Researchers with Industry Ties Respond to Evidence of Accelerated Skin Damage With Faulty, Misleading Commentary

In June 2010 the Environmental Working Group (EWG) published an analysis of new government data indicating that a form of vitamin A, retinyl palmitate, when applied to skin exposed to sunlight, may speed the development of skin tumors and lesions (NTP 2009). EWG considered this evidence troubling because the sunscreen industry adds vitamin A to 41 percent of all sunscreens.

EWG's analysis of data from ongoing research conducted jointly by the Food and Drug Administration (FDA) and the National Toxicology Program (NTP) apparently hit a nerve in the sunscreen industry. On Aug. 6, three researchers, including two who have been industry consultants, published a highly misleading "commentary" attacking EWG's analysis in the Journal of the American Academy of Dermatology (Wang et al. 2010).

Among other flaws, the commentary omits a crucial endpoint in the government data that shows skin tumors and lesions developing more quickly in Vitamin A-exposed animals, dismisses the relevance of animal research that is widely considered the "gold standard" in assessing carcinogenicity in humans, ignores previously published studies and conflates the effects of vitamin A taken orally with its effects when applied to the skin.

The commentary's authors are at the dermatology divisions of New York's Memorial Sloan-Kettering Cancer Center and Detroit's Henry Ford Hospital.

Incomplete analysis, exclusion of critical endpoints

In the FDA/NTP study, tumors and lesions developed up to 21 percent sooner in lab animals coated in a vitamin A-laced cream (at concentrations of 0.1 percent to 0.5 percent) than in control animals treated with a vitamin-free cream, according to EWG's analysis of study data released by the agencies (NTP 2009, EWG 2010). Animals were exposed to the equivalent of up to nine minutes of noontime Florida sunlight each day for a year. The effects were statistically significant for all dose groups tested.

Source: EWG analysis of data from FDA photocarcinogenicity study of retinyl palmitate (NTP 2009).

The commentary's authors, however, present a narrow analysis that examines only the absolute number of tumors (neoplasms) in exposed animals. They ignore data on the rate at which animals developed tumors and other skin damage, even though this study endpoint most clearly shows vitamin A's influence on skin. Their analysis simply sidesteps data showing that vitamin A-exposed animals developed skin lesions and tumors up to two months sooner than control animals.

Misleading conclusions

The commentary's authors acknowledge a statistically significant increase in malignant lesions among animals exposed to vitamin A combined with low doses of UV radiation. The authors discount this increase, however, with the observation that a similar, statistically significant increase was not observed in vitamin A-exposed animals subjected to higher doses of UV. This conclusion disregards the fact that animals in the higher-UV group were massively sickened by the sheer amount of UV radiation (with 82 percent developing malignant tumors, compared to 38 percent in the lower-dose group), potentially overwhelming the impact of skin damage caused by vitamin A.

In contrast, there remained an unambiguous dose-response relationship in the speed at which skin damage developed for both lower- and higher-dose UV radiation, as reported by EWG (EWG 2010).

Casting doubt on established, "gold standard" methods

The commentary's authors write, "Extreme caution is needed when extrapolating these animal study results to human beings." In media reports, one author has discounted the relevance of FDA's standard animal-based photocarcinogenicity studies. He is quoted in one case as saying, "It's dangerous to apply a finding in mice to humans" (Kotz 2010).

This position conflicts with established science and policy. As the FDA itself says, "testing for photocarcinogenicity in humans is unethical; animal testing has been used as a surrogate" (FDA 2003). Scientists consider research like the FDA/NTP study to be the "gold standard" for assessing human carcinogenicity risks (Ball 2009; Bucher 2002). Scientists at the M.D. Anderson Cancer Center have written that "SKH1 [hairless] mice are the most widely used in dermatologic research… tumors induced in these mice resemble, both at the morphologic and molecular levels, UVR-induced skin malignancies in man" (Benavides 2009). These are the type of mice used in the FDA/NTP study. FDA's experts would surely not choose for their research animals as unsuited for the purpose as the commentary's authors suggest.

Faulty logic

The commentary's authors justify the use of vitamin A in sunscreen by saying that in the "more than 40 years that topical retinoids have been in use, millions of patients used them…"

In fact, many retinoid-based products are creams applied at night. Patients treated with such creams are advised to minimize sun exposure because the cream "may make your skin more susceptible to sunburn and other adverse effects of the sun…" of vitamin-A based skin creams warn that "unprotected exposure to natural or artificial sunlight should be minimized" and that "patients should minimize exposure to sunlight or artificial ultraviolet irradiation sources" ( 2010)

Errors of fact

The study authors assert that there is "no published evidence to suggest that topical retinoids increase the risk of photocarcinogenisis." This is not true. At EWG, we are surprised that our colleagues overlooked the published science on this topic.

