Risk Factors for Melanoma and Genetic Testing

January 7, 2016

While one German study has shown that population-wide screening for melanoma could dramatically reduce mortality, potential problems with the strategy include cost due to the overall rarity of the disease, patient compliance, and difficulties identifying patients likely to develop melanoma. Therefore, scientists are developing prediction models to identify those at high risk for melanoma, which can include genetic factors. Learn more

Although cutaneous melanoma results in more years of life lost than any adult cancer except breast cancer, 98% of local cases are cured. In fact, a 10-year study conducted in Schleswig-Holstein, an area of Germany, found that a screening program reduced mortality from 1.9 per 100,000 people to one death per 100,000.

 “You might say, ‘Why should we worry about assessing risk for melanoma, why not just screen the entire population?’” says David Polsky, M.D., Ph.D., who spoke to colleagues at the summer meeting of the American Academy of Dermatology (New York, 2015).

READ: Melanoma subtypes challenging to diagnose, treat

One problem is cost. In order to prevent a single death from melanoma, the researchers in Germany had to screen 111,000 men (compared with only 67 men in colon cancer, based on other published studies, for example).

But dermatologists face another problem when it comes to screening: compliance. In a 2009 study conducted in western France, researchers used a self-assessment score to identify participants at high risk for melanoma, and referred 30% of them to dermatologists for a full body examination.

ALSO READ: Skin cancer screenings target marathoners

“But out of 2,400 high-risk patients, only 1,000 were compliant and actually showed up,” according to Dr. Polsky, who is the Alfred W. Kopf professor of cutaneous oncology, and director of the Pigmented Lesion Section in The Ronald O. Perelman Department of Dermatology, New York University School of Medicine, Langone Medical Center. “It’s a bit of a problem. If we improve our risk prediction models, maybe our message to patients will be more impactful.”

On follow-up, the physicians performed 95 biopsies, and found ten melanomas.

“They found an eleven-fold enrichment in melanomas by going to this process of the risk score, which means you could reduce the number needed to screen by a factor of about ten,” Dr. Polsky explains.

NEXT: Fine-tuning risk assessment

 

Fine-tuning risk assessment

During his presentation, Dr. Polsky described factors and how they can be combined into prediction models to identify those at high risk for melanoma.

While a strong family history is an important risk factor, clinicians should be aware of other factors:

“Ninety percent of melanoma patients lack a family history, and 40% of family histories may be inaccurate. If somebody comes in the door and says, ‘I have a family history of melanoma,’ their risk is only 1.7 greater than normal, perhaps because people confuse melanoma with their aunt who had a basal cell,’” Dr. Polsky explains.

READ: Viruses may lead to skin disease, cancer

 “Lacking a strong family history or personal history of melanoma, a patient’s mole phenotype has the greatest impact on their risk,” he continues. According to a recent study, having a large number of moles gives a patient a risk factor of 5.4. Those with one or two large nevi have about double the average risk for developing melanoma, he adds.

Indoor tanning is another factor. In a study published by Dr. Polsky and colleagues in a study population of 1,640 subjects, 10 or more hours of indoor tanning was associated with an increased odds ratio for melanoma of 2.32.

By contrast, lifetime number of severe sunburns was only associated with a significant increase in melanoma risk in univariate analysis, where the odds ratio was 3.23 (p=0.01), but not multivariate analysis (odds ratio 1.6, p=0.15), the article said.[1]

ALSO READ: Survival with melanoma subtypes varies

Among other factors, immune suppression (such as may occur in transplant patients) is also a risk factor. Other important factors include the patient’s pigmentation, tanning ability, and freckling. However, dermatologists should beware, because as many as half of melanoma patients lack these risk factors.

NEXT: Genetics: The missing puzzle piece?

 

Genetics: The missing puzzle piece?

In addition, Dr. Polsky says, while red hair is associated with having a greater genetic predisposition to melanoma, as many as 35% of people with dark brown or black hair actually have one of the red hair genes, he adds.

READ: Protein p15 may influence melanoma progression

Having red hair can increase an individual’s risk by a factor of four in some studies, according to Dr. Polsky. Furthermore, certain genes that are associated with developing melanoma are also associated, more or less robustly, with red hair.

 “Genetic variants that are very strongly associated with red hair (the ‘big R’) have an odds ratio of 50 to 118 that the individual has red hair, and that’s a real strong association. Less associated are the ‘little r’ variants, which have odds ratios of two to six that the person has red hair,” he explains, although as many as one-third of those with darker hair also have some of the variants.

 “Red hair is associated with genetic variants in the melanocortin-1 receptor (MC1R),” Dr. Polsky continues. “When the receptor is stimulated by melanocyte-stimulating hormone, it activates signal transduction, leading, in part, to an increase in pigmentation. Interestingly, this signaling is also involved in DNA repair. Best evidence regarding weak signaling demonstrates that receptors encoded by the red hair variants such as R151C, R160W, and D294H don’t signal so well[1]  and are associated with an increased odds ratio for melanoma of up to 2.4 (p=0.008), he says. Other associated variants are D84E, V60L, V92M, R142H, I155T, and R163Q.

ALSO READ: Collaborative care challenges

As part of an attempt to develop a predictive formula, Dr. Polsky and colleagues conducted a retrospective review of data collected from invasive cutaneous melanoma patients between 25 and 59 years old between July of 2004 and December of 2007, as part of the Minnesota Skin Health Study. Controls were randomly selected from state driver’s license holders, and all participants provided mouthwash samples and received a self-administered questionnaire, the article said.

The researchers found that taking into account the variants in MC1R increased the predictive ability of the model to an extent that was statistically significant, he adds. This suggests that adding more genetic variants to a risk prediction model may improve its ability to identify patients at high risk for melanoma. The findings were described in the article referenced above.

NEXT: The future of screening

 

The future of screening

 “A future risk assessment paradigm might start by giving the patient a mole diagram with four categories: People with no moles, very few moles, some moles, or many moles. Then they spit into a cup so you can isolate their inherited DNA, and you can run your genetic risk model.

READ: Newly discovered tumor-suppressor gene impacts melanoma survival

 “You might say, ‘OK you don’t have a lot of moles, and you’ve got a low-risk genetic profile, so we’re not going to formally screen you every year.’ And then you have these people who are phenotypically low-risk but genetically high-risk, and you say, ‘You ought to come in once a year.’ Then, for people who are phenotypically and genetically high risk, you say, ‘You know what? We definitely need to screen you two or three times a year, because your risk is so high.’ Surveillance might be increased, by increasing the frequency of skin exams, total body photos, and all of the technologies - throw it all at them so you can catch melanoma early.

“Maybe patients will be more compliant if we say, ‘Hey, this test is actually pretty good,’ and maybe we can tell people more effectively to be more serious about sun protection,” he adds.

Dr. Polsky reports no relevant conflicts related to this article.

 

[1] Penn LA, Qian M, Zhang E, et al. Development of a melanoma risk prediction model incorporating MC1R genotype and indoor tanning exposure: impact of mole phenotype on model performance. PLoS ONE. 2014;9(7):e101507. (Polsjky was senior author).

 [1]

Polsky 9/4/15, 1:52 PM

Best evidence for weak signaling is for these 3 variants.

 

Related Content:

News