Oral Presentation 2015 Annual Meeting of the Australasian Society for Dermatology Research

Genetic association of dermoscopic nevus patterns in a Queensland case-control study of melanoma (#40)

Richard A Sturm 1 , Kasturee Jagirdar 1 , Katie Lee 1 , Phil McClenahan 1 , Elizabeth McEniery 1 , Sam Beh 1 , Blair Burke 1 , Helmut Schaider 1 , David Duffy 2 , H. Peter Soyer 1 3
  1. Dermatology Research Center, School of Medicine, University of Queensland, Brisbane, Queensland, Australia
  2. QIMR-Berghofer Medical Research Institute, Brisbane, Qld, Australia
  3. Department of Dermatology, Princess Alexandra Hospital, Brisbane, Qld, Australia

Total number of nevi is recognized as an important risk factor for melanoma, however our understanding of this relationship for the general population is still superficial in relation to mole and melanoma morphology.  We report on a case-control study of melanoma and nevi in South-east Queensland, combining an assessment of pigmentation characteristics, nevus phenotype including dermoscopic nevus subtypes with genotypic comparisons and melanoma risk for 573 individuals.  This study is referred to as the Brisbane Nevus Morphology Study (BNMS) and includes 238 melanoma patients, 76 individuals with familial association and 259 non-melanoma/healthy controls.  DNA prepared from saliva samples from each volunteer has been subject to complete MC1R gene sequencing and genotyping at over 500,000 SNPs using an Illumina CoreExome Chip. We have captured dermoscopic images of 13,587 nevi >5mm in diameter this collection, among these the dominant nevus subtype pattern was nonspecific (62.5%) then reticular (21.9%), with a minority having a globular pattern (15.6%).  There was a significant association between the non-specific dermoscopic signature pattern and MC1R genotype (P=0.002).  We have also found 9 participants in our cohort who carry the SUMOylation deficient MITF E318K mutation that has recently been described as a medium-penetrance melanoma gene.  Of these 9 heterozygote carriers 7 are melanoma cases, and all have more nevi on average than the controls.  Analysis of SNPs within 13 loci previously reported in GWAS studies of melanoma showed a statistically significance association with melanoma for 10 of these loci in our cohort.  Of six genes identified in a meta-analysis of nevus GWAS, we replicated four.  The most striking of these was the IRF4 SNP rs12203592*C/T (P<10-7).  Notably, IRF4 was also the lead gene associated with dermoscopic nevus pattern when a multivariate analysis was performed, with strongest effect of the rs12203592*T variant allele decreasing globular nevus counts (P<10-6).