Factors in Early Adolescence Associated With a Mole-Prone Phenotype in Late Adolescence | Adolescent Medicine | JAMA Dermatology | JAMA Network
[Skip to Navigation]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Navigation Landing]
Table 1.  Host, Sun, and Dermoscopic Factors Associated With a Mole-Prone Phenotype
Host, Sun, and Dermoscopic Factors Associated With a Mole-Prone Phenotype
Table 2.  Adjusted Sun and Dermoscopic Factors Associated With a Mole-Prone Phenotype
Adjusted Sun and Dermoscopic Factors Associated With a Mole-Prone Phenotype
1.
Siegel  RL, Miller  KD, Jemal  A.  Cancer statistics, 2016.  CA Cancer J Clin. 2016;66(1):7-30.PubMedGoogle ScholarCrossref
2.
Helvind  NM, Hölmich  LR, Smith  S,  et al.  Incidence of in situ and invasive melanoma in Denmark from 1985 through 2012: a national database study of 24,059 melanoma cases.  JAMA Dermatol. 2015;151(10):1087-1095.PubMedGoogle ScholarCrossref
3.
Gandini  S, Sera  F, Cattaruzza  MS,  et al.  Meta-analysis of risk factors for cutaneous melanoma, I: common and atypical naevi.  Eur J Cancer. 2005;41(1):28-44.PubMedGoogle ScholarCrossref
4.
Bauer  J, Büttner  P, Wiecker  TS, Luther  H, Garbe  C.  Risk factors of incident melanocytic nevi: a longitudinal study in a cohort of 1,232 young German children.  Int J Cancer. 2005;115(1):121-126.PubMedGoogle ScholarCrossref
5.
English  DR, Milne  E, Simpson  JA.  Ultraviolet radiation at places of residence and the development of melanocytic nevi in children (Australia).  Cancer Causes Control. 2006;17(1):103-107.PubMedGoogle ScholarCrossref
6.
Green  A, Siskind  V, Green  L.  The incidence of melanocytic naevi in adolescent children in Queensland, Australia.  Melanoma Res. 1995;5(3):155-160.PubMedGoogle ScholarCrossref
7.
Harrison  SL, MacLennan  R, Speare  R, Wronski  I.  Sun exposure and melanocytic naevi in young Australian children.  Lancet. 1994;344(8936):1529-1532.PubMedGoogle ScholarCrossref
8.
Luther  H, Altmeyer  P, Garbe  C,  et al.  Increase of melanocytic nevus counts in children during 5 years of follow-up and analysis of associated factors.  Arch Dermatol. 1996;132(12):1473-1478.PubMedGoogle ScholarCrossref
9.
Milne  E, Simpson  JA, English  DR.  Appearance of melanocytic nevi on the backs of young Australian children: a 7-year longitudinal study.  Melanoma Res. 2008;18(1):22-28.PubMedGoogle ScholarCrossref
10.
Siskind  V, Darlington  S, Green  L, Green  A.  Evolution of melanocytic nevi on the faces and necks of adolescents: a 4 y longitudinal study.  J Invest Dermatol. 2002;118(3):500-504.PubMedGoogle ScholarCrossref
11.
Wu  S, Han  J, Laden  F, Qureshi  AA.  Long-term ultraviolet flux, other potential risk factors, and skin cancer risk: a cohort study.  Cancer Epidemiol Biomarkers Prev. 2014;23(6):1080-1089.PubMedGoogle ScholarCrossref
12.
US Preventive Services Task Force.  Screening for skin cancer: US Preventive Services Task Force recommendation statement.  Ann Intern Med. 2009;150(3):188-193.PubMedGoogle ScholarCrossref
13.
Watts  CG, Dieng  M, Morton  RL, Mann  GJ, Menzies  SW, Cust  AE.  Clinical practice guidelines for identification, screening and follow-up of individuals at high risk of primary cutaneous melanoma: a systematic review.  Br J Dermatol. 2015;172(1):33-47.PubMedGoogle ScholarCrossref
14.
Crane  LA, Mokrohisky  ST, Dellavalle  RP,  et al.  Melanocytic nevus development in Colorado children born in 1998: a longitudinal study.  Arch Dermatol. 2009;145(2):148-156.PubMedGoogle ScholarCrossref
15.
Darlington  S, Siskind  V, Green  L, Green  A.  Longitudinal study of melanocytic nevi in adolescents.  J Am Acad Dermatol. 2002;46(5):715-722.PubMedGoogle ScholarCrossref
16.
Gallagher  RP, McLean  DI, Yang  CP,  et al.  Suntan, sunburn, and pigmentation factors and the frequency of acquired melanocytic nevi in children: similarities to melanoma: the Vancouver Mole Study.  Arch Dermatol. 1990;126(6):770-776.PubMedGoogle ScholarCrossref
17.
Valiukeviciene  S, Miseviciene  I, Gollnick  H.  The prevalence of common acquired melanocytic nevi and the relationship with skin type characteristics and sun exposure among children in Lithuania.  Arch Dermatol. 2005;141(5):579-586.PubMedGoogle ScholarCrossref
18.
Oliveria  SA, Geller  AC, Dusza  SW,  et al.  The Framingham school nevus study: a pilot study.  Arch Dermatol. 2004;140(5):545-551.PubMedGoogle ScholarCrossref
19.
Oliveria  SA, Satagopan  JM, Geller  AC,  et al.  Study of Nevi in Children (SONIC): baseline findings and predictors of nevus count.  Am J Epidemiol. 2009;169(1):41-53.PubMedGoogle ScholarCrossref
