Skip to content

Germline predisposition to melanoma

TL;DR — About 10% of people with melanoma report an affected relative, and mutations in the seven established high-penetrance genes (CDKN2A, CDK4, BAP1, POT1, ACD, TERF2IP, TERT promoter) explain roughly half of high-density melanoma families, leaving the remainder unexplained and probably polygenic (Read 2016, PMID 26337759). CDKN2A is the dominant gene, accounting for ~40% of familial melanoma in the sequencing series that discovered POT1 (Robles-Espinoza 2014, PMID 24686849), and its clinical importance extends beyond skin: prospective Swedish follow-up of p.Arg112dup carriers found pancreatic cancer RR 43.8 (95% CI 13.8–139.0), upper-digestive 17.1 (6.3–46.5) and respiratory 15.6 (5.4–46.0), concentrated in ever-smokers (Helgadottir 2014, PMID 24935963). A genome-first analysis of 696,665 people found pathogenic variants in the eight major genes at 0.5–0.9% population prevalence, exceeding the 2.5% testing threshold only among people with multiple primaries or a first melanoma before 40 (Goldstein 2026, PMID 42201696). Germline testing changes management mainly through non-cutaneous surveillance — 20-year prospective pancreatic surveillance in 347 CDKN2A carriers found 20.7% cumulative PDAC incidence by age 70 with 83.3% of tumours resectable and 5-year survival 32.4% (Klatte 2022, PMID 35658523) — not through melanoma outcomes, for which no randomised evidence exists.

The gene set

Gene Mechanism Core cancer spectrum Anchor evidence
CDKN2A (p16INK4a / p14ARF) Cell-cycle checkpoint loss Cutaneous melanoma, pancreatic adenocarcinoma, tobacco-related cancers, brain, head and neck PMID 24686849; PMID 24935963; PMID 42201696
CDK4 Gain-of-function at the p16-binding site Cutaneous melanoma PMID 26337759
BAP1 Tumour-suppressor deubiquitinase Uveal melanoma, mesothelioma, cutaneous melanoma, renal cell carcinoma, BAP1-inactivated melanocytic tumours Walpole 2018, PMID 30517737
POT1 Shelterin component; telomere elongation Cutaneous melanoma, glioma, haematologic, thyroid PMID 24686849; PMID 42201696
ACD, TERF2IP Shelterin components Cutaneous melanoma PMID 26337759
TERT promoter Telomerase up-regulation Familial melanoma (rare) PMID 26337759
MITF p.E318K SUMOylation-defective transcription factor Cutaneous melanoma, renal cell carcinoma Bertolotto 2011, PMID 22012259
MC1R Pigmentation receptor; moderate-penetrance modifier Cutaneous melanoma Raimondi 2008, PMID 18366057

Prevalence and testing threshold. In two genomically ascertained cohorts — 227,286 Geisinger MyCode and 469,379 UK Biobank participants — the combined prevalence of pathogenic variants across ACD, BAP1, CDKN2A, CDK4, MITF E318K, POT1, TERF2IP and the TERT promoter was 0.5% (MyCode) to 0.9% (UK Biobank). Prevalence exceeded the 2.5% threshold conventionally used to justify germline testing only among people with multiple cutaneous melanomas or a first melanoma before age 40. Case-control analysis replicated CDKN2A (brain, cutaneous melanoma, head and neck, pancreas), MITF E318K (melanoma, kidney) and POT1 (melanoma, haematologic, thyroid), and added inconsistently reported associations: BAP1–prostate, CDKN2A–biliary tract, breast, non-melanoma skin and small intestine, MITF E318K–cervix and nasal cavity, POT1–myeloma. Both cohorts are >93% European-ancestry, which bounds transportability (Goldstein 2026, PMID 42201696).

CDKN2A: penetrance, modifiers and the extracutaneous spectrum

  • Share of familial melanoma. Deleterious germline CDKN2A variants account for around 40% of familial melanoma (PMID 24686849); across melanoma-prone families the range is quoted as 5–40% depending on ascertainment (Taylor 2019, PMID 30731170).
  • Non-melanoma cancer risk, prospectively. In 117 mutation-positive and 136 mutation-negative members of 15 melanoma families followed 4–26 years, carriers had significantly increased risk of all cancers combined (Obs/Exp 2.2, 95% CI 1.1–3.8), driven by digestive-system and particularly pancreatic tumours; non-carriers showed no excess (Goldstein 2004, PMID 15173226).
  • The Swedish founder cohort. Among 120 p.Arg112dup carriers, 111 non-carriers, 275 first-degree and 321 second-degree relatives against 3,976 controls, carriers had prospective RR 5.0 (3.7–7.3) for non-melanoma cancer, 43.8 (13.8–139.0) for pancreatic cancer, 17.1 (6.3–46.5) for upper-digestive and 15.6 (5.4–46.0) for respiratory cancers. Ever-smoking carriers versus never-smoking carriers had OR 9.3 (1.9–44.7) for those cancers combined, supporting increased sensitivity to tobacco carcinogens — the clearest gene–environment interaction in this literature (PMID 24935963).
  • Genotype–phenotype within the gene. Across 172 Dutch families carrying 15 pathogenic variants (649 carriers), pancreatic cancer occurred in 58% of the 163 families whose variant affected p16INK4a, and in none of the 9 families (20 carriers) whose variant affected only p14ARF. For the c.67G>C variant the pancreatic SIR was 19.1 (8.3–33.6) and cumulative incidence to age 75 was 19% (7.5–30.1) (Overbeek 2021, PMID 32482799). This is a directly actionable distinction: p14ARF-only variants have no clinical evidence of raised pancreatic risk.
  • Penetrance modifiers. In 815 CDKN2A carriers from 186 GenoMEL families across three continents, carrying any of the four commonest MC1R variants raised melanoma risk (1.24 × 10⁻⁶ ≤ P ≤ .0007); ≥2 MC1R variants conferred 2.6-fold higher risk than one (OR 5.83, 3.60–9.46 vs 2.25, 1.44–3.52; Ptrend = 1.86 × 10⁻⁸). Hair colour and nevus count contributed jointly (Demenais 2010, PMID 20876876). Penetrance is therefore not a property of CDKN2A alone.
  • Predicