Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Sep 15.
Published in final edited form as: Cancer. 2018 Sep 12;124(18):3715–3723. doi: 10.1002/cncr.31641

Pediatric Melanoma in Melanoma-prone Families

Alisa M Goldstein 1,*, Kelsey C Stidd 1, Xiaohong R Yang 1, Mary C Fraser 1, Margaret A Tucker 1
PMCID: PMC6214720  NIHMSID: NIHMS974862  PMID: 30207590

Abstract

Background:

In the United States, only 0.4% of all melanomas are diagnosed in patients <20 years. Melanoma in pediatric members of melanoma-prone families has not been fully investigated. The study goal was to evaluate pediatric melanoma patients with extensive follow-up in melanoma-prone families with and without CDKN2A mutations.

Methods:

For this non-population-based study, families were followed prospectively for up to 40 years. Sixty families with >3 melanoma patients were included for analysis: 30 CDKN2A mutation positive (CDKN2A+) and 30 CDKN2A negative (CDKN2A-). Age at first melanoma and number of melanomas were obtained for each patient and summarized by family or sets (CDKN2A+ vs CDKN2A-). For set comparisons and categorical variables (occurrence of melanoma in pediatric patients, number of melanomas, number of patients with single or multiple melanomas), Pearson chi-square or Fisher exact test was used.

Results:

Regardless of CDKN2A status, melanoma-prone families had 6–28-fold higher proportions of patients with pediatric melanoma compared to the general population of melanoma patients in the United States. Within CDKN2A+ families, pediatric melanoma patients were significantly more likely to have multiple melanomas than their relatives who were diagnosed at >20 years (71% vs 38%, respectively; p=0.004). CDKN2A+ families had significantly higher percentages of pediatric melanoma patients (11.1% versus 2.5%, p=0.004) compared to CDKN2A- families.

Conclusions:

These observations have implications for prevention of melanoma as well as clinical care for early detection of melanoma. Children in melanoma-prone families should have careful sun protection from an early age and skin surveillance to reduce their risk of melanoma.

Keywords: melanoma, CDKN2A, pediatrics, family research, genetics, sun protection

Precis:

Regardless of CDKN2A status, melanoma-prone families had 6–28-fold higher proportions of patients with pediatric melanoma compared to the general population of melanoma patients in the United States. Children in melanoma-prone families should have careful sun protection from an early age and skin surveillance to reduce their risk of melanoma.

Introduction

Cutaneous malignant melanoma (CMM) is a potentially fatal form of skin cancer, resulting from a combination of environmental, host, and genetic factors.1–3 Multiple high-, intermediate- and low-risk susceptibility genes are linked to CMM, with cyclin-dependent kinase inhibitor 2A (CDKN2A) being the major high-risk susceptibility gene.4 CDKN2A, a tumor suppressor gene, located on chromosome 9p21, encodes two distinct proteins translated in alternate reading frames (ARF), from alternatively spliced transcripts. The alpha transcript encodes p16, a protein that inhibits the cyclin-D1-cyclin-dependent kinase 4 (CDK4) or cyclin D1-CDK6 complex. The p16 protein arrests cell growth at the G1 stage of the cell cycle, acting as a tumor suppressor.1,4 The smaller beta transcript specifies p14ARF; this protein induces cell cycle arrest or apoptosis acting through the p53 pathway. Germline mutations in CDKN2A have been found in ~20–40% of melanoma-prone families worldwide.1,4

In the United States general population, melanoma is the fifth most common cancer in men and sixth most common in women. The median age at melanoma diagnosis is 64 years and only 0.4% of all melanomas are diagnosed in patients less than 20 years of age.5 The major environmental risk factor for melanoma is exposure to ultraviolet (UV) radiation, either natural exposure or artificial exposure usually via tanning beds. Host factors associated with melanoma include fair skin, hair, and eye color, poor tanning ability, and the presence of multiple melanocytic nevi.1

Compared to the general population, melanoma-prone families have a reduced age at melanoma diagnosis and increased frequency of multiple primary melanomas (MPM).6,7 However, the occurrence of melanoma in pediatric (<20 years) members of these families has not been fully investigated. The goal of the current study was to evaluate the occurrence and clinical and genetic characteristics of young onset (i.e. pediatric) melanoma cases with extensive follow-up in melanoma-prone families with and without CDKN2A mutations. Better understanding of pediatric melanoma will help to ensure UV protection from an early age and classification of nevus status; it would also be useful for helping to decide surveillance approaches.