Ten years ago, FDA nominated retinyl palmitate for testing because of concerns about the compound's possible photocarcinogenic effects: "Retinyl palmitate was selected by the [FDA's] Center for Food Safety and Applied Nutrition for phototoxicity and photocarcinogenicity testing based on the increasingly widespread use of this compound in cosmetic retail products for use on sun-exposed skin, the biochemical and histological cutaneous alterations elicited by retinyl palmitate, and the association between topical application of retinoids and enhancement of photocarcinogenesis."

Since that time, FDA researchers have published 17 studies and reviews on the toxicity and chemistry of retinyl palmitate on the skin. According to FDA scientists, the study findings suggest that: retinyl palmitate breaks down in sunlight to photomutagenic compounds that trigger mutation in the presence of light; forms cell-damaging free radicals in the presence of UVA and UVB radiation; and "[causes] events that affect a large segment of the chromosome" (e.g., Mei et al. 2005, 2006; Xia et al. 2006, EWG 2010b).

Why would the FDA undertake a significant, multi-year research effort on this issue if it did not have serious concerns about retinyl palmitate's potential to be photocarcinogenic?

University and pharmaceutical industry researchers have also noted photocarcinogenic effects in animals exposed to vitamin A-based creams:

" -In a 91-week dermal study in which CD-1 mice were administered 0.017% and 0.035% formulations of Tretinoin, cutaneous squamous cell carcinomas and papillomas in the treatment area were observed in some female mice. A dose-related incidence of liver tumors in male mice was observed at those same doses."

"Studies in hairless albino mice suggest that concurrent exposure to Tretinoin may enhance the tumorigenic potential of carcinogenic doses of UVB and UVA light from a solar simulator. This effect has been confirmed in a later study in pigmented mice, and dark pigmentation did not overcome the enhancement of photocarcinogenesis by 0.05% Tretinoin. Although the significance of these studies to humans is not clear, patients should minimize exposure to sunlight or artificial ultraviolet irradiation sources."

Justifying vitamin A in skin cream by citing its use in oral drugs.

The commentary's authors support the use of vitamin A in sunscreen by noting that "…oral retinoids… are used for chemoprevention of skin cancers…" This argument is specious. FDA's data raises concerns about effects from vitamin A creams in the presence of sunlight. Sunlight does not reach the stomach or bloodstream, where a vitamin-based oral drug would circulate.

Other misleading arguments

The authors write that, "On a casual glance, the generation of oxygen radicals and their potential for photomutagenicity may be a concern," but then discount this issue by noting that human skin contains vitamins C and E and other compounds that may counteract the potential for vitamin A-induced damage. The fact is that all cells and all animals, including the mice in the FDA/NTP research, contain vitamins C and E and other antioxidants. The FDA/NTP photocarcinognicity data and other available studies account for beneficial antioxidants counteracting the damaging effects of UV radiation in combination with vitamin A. These protective effects were not enough to dampen the statistically significant accelerated skin damage.

Conflicts of interest

The commentary's authors – Steven Q. Wang, Stephen W. Dursza and Henry W. Lim --declare no conflicts of interest. In other publications, however, author Dr. Wang has disclosed that he has been a consultant to La Roche-Posay (Wang et al. 2008), a company that markets sunscreens and that uses vitamin A in eye creams. Dr. Lim has disclosed a consulting relationship with Johnson & Johnson (maker of vitamin A-containing Neutrogena and Aveeno sunscreens and La Roche-Posay (Lim and Rigel, 2007).

Drs. Wang and Lim also volunteer on the Photobiology Committee of the Skin Cancer Foundation, a trade organization supported by the sunscreen industry. This committee sets criteria for the foundation's Seal of Recommendation and reviews the applications of companies requesting to use the Seal on their products. Each company seeking the Foundation's endorsement must pay an annual $10,000 fee.


EWG's analysis of the FDA/NTP preliminary data raises concerns about the widespread use of vitamin A in sunscreens. NTP plans to publish its assessment of the data in a report due in the fall of 2010. In the interim, EWG recommends that consumers select from among the nearly 60 percent of sunscreens that are free of the compound until more conclusive information is available.