Skin Cancer Prevention: Recent Evidence from Randomized Controlled Trials
- 1.3k Downloads
Despite the billions of health care dollars spent each year on treating skin cancer, there is a dearth of randomized controlled trials (RCTs) that have evaluated skin cancer prevention. RCTs published in the last 3 years that have directly assessed skin cancer prevention as their primary aim suggest that regular use of sunscreen is cost effective, but prolonged use of topical therapies such as tretinoin and 5-fluorouracil may not be. Sirolimus-based immunosuppression for secondary skin cancer prevention in long-term renal transplant recipients appears effective, but benefits may be offset by the adverse effects. Many RCTs using pre-invasive actinic keratoses (AKs) as endpoints are too small and/or too short to provide evidence on skin cancer prevention. Another stumbling block is the difficulty in reproducibly diagnosing and counting AKs in response to preventive agents. Longer term and better surveillance methods are urgently required to improve the quality of evidence from future RCTs.
KeywordsSkin cancer Prevention Randomized controlled trials RCTs Clinical trials Outcome measures Interventions Sunscreen Topical treatment Basal cell carcinoma BCC Squamous cell carcinoma SCC Actinic keratosis AK 5-fluorouracil 5-FU Digital photography Dermoscopy
Basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) are by far the commonest cancers in white-skinned populations. It is estimated that in the United States alone in 2006, a total of 2,152,500 people were treated for around 3,507,700 such cancers . Despite their relatively low mortality rates, such high-incidence rates mean that the associated health care costs impose a substantial financial burden on health care systems . Cutaneous melanoma, the other major type of skin cancer, affected around 10,342 people in Australia in 2007  and some 68,700 people in the United States in 2009, with around 8,600 deaths . Collective management of these skin cancers drives an expenditure that exceeds the total amount spent on other major cancers combined [5, 6, 7]. In the United States, skin cancer treatments cost an estimated $2 billion each year [8••]. Beyond these are further billions—around $1.2 billion in the United States [9•]—expended treating the related and highly prevalent skin lesions, actinic keratoses (AKs). AKs affect 5 % to 25 % of people in the United Kingdom and United States and up to 60 % in Australia [10, 11], and are one of the strongest risk factors for skin cancer.
Exposure to solar UV radiation is the major environmental cause of skin cancer, and, recently—especially in temperate climates—exposure to artificial UV light such as sunbeds has contributed to carcinogenic exposure of the skin . Theoretically, it should be straightforward to control a large proportion of skin cancers and their massive costs by implementing measures to decrease susceptible people’s level of UV exposure. Adoption of sun-avoidance behaviors by fair-skinned people is an obvious and fundamental preventive measure. The U.S. Preventive Services Task Force has systematically reviewed all counseling measures aimed at preventing skin cancer [13•]. It found few rigorous counseling trials and concluded that, of all counseling strategies assessed, only those relevant to primary care settings increased skin-protective behaviors. Other diverse approaches to achieving skin-protective behavior changes have ranged from shade provision interventions in schools  to appearance-focused interventions based on decision-theoretical models of health behavior in young female indoor tanners . A more direct interventional approach is to focus on skin cancer as the primary outcome, and trials that have taken this direct approach are the focus of this paper.
Scope of This Review
Randomized controlled trials (RCTs) published in the last 3 years that followed up 100 or more patients and were conducted with the explicit primary aim of evaluating the effectiveness of preventive measures in decreasing the incidence of BCC, SCC, or melanoma have been reviewed. Intervention studies that aimed to prevent AKs in the long term (defined as at least 1 year) or, based on the premise that a proportion of AKs are premalignant SCC precursors, aimed to diminish AK burden as a long-term outcome and thus a means of achieving SCC prevention (although this was mostly implicit) have also been mentioned. Systematic reviews of relevant trials, including RCTs with outcomes assessed at less than 1 year given the high level of review evidence, have also been summarized.
We begin by briefly considering findings from prevention trials that were published prior to 2009 and then review the most recent substantive published evidence. We conclude by discussing the controversies regarding the natural history of AKs and the difficulties of assessing the prevalence of AKs in large-scale clinical settings.