Subjects and Methods

Data for this study came from a non-population-based family study from the Division of Cancer Epidemiology and Genetics at the National Cancer Institute (NCI). Families were ascertained through self or health professional referrals, and have been followed prospectively for up to 40 years, starting in the 1970s. For this analysis, eligibility criteria included documented cutaneous melanoma, invasive and/or in situ, in at least 3 family members, with at least two of the required melanoma patients being first-degree relatives. Age at melanoma diagnosis was not a criterion for ascertainment. After initial confirmation of family eligibility, all identified living family members were invited to the NIH Clinical Center for detailed skin examinations to document susceptibility phenotypes. If families/individuals could not travel to the NIH, we organized field trips near individuals’ homes to collect biospecimens, conduct skin examinations, and photograph nevi. Written informed consent for each participant or each participant’s guardian for this observational study was obtained prior to participation under an NCI Institution Review Board approved protocol (NCI 02–0211; Clinicaltrials.gov ID NCT00040352). Willing participants provided blood primarily for genetic studies. All participating families were Caucasian and resided in various regions of the United States. Variables collected and examined in this study included gender, study period, age at melanoma diagnosis, melanoma subtype and site, thickness, precursor nevus status, number of melanomas, and CDKN2A mutation status. Study period for time of melanoma occurrence was defined as “retrospective” if the melanoma in a study participant occurred prior to the participant’s initial clinical examination and “prospective” if the melanoma occurred after the initial clinical examination of the participant. All diagnoses of melanoma were confirmed by review of pathology materials/reports, medical records, or death certificates. All melanoma diagnoses were confirmed using the above review strategy except for one pediatric patient (patient #23 from family D7) for whom it was not possible to retrieve medical records or death certificate. Sixty families were included in this study, 30 of which were CDKN2A mutation positive (denoted CDKN2A+) and 30, CDKN2A mutation negative (CDKN2A-). These 60 families included more than 1300 clinically evaluated participants (melanoma cases, unaffected relatives, and spouses).

The Surveillance Epidemiology and End Results (SEER) Program was used to obtain data regarding the general population (https://seer.cancer.gov/statfacts/html/melan.html; https://seer.cancer.gov/csr/1975_2014)5 of melanoma patients for comparison with proportion of pediatric melanoma patients and age at diagnosis in the melanoma-prone families.

The median age at first diagnosis of melanoma and number of melanomas were obtained for each subject with melanoma and then summarized by family or sets of families (CDKN2A+ vs CDKN2A-). For comparisons between sets (CDKN2A+ vs. CDKN2A- families; melanoma patients from CDKN2A+ vs. CDKN2A- families) and different categorical variables (occurrence of melanoma in pediatric patients, number of melanomas, number of patients with single or multiple melanomas), the Pearson chi-square or Fisher exact test was used depending on sample sizes. Because of the relatively small number of patients available for analysis, we assumed independence of melanoma patients within families for patient-level analyses. As a sensitivity analysis to reduce effects of bias from family referral, selection, or correlation, we performed the same comparisons after exclusion of the probands (the melanoma patients who led to ascertainment of a family) from each family. All statistical analyses were conducted using Excel, StataSE 11.2, or SAS 9.4. All p-values were two-sided and considered significant at the 0.05 level.

Results

Among the 60 families (30 CDKN2A+, 30 CDKN2A-) in the current study, there were 311 confirmed melanoma patients, of whom 24 (7.7%) were diagnosed before age 20 years. Table 1 shows the clinical characteristics of the 24 pediatric melanoma patients, 21 from thirteen CDKN2A+ families and 3 from three CDKN2A- families. All pediatric melanoma patients from CDKN2A+ families for whom biologic specimens were available (n=19) carried their respective family’s mutation. The median age at first melanoma diagnosis was 16.5 years (range: 9–19 years). Thirteen patients (54%) were female. For most patients (16/20 with known classification), the first melanoma was a superficial spreading melanoma; more than half of the melanomas had a precursor nevus, the clear majority being a dysplastic nevus (DN). Overall, the 24 pediatric patients had a median of two melanomas (range: 1–32 over the time of follow-up). Fifteen patients, all from CDKN2A+ families, had multiple primary melanomas (MPM). Seven of the 15 MPM patients had multiple melanomas before age 20 years.

Table 1. Clinical characteristics of first melanoma and total number of melanomas in pediatric melanoma patients from melanoma-prone families with (Positive) and without (Negative) CDKN2A mutations.

Patient No. Family ID Gender Age Melanoma Type Site Thickness (mm) Precursor Lesion
(Pathology)		 Period Multiple Primary Melanoma No. Mel before age 20 years Total No. of Mel Family Mutation Status
1 AN Male 16 NM Chest 2.10 No Retrospective Yes 1 2 Positive
2 AN Female 13 SSM Back MIS Yes/DN Retrospective Yes 9 13 Positive
3 A Female 19 SSM Knee 0.72 Yes/DN Prospective Yes 2 2 Positive
4 E Female 13 SSM Calf MIS Yes/DN Prospective Yes 1 2 Positive
5 D Male 14 Unclassified Back Unknown Yes/DN Retrospective Yes 1 23 Positive
6 BB Female 19 SSM Foot 1.50 Yes/Not DN Retrospectivea No 1 1 Negative
7 F Male 19 SSM Scalp 2.42 Yes/DN Retrospective Yes 1 4 Positive
8 F Male 17 Unknown Neck Unknown Unknown Retrospective No 1 1 Positive
9 U Male 12 Unclassified Face 0.88 Yes/Not DN Retrospective No 1 1 Negative
10 G Female 18 SSM Forearm Microinvasive Yes/DN Retrospective Yes 1 32 Positive
11 G Female 19 NM Forearm 1.65 Yes/DN Prospective Yes 1 7 Positive
12 G Female 17 SSM Shoulder 0.30 No Prospective Yes 3 3 Positive
13 G Female 16 SSM Scalp 0.70 No Prospective Yes 2 2 Positive
14 J Male 17 NM Upper arm 1.66 No Retrospective Yes 1 2 Positive
15 J Female 17 SSM Calf MIS De Novo Retrospective Yes 1 2 Positive
16 O Female 14 SSM Back 0.52 Yes/DN Retrospective Yes 5 8 Positive
17 AH Male 9 SSM Shoulder 0.34 Yes/DN Prospective Yes 2 3 Positive
18 AH Male 11 SSM Back 0.47 No Prospective No 1 1 Positive
19 P Female 13 LMM Hip MIS No Prospective No 1 1 Positive
20 P Male 18 SSM Scalp 6.50 Yes/DN Prospective No 1 1 Positive
21 AS Male 18 SSM Breast MIS No Prospective No 1 1 Negative
22 AP Male 15 SSM Face 0.32 Yes/DN Retrospectivea Yes 5 5 Positive
23 D7 Female 13 Unknown Unknown Unknown Unknown Retrospectivea No 1 1 Positive
24 A9 Female 17 SSM Back 0.60 Yes/DN Retrospectivea No 1 1 Positive
Open in a new tab

Abbreviations: No., Number; NM, nodular melanoma; SSM, superficial spreading melanoma; LMM, lentigo maligna melanoma; mm, millimeters; DN, dysplastic nevus; Mel, melanoma; MIS, melanoma-in-situ

a

Patient was a proband, i.e., one of the melanoma patients who led to ascertainment of family

Table 2 shows the age at diagnosis, thickness, study period, and nevus status for each melanoma, and clinical DN and age at last status for the 24 pediatric melanoma patients. Similar to the situation for first melanomas, most subsequent melanomas also had precursor lesions (67/83=80.7%) with about 70% being DN. Among the 21 examined patients, all but one patient (#21) had clinical DN. Patient #21, however, had small atypical nevi that did not meet criteria for DN. For five patients (#2, 4, 15, 19, 21), their first tumor was melanoma-in-situ (MIS). Two of these patients (#19, 21) have not yet developed any additional melanomas although both are still young (age 22 and 33 years, respectively, at last follow-up) and therefore may develop additional melanomas in the future. Of the 15 MPM patients, half developed additional melanomas at least ten years after their initial melanomas. Three patients (#2, 5, 10) were diagnosed with >10 melanomas, invasive or in-situ, during their follow-ups of more than 30 years. Of interest, the two patients who developed the most melanomas used tanning beds during part of the follow-up period.8

Table 2.

Clinical characteristics of each melanoma, overall clinical dysplastic nevus (DN) status, and age at last status in the 24 pediatric melanoma patients.

Patient No. Mel No. Age at Mel Diagnosis, years Thickness, mm Study Period Precursor Lesion Patient Clinical DN
Status		 Age at last status, years
1 1 16 2.1 Retrospective No Yes 47
2 34 2.85 At Exam Yes/not DN
2 1 13 MIS Retrospective Yes/DN Yes 49
2 13 MIS Retrospective Yes/DN
3 16 0.43 Prospective Yes/DN
4 16 MIS Prospective Yes/DN
5 16 MIS Prospective Yes/DN
6 16 MIS Prospective Yes/DN
7 16 0.35 Prospective Yes/DN
8 18 MIS Prospective Yes/DN
9 18 MIS Prospective Yes/DN
10 21 Microinvasive Prospective Yes/DN
11 42 MIS Prospective Yes/DN
12 45 MIS Prospective Clinical nevusb
13 46 0.28 Prospective Yes/DN
3 1 19 0.72 Prospective Yes/DN Yes 45
2 19 0.94 Prospective No
4 1 13 MIS Prospective Yes/DN Yes 36
2 36 MIS Prospective Yes/DN
5 1 14 Unknown Retrospective Yes/DN Yes 63
2 24 0.4 Retrospective Yes/DN
3 24 0.22 Retrospective No
4 26 0.41 Prospective Yes/DN
5 26 Microinvasive Prospective Yes/DN
6 27 MIS Prospective Yes/DN
7 31 0.25 Prospective Yes/DN
8 31 0.39 Prospective Yes/DN
9 32 0.38 Prospective Yes/DN
10 33 0.35 Prospective No
11 35 Microinvasive Prospective No
12 37 0.5 Prospective No
13 41 0.45 Prospective Yes/not DN
14 41 0.55 Prospective Yes/DN
15 41 0.33 Prospective No
16 42 MIS Prospective No
17 48 0.42 Prospective No
18 48 MIS Prospective Yes/DN
19 48 MIS Prospective No
20 55 MIS Prospective Yes/DN
21 55 MIS Prospective Unknown
22 56 MIS Prospective Yes/DN
23 62 0.84 Prospective Clinical nevusb
6 1 19 1.5 Retrospectivea Yes/not DN Unknown (Not Examined) 21
7 1 19 2.42 Retrospective Yes/DN Yes 25
2 22 1 Retrospective Yes/DN
3 22 0.4 Retrospective Yes/DN
4 23 0.54 Retrospective Yes/DN
8 1 17 Unknown Retrospective Unknown Unknown (Not Examined) 20
9 1 12 0.88 Retrospective Yes/not DN Yes 25
10 1 18 Microinvasive Retrospective Yes/DN Yes 60
2 21 Microinvasive Retrospective Yes/DN
3 21 0.88 Retrospective Yes/DN
4 21 0.15 Retrospective Yes/DN
5 22 Microinvasive Retrospective Yes/not DN
6 23 0.56 Retrospective Yes/DN
7 24 Microinvasive Prospective Yes/DN
8 24 MIS Prospective No
9 26 MIS Prospective Yes/not DN
10 26 0.59 Prospective Yes/DN
11 27 MIS Prospective No
12 28 MIS Prospective No
13 28 MIS Prospective Yes/DN
14 28 Microinvasive Prospective Yes/DN
15 29 0.33 Prospective Yes/DN
indeterminate
16 29 Microinvasive Prospective No
17 29 Microinvasive Prospective No
18 29 MIS Prospective Yes/not DN
19 35 MIS Prospective No
20 36 Unknown Prospective Unknown
21 38 0.33 Prospective Yes/DN
22 42 MIS Prospective Clinical nevusb
23 45 MIS Prospective Unknown
24 47 0.24 Prospective Yes/DN
25 48 MIS Prospective Yes/DN
26 50 0.4 Prospective Yes/DN
indeterminate
27 51 MIS Prospective Unknown
28 52 MIS Prospective Unknown
29 52 0.25 Prospective Unknown
30 54 0.5 Prospective Unknown
31 56 0.34 Prospective Clinical nevusb
32 57 0.2 Prospective Clinical nevusb
11 1 19 1.65 Prospective Yes/DN Yes 51
2 30 MIS Prospective Yes/not DN
3 35 MIS Prospective Unknown
4 40 MIS Prospective Yes/DN
5 42 MIS Prospective Yes/DN
6 46 0.4 Prospective Unknown
7 49 MIS Prospective Unknown
12 1 17 0.3 Prospective No Yes 26
2 18 0.35 Prospective Yes/DN
3 19 0.33 Prospective Yes/DN
13 1 16 0.7 Prospective No Yes 20
2 18 0.55 Prospective Yes/DN
indeterminate
14 1 17 1.66 Retrospective No Yes 27
2 20 MIS Prospective Yes/DN
15 1 17 MIS Retrospective No Yes 44
2 39 0.53 Prospective Yes/not DN
16 1 14 0.52 Retrospective Yes/DN Yes 25
2 14 0.46 Retrospective Unknown
3 16 0.37 Retrospective Yes/DN
4 16 0.4 Retrospective Yes/DN
5 16 0.32 Retrospective Yes/DN
6 20 2.4 Retrospective Yes/DN
7 21 1 Retrospective Yes/DN
8 21 MIS Retrospective Unknown
17 1 9 0.34 Prospective Yes/DN Yes 30
2 10 0.49 Prospective Yes/not DN
3 30 MIS Prospective Clinical nevusb
18 1 11 0.47 Prospective No Yes 23
19 1 13 MIS Prospective No Yes 22
20 1 18 6.5 Prospective Yes/DN Yes 32
21 1 18 MIS Prospective No Indeterminate 33
22 1 15 0.32 Retrospectivea Yes/DN Yes 31
2 15 0.64 Retrospective No
3 15 MIS Retrospective Yes/not DN
4 16 0.43 Retrospective Yes/not DN
5 16 0.45 Retrospective Yes/not DN
23 1 13 Unknown Retrospectivea Unknown Unknown (Not Examined) 39
24 1 17 0.6 Retrospectivea Yes/DN Yes 44
Open in a new tab

Abbreviations: No., Number; Mel, melanoma; mm, millimeters; DN, dysplastic nevus; MIS, melanoma-in-situ

a

Patient was a proband, i.e., one of the melanoma patients who led to ascertainment of family

b

Clinical nevus means that on a previous exam, or by history, a nevus was present but it was not detected in pathology report

Among the 60 families evaluated, 43% of CDKN2A+ families had at least one pediatric CMM case and 10% of CDKN2A- families had at least one pediatric CMM case (Table 3). Further, one-third (10/30) of CDKN2A+ families had pediatric CMM cases who developed multiple primary melanoma tumors. In contrast, no CDKN2A- families had pediatric MPM cases. Eleven percent of the CMM cases (21/189) in CDKN2A+ families and 2.5% of CMM cases (3/122) in CDKN2A- families were diagnosed before age 20 years (Table 3). Although these percentages differed significantly (p=0.004), both percentages were substantially greater than in the United States general population (0.4% occurring in patients diagnosed before 20 years).5 Specifically, the melanoma-prone families in this study had 6–28-fold (.025/.004 for pediatric melanoma patients from CDKN2A- families; .111/.004 for pediatric melanoma patients from CDKN2A+ families) higher proportions of patients with pediatric melanoma compared to the NCI’s SEER-based melanoma patient population.5 The sensitivity analyses excluding probands showed similar results to analyses that included all melanoma patients (including probands) with significant differences between family sets for the pediatric cases (Table 3). Finally, within CDKN2A+ families, pediatric melanoma patients were significantly more likely to have MPM than their relatives who were diagnosed at greater than 20 years of age (71% vs 38%, respectively; p=0.004).

Table 3.

Proportion of pediatric melanoma patients by family and by melanoma patients in CDKN2A+ and CDKN2A- families

CDKN2A+ Families CDKN2A- Families P value
No. families with pediatric cases (%) 13/30 (43.3) 3/30 (10.0) 0.007
No. pediatric cases among all melanoma patients (%) 21/189 (11.1) 3/122 (2.5) 0.004
Excluding Probands:
 No. families with pediatric cases (%) 10/24 (41.7) 1/16 (6.2) 0.027
 No. pediatric cases among all melanoma patients (%) 18/109 (16.5) 1/39 (2.6) 0.026
Open in a new tab

Discussion

Cutaneous melanoma is predominantly an adult onset disease with median age at diagnosis of 64 years in the United States.5 Pediatric melanoma, defined here as occurring before 20 years of age, is rare and found in only 0.4% of melanoma patients in the United States. In contrast, regardless of CDKN2A mutation status, the melanoma-prone families in this study had 6–28-fold higher proportions of patients with pediatric melanoma compared to the NCI’s SEER-based melanoma patient population.5 Moreover, in the general population, melanoma occurring in this age range shows a female predominance, whereas in the families the occurrence of pediatric melanoma was close to equal in both genders.5,9 In addition, 71% of pediatric melanoma patients in CDKN2A+ families had multiple melanomas with almost half of these patients (7/15) having multiple melanomas before age 20 years. Pediatric melanoma patients from CDKN2A+ families were also significantly more likely than their non-pediatric melanoma relatives (i.e. diagnosed at 20 years or greater) to have multiple melanomas. This observation is consistent with what is seen in the general population of melanoma patients in which individuals who develop melanoma at an early age are also at higher risk of developing subsequent melanomas than older individuals. The risk of second primary melanoma in individuals less than 30 years of age is 15.4-fold increased; in those 30–49 years, it is 9.6-fold increased, and in those over age 50 years, 8-fold increased.10

The incidence of cutaneous melanoma has been continuing to rise in the United States for the past several decades with increasing incidence in young onset melanoma in recent years.5,11,12 Among patients diagnosed before 20 years of age, the clear majority (>90%) are diagnosed in the teen years13 similar to what was observed in the melanoma-prone families in the current study. Although numerous studies have investigated the prevalence of CDKN2A mutations in adult onset melanoma cases from families and the general population14–18 less is known about the prevalence of mutations in CDKN2A in young onset melanoma cases. A study using the Swedish Cancer Register evaluated 51 histopathologically confirmed melanoma patients diagnosed before age 20 years for germline mutations in CDKN2A and found only one CDKN2A mutation in a patient with a strong family history of melanoma.19 A recent study of 23 clinic-based American, Spanish, and Dutch melanoma patients diagnosed before 20 years of age found no germline mutations in the known melanoma-predisposing genes including CDKN2A.13 Although based on relatively small numbers, these studies suggest that germline CDKN2A mutations are rare in pediatric melanoma patients outside of the familial melanoma setting. The current study could not evaluate the prevalence of CDKN2A mutations in pediatric melanoma patients from the general population but showed that melanoma-prone families, with and without CDKN2A mutations, have an increased frequency of pediatric melanoma compared to the general population of melanoma patients.

All but one of the pediatric melanomas in the current study were confirmed by review of pathology material/reports, medical records, or death certificates. For the pediatric melanoma patient from family D7 (patient #23), it was not possible to confirm her melanoma diagnosis because we were unable to obtain medical records or death certificate. Exclusion of this patient and family from the study had minimal effect on the results. Family U was included among the families without a CDKN2A mutation. The pediatric melanoma patient in this family was diagnosed with melanoma at age 12 years. His father was diagnosed with melanoma at age 23 years and his paternal grandmother was diagnosed at age 36 years. The paternal grandmother was negative for a CDKN2A mutation. However, the unaffected paternal grandfather, who self-identified as Latino, carried a CDKN2A variant p.I49T that has conflicting interpretations of pathogenicity (ClinVar20) and occurs at an allele frequency in the Latino population from gnomAD (0.0044) that is above the threshold considered too common to plausibly cause disease.21,22 Therefore, this variant was excluded from consideration as disease-causing and family U was considered CDKN2A- for this (and all our previous) studies.

In general, melanocytic neoplasms in children (diagnosed before 20 years of age) have been classified into three main subtypes: in association with a large congenital melanocytic nevus, spitzoid melanocytic tumors which include spitzoid melanoma and atypical Spitz tumours, and adult-like (or ‘conventional) melanoma.9,13 Most melanomas developing in pediatric subjects after puberty, designated adolescents, tend to show clinical features consistent with adult melanoma.9,23 About one-quarter of the pediatric melanoma patients in the current study developed their initial melanomas before/during puberty with the youngest pediatric melanoma patient being diagnosed at nine years of age, however, all the melanoma tumors in pediatric patients were adult-like in their clinical and histologic characteristics, with the predominant histologic type being superficial spreading melanoma. Further, about one-quarter of the initial tumors were MIS/microinvasive and thus associated with excellent prognosis, whereas one-quarter were at least 1.50 mm thick (three being >2 mm thick) with a much less favorable prognosis. Children in melanoma-prone families should be protected from UV exposure from birth to reduce their risk of melanoma. Further, given the earliest diagnosis of melanoma at only nine years of age, regular skin surveillance including the scalp for early detection of features associated with increased risk for melanoma including occurrence of DN should be considered for children from melanoma-prone families before they enter their teen years.23,24 Usually, the first indication that children may develop DN is an increased number of nevi, some of which may be slightly irregular in outline or variable in color. Anecdotally, dysplastic nevi may become more apparent around the time of puberty. Some of these pediatric patients also continue to develop melanomas throughout their lives and therefore need to practice sun protection, carefully monitor their skin for changes both in nevi and normal skin that occur in a manner worrisome for melanoma, and continue to have regular professional skin surveillance throughout their lives.

Whether children from melanoma-prone families should undergo genetic testing is complicated. In a position paper published in 2015, the American Society of Human Genetics recommended use of predictive genetic testing in minors only for conditions in which a clinical intervention could be delivered in childhood and for which benefits of early intervention outweighed potential psychological harms.25,26 Stump et al26 recently investigated whether genetic counseling and test reporting for CDKN2A mutations improved sun protection without inducing distress. In a very small initial study of nine carriers and nine noncarriers, the authors reported significantly fewer sunburns and a greater proportion reporting sun protection adherence between baseline and one-year post-disclosure. The results, however, did not differ by mutation status suggesting that the education provided during the counseling session may have itself contributed to the changes in behavior. Although the authors suggested that this small initial study provided support for the clinical utility of genetic testing and counseling for melanoma risk among minors from families with known familial predisposition mutations, they also noted the importance of confirming the findings in a much larger sample.26 Although mutation positive members of CDKN2A+ families have increased risks for melanoma, we recently reported that among American and Swedish CDKN2A+ families, members who tested negative for their family’s mutation remained at moderately increased risk for melanoma.27 Therefore, mutation negative family members should also be encouraged to follow sun safety recommendations and practice skin self-exams in addition to being considered for continuing dermatologic surveillance.27 Given that the occurrence of pediatric melanoma in both CDKN2A+ and CDKN2A- families in the current study was significantly increased compared to the general population of US melanoma patients, education and counseling plus skin surveillance for prevention/early detection of melanoma may be the most important proposed strategies for children in melanoma-prone families.