20.
Scope  A, Marghoob  AA, Chen  CS, Lieb  JA, Weinstock  MA, Halpern  AC; SONIC Study Group.  Dermoscopic patterns and subclinical melanocytic nests in normal-appearing skin.  Br J Dermatol. 2009;160(6):1318-1321.PubMedGoogle ScholarCrossref
21.
Dusza  SW, Halpern  AC, Satagopan  JM,  et al.  Prospective study of sunburn and sun behavior patterns during adolescence.  Pediatrics. 2012;129(2):309-317.PubMedGoogle ScholarCrossref
22.
Scope  A, Dusza  SW, Marghoob  AA,  et al.  Clinical and dermoscopic stability and volatility of melanocytic nevi in a population-based cohort of children in Framingham school system.  J Invest Dermatol. 2011;131(8):1615-1621.PubMedGoogle ScholarCrossref
23.
Oliveria  SA, Scope  A, Satagopan  JM,  et al.  Factors associated with nevus volatility in early adolescence.  J Invest Dermatol. 2014;134(9):2469-2471.PubMedGoogle ScholarCrossref
24.
Fonseca  M, Marchetti  MA, Chung  E,  et al.  Cross-sectional analysis of the dermoscopic patterns and structures of melanocytic naevi on the back and legs of adolescents.  Br J Dermatol. 2015;173(6):1486-1493.PubMedGoogle ScholarCrossref
25.
Scope  A, Marghoob  AA, Dusza  SW,  et al.  Dermoscopic patterns of naevi in fifth grade children of the Framingham school system.  Br J Dermatol. 2008;158(5):1041-1049.PubMedGoogle ScholarCrossref
26.
Geller  AC, Oliveria  SA, Bishop  M, Buckminster  M, Brooks  KR, Halpern  AC.  Study of health outcomes in school children: key challenges and lessons learned from the Framingham Schools’ Natural History of Nevi Study.  J Sch Health. 2007;77(6):312-318.PubMedGoogle ScholarCrossref
27.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194.PubMedGoogle ScholarCrossref
28.
Seg3D Segmentation [computer program]. Salt Lake City, UT: Center for Integrative Biometical Computing; 2014.
29.
Hofmann-Wellenhof  R, Marghoob  AA, Zalaudek  I.  Large acquired nevus or dysplastic nevus: what’s in the name of a nevus?  JAMA Dermatol. 2016;152(6):623-624.PubMedGoogle ScholarCrossref
30.
English  DR, Armstrong  BK.  Identifying people at high risk of cutaneous malignant melanoma: results from a case-control study in Western Australia.  BMJ (Clin Res Ed). 1988;296(6632):1285-1288.PubMedGoogle ScholarCrossref
31.
Newton-Bishop  JA, Chang  YM, Elliott  F,  et al.  Relationship between sun exposure and melanoma risk for tumours in different body sites in a large case-control study in a temperate climate.  Eur J Cancer. 2011;47(5):732-741.PubMedGoogle ScholarCrossref
32.
Veierød  MB, Weiderpass  E, Thörn  M,  et al.  A prospective study of pigmentation, sun exposure, and risk of cutaneous malignant melanoma in women.  J Natl Cancer Inst. 2003;95(20):1530-1538.PubMedGoogle ScholarCrossref
33.
Vuong  K, Armstrong  BK, Weiderpass  E,  et al; Australian Melanoma Family Study Investigators.  Development and external validation of a melanoma risk prediction model based on self-assessed risk factors.  JAMA Dermatol. 2016;152(8):889-896.PubMedGoogle ScholarCrossref
34.
Gandini  S, Sera  F, Cattaruzza  MS,  et al.  Meta-analysis of risk factors for cutaneous melanoma, II: sun exposure.  Eur J Cancer. 2005;41(1):45-60.PubMedGoogle ScholarCrossref
35.
Lee  TK, Rivers  JK, Gallagher  RP.  Site-specific protective effect of broad-spectrum sunscreen on nevus development among white schoolchildren in a randomized trial.  J Am Acad Dermatol. 2005;52(5):786-792.PubMedGoogle ScholarCrossref
36.
Bassoli  S, Maurichi  A, Rodolfo  M,  et al.  CDKN2A and MC1R variants influence dermoscopic and confocal features of benign melanocytic lesions in multiple melanoma patients.  Exp Dermatol. 2013;22(6):411-416.PubMedGoogle ScholarCrossref
37.
Lipoff  JB, Scope  A, Dusza  SW, Marghoob  AA, Oliveria  SA, Halpern  AC.  Complex dermoscopic pattern: a potential risk marker for melanoma.  Br J Dermatol. 2008;158(4):821-824.PubMedGoogle ScholarCrossref
38.
Douglas  NC, Borgovan  T, Carroll  MJ,  et al.  Dermoscopic naevus patterns in people at high versus moderate/low melanoma risk in Queensland.  Australas J Dermatol. 2011;52(4):248-253.PubMedGoogle ScholarCrossref
39.
Wei  EX, Qureshi  AA, Han  J,  et al.  Trends in the diagnosis and clinical features of melanoma in situ (MIS) in US men and women: a prospective, observational study.  J Am Acad Dermatol. 2016;75(4):698-705.PubMedGoogle ScholarCrossref
40.
Clark  LN, Shin  DB, Troxel  AB, Khan  S, Sober  AJ, Ming  ME.  Association between the anatomic distribution of melanoma and sex.  J Am Acad Dermatol. 2007;56(5):768-773.PubMedGoogle ScholarCrossref
Original Investigation
October 2017

Factors in Early Adolescence Associated With a Mole-Prone Phenotype in Late Adolescence

Author Affiliations
  • 1Dermatology Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, New York
  • 2Department of Dermatology, Sheba Medical Center, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
  • 3Department of Social and Behavioral Sciences, Harvard T.H. Chan School of Public Health, Harvard University, Boston, Massachusetts
  • 4School Health Services, Framingham Public Schools, Framingham, Massachusetts
JAMA Dermatol. 2017;153(10):990-998. doi:10.1001/jamadermatol.2017.1547
Key Points

Question  Can factors associated with a mole-prone phenotype in late adolescence be identified in early adolescence?

Findings  In this cohort study, baseline total nevus count and variability in nevus dermoscopic pattern in early adolescence was significantly associated with a mole-prone phenotype in late adolescence.

Meaning  Clinically recognizable factors associated with a mole-prone phenotype may facilitate the identification of a population at risk for melanoma and have implications for primary prevention strategies and skin self-examination practices.

Abstract

Importance  Nevi are important phenotypic risk factors for melanoma in adults. Few studies have examined the constitutional and behavioral factors associated with a mole-prone phenotype in adolescents.

Objective  To identify host, behavioral, and dermoscopic factors in early adolescence (age, 14 years) that are associated with a mole-prone phenotype in late adolescence (age, 17 years).

Design, Setting, and Participants  A prospective observational cohort study from the Study of Nevi in Children was conducted from January 1, 2009, to December 31, 2014, with a 2- to 3-year follow-up. A total of 569 students from the school system in Framingham, Massachusetts, were enrolled in the 8th or 9th grade (baseline; mean [SD] age, 14.4 [0.7] years). The overall retention rate was 73.3%, and 417 students were reassessed in the 11th grade.

Main Outcome and Measures  Mole-prone phenotype in the 11th grade, defined as total nevus count of the back and 1 randomly selected leg in the top decile of the cohort or having any nevi greater than 5 mm in diameter.

Results  Of the 417 students assessed at follow-up in the 11th grade (166 females and 251 males; mean [SD] age, 17.0 [0.4] years), 111 participants (26.6%) demonstrated a mole-prone phenotype: 69 students (62.2%) with 1 nevus greater than 5 mm in diameter, 23 students (20.7%) with total nevus count in the top decile, and 19 students (17.1%) with both characteristics. On multivariate analysis, baseline total nevus count (adjusted odds ratio, 9.08; 95% CI, 4.0-23.7; P < .001) and increased variability of nevus dermoscopic pattern (adjusted odds ratio, 4.24; 95% CI, 1.36-13.25; P = .01) were associated with a mole-prone phenotype.

Conclusions and Relevance  This study found clinically recognizable factors associated with a mole-prone phenotype that may facilitate the identification of individuals at risk for melanoma. These findings could have implications for primary prevention strategies and help target at-risk adolescents for higher-intensity counseling about sun protection and skin self-examination.

Introduction

The incidence of and mortality from cutaneous melanoma continue to rise.1,2 Phenotypic characteristics of melanocytic nevi in adulthood, such as total body nevus count and the presence of atypical or dysplastic nevi, are among the strongest known risk factors of melanoma.3 Childhood and adolescence are critical periods for the appearance and evolution of nevi.4-10 A more precise understanding of the natural history of nevi, particularly in young individuals, may have important implications for primary melanoma prevention.11 The US Preventive Services Task Force currently recommends educational interventions for fair-skinned individuals 10 to 24 years of age related to minimizing exposure to UV radiation to reduce the risk for skin cancer.12 Several expert groups in the United States and worldwide recommend that patients at high risk of developing melanoma receive regular physician screenings and education about skin self-examination and sun protection.13 Previous studies4-10,14-17 have investigated factors associated with a mole-prone phenotype during childhood and adolescence, but few have longitudinally examined the dermoscopic features of individual nevi in children, which may help inform melanoma prevention strategies.

The Study of Nevi in Children (SONIC) is a population-based study designed to prospectively document the evolution of individual nevi in children using clinical and dermoscopic photography.18-21 A previous report on the stability and volatility of nevi in this cohort found a median increase of 2 new back nevi during a 3-year period from the 5th to 8th grade, with 75% of participants having 1 new back nevus.22 Previous analyses also suggested that factors such as high baseline nevus count at 11 years of age, globular nevi dermoscopic pattern, and sunburn were associated with an increase in total nevus count by 14 years of age.23 Herein, we aimed to identify factors in early adolescence (age, 14 years) associated with a mole-prone phenotype in late adolescence (age, 17 years).

Methods

This study was conducted from January 1, 2009, to December 31, 2014, with a 2- to 3-year follow-up. Details about participants and protocols regarding implementation and data collection have been previously described.22,24-26 The study population included the graduating classes of 2012, 2014, and 2015 from high schools in Framingham, Massachusetts. A list of all eligible participants aged 13 to 15 years in the 8th or 9th grade was obtained from the school system, and a description of the study as well as consent and assent forms were mailed to families requesting participation. Consenting participants underwent skin examinations, digital photography of the back and legs, and completed a self-administered questionnaire at baseline in the 8th (graduating classes of 2014 and 2015) or 9th grade (graduating class of 2012) and at follow-up in the 11th grade. The study was approved by the Boston University and Harvard University institutional review boards and was conducted in accordance with the Declaration of Helsinki.27

Data Collection

Clinical overview images of the participants’ back and legs as well as contact, nonpolarized dermoscopic images of select back and leg nevi were obtained. Up to 8 back nevi were imaged per participant: the largest nevus and up to 3 random nevi were selected from the upper back and the largest nevus and up to 3 random nevi were selected from the lower back, if present. In addition, up to 8 leg nevi were imaged per participant: the largest nevus and up to 3 random nevi were selected from the upper leg and the largest nevus and up to 3 random nevi were selected from the lower leg, if present. Owing to time constraints imposed by the school, leg nevi were imaged from only 1 randomly selected leg that was determined prior to the date of imaging. Definitions for anatomical sites and method of selection of the random nevi have been previously published.24-26

Digital photography was performed with a digital camera back (Phase One P25; Hasselblad), Hasselblad 403w camera system (403w; Hasselblad), 1-kW flash system (Studio Flash; Canfield Scientific Inc). Dermoscopic images were obtained using a digital camera (Nikon D90; Nikon) and 60-mm macro lens (Nikkor; Nikon) with a dermoscopy attachment (Epi-Flash; Canfield Scientific Inc). Images were stored in Mirror software, a clinical imaging database that permits a resolution of 3 million pixels (Canfield Scientific Inc).

A study nurse (M.B.) assessed participant characteristics on the day of imaging. Data on sex and race/ethnicity were obtained from the school district. In addition, the children completed surveys that included questions on sun sensitivity; total sun exposure; sun protection practices, including use of hats and sunscreen; frequency of sunburns and painful sunburns; and tanning behavior (eg, like or dislike tanning, spending time to get a tan). From these data, composite variables, sun sensitivity index (SSI),22 and outdoor sun exposure were created. The SSI is a composite variable representing skin color, hair color, and tendency to burn with sun exposure rather than tan whereby a higher score represents fair skin, light hair, and increased tendency to burn. Outdoor sun exposure is a composite variable quantifying total outdoor sun exposure for a typical week by combining weekday and weekend responses for each participant by his or her relative contributions. This variable is categorized by low, medium, and high levels of exposure by tertiles of the distribution.

Image Analysis

Dermoscopic assessments were performed by 2 trained observers (M.A.M. and M.F.) for consensus agreement and were blinded to anatomical location. Lesions were assessed for global dermoscopic pattern, color, and symmetry, then grouped into 1 of 4 categories: reticular, globular, homogeneous, and complex. Definitions for these patterns have been previously published.24 Each nevus was categorized as flat or raised; if any component of the nevus was raised, the lesion was considered raised. Nevi with a distinct peripheral rim of globules or a starburst pattern were classified as complex. Global dermoscopic pattern concordance between reviewers has previously been shown to be high (κ, 0.77).24 Because an individual could exhibit many nevus patterns, we created a categorical variable (host dermoscopy pattern) based on all the dermoscopic patterns of an individual’s indexed nevi. Definitions of host dermoscopy patterns are provided in eTable 1 in the Supplement. In addition, we looked at the association of nevus pattern variability with a mole-prone phenotype by examining the number of different nevus patterns present in an individual.

Total nevus counts were performed semi-automatically using a color-based segmentation method and Seg3D,28 an open-source manual mark-up software tool. Nevus surface area was calculated from dermoscopic images using the same method and software. Lesion borders were visually identified, and then the number of pixels in the circumscribed area was obtained and converted to the number of pixels per millimeter.

A total of 3293 lesions of the baseline cohort were dermoscopically imaged, of which 1005 (30.5%) were excluded because they were associated with participants lost to follow-up, they were not nevi (as determined in image analyses), or their images were of poor quality. In sum, 2288 nevi from the back (n = 1777) and legs (n = 511) were included in the study: 972 (42.5%) were homogeneous, 922 (40.3%) were reticular, 280 (12.2%) were globular, and 83 (3.6%) were complex. Dermoscopic images were not available for 31 lesions. The method used for image analyses was the same for baseline and follow-up.

Statistical Analysis

Descriptive statistics were used to characterize the study population. Descriptive frequencies were calculated to assess the distribution of host, sun, and dermoscopic characteristics. Quiz Ref IDParticipants who completed the follow-up evaluation were classified as having a mole-prone phenotype if, in the 11th grade, they had a total nevus count of the back and 1 randomly selected leg in the top decile of the cohort, or they had any nevi greater than 5 mm in diameter in these anatomical locations. We selected nevus size as an objective criterion for a mole-prone phenotype given the lack of consensus regarding the features of atypical or dysplastic nevi.29 Multiple imputation using multinomial logistic regression was used to handle missing survey data. Univariate and multivariate analyses of baseline host factors and sun behaviors were evaluated by mole-prone phenotype using 2-tailed t tests, χ2 tests, and logistic regression. As nevi were nested within students, random-effects logistic models were used to evaluate dermoscopic factors. In these models, a variable for the student was entered as a random effect. The dependent variable was mole-prone phenotype in the 11th grade. Odds ratio (OR) estimates along with 95% CIs were obtained from both univariate and multivariate models. Student sex, sun sensitivity index, race/ethnicity, and baseline nevus count were included in multivariate regression models as potential confounding factors. P < .05 was considered significant. All statistical analyses were performed using Stata software, version 14.1 (Stata Corp).

