalence of adverse events will increase. This may not be a reflection of product photostability. A careful analysis of the population data, that is, case reports, photopatch testing, as currently performed in United States and Europe (40, 41), will be needed. REFERENCES 1. Food and Drug Administration. Labeling and effectiveness testing; sunscreen drug products for over-the-counter human use. Final rule. Fed Regist 2011; 76: 35620 35665. 2. Bonda CA. The photostability of organic sunscreen actives: a review. In: Shaath NA, ed. Sunscreens: regulations and commercial development. Boca Raton, FL: Taylor & Francis, 2005, 321 349. 3. Couteau C, Faure A, Fortin J, Paparis E, Coiffard LJ. Study of the photostability of 18 sunscreens in creams by measuring the SPF in vitro. J Pharm Biomed Anal 2007; 44: 270 273. 4. Kockler J, Oelgemöller M, Robertson S, Glass BD. Photostability of sunscreens. J Photochem Photobiol C Photochem Rev 2012; 13: 91 110. 5. Diffey BL, Tanner PR, Matts PJ, Nash JF. In vitro assessment of the broad-spectrum ultraviolet protection of sunscreen products. J Am Acad Dermatol 2000; 43: 1024 1035. 6. Moyal D, Alard V, Bertin C et al. The revised COLIPA in vitro UVA method. Int J Cosmet Sci 2013; 35: 35 40. 7. Tanner PR. Sunscreen product formulation. Dermatol Clin 2006; 24: 53 62. 8. Wang SQ, Tanner PR, Lim HW, Nash JF. The evolution of sunscreen products in the United States a 12-year cross sectional study. Photochem Photobiol Sci 2013; 12: 197 202. 9. Yaar M, Gilchrest BA. Photoageing: mechanism, prevention and therapy. Br J Dermatol 2007; 157: 874 887. 10. Rünger TM, Kappes UP. Mechanisms of mutation formation with long-wave ultraviolet light (UVA). Photodermatol Photoimmunol Photomed 2008; 24: 2 10. 11. Lhiaubet-Vallet V, Marin M, Jimenez O, Gorchs O, Trullas C, Miranda MA. Filterfilter interactions. Photostabilization, triplet quenching and reactivity with singlet oxygen. Photochem Photobiol Sci 2010; 9: 552 558. 12. Benevenuto CG, Sala Di Matteo MA, Maia Campos PMBG, Gaspar LR. Influence of the photostabilizer in the photoprotective effects of a formulation containing UV-filters and vitamin A. Photochem Photobiol 2010; 86: 1390 1396. 13. Gasparro FP, Mitchnick M, Nash JF. A review of sunscreen safety and efficacy. Photochem Photobiol 1998; 68: 243 256. 14. Nash JF. Human safety and efficacy of ultraviolet filters and sunscreen products. Dermatol Clin 2006; 24: 35 51. 15. Nohynek GJ, Schaefer H. Benefit and risk of organic ultraviolet filters. Regul Toxicol Pharmacol 2001; 33: 285 299. 16. Gasparro FP. UV-induced photoproducts of paraaminobenzoid acid. Photodermatol 1985; 2: 151 157. 17. Gasparro FP. PABA: friend or foe? Photodermatol 1986; 3: 61 63. 18. Flindt-Hansen H, Nielsen CJ, Thune P. Measurements of the photodegradation of PABA and some PABA derivatives. Photodermatol 1988; 5: 257 261. 19. Horio T, Higuchi T. Photocontact dermatitis from P-aminobenzoic acid. Dermatologica 1978; 156: 124 128. 20. Wennersten G, Thune P, Jansen CT, Brodthagen H. Photocontact dermatitis: current status with emphasis on allergic contact photosensitivity (CPS) occurrence, allergens, and practical phototesting. Semin Dermatol 1986; 5: 277 289. 21. Thune P, Jansen C, Wennersten G, Rystedt I, Brodthagen H, McFadden N. The Scandinavian multicenter photopatch study 1980 1985: final report. Photodermatol 1988; 5: 261 269. 94

Photostability and human safety 22. Dromgoole SH, Maibach HI. Sunscreening agent intolerance: contact and photocontact sensitization and contact urticaria. J Am Acad Dermatol 1990; 22: 1068 1078. 23. Holzle E, Neumann N, Hausen B et al. Photopatch testing: the 5-year experience of the German, Austrian, and Swiss photopatch test group. J Am Acad Dermatol 1991; 25: 59 68. 24. Neumann NJ, Hölzle E, Plewig G et al. Photopatch testing: the 12-year experience of the German, Austrian, Swiss photopatch test group. J Am Acad Dermatol 2000; 42: 183 192. 25. Bryden AM, Moseley H, Ibbotson SH et al. Photopatch testing of 1155 patients: results of the U.K. multicentre photopatch study group. Br J Dermatol 2006; 155: 737 747. 26. Gerberick GF, Ryan CA. Contact photoallergy testing of sunscreens in guinea pigs. Contact Dermatitis 1989; 20: 251 259. 27. Gerberick GF, Ryan CA. A predictive mouse ear-swelling model for investigating topical photoallergy. Food Chem Toxicol 1990; 28: 361 368. 28. Wang LH, Huang WS, Tai HM. Simultaneous determination of p-aminobenzoic acid and its metabolites in the urine of volunteers, treated with p-aminobenzoic acid sunscreen formulation. J Pharm Biomed Anal 2007; 43: 1430 1436. 29. Bissett DL, McBride JF, Hannon DP, Patrick LF. Time-dependent decrease in sunscreen protection against chronic photodamage in UVB-irradiated hairless mouse skin. J Photochem Photobiol B 1991; 9: 323 334. 30. Flindt-Hansen H, Thune P, Nielsen CJ. Photocarcinogenesis is retarded by a partly photodegraded solution of paraaminobenzoic acid. Photodermatol 1989; 6: 263 267. 31. Flindt-Hansen H. Photocarcinogenesis in mice and the effect of a sunscreen, paraaminobenzoic acid. Dan Med Bull 1992; 39: 355 362. 32. Gonzenbach H, Hill TJ, Truscott TG. The triplet energy levels of UVA and UVB sunscreens. J Photochem Photobiol B 1992; 16: 377 379. 33. Schwack W, Rudolph T. Photochemistry of dibenzoyl methane UVA filters. Part 1. J Photochem Photobiol B 1995; 28: 229 234. 34. Food and Drug Administration. Marketing status of products containng avobenzone; enforcement policy. Fed Regist 1997; 62: 23350 23356. Ref Type: Generic. 35. Schauder S, Ippen H. Photoallergic and allergic contact dermatitis from dibenzoylmethanes. Photodermatol 1986; 3: 140 147. 36. Motley RJ, Reynolds AJ. Photocontact dermatitis due to isopropyl and butyl methoxy dibenzoylmethanes (Eusolex 8020 and Parsol 1789). Contact Dermatitis 1989; 21: 109 110. 37. Gaspar LR, Tharmann J, Maia Campos PM, Liebsch M. Skin phototoxicity of cosmetic formulations containing photounstable and photostable UV-filters and vitamin A palmitate. Toxicol In Vitro 2013; 27: 418 425. 38. Karlsson I, Hillerström L, Stenfeldt AL, Mårtensson J, Börje A. Photodegradation of dibenzoylmethanes: potential cause of photocontact allergy to sunscreens. Chem Res Toxicol 2009; 22: 1881 1892. 39. Victor FC, Cohen DE, Soter NA. A 20-year analysis of previous and emerging allergens that elicit photoallergic contact dermatitis. J Am Acad Dermatol 2010; 62: 605 610. 40. Shaw T, Simpson B, Wilson B, Oostman H, Rainey D, Storrs F. True photoallergy to sunscreens is rare despite popular belief. Dermatitis 2010; 21: 185 198. 41. Kerr AC, Ferguson J, Ibbotson SH et al.a European multicentre photopatch test study. Br J Dermatol 2012; 166: 1002 1009. 42. Harrison JA, Walker SL, Plastow SR, Batt MD, Hawk JLM, Young AR. Sunscreens with low sun protection factor inhibit ultraviolet B and A photoaging in the skin of the hairless albino mouse. Photodermatol Photoimmunol Photomed 1991; 8: 12 20. 43. Maier H, Schauberger G, Brunnhofer K, Honigsmann H. Change of ultraviolet absorbance of sunscreens by exposure to solar-simulated radiation. J Invest Dermatol 2001; 117: 256 262. 44. Serpone N, Salinaro A, Emeline AV, Horikoshi S, Hidaka H, Zhao J. An in vitro systematic spectroscopic examination of the photostabilities of a random set of commercial sunscreen lotions and their chemical UVB/UVA active agents. Photochem Photobiol Sci 2002; 1: 970 981. 45. Hojerova J, Medovcikova A, Mikula M. Photoprotective efficacy and photostability of fifteen sunscreen products having the same label SPF subjected to natural sunlight. Int J Pharm 2011; 408: 27 38. 46. Dondi D, Albini A, Serpone N. Interactions between different solar UVB/ UVA filters contained in commercial suncreams and consequent loss of UV protection. Photochem Photobiol Sci 2006; 5: 835 843. 47. Scalia S, Mezzena M. Photostabilization effect of quercetin on the UV filter combination, butyl methoxydibenzoylmethane-octyl methoxycinnamate. Photochem Photobiol 2010; 86: 273 278. 48. Green A, Battistutta D, Hart V et al. The nambour skin cancer and actinic eye disease prevention trial: design and baseline characteristics of participants. Control Clin Trials 1994; 15: 512 522. 49. Green A, Williams G, Neale R et al. Daily sunscreen application and betacarotene supplementation in prevention of basalcell and squamous-cell carcinomas of the skin: a randomised controlled trial. Lancet 1999; 354: 723 729. 50. Green AC, Williams GM, Logan V, Strutton GM. Reduced melanoma after regular sunscreen use: randomized trial follow-up. J Clin Oncol 2011; 29: 257 263. 51. Hughes MCB, Williams GM, Baker P, Green AC. Sunscreen and prevention of skin aging: a randomized trial. Ann Intern Med 2013; 158: 781 790. 52. Hillebrand GG, Winslow MS, Heitmeyer DA, Bissett DL. Chronic ultraviolet radiation-induced skin tumors and associated changes in basal epidermal ornithine decarboxylase activity. J Soc Cosmet Chem 1990; 41: 187 195. 53. Lewicka ZA, Yu WW, Oliva BL, Contreras EQ, Colvin VL. Photochemical behavior of nanoscale TiO 2 and ZnO sunscreen ingredients. J Photochem Photobiol A Chem 2013; 263: 24 33. 54. Roscher NM, Lindemann MKO, Bin Kong S, Cho CG, Jiang P. Photodecomposition of several compounds commonly used as sunscreen agents. J Photochem Photobiol A Chem 1994; 80: 417 421. 55. Chatelain E, Gabard B. Photostabilization of butyl methoxydibenzoylmethane (Avobenzone) and ethylhexyl methoxycinnamate by bis-ethylhexyloxyphenol methoxyphenyl triazine (Tinosorb S), a new UV broadband filter. Photochem Photobiol 2001; 74: 401 406. 56. Cantrell A, McGarvey DJ. Photochemical studies of 4-tert-butyl-4'- methoxydibenzoylmethane (BM-DBM). J Photochem Photobiol Biol 2001; 64: 117 122. 95