The current study was not population-based and therefore might be prone to referral bias. It was also limited by relatively small number of patients with melanoma diagnosed before 20 years. Given the relatively small number of pediatric melanoma patients, analyses comparing melanoma patients in families with and without CDKN2A mutations were conducted assuming independence of these patients. Results were similar, however, with (patient-based) or without (family-based) this independence assumption. To reduce any effects of bias from family referral, selection, or correlation, we conducted a sensitivity analysis from which the melanoma patients who led to ascertainment of each family, i.e. the probands, were excluded from the analyses. The sensitivity analyses showed results consistent with the full analyses suggesting that ascertainment did not bias the results. Since this study was a non-population-based family study, it was not possible to examine the prevalence of CDKN2A mutations in pediatric melanoma patients from the general population. Large population-based studies would be required for such an evaluation.

The occurrence of pediatric melanoma is significantly increased in melanoma-prone families, with and without CDKN2A mutations, compared to the general population of melanoma patients in the United States. These observations have implications for prevention of melanoma as well as clinical care for early detection of melanoma. Children in melanoma-prone families should have careful sun protection from an early age and skin surveillance to reduce their risk of melanoma.

Acknowledgements:

We are indebted to the participating families, whose generosity and cooperation have made this study possible. We acknowledge the contributions to this work that were made by Virginia Pichler, Deborah Zametkin, and Laura Fontaine.

Funding: This research was supported entirely by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Division of Cancer Epidemiology and Genetics.

Footnotes

Author Contributions: Alisa Goldstein: Conceptualization, funding acquisition, formal analysis, project administration, supervision, writing – original draft, writing – review and editing. Kelsey Stidd: formal analysis, data curation, writing – original draft, writing – review and editing. Xiaohong R. Yang: funding acquisition, methodology, writing – review and editing. Mary C. Fraser: data curation, project administration, writing – review and editing. Margaret Tucker: Conceptualization, funding acquisition, data curation, project administration, writing – review and editing.

Conflict of Interest: The authors have no conflicts of interest to disclose.