Results
Description of Cohort

The baseline cohort included 569 children who underwent imaging in 8th or 9th grade (mean [SD] age, 14.4 [0.7] years). The consent rate was 32.0% of eligible students (569 of 1778). Of these children, 417 (251 [60.2%] male, 372 [89.2%] white) underwent repeat imaging in 11th grade (mean [SD] age, 17.0 [0.4] years); these participants were used in the analysis. The overall retention rate was 73.3%. The median nevus count of the back and 1 randomly selected leg at baseline was 15 nevi (interquartile range, 5) and at follow-up was 21 nevi (interquartile range, 28). The distributions of total nevus counts at baseline and follow-up are in eTable 2 in the Supplement. The median number of dermoscopically imaged nevi per participant was 7 (interquartile range, 5). Quiz Ref IDThe median nevus count for the top decile of the cohort at follow-up was 73 nevi (interquartile range, 21). At follow-up, 111 students (26.6%) demonstrated a mole-prone phenotype. Of the students with a mole-prone phenotype, 69 (62.2%) had at least 1 nevus that was greater than 5 mm in diameter, 23 (20.7%) had a total nevus count in the top decile, and 19 (17.1%) exhibited both characteristics.

Host Characteristics

Host characteristics, such as white race/ethnicity and SSI, were associated with a mole-prone phenotype in the 11th grade (Table 1). Of the 111 participants with a mole-prone phenotype, 107 (96.4%) were of white race/ethnicity and 56 (50.5%) exhibited the highest SSI. White participants were more likely to be categorized as having a mole-prone phenotype compared with nonwhite participants (OR, 4.14; 95% CI, 1.45-11.84; P = .008). Participants with the highest SSI were also more likely to have a mole-prone phenotype compared with those with the lowest SSI (OR, 7.64; 95% CI, 1.76-33.12; P = .007). A higher geometric mean baseline nevus count (38.4 vs 15.3; P < .001) in the 8th grade was significantly associated with a mole-prone phenotype in the 11th grade. To better characterize this association, we conducted analyses by quintiles of baseline nevus count (Table 1). Compared with students with between 0 and 6 nevi at baseline (first quintile), students with greater than 12 nevi (third quintile and higher) were more likely to be categorized as having a mole-prone phenotype at follow-up (third quintile: OR, 5.08; 95% CI, 1.95-13.23; P = .001; fourth quintile: OR, 5.53; 95% CI, 2.10-14.52; P = .001; and fifth quintile: OR, 26.79; 95% CI, 10.39-69.06; P < .001).

Sun Behaviors

A history of any sunburn was significantly associated with a mole-prone phenotype in the 11th grade: 89 participants with a mole-prone phenotype (80.2%) reported 1 or more sunburn in the previous summer compared with 194 of 306 participants without the mole-prone phenotype (63.4%). Participants with 3 or more sunburns were more likely to be categorized as having a mole-prone phenotype compared with those who did not have any sunburns (OR, 2.97; 95% CI, 1.60-5.51; P = .001), as were those with 3 or more painful sunburns in the past summer (OR, 2.45; 95% CI, 1.15-5.21; P = .02). Conversely, liking to get a tan was associated with decreased odds of having a mole-prone phenotype (OR, 0.58; 95% CI, 0.34-0.99; P = .046). Outdoor sun exposure, spending time to get a tan, and sun protection behaviors, such as wearing sunscreen and limiting time outdoors, were not associated with having a mole-prone phenotype at follow-up (Table 1).

Dermoscopic Features

Univariate analyses of dermoscopic pattern by mole-prone phenotype in the 11th grade demonstrated an overall significant association (Pearson χ2 = 34.1; P < .001). Compared with homogeneous nevi, globular nevi at baseline were associated with increased odds of having a mole-prone phenotype at follow-up evaluation (OR, 1.64; P = .003), as were complex nevi at baseline (OR, 2.28; P = .002). Raised lesions were also more likely to be associated with a mole-prone phenotype (OR, 2.18; P = .005) (Table 2). The number of colors in a single nevus or the presence of lesion asymmetry was not significantly associated with having a mole-prone phenotype. Descriptive frequencies of dermoscopic factors are given in eTable 3 in the Supplement. Quiz Ref IDA reticular-globular host dermoscopy pattern was more likely to be associated with a mole-prone phenotype compared with a homogeneous host dermoscopy pattern (OR, 2.51; 95% CI, 1.00-6.32; P = .051), as was a complex host dermoscopy pattern (OR, 3.86; 95% CI, 1.48-10.05; P = .006). In addition, the presence of 3 (OR, 4.64; 95% CI, 2.11-10.21; P < .001) or all 4 (OR, 13.11; 95% CI, 4.52-38.00; P < .001) nevus dermoscopic patterns in an individual was significantly associated with a mole-prone phenotype compared with having only 1 pattern.

Multivariate Analyses

Following adjustment for sex, SSI, and baseline nevus count, a history of sunburn or painful sunburn was not associated with a mole-prone phenotype (Table 2). Participants who had 3 or more sunburns in the past summer exhibited increased odds of mole-prone phenotype that were not significant (adjusted OR, 1.92; 95% CI, 0.94-3.89; P = .07). Baseline dermoscopic factors including nevus pattern, raised lesions, and host dermoscopy pattern were also not associated with a mole-prone phenotype. Quiz Ref IDHaving all 4 types of nevus dermoscopic patterns remained associated with a mole-prone phenotype (adjusted OR, 4.24; 95% CI, 1.36-13.25; P = .01). Controlling for race/ethnicity and number of dermoscopically imaged nevi per participant did not change our results. Finally, individuals in the fifth quintile of baseline total nevus count (>31 nevi) were significantly more likely to exhibit a mole-prone phenotype at follow-up compared with the individuals in the first quintile (<6 nevi), following adjustment for all host, sun, and dermoscopic factors with a significant association on univariate analysis (adjusted OR, 9.08; 95% CI, 4.0-23.7; P < .001).

Discussion

We report on factors in early adolescence associated with a mole-prone phenotype in late adolescence. Our study longitudinally followed up a cohort of children in Framingham, Massachusetts, and identified host, sun, and dermoscopic characteristics in the 8th or 9th grade associated with a mole-prone phenotype in the 11th grade. In multivariate analyses, we found clinically recognizable baseline characteristics, specifically, total nevus count and variability of nevus dermoscopic pattern, to be associated with having a mole-prone phenotype at follow-up. Identification of a population with a mole-prone phenotype at an early age may have implications for primary prevention strategies and skin self-examination practices.11

Previous work in identifying risk factors for cutaneous melanoma has largely been conducted in adult populations and focused on host and sun exposure characteristics. Our findings are consistent with those of previous studies examining the factors associated with risk of melanoma, which have demonstrated that race/ethnicity, skin phenotype, and total nevus count are significant risk factors for the development of melanoma.3,30-33 Furthermore, a history of sunburn, which was a significant factor associated with a mole-prone phenotype on univariate analysis, has also been previously shown to be positively correlated with increased risk of melanoma.11,34 We did not find any association between outdoor sun exposure or the use of sun protection and a mole-prone phenotype. However, these findings must be interpreted with caution as our survey questions were self-reported and addressed exposures only in the 1 year prior to administration; as such, they do not represent cumulative lifetime sun behaviors. In fact, a previous randomized intervention trial demonstrated that sunscreen use may attenuate the development of new nevi in white school-aged children.35 Future long-term prospective cohort studies are needed to definitively examine the association of sun exposure and photoprotection habits with nevus counts in children and adolescents, ideally starting with a younger cohort and using more frequent follow-up intervals. In addition, liking to get tan, but not spending time to get tan, was associated with decreased odds of having a mole-prone phenotype, even after adjusting for SSI. “Liking to get tan” is an attitude, not a behavior, which is likely affected by many social and cultural influences. We advocate caution in drawing significant conclusions from these findings.

To our knowledge, few studies have examined whether dermoscopic patterns of nevi are associated with risk of melanoma. Two previous case-control studies of adults showed that a complex dermoscopic pattern was observed more frequently in patients with melanoma than in controls.36,37 Another cross-sectional study demonstrated that dermoscopic nevus patterns were similar for age and body site in people with different levels of risk of melanoma.38 On multivariate analysis, neither dermoscopic pattern of individual nevi nor host dermoscopy pattern was significantly associated with a mole-prone phenotype. However, we did find on multivariate analysis that individuals possessing all 4 nevus dermoscopic patterns at baseline were more likely to be categorized with a mole-prone phenotype at follow-up. These findings suggest that variability in the dermoscopic pattern of an individual’s nevi may play a role in assessing risk of melanoma. However, the clinical significance of our findings requires further validation.

Strengths and Limitations

Our study has multiple strengths. The longitudinal design allows for the prospective documentation of individual nevi over time. This design affords a better understanding of the natural history and evolution of nevi under the lens of dermoscopy. In addition, our outcome variable incorporates 2 factors (total nevus count and nevi of diameter >5 mm) previously shown to be positively correlated with risk of melanoma.3 Finally, the fact that our study was conducted in a population-based cohort enhances the generalizability of our findings.

There are also significant limitations to our study. First, the study population was predominantly white, which hinders the applicability of our findings to other racial groups. Second, imaging of nevi was limited to the back and legs, so our results may not be representative of nevi present in all anatomical sites. Although these areas do not represent all of the constitutional nevi, the use of the back and legs provides a sampling of nevi from sites that frequently develop melanoma.39,40 Third, our study analyzed changes in host and dermoscopic factors in a relatively small cohort during a short follow-up period of 2 or 3 years. Fourth, our outcome of interest (mole-prone phenotype) was a surrogate marker for the possible development of melanoma. Not all individuals with a mole-prone phenotype will develop melanoma, nor do all individuals who develop melanoma express a mole-prone phenotype. In addition, it is unclear if a mole-prone phenotype at 17 years of age is associated with a nevus phenotype at high risk for melanoma as an adult; therefore, the results of our study cannot be extrapolated past the 11th grade. However, since individuals with a mole-prone phenotype may be an important target population for primary prevention measures and skin self-examination, we believe that our use of this surrogate marker is valid, especially since a study in adolescence that uses melanoma as the primary outcome is not feasible.

Conclusions

We found clinical and dermoscopic features in early adolescence that are associated with a mole-prone phenotype in late adolescence. Participants with a mole-prone phenotype constituted less than one-third of the total study population at follow-up and were more likely to have a higher baseline total nevus count and increased variability of dermoscopic patterns. Combined with earlier findings in preadolescent individuals,21-23Quiz Ref ID the data suggest that a segment of the adolescent population is likely to be at an increased risk for melanoma and can be identified and targeted for enhanced melanoma prevention efforts before the initiation of tanning behaviors.

Back to top
Article Information

Corresponding Author: Allan C. Halpern, MD, Dermatology Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, 16 E 60th St, New York, NY 10022 (halperna@mskcc.org).

Accepted for Publication: April 6, 2017.

Published Online: June 7, 2017. doi:10.1001/jamadermatol.2017.1547

Author Contributions: Mr Xu and Dr Dusza had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Xu, Marchetti, Dusza, Fonseca, Geller, Halpern.

Acquisition, analysis, or interpretation of data: Xu, Marchetti, Dusza, Chung, Fonseca, Scope, Bishop, Marghoob.

