Advertisement

Combining Genetic and Sun Exposure Data May Improve Prediction of Skin Cancer Risk

Advertisement

Key Points

  • The best-performing models incorporated a genetic risk score composed of data on up to 50 genetic variants, along with survey data on family history, skin pigmentation and sensitivity, number of moles, sun exposure estimates, and body mass index.
  • The genetic risk scores, computed from genome-wide associate studies significant loci, accounted for 8% to 14% of the total variance explained by the best models. 
  • The out-of-sample area under the curve for the best models was 0.79, 0.80, and 0.78, for BCC, SCC, and melanoma, respectively.

By combining data on individuals’ lifetime sun exposure and their genetics, researchers may be able to generate improved predictions of their risk of skin cancer, according to findings presented by Fontanillas et al at the 2018 American Society of Human Genetics Annual Meeting (Abstract PgmNr 10).

Pierre Fontanillas, PhD, and colleagues collected genetic and survey data from over 210,000 consented 23andMe research participants of European descent. They analyzed the data to identify correlations between previously known and potentially novel skin cancer risk factors and the occurrence of three forms of skin cancer: melanoma, basal cell carcinoma (BCC), and squamous cell carcinoma (SCC). Past studies had found that exposure to ultraviolet (UV) light increases skin cancer risk, as do other environmental factors such as living in a sunnier climate or at a higher altitude, and personal factors such as lighter skin pigmentation, higher numbers of moles on the skin, and family history of skin cancer.

“We aimed to validate previously known skin cancer risk factors in a large cohort, add detail to these and explore potential new ones, and find out whether and how these factors might interact with genetic risk,” said Dr. Fontanillas in a statement.

Study Findings

The researchers found that while each single factor was not particularly significant on its own, multiple factors could be combined into statistical models that were more informative. The best-performing models incorporated a genetic risk score composed of data on up to 50 genetic variants, along with survey data on family history, skin pigmentation and sensitivity, number of moles, sun exposure estimates, and body mass index.

The genetic risk scores, computed from genome-wide associate studies significant loci, accounted for 8% to 14% of the total variance explained by the best models. The out-of-sample area under the curve for the best models was 0.79, 0.80, and 0.78, for BCC, SCC, and melanoma, respectively.

While the self-reported nature of the survey data permitted researchers to collect a large dataset, it also presented some challenges, Dr. Fontanillas noted. “Measuring lifetime exposure is generally challenging. It is particularly hard to capture sun exposure and when in life it happened, and it may be that some of the other correlates we found, like higher body mass index, reflect a lack of outdoor activity rather than being directly correlated with risk of skin cancer,” he said.

Moving forward, the researchers plan to expand their sample to groups with non-European ancestry and are exploring additional methods of calculating genetic risk score and measuring sun exposure. They hope to eventually obtain risk estimates accurate enough to be used by individuals and clinicians.

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.


Advertisement

Advertisement



Advertisement

By continuing to browse this site you permit us and our partners to place identification cookies on your browser and agree to our use of cookies to identify you for marketing. Read our Privacy Policy to learn more.