References

  • 1.Hill VK, Gartner JJ, Samuels Y, Goldstein AM. The genetics of melanoma: recent advances. Annu Rev Genomics Hum Genet . 2013;14:257–279. [DOI] [PubMed] [Google Scholar]
  • 2.Demenais F, Mohamdi H, Chaudru V, et al. Association of MC1R variants and host phenotypes with melanoma risk in CDKN2A mutation carriers: a GenoMEL study. J Natl Cancer Inst . 2010;102(20):1568–1583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Goldstein AM, Tucker MA. Genetic epidemiology of cutaneous melanoma: a global perspective. Arch Dermatol . 2001;137(11):1493–1496. [DOI] [PubMed] [Google Scholar]
  • 4.Read J, Wadt KA, Hayward NK. Melanoma genetics. J Med Genet . 2016;53(1):1–14. [DOI] [PubMed] [Google Scholar]
  • 5.Howlader N, Noone AM, Krapcho M, et al. SEER Cancer Statistics Review, 1975–2014, National Cancer Institute. Bethesda, MD https://seer.cancer.gov/csr/1975_2014/, based on November 2016 SEER data submission, posted to the SEER web site, April 2017. [Google Scholar]
  • 6.Barnhill RL, Roush GC, Titus-Ernstoff L, Ernstoff MS, Duray PH, Kirkwood JM. Comparison of nonfamilial and familial melanoma. Dermatology . 1992;184(1):2–7. [DOI] [PubMed] [Google Scholar]
  • 7.Kopf AW, Hellman LJ, Rogers GS, et al. Familial malignant melanoma. JAMA . 1986;256(14):1915–1919. [PubMed] [Google Scholar]
  • 8.Buckel TB, Goldstein AM, Fraser MC, Rogers B, Tucker MA. Recent tanning bed use: a risk factor for melanoma. Arch Dermatol . 2006;142(4):485–488. [DOI] [PubMed] [Google Scholar]
  • 9.Barnhill RL. Childhood Melanoma In: LeBoit PE, Burg G, Weedon D, Sarasin A, eds. WHO Classification of Tumors: Pathology & Genetics of Skin Tumours. Lyon: IARC Press; 2006:84–85. [Google Scholar]
  • 10.In: Curtis REFD, Ron E, Ries LAG, Hacker DG, Edwards BK, Tucker MA, Fraumeni JF Jr., ed. New Malignancies Among Cancer Survivors: SEER Cancer Registries, 1973–2000 . Bethesda, MD: National Cancer Institute, NIH Publ No. 05–5302; 2006:339–362. [Google Scholar]
  • 11.Purdue MP, Freeman LE, Anderson WF, Tucker MA. Recent trends in incidence of cutaneous melanoma among US Caucasian young adults. J Invest Dermatol . 2008;128(12):2905–2908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Wong JR, Harris JK, Rodriguez-Galindo C, Johnson KJ. Incidence of childhood and adolescent melanoma in the United States: 1973–2009. Pediatrics . 2013;131(5):846–854. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Rabbie R, Rashid M, Arance AM, et al. Genomic analysis and clinical management of adolescent cutaneous melanoma. Pigment Cell Melanoma Res . 2017;30(3):307–316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Berwick M, Orlow I, Hummer AJ, et al. The prevalence of CDKN2A germ-line mutations and relative risk for cutaneous malignant melanoma: an international population-based study. Cancer Epidemiol Biomarkers Prev . 2006;15(8):1520–1525. [DOI] [PubMed] [Google Scholar]
  • 15.Goldstein AM, Chan M, Harland M, et al. High-risk melanoma susceptibility genes and pancreatic cancer, neural system tumors, and uveal melanoma across GenoMEL. Cancer Res . 2006;66(20):9818–9828. [DOI] [PubMed] [Google Scholar]
  • 16.Harland M, Cust AE, Badenas C, et al. Prevalence and predictors of germline CDKN2A mutations for melanoma cases from Australia, Spain and the United Kingdom. Hered Cancer Clin Pract . 2014;12(1):20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Monzon J, Liu L, Brill H, et al. CDKN2A mutations in multiple primary melanomas. N Engl J Med . 1998;338(13):879–887. [DOI] [PubMed] [Google Scholar]
  • 18.Orlow I, Begg CB, Cotignola J, et al. CDKN2A germline mutations in individuals with cutaneous malignant melanoma. J Invest Dermatol . 2007;127(5):1234–1243. [DOI] [PubMed] [Google Scholar]
  • 19.Berg P, Wennberg A-M, Tuominen R, et al. Germline CDKN2A mutations are rare in child and adolescent cutaneous melanoma. Melanoma Res . 2004;14(4):251–255. [DOI] [PubMed] [Google Scholar]
  • 20.ClinVar. https://www.ncbi.nlm.nih.gov/clinvar/variation/127523/#clinical-assertions Accessed November 4, 2017.
  • 21.Whiffin N, Minikel E, Walsh R, et al. Using high-resolution variant frequencies to empower clinical genome interpretation. Genetics in medicine : official journal of the American College of Medical Genetics . 2017;19(10):1151–1158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Lek M, Karczewski KJ, Minikel EV, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature . 2016;536(7616):285–291. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Novakovic B, Clark WH Jr., Fears TR, Fraser MC, Tucker MA. Melanocytic nevi, dysplastic nevi, and malignant melanoma in children from melanoma-prone families. J Am Acad Dermatol . 1995;33(4):631–636. [DOI] [PubMed] [Google Scholar]
  • 24.Tucker MA, Greene MH, Clark WH Jr., Kraemer KH, Fraser MC, Elder DE. Dysplastic nevi on the scalp of prepubertal children from melanoma-prone families. J Pediatr . 1983;103(1):65–69. [DOI] [PubMed] [Google Scholar]
  • 25.Botkin JR, Belmont JW, Berg JS, et al. Points to consider: Ethical, legal, and psychosocial implications of genetic testing in children and adolescents. Am J Hum Genet . 2015;97(1):6–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Stump TK, Aspinwall LG, Kohlmann W, et al. Genetic test reporting and counseling for melanoma risk in minors may improve sun protection without inducing distress. J Genet Couns . 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Helgadottir H, Olsson H, Tucker MA, Yang XR, Hoiom V, Goldstein AM. Phenocopies in melanoma-prone families with germline CDKN2A mutations. Genetics in medicine : official journal of the American College of Medical Genetics . [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES