Total body naevus count (TBNC) has been shown to be predictive of melanoma risk, but is time-consuming to perform, particularly in general practice. The authors’ recent study has shown that counting naevi on one arm can quickly estimate TBNC, making it a useful tool for assessment of melanoma risk.
Melanoma, mole count, naevi, naevus, risk assessment
The incidence of malignant melanoma is increasing and this is explained in part by the removal of very early melanomas.1
Since 2000, two-week wait clinics have been set up in the UK to speed up the referral of suspected cancer.2 However, many of these clinics see a large number of benign lesions and the ratio
of malignant to benign can be very low. Patients attending the clinics are often at very low risk of melanoma, for example, those with dark skin types, those from non-white populations or very young patients, including children.
When assessing skin cancer risk, GPs often ask patients about previous sunburn and lifetime sun exposure. This does not discriminate well between low- and high-risk patients, because sun exposure is so ubiquitous.
TBNC is considered one of the most important risk factors for melanoma, with much higher relative risks than environmental factors, such as sun exposure.3
Naevi are also under genetic control, as shown by family and twin studies, and a smaller percentage of the variance in naevus counts can be explained by environmental factors (namely sun exposure).4,5 Naevi genetics has helped in discovering new melanoma genes, but also unravels some associations between melanoma risk factors and other phenotypes, such as reduced ageing and longer telomeres.6
In white populations, naevi confer the same magnitude of risk for melanoma at all latitudes, suggesting that sun exposure is not an important factor for this association.7 Despite this, only 20-40% of melanoma arises from an existing naevus, so it is clear that a high number of naevi is a marker of risk and individual lesions have a very low chance of transforming into melanoma.8
In the UK population, the average total number of naevi is 33, but people in the at-risk population for melanoma often have more than 100 naevi.9 In clinical practice, naevus count is a very important clinical marker for melanoma risk and is easy to document.10 However, total body skin examinations and mole checks are rarely performed during GP consultations.11
Melanomas do occur in patients with very few naevi, but these patients usually have very fair skin, with red or blond hair and freckles. Lentigo maligna on the face and scalp is seen in older patients, is not associated with numerous naevi and is seen in patients with chronic sun damage.
In white populations, naevi typically involute after the fourth decade of life and are rare in the elderly. However, individuals with a susceptibility to melanoma often have large numbers of common and atypical naevi that persist until middle age or later.12 So individuals at risk of melanoma have a delayed senescence and keep their naevi longer with age. This means assessing naevus counts in patients aged 45 years and over is a very sensitive way to detect patients most at risk of melanoma. It is therefore important for GPs to be aware of the natural biology of naevi over a lifetime, because this will also help them to decide if lesions are suspicious.
Body sites are very important when assessing naevi. For example, intradermal naevi are common on the face, but rarer on the limbs. Atypical naevi (figure 1), on the contrary, are common on the trunk and limbs, but rare on the face, so body distribution is also important, because this will raise the index of suspicion if odd lesions are seen in unusual sites.
Figure 1: (above) Atypical mole syndrome in a patient with a history of multiple melanomas; (below) Atypical naevi (Images: Dr Simone Ribero)
Many studies have used naevus count on selected body sites to identify patients at risk of melanoma, because TBNC is time-consuming in general practice.13,14 Most of these studies have a small sample size. In addition, some are based on patients attending dermatology clinics for melanoma screening, so may not reflect population-based data.
Whole arm naevus count has previously been reported to be the most predictive for TBNC in children and adults.15
Predictive body site
In a large cohort of healthy UK white female subjects, we described the body site most predictive of TBNC out of the 17 sites analysed.16 Findings were replicated in a separate group from a melanoma case-control study of UK white subjects.
We demonstrated that having an arm naevus count of more than 11 is associated with a higher risk of having more than 100 naevi, which is a strong risk factor for melanoma.
This tool can be used for a quick estimation of melanoma risk in general practice. Moreover, patients can count their own moles and ask their GP to confirm their risk.
GPs could then examine patients and on the basis of a quick arm count (or TBNC if time allows), plus any family history of melanoma and other cancers, could decide which patients are most at risk of melanoma and refer if appropriate.
Figure 2: (above) Dermatofibroma; (below) Dermoscopic view (Images: Dr Simone Ribero)
GPs should also be aware that melanoma is very rare in non-white populations and children, and other diagnoses should be considered, such as benign naevi in children and adults or dermatofibromas (figure 2) and seborrhoeic keratoses (figure 3) in adults, when examining a pigmented lesion in these groups.
Figure 3: Multiple seborrhoeic keratoses (Image: Dr Simone Ribero)
Family history of cancers is important, because melanoma and the presence of multiple atypical naevi are more likely to occur in patients with cancer in the family, including brain, pancreas, breast, colon, kidney and many other types of cancer.
More importantly, by taking a family history in patients with a high number of naevi, susceptibility to non-skin neoplasms, such as breast or bowel cancer, can be detected. Patients can then be referred for screening via cancer genetics clinics, which now exist in each cancer network in the UK. Cancers in these families are also more likely to occur at a younger age, so the age of onset of cancer is also a useful guide to genetic susceptibility.