Drafting of the manuscript: Xu, Dusza.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Xu, Dusza.

Obtained funding: Geller, Halpern.

Administrative, technical, or material support: Marchetti, Chung, Fonseca, Scope.

Study supervision: Dusza, Fonseca, Halpern.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was funded in part through Cancer Center Support Grant P30-CA008748 from the National Institutes of Health (NIH)/National Cancer Institute and supported by award number R01-AR049342 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the NIH as well as grant P41-GM103545 from the National Institute of General Medical Sciences of the NIH.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

References
1.
Siegel  RL, Miller  KD, Jemal  A.  Cancer statistics, 2016.  CA Cancer J Clin. 2016;66(1):7-30.PubMedGoogle ScholarCrossref
2.
Helvind  NM, Hölmich  LR, Smith  S,  et al.  Incidence of in situ and invasive melanoma in Denmark from 1985 through 2012: a national database study of 24,059 melanoma cases.  JAMA Dermatol. 2015;151(10):1087-1095.PubMedGoogle ScholarCrossref
3.
Gandini  S, Sera  F, Cattaruzza  MS,  et al.  Meta-analysis of risk factors for cutaneous melanoma, I: common and atypical naevi.  Eur J Cancer. 2005;41(1):28-44.PubMedGoogle ScholarCrossref
4.
Bauer  J, Büttner  P, Wiecker  TS, Luther  H, Garbe  C.  Risk factors of incident melanocytic nevi: a longitudinal study in a cohort of 1,232 young German children.  Int J Cancer. 2005;115(1):121-126.PubMedGoogle ScholarCrossref
5.
English  DR, Milne  E, Simpson  JA.  Ultraviolet radiation at places of residence and the development of melanocytic nevi in children (Australia).  Cancer Causes Control. 2006;17(1):103-107.PubMedGoogle ScholarCrossref
6.
Green  A, Siskind  V, Green  L.  The incidence of melanocytic naevi in adolescent children in Queensland, Australia.  Melanoma Res. 1995;5(3):155-160.PubMedGoogle ScholarCrossref
7.
Harrison  SL, MacLennan  R, Speare  R, Wronski  I.  Sun exposure and melanocytic naevi in young Australian children.  Lancet. 1994;344(8936):1529-1532.PubMedGoogle ScholarCrossref
8.
Luther  H, Altmeyer  P, Garbe  C,  et al.  Increase of melanocytic nevus counts in children during 5 years of follow-up and analysis of associated factors.  Arch Dermatol. 1996;132(12):1473-1478.PubMedGoogle ScholarCrossref
9.
Milne  E, Simpson  JA, English  DR.  Appearance of melanocytic nevi on the backs of young Australian children: a 7-year longitudinal study.  Melanoma Res. 2008;18(1):22-28.PubMedGoogle ScholarCrossref
10.
Siskind  V, Darlington  S, Green  L, Green  A.  Evolution of melanocytic nevi on the faces and necks of adolescents: a 4 y longitudinal study.  J Invest Dermatol. 2002;118(3):500-504.PubMedGoogle ScholarCrossref
11.
Wu  S, Han  J, Laden  F, Qureshi  AA.  Long-term ultraviolet flux, other potential risk factors, and skin cancer risk: a cohort study.  Cancer Epidemiol Biomarkers Prev. 2014;23(6):1080-1089.PubMedGoogle ScholarCrossref
12.
US Preventive Services Task Force.  Screening for skin cancer: US Preventive Services Task Force recommendation statement.  Ann Intern Med. 2009;150(3):188-193.PubMedGoogle ScholarCrossref
13.
Watts  CG, Dieng  M, Morton  RL, Mann  GJ, Menzies  SW, Cust  AE.  Clinical practice guidelines for identification, screening and follow-up of individuals at high risk of primary cutaneous melanoma: a systematic review.  Br J Dermatol. 2015;172(1):33-47.PubMedGoogle ScholarCrossref
14.
Crane  LA, Mokrohisky  ST, Dellavalle  RP,  et al.  Melanocytic nevus development in Colorado children born in 1998: a longitudinal study.  Arch Dermatol. 2009;145(2):148-156.PubMedGoogle ScholarCrossref
15.
Darlington  S, Siskind  V, Green  L, Green  A.  Longitudinal study of melanocytic nevi in adolescents.  J Am Acad Dermatol. 2002;46(5):715-722.PubMedGoogle ScholarCrossref
16.
Gallagher  RP, McLean  DI, Yang  CP,  et al.  Suntan, sunburn, and pigmentation factors and the frequency of acquired melanocytic nevi in children: similarities to melanoma: the Vancouver Mole Study.  Arch Dermatol. 1990;126(6):770-776.PubMedGoogle ScholarCrossref
17.
Valiukeviciene  S, Miseviciene  I, Gollnick  H.  The prevalence of common acquired melanocytic nevi and the relationship with skin type characteristics and sun exposure among children in Lithuania.  Arch Dermatol. 2005;141(5):579-586.PubMedGoogle ScholarCrossref
18.
Oliveria  SA, Geller  AC, Dusza  SW,  et al.  The Framingham school nevus study: a pilot study.  Arch Dermatol. 2004;140(5):545-551.PubMedGoogle ScholarCrossref
19.
Oliveria  SA, Satagopan  JM, Geller  AC,  et al.  Study of Nevi in Children (SONIC): baseline findings and predictors of nevus count.  Am J Epidemiol. 2009;169(1):41-53.PubMedGoogle ScholarCrossref
20.
Scope  A, Marghoob  AA, Chen  CS, Lieb  JA, Weinstock  MA, Halpern  AC; SONIC Study Group.  Dermoscopic patterns and subclinical melanocytic nests in normal-appearing skin.  Br J Dermatol. 2009;160(6):1318-1321.PubMedGoogle ScholarCrossref
21.
Dusza  SW, Halpern  AC, Satagopan  JM,  et al.  Prospective study of sunburn and sun behavior patterns during adolescence.  Pediatrics. 2012;129(2):309-317.PubMedGoogle ScholarCrossref
22.
Scope  A, Dusza  SW, Marghoob  AA,  et al.  Clinical and dermoscopic stability and volatility of melanocytic nevi in a population-based cohort of children in Framingham school system.  J Invest Dermatol. 2011;131(8):1615-1621.PubMedGoogle ScholarCrossref
23.
Oliveria  SA, Scope  A, Satagopan  JM,  et al.  Factors associated with nevus volatility in early adolescence.  J Invest Dermatol. 2014;134(9):2469-2471.PubMedGoogle ScholarCrossref
24.
Fonseca  M, Marchetti  MA, Chung  E,  et al.  Cross-sectional analysis of the dermoscopic patterns and structures of melanocytic naevi on the back and legs of adolescents.  Br J Dermatol. 2015;173(6):1486-1493.PubMedGoogle ScholarCrossref
25.
Scope  A, Marghoob  AA, Dusza  SW,  et al.  Dermoscopic patterns of naevi in fifth grade children of the Framingham school system.  Br J Dermatol. 2008;158(5):1041-1049.PubMedGoogle ScholarCrossref
26.
Geller  AC, Oliveria  SA, Bishop  M, Buckminster  M, Brooks  KR, Halpern  AC.  Study of health outcomes in school children: key challenges and lessons learned from the Framingham Schools’ Natural History of Nevi Study.  J Sch Health. 2007;77(6):312-318.PubMedGoogle ScholarCrossref
27.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194.PubMedGoogle ScholarCrossref
28.
Seg3D Segmentation [computer program]. Salt Lake City, UT: Center for Integrative Biometical Computing; 2014.
29.
Hofmann-Wellenhof  R, Marghoob  AA, Zalaudek  I.  Large acquired nevus or dysplastic nevus: what’s in the name of a nevus?  JAMA Dermatol. 2016;152(6):623-624.PubMedGoogle ScholarCrossref
30.
English  DR, Armstrong  BK.  Identifying people at high risk of cutaneous malignant melanoma: results from a case-control study in Western Australia.  BMJ (Clin Res Ed). 1988;296(6632):1285-1288.PubMedGoogle ScholarCrossref
31.
Newton-Bishop  JA, Chang  YM, Elliott  F,  et al.  Relationship between sun exposure and melanoma risk for tumours in different body sites in a large case-control study in a temperate climate.  Eur J Cancer. 2011;47(5):732-741.PubMedGoogle ScholarCrossref
32.
Veierød  MB, Weiderpass  E, Thörn  M,  et al.  A prospective study of pigmentation, sun exposure, and risk of cutaneous malignant melanoma in women.  J Natl Cancer Inst. 2003;95(20):1530-1538.PubMedGoogle ScholarCrossref
33.
Vuong  K, Armstrong  BK, Weiderpass  E,  et al; Australian Melanoma Family Study Investigators.  Development and external validation of a melanoma risk prediction model based on self-assessed risk factors.  JAMA Dermatol. 2016;152(8):889-896.PubMedGoogle ScholarCrossref
34.
Gandini  S, Sera  F, Cattaruzza  MS,  et al.  Meta-analysis of risk factors for cutaneous melanoma, II: sun exposure.  Eur J Cancer. 2005;41(1):45-60.PubMedGoogle ScholarCrossref
35.
Lee  TK, Rivers  JK, Gallagher  RP.  Site-specific protective effect of broad-spectrum sunscreen on nevus development among white schoolchildren in a randomized trial.  J Am Acad Dermatol. 2005;52(5):786-792.PubMedGoogle ScholarCrossref
36.
Bassoli  S, Maurichi  A, Rodolfo  M,  et al.  CDKN2A and MC1R variants influence dermoscopic and confocal features of benign melanocytic lesions in multiple melanoma patients.  Exp Dermatol. 2013;22(6):411-416.PubMedGoogle ScholarCrossref
37.
Lipoff  JB, Scope  A, Dusza  SW, Marghoob  AA, Oliveria  SA, Halpern  AC.  Complex dermoscopic pattern: a potential risk marker for melanoma.  Br J Dermatol. 2008;158(4):821-824.PubMedGoogle ScholarCrossref
38.
Douglas  NC, Borgovan  T, Carroll  MJ,  et al.  Dermoscopic naevus patterns in people at high versus moderate/low melanoma risk in Queensland.  Australas J Dermatol. 2011;52(4):248-253.PubMedGoogle ScholarCrossref
39.
Wei  EX, Qureshi  AA, Han  J,  et al.  Trends in the diagnosis and clinical features of melanoma in situ (MIS) in US men and women: a prospective, observational study.  J Am Acad Dermatol. 2016;75(4):698-705.PubMedGoogle ScholarCrossref
40.
Clark  LN, Shin  DB, Troxel  AB, Khan  S, Sober  AJ, Ming  ME.  Association between the anatomic distribution of melanoma and sex.  J Am Acad Dermatol. 2007;56(5):768-773.PubMedGoogle ScholarCrossref
×