Pre-2009 Skin Cancer Prevention Trials and Systematic Reviews
RCTs published before April 2007 and conducted among people at high risk of developing keratinocyte cancer have been reviewed by the Cochrane Collaboration . The authors noted that, because only 10 such studies had been conducted, data were very limited, and they summarized the results as follows. One trial that had assessed topical T4N5 liposome lotion containing DNA repair enzymes in patients with xeroderma pigmentosum, the rare inherited disorder , showed that annual rates of new AKs and BCCs were reduced, whereas annual rates of SCCs and melanomas were non-significantly raised in the intervention compared with placebo group. No significant adverse effects were seen. Six trials evaluated systemic retinoid therapies to prevent skin cancers in renal transplant recipients: half assessed acitretin and half other oral retinoids. Of the three that assessed acitretin, only one provided standard measures of outcome. It showed no difference in time to first skin cancer when 30 mg per day acitretin was compared to placebo over 6 months, but there was a reduction in number of new skin cancers in the first 12 months after intervention. All three acitretin trials reported frequent mucocutaneous side effects, and when adverse event data (evaluable in two of the trials) were pooled, an increased risk of having an adverse event (unspecified) with acitretin therapy could not be excluded (RR, 1.80; 95 % CI, 0.70–4.61) . Among people with past keratinocyte cancer, results regarding new BCCs after oral retinoid or isoretinoin treatment were inconclusive, but risk of new SCC was raised in one trial and of adverse events in another . In two trials of antioxidant supplements among people with a history of skin cancer, one found that selenium supplements significantly elevated the risk of a new SCC, whereas the other gave inconclusive evidence regarding beta-carotene supplements. A trial of a reduced fat diet showed a trend toward fewer new keratinocyte cancers , although BCC and SCC endpoints could not be distinguished. The authors of this systematic review concluded that, although preventive treatments may benefit people at high risk of developing skin cancer, firm conclusions could not be drawn because of the very small number of trials addressing each type of intervention.
A systematic review and meta-analysis of five short-term RCTs (1293 patients) evaluated the efficacy and safety of an immune-response modifier, imiquimod 5 % cream, in the treatment of AKs and found that complete clearance occurred in 50 % of patients treated with imiquimod compared with 5 % treated with vehicle only . The proportion of patients with any adverse event, most often erythema (27 %) and scabbing or crusting (21 %), was substantially higher with imiquimod than with the vehicle. The authors concluded that imiquimod 5 % cream was effective in the treatment of AK and thus may potentially prevent the development of SCC .
Finally, a single community-based skin cancer prevention trial was conducted in Australia . The Nambour Skin Cancer Prevention Trial evaluated daily sunscreen use and beta-carotene supplementation in the prevention of BCC and SCC. The trial showed that, in comparison with people randomized to using sunscreen at their discretion, if at all, people randomized to daily use of a broad-spectrum SPF15+ sunscreen had no reduction in BCC but a 40 % reduction in SCC tumors at the conclusion of the trial —a reduction maintained 8 years later . Rate of acquisition of AKs was also reduced in the daily sunscreen group , as was the time to subsequent BCCs after the first BCC  in the trial period. There was no effect of beta-carotene supplementation on development of actinic tumors, malignant or benign [20, 22].