Self-examination and screening
Patient self-examination is reported to be a good, practical approach to screening for early detection in high-risk populations for melanoma in the US.17 The arm count is also helpful in educating the public about melanoma risk factors.
Self-examination is rare in the UK and the general population is often unaware that naevus count is the most important risk factor for melanoma. Offering a screening visit to the whole population above the age of 45 years without selection, as is the practice in Germany,18 is not feasible in the UK and the pick-up rate for melanoma would be too low for this to be worthwhile.
Another aspect of screening is its impact on mortality. So far, melanoma screening campaigns in several countries have failed to decrease melanoma mortality, which has remained stable over many years, despite many different strategies.19
Screening campaigns increase the incidence of very early melanomas, as has been observed in Australia, for example.20 These borderline melanocytic lesions may not have a huge impact on mortality, because the surge in the diagnosis of very early melanomas is not mirrored by a reduction in mortality.
The only exception would be some eastern European countries, where thick melanomas are still being diagnosed and earlier detection would save lives.21 More aggressive and advanced tumours are not being adequately addressed by these campaigns, because patients do not always self-select well for melanoma risk factors. Thick melanomas are more likely to occur in elderly males, who may have no partner and rarely perform self-examination.
It is therefore important to educate patients that moles disappear with age, so if they remain numerous after the age of 50 years, this should be regarded as a significant risk factor for melanoma. The most common message given in public health campaigns for skin cancer, and especially melanoma, is that sun exposure is the most important risk factor. This does not pick up patients most at risk, because it is not discriminatory.
Fair skin that burns easily is too common in the UK to be a reliable at-risk phenotype. Fair skin predicts non-melanoma skin cancers better, but these are very common, and are not lethal. Patients have time to notice changes in these lesions before this becomes a significant health problem.
Adding the arm-counting tool in GP practices and in patient education messages, with consideration of the family history, would help to shift the emphasis towards mole count, rather than excessive sun exposure alone.
NICE already recommends that GPs should be trained in dermoscopy for the diagnosis of skin tumours.22 By teaching GPs how to select high-risk groups, it is likely the detection of melanoma would increase, following appropriate referrals.
This would also sensitise patients to the most important risk factors and encourage self-examination.
- Dr Simone Ribero is clinical research fellow and Dr Veronique Bataille is consultant dermatologist at the Twin Research and Genetic Epidemiology Unit, St Thomas’ Hospital, London, and Dermatology Department, West Herts NHS Trust
Competing interests: None declared
1. Little E, Eide M. Dermatol Clin 2012; 30(3): 355-61.
2. Dodds W, Morgan M, Wolfe C et al. Eur J Cancer Care (Engl) 2004; 13(1): 82-7.
3. Garbe C, Büttner P, Weiss J et al. J Invest Dermatol 1994; 102(5): 700-5.
4. Bataille V, Snieder H, MacGregor A et al. J Natl Cancer Inst 2000; 92(6): 457-63.
5. Wachsmuth R, Turner F, Barrett J et al. J Invest Dermatol 2005; 124(1): 56-62.
6. Ribero S, Glass D, Aviv A et al. PLoS One 2015; 10(1): e0116863.
7. Chang Y, Newton-Bishop J, Bishop D. Int J Cancer 2009; 124: 420-8.
8. Shitara D, Nascimento M, Puig S et al. Am J Clin Pathol 2014; 142(4): 485-91.
9. Gandini S, Sera F, Cattaruzza M et al. Eur J Cancer 2005; 41(1): 28-44.
10. Smith A, Cokkinides V, Brawley W. CA Cancer J Clin 2009; 59: 27-41.
11. Federman D, Kirsner R, Viola K. Clin Dermatol 2013; 31(6): 666-70.
12. Newton J, Bataille V, Griffiths K et al. J Am Acad Dermatol 1993; 29(6): 989-96.
13. Echeverría B, Bulliard JL, Guillén C et al. Br J Dermatol 2014; 170(1): 144-9.
14. Argenziano G, Giacomel J, Zalaudek I et al. Eur J Cancer Prev 2014; 23(5): 458-63.
15. Gallus S, Naldi L, Carli P et al. Am J Epidemiol 2007; 166(4): 472-8.
16. Ribero S, Zugna D, Osella-Abate S et al. Br J Dermatol 2016; 174(2): 312-18.
17. Swetter S, Pollitt R, Johnson T et al. Cancer 2012; 118(15): 3725-34.
18. Breitbart E, Waldmann A, Nolte S et al. J Am Acad Dermatol 2012; 66(2): 201-11.
19. Barbaric J, Sekerija M, Agius D et al. Eur J Cancer 2016; 55: 47-55.
20. Coory M, Baade P, Aitken J et al. Cancer Causes Control 2006; 17(1): 21-7.
21. De Vries E, Boniol M, Dore JF. Eur J Cancer 2004; 40: 1045-52.
22. Macbeth F, Newton-Bishop J, O’Connell S et al. BMJ 2015; 351: h3708.