Background

Ultraviolet radiation (UVR) as ultraviolet A (UVA; wavelength around 315-400nm) or ultraviolet B (UVB; wavelength around 280-315nm) can penetrate the epidermis of human skin. UVR interacts with chromophores (molecular components which absorb visible light or UV radiation) and can then damage DNA, RNA, proteins and lipids. This results in erythema ('sunburn') in the short term, and photoageing (roughness and fine wrinkling) and carcinogenesis with cumulative exposure. The specific roles of UVA and UVB in skin cancer are not clear.

The three major types of skin cancer are malignant melanoma, basal cell carcinoma (BCC) and squamous cell carcinoma (SCC). Previously, BCC and SCC have been collectively termed non-melanoma skin cancer (NMSC). However, there is increasing genetic and epidemiological evidence that they should be considered separately.2^,3 The incidence of all three types of skin cancer in 'white-skinned' people is greater the nearer they live to the equator (where UVR is more intense). Childhood sun exposure involving severe sunburn increases the risk of melanoma as an adult.4^,5 Studies indicate that BCC is more likely with intermittent intense sun exposure, particularly in those who tend to burn rather than tan.2^,3 This suggests that the main risk factor for both malignant melanoma and BCC is sun exposure that is intermittent and severe, and causes burning. However, differences in the commonest sites of development of melanoma (the legs of women, the backs of men) and BCC (the face in both sexes) suggest that other factors also contribute.

Squamous cell carcinoma follows a different pattern; incidence is highest in those with high cumulative lifetime UVR exposure,6 typically people in outdoor occupations such as sailors or farmers.7 The commonest sites of development of SCC and actinic keratosis (a benign precursor skin lesion with a low rate of transformation to SCC8) are those which receive the greatest lifetime sun exposure (e.g. the face, the scalp and the backs of hands).

Genetic factors also influence susceptibility to skin cancer. Those at particular risk include pale-skinned, light or red haired people with a tendency to freckle or burn. Having many benign melanocytic naevi (pigmented moles) is also a risk factor,8^,9 but this may be partly determined by childhood UVR exposure as well as being genetic in origin.

Pharmacology of sunscreens

Older sunscreens, which were designed to prevent sunburn, only protected against UVB. By reducing the natural deterrent effect of sunburn, these sunscreens may have allowed exposure to much higher doses of UVA. Current sunscreens block both UVA and UVB (broad-spectrum block). The ability to protect against UVB-induced erythema is the basis of the standardised measure of sunscreen potency - the sun protection factor (SPF). A correctly applied sunscreen with an SPF of 15 reduces the UVR exposure to the skin to 1/15th of its original value. However, because application is rarely adequate, the SPF achieved is typically between 20-50% of that expected.10 The ability to protect against UVA is indicated by a star rating system graded 1-4. A product with 4 stars protects equally against UVA and UVB, 1 star indicates the poorest level of protection against UVA compared with UVB. Chemical sunscreens (e.g. those containing cinnamates, benzophenones or dibenzoylmethanes) absorb UVR. Physical sunscreens (e.g. those containing titanium dioxide) protect mainly by reflecting UVR at the skin surface.

Malignant melanoma

No prospective controlled trial has tested whether sunscreens prevent melanoma. However, there is other evidence that sun protection in children helps prevent the development of melanoma in later life. Retrospective case-control studies indicate that patients with melanoma are less likely to have been protected from sunlight during childhood11 or to use a sunscreen12 compared with controls. By contrast, several large questionnaire-based surveys suggest that the use of sunscreens affords no protection against melanoma5^,13 and might even increase the risk of developing melanoma.14^,15 However, some of these studies may not have been controlled adequately5^,14 and could have been influenced by recall bias or unreliable use of the sunscreen. It has also been suggested that effective broad-spectrum sunscreens were not commonly available in the study period.5

Non-melanoma skin cancer

There is some evidence that sunscreens may protect against NMSC. Studies have measured the appearance and the disappearance of actinic keratoses. In a double-blind, placebo-controlled trial of an SPF 17 broad-spectrum sunscreen, 588 white Australians (mean age 63 years) with 1-30 actinic keratoses applied 1.5mL of cream to the head, neck and each forearm and hand once each morning.8 They were asked to avoid the midday sun, to wear hats and clothing to cover vulnerable skin and to record all applications throughout the day. One hundred and thirty-seven people withdrew for study-related reasons (e.g. 60 participants had skin reactions). The sunscreen group had fewer new actinic keratoses (rate ratio, 0.62; 95% CI, 0.54-0.71) and more remission (odds ratio 1.53; 95% CI, 1.29-1.80) than the placebo group over a 7-month period. In a smaller randomised, placebo-controlled study, 53 volunteers (mean age 64 years) with actinic keratoses or non-melanoma skin cancer and not already using a sunscreen were encouraged to apply cream liberally and to wear hats and other protective clothing where necessary, over a 2-year period.16 The active cream (an SPF 29 sunscreen filtering mainly UVB) reduced the rate of appearance of new actinic keratoses. The design of the studies did not permit demonstration of a reduction in NMSC.

Unwanted effects

Sunscreens can cause stinging, contact urticaria or dermatitis. Probably more than 90% of reactions represent an irritant effect of the preparations,17 although all the chemicals in sunscreens (particularly fragrances and preservatives) can produce true allergic reactions. Some sunscreens may also cause photo-allergic reactions. The physical sunscreens produce fewer cutaneous reactions than the chemical sunscreens but can leave a white film on the skin which may be cosmetically unacceptable.

Some theoretical arguments have prompted speculation that sunscreen use accounts for the increase in incidence of skin cancer. These include suggestions that ingredients of chemical sunscreens are weakly genotoxic in bacteria, and that inhibition of UVR-induced cutaneous synthesis of vitamin D by sunscreens might promote the development of melanoma.18 However, the increased frequency of skin cancer is more likely to be due to other factors, including the popularity of suntans, inadequate application of sunscreens and higher rates of reporting following greater public awareness.

General protective measures

Several general measures can reduce total UVR exposure. The most important is to avoid direct sun exposure by seeking natural shade (especially between 11am and 3pm in the UK) and wearing appropriate clothing (close weave materials in darker colours) and hats. The Health Education Authority recommends that school playgrounds have shaded areas and that where possible, outdoor activities are scheduled to avoid excessive UVR exposure. Regular and adequate application of a broad-spectrum sunscreen is complementary but secondary to these measures. Sunscreens should be used to decrease sun exposure while outdoors and not to encourage prolonged sun exposure. This advice is particularly important for children, people who tan poorly and those at particularly high risk of developing skin cancer (e.g. those with many benign naevi, atypical naevi or a family history of melanoma). It is important to advise those who have had skin cancers and precancerous lesions to minimise their exposure to UVR.

Prescribing sunscreens

Sunscreens with a SPF of 15 or over can be prescribed on the NHS (on a FP10) for photodermatoses, an indication that could conceivably include the primary prevention of chronic photodamage. The approximate cost per day of using sunscreens ranges from £2.00 (Spectraban Lotion UVB, SPF 25) to £9.90 (Ambre Solaire Total Sunscreen, SPF 60).*

* Assuming 30mL of sunscreen is required for total body coverage, 3 times daily.

Conclusion

The increased incidence of skin cancer in the UK re-emphasises the need for effective prevention, particularly in children and people known to be at higher risk of skin cancer. It is possible that modern, broad-spectrum sunscreens protect against skin cancer provided that they are used appropriately, but this is not proven. Sunscreens must not encourage users to stay in the sun for longer than they otherwise would. Protective measures such as avoiding midday sun and wearing protective clothing and hats are crucial and are much cheaper than using sunscreens. Adjunctive use of sunscreens (with a SPF of at least 15) as part of a programme to reduce exposure to UVR seems sensible.