Recent Skin Cancer Prevention Trials and Systematic Reviews
Randomized controlled trials and systematic reviews of skin cancer prevention/treatment of AKs published since 2009
Intervention and control groups
Nambour Skin Cancer Trial (Green et al. [24••])
1621 community members randomly selected; 44 % male
Daily application of sunscreen to head and arms vs discretionary sunscreen use; beta-carotene 30 mg vs placebo from 1992 to 1996
Reduced new primary skin melanoma in daily sunscreen group from 1993 to 2006 (HR, 0.50; 95 % CI, 0.24–1.02); beta-carotene no effect
Melanomas histologically reviewed by two pathologists blinded to treatment allocation
CONVERT Trial (Alberu et al. [25••])
830 renal transplant patients; ages 25–75 y; 69 % male
Conversion to SRL-based vs continued CNI immunotherapy for 2 y
Lower keratinocyte skin cancer rates after SRL-based, CNI-free immunotherapy (1.2 vs 4.3, P < 0.001)
Adverse events and withdrawal rates higher in SRL-based immunotherapy at 2 y (26 % vs 20 %, P < 0.07)
VATTC Trial (Weinstock et al. [26••])
1131 U.S. veterans with 2+ BCC or SCC in past 5 y; 59 % ages > 70 y
Topical 0.1 % tretinoin vs matching placebo for 1.5–5.5 y
No difference in times to first BCC (P = 0.3) or first SCC (P = 0.4) between groups; no difference in AK counts
Worse symptoms in intervention group at 1 y
Systematic review of 5-FU and AKs (Askew et al. [27•])
13 studies involving 865 patients with multiple prevalent AKs
Topical 5 % 5-FU (5 % or 0.5 %) vs variety of treatments* for 4 w–12 m
Average reduction in mean number AK: 80 % (range, 59 %–100 %)(5 % 5-FU); 86 % (0.5 % 5-FU) vs 95 % (laser resurfacing); 28.0 % (placebo)
66 % required retreatment after 1 y; up to 50 % unable to complete treatment because of adverse effects
Systematic review of photodynamic therapy and AKs (Fayter [28•])
28 studies involving 2611 patients with multiple prevalent AKs
PDT vs cryotherapy; placebo PDT (cream); cryotherapy; 5-FU; imiquimod; different PDT parameters for various durations
MAL-PDT superior to placebo (pooled odds of clearance at 3 m = OR, 8.05; 95 % CI, 5.50–11.79)
Insufficient evidence regarding PDT compared with other treatments
Results of the long-term follow-up of participants of the aforementioned Nambour Skin Cancer Prevention Trial were recently published in relation to sunscreen allocation and prevention of new primary melanoma [24••]. It was shown that regular use of sunscreen by people in the high sun exposure setting of Queensland reduced the development of melanoma compared with discretionary sunscreen use. Investigation of the lifetime health costs and benefits of sunscreen promotion in the primary prevention of skin cancers, including melanoma, showed that routine sunscreen use by white populations residing in sunny settings is likely to be a highly cost-effective investment for governments and consumers over the long term [8••].
Trial in Renal Transplant Patients
The CONVERT open-label, randomized, multicenter trial found keratinocyte skin cancer rates were reduced in long-term renal transplant recipients at 2 years after their conversion to sirolimus-based immunosuppression compared with continued calcineurin inhibitor immunotherapy [25••] (Table 1).
Trial in Patients with Multiple Past Basal Cell Carcinomas and Squamous Cell Carcinomas
The Veterans Affairs Topical Tretinoin Chemoprevention Trial found no difference in times to first BCC or SCC and no difference in AK counts between patients treated with topical 0.1 % tretinoin compared with matching placebo over 1.5 to 5.5 years [26••].
Systematic Reviews of Treatment of Multiple Prevalent Actinic Keratoses
A systematic review of 5-FU in the treatment of prevalent AKs found greater reduction in mean number AKs with both 5 % 5-FU and 0.5 % 5-FU compared with placebo, but not compared with laser resurfacing [27•]. However, two thirds of patients required retreatment after 1 year, and up to half of those treated were unable to complete treatment because of adverse effects. The systematic review of photodynamic therapy in the treatment of prevalent AKs found photodynamic therapy to be superior to placebo, but there was insufficient evidence regarding its comparison with other treatments [28•] (Table 1).
Ongoing Skin Cancer Prevention Trials
The U.S. National Institutes of Health Clinical Trials registry  currently lists 66 RCTs of AK treatment. Of those that have recently updated information available, only two appear to have started in the last 5 years, with 100 or more patients followed-up for at least 12 months. In a phase II trial, the efficacy of afamelanotide (formerly CUV1647), a chemical analogue of alpha-melanocyte stimulating hormone, in reducing AKs and SCCs in around 200 organ transplant patients is being investigated (NCT00829192). Another RCT is studying the long-term effects of treatment of skin areas with 5 to 10 AKs with imiquimod 5 % cream versus diclofenac 3 % gel with respect to the risk of progression to in situ and invasive SCC at 3 years in around 250 immunocompetent patients (NCT01453179).
There are at least two other ongoing randomized controlled trials of skin cancer prevention, both in immunosuppressed patients (N > 100), of which the authors are aware and/or directly involved. A multicenter-sponsored (Spirig Pharma Ltd., Switzerland) RCT examining the role of sunscreen in skin cancer prevention is currently in progress in Turkey and 11 countries across Europe. In this study, 300 organ transplant recipients at risk of skin cancer are being randomized to discretionary use of sunscreen in a galenically improved, dosable version of the liposomal sunscreen reported in a smaller study  of organ transplant recipients in Germany. This current 2-year study will examine reduction of AK and development of new skin cancers (as well as viral warts) as primary outcome measures (C. Ulrich, personal communication). The aims of another scheduled trial, the O3A Trial, are to evaluate omega-3 polyunsaturated fatty acid supplementation (4 g per day) versus placebo (double blind), in the prevention of BCC and SCC in 340 organ transplant patients in a 2-year intervention period. It is being conducted by a multicenter team, including some of the authors (A. C. Green, C. A. Harwood, J. Lear, and H. P. Soyer) in Australia and United Kingdom.
Controversies Regarding Actinic Keratoses
Since the majority of skin cancer prevention trials have been aimed at secondary prevention of SCC, many RCTs have focused on the treatment of AKs as trial endpoint, since AKs are presumed to be surrogate biomarkers of SCC, as noted previously. This is a difficult area in many ways given the controversies surrounding the malignant potential of AKs and the assumption—but lack of conclusive evidence—that clearance of AK prevents SCC. In addition, even if the AK intervention may be assumed to prevent SCC, it is not known whether complete clearance is required and/or whether persistent clearance or maintenance of clearance over an unknown period is required to achieve this. There are also controversies at the clinical level; that is, it is not known what type of AKs are most closely associated with SCC. For example, are small, barely palpable ones less of a risk than large, inflamed hyperkeratotic ones? The challenges around clinical diagnosis and quantification have not been addressed often in published studies but have important implications for interpreting study validity in terms of intervention agents’ long-term efficacy in treating and/or preventing AK, as well as their putative role in SCC prevention.
Natural History and Rate of Malignant Transformation
The natural history of AK is poorly understood but is believed to involve turnover of prevalent individual lesions (regression and recurrence) as well as development of new lesions . Although some AKs clearly evolve into SCCs, the majority do not and the risk of progression to invasive malignancy is unknown  but much debated. A systematic review of 15 studies showed estimated progression rates of between 0.025 % and 20 % per year, per lesion . A more recent study from the United States that prospectively followed 169 patients with 7784 AKs found that the risk of progression for a specific AK was 2.57 % at 4 years, and that 65 % of all primary SCCs arose directly in lesions previously diagnosed as AKs . Cited rates of transformation into SCC are almost certainly overestimates, since even the most conservative have been based on successive clinical counts of AK of subjects’ skin at widely varying intervals, using counting techniques of uncertain reliability and validity . Indeed, there remains lack of agreement among histopathologists regarding the diagnosis of SCC and AK [36••] because of the continuous clinicopathologic spectrum from benign atypical keratotic lesions to invasive malignant lesions .
Monitoring Actinic Keratoses
Evaluating the efficacy of AK therapies requires quantification of AK burden before and after treatment. For the purposes of clinical studies, distinguishing features for AK have not been well established. In immunosuppressed organ transplant recipients, the clinical picture is further complicated by the increased prevalence of verrucokeratotic lesions, most of them viral warts, which are often present in areas of field cancerization and are clinically (or sometimes histologically) indistinguishable from AK . Many studies have validated clinical diagnosis of AK by biopsying lesions, but how target lesions were selected has not been stated often. Diagnosis of discrete, clinically typical AK is relatively easy, but in patients with multiple and atypical lesions—particularly where lesions merge to give a field of skin abnormality (Fig. 1)—diagnosis and counting of AKs is far from straightforward and may be almost impossible [36••].
Previous studies of treatment and natural history of AK have usually quantified lesions by counting alone, yet counting of AK as a technique for evaluating therapeutic efficacy may be unreliable, even when performed by experienced dermatologists, since interobserver reproducibility is poor [26••] and may inadequately account for spontaneous regression of AK . Joint discussion of discrepancies by observers may enhance the reliability of these counts, although substantial variation remains . The importance of developing and validating techniques to enhance reliability of AK assessment has been highlighted, but the problem remains unresolved [30, 38].
Role of Digital Photography and Dermoscopy for Imaging of Actinic Keratoses
Standard photographic evaluation, blinded, has been suggested as perhaps the most reliable technique, but this is also untested . More recently, digital photography of regions of the body such as the face, scalp, and dorsa of hands and forearms has been advocated for early detection and surveillance of keratinocyte skin cancer. Advantages of imaging AKs with digital photographs over counting AKs, especially on severely photodamaged skin (Fig. 1), are the relative ease and speed that an established imaging protocol can be carried out. Digital photography also allows clinicians to detect smaller or nonspecific lesions, which may otherwise be missed or passed over, especially in a busy outpatient clinic. Digital images can be stored for successive comparisons during repeated follow-up, as in clinical trials, using similar protocols to those now established for naevus surveillance .
On the other hand, the role of dermoscopy in assessing large number of AKs over time is not established. To date, most of the dermoscopic literature on AK has focused on differentiating pigmented AKs from lentigo maligna , although recently a progression model of facial AK has been proposed based on a large series of dermoscopic images of AKs developing into invasive SCCs . However, the added value of dermoscopic examination of individual AKs among some thousands of AKs being followed during large skin cancer prevention trials seems quite low, and even lower for confocal microscopy. Ultimately, it is hoped that high-definition regional photography in combination with available image-analysis tools will enhance clinical examination to achieve accurate and reproducible counts and assessments of AKs.
The number of recently published skin cancer prevention trials is small in inverse proportion to the need for them to better control the vast health costs associated with skin cancer treatment. Sunscreen emerges as a cost-effective preventive agent for people living in sunny places. There is inconsistent evidence regarding the long-term efficacy of topical treatments such as retinoids and 5-FU, and AK clearance in the short term is often associated with side effects. Several skin cancer prevention trials evaluating a range of possible preventive agents in immunosuppressed patients are currently underway, but the lack of validity and reproducibility of standard clinical and histopathological assessment of AK as a trial endpoint is an ongoing limitation. There is an urgent need to address these methodological challenges of diagnosis, quantification, and surveillance of AK burden as well as a need for future adequately powered, well-designed, long-term RCTs for strengthened evidence regarding skin cancer prevention.
A. C. Green: grants from National Health and Medical Research Council and L’Oreal; C. A. Harwood:; J. Lear: payment for giving lectures from Galderma, Leo, Almirall and payment for development of educational presentations from Galderma; C. Proby: none; S. Sinnya: none; H. P. Soyer: consulting fees and stock options from MoleMap Australia.
Papers of particular interest, published recently, have been highlighted as: • Of importance •• Of major importance
- 2.Soyer HP, Rigel DS, Wurm EMT Dermatology. In: Bolognia JL, Jorizzo JL, Schaffer JV (editors) Vol 1. Actinic keratosis, basal cell carcinoma and squamous cell carcinoma. 3rd edn. Mosby/Elsevier, 2012.Google Scholar
- 3.AIHW. Cancer survival and prevalence in Australia: cancers diagnosed from 1982 to 2004. In. Canberra; 2008.Google Scholar
- 4.Jemal A, Saraiya M, Patel P, Cherala SS, Barnholtz-Sloan J, Kim J, et al. Recent trends in cutaneous melanoma incidence and death rates in the United States, 1992–2006. J Am Acad Dermatol. 2011;65(5):S17–S25.e13.Google Scholar
- 6.Staples MP, Elwood M, Burton RC, Williams JL, Marks R, Giles GG. Non-melanoma skin cancer in Australia: the 2002 national survey and trends since 1985. Med J Aust. 2006;184(1):6–10.Google Scholar
- 8.•• Hirst NG, Gordon LG, Scuffham PA, Green AC. Lifetime cost-effectiveness of skin cancer prevention through promotion of daily sunscreen use. Value Health. 2012;15(2):261–8. This paper provides a newly published analysis of the cost-effectiveness of sunscreen application for skin cancer preventon based on RCT evidence.Google Scholar
- 9.• Neidecker MV, Davis-Ajami ML, Balkrishnan R, Feldman SR: Pharmacoeconomic considerations in treating actinic keratosis. Pharmacoeconomics 2009;27(6):451–64. This paper illustrates the economic importance of preventing rather than treating AKs.Google Scholar
- 10.Frost C, Williams G, Green A. High incidence and regression rates of solar keratoses in a Queensland Community. J Invest Dermatol. 2000;115(2):273–7.Google Scholar
- 12.El Ghissassi F, Baan R, et al. A review of human carcinogens--part D: radiation. Lancet Oncol. 2009;10:751–2.Google Scholar
- 13.• Lin JS, Eder M, Weinmann S: Behavioral counseling to prevent skin cancer: a systematic review for the U.S. Preventive Services Task Force. Annals of Internal Medicine. 2011;154(3):190–201. This systematic review of studies of behavioral counseling to prevent skin cancer complements the present review, which does not cover behavioral counseling RCTs, since none have skin cancer as primary endpoint.Google Scholar
- 14.Dobbinson SJ, White V, Wakefield MA, Jamsen KM, White V, Livingston PM, English DR, Simpson JA. Adolescents’ use of purpose built shade in secondary schools: cluster randomised controlled trial. BMJ 2009;338.Google Scholar
- 16.Bath-Hextall F, Leonardi-Bee J, Somchand N, Webster A, Delitt J, Perkins W. Interventions for preventing non-melanoma skin cancers in high-risk groups. Cochrane Database of Systematic Reviews (Online) (4). 2007.Google Scholar
- 20.Green A, Williams G, Nèale R, Hart V, Leslie D, Parsons P, Marks GC, Gaffney P, Battistutta D, Frost C, et al. Daily sunscreen application and betacarotene supplementation in prevention of basal-cell and squamous-cell carcinomas of the skin: a randomised controlled trial. Lancet. 1999;354(9180):723–9.PubMedCrossRefGoogle Scholar
- 24.•• Green AC, Williams GM, Logan V, Strutton GM. Reduced melanoma after regular sunscreen use: randomized trial follow-up. Journal Clinical Oncology. 2011;29(3):257–63. This report is the first RCT evidence that sunscreen application can prevent melanoma.Google Scholar
- 25.•• Alberu J, Pascoe MD, Campistol JM, Schena FP, del Carmen Rial M, Polinsky M, et al. Lower malignancy rates in renal allograft recipients converted to sirolimus-based, calcineurin inhibitor-free immunotherapy: 24-month results from the CONVERT Trial. Transplantation 2011;92:303–10. This report provides RCT evidence that sirolimus-based immunosuppression can prevent skin cancer in renal transplant recipients who are at high risk of skin cancer.Google Scholar
- 26.•• Weinstock MA, Bingham SF, DiGiovanna JJ, Rizzo AE, Marcolivio K, Hall R, Eilers D, Naylor M, Kirsner R, Kalivas J et al. Tretinoin and the prevention of keratinocyte carcinoma (basal and squamous cell carcinoma of the skin): a veterans affairs randomized chemoprevention trial. J Invest Dermatol. 2012. This paper provides the results of a well-conducted trial of tretinoin use for skin cancer prevention.Google Scholar
- 27.• Askew DA, Mickan SM, Soyer HP, Wilkinson D. Effectiveness of 5-fluorouracil treatment for actinic keratosis – a systematic review of randomized controlled trials. International Journal Dermatology. 2009;48(5):453–63. This systematic review of 5-FU and AK treatment includes both small, short-term studies and large, longer term RCTs.Google Scholar
- 28.• Fayter D, Corbett M, Heirs M, Fox D, Eastwood A. A systematic review of photodynamic therapy in the treatment of pre-cancerous skin conditions, Barrett’s oesophagus and cancers of the biliary tract, brain, head and neck, lung,oesophagus and skin. 2010. Health technology assessment (Winchester, England) 14(37):1–288. This quality systematic review of photodynamic therapy includes treatment of AKs.Google Scholar
- 29.ClinicalTrials.gov, US National Instit Health [cited 2012 April]; Available from: [www.clinicaltrials.gov].
- 35.Criscione VD WM, et al. Actinic keratoses: natural history and risk of malignant transformation in the Veterans Affairs Topical Tretinoin Chemoprevention Trial. Cancer 2009;115:2523–30.Google Scholar
- 36.•• Heal CF, Weedon D, Raasch BA, Hill BT, Buettner PG: Agreement between histological diagnosis of skin lesions by histopathologists and a dermato-histopathologist. International Journal of Dermatology 2009;48(12):1366–9. This paper, which discusses many of the difficult issues relating to AK assessment, is important for its discussion of histopathological as well as clinical aspects.Google Scholar
- 39.Wurm E. SH: Scanning for melanoma. Australian Prescriber. 2010;33:150–5.Google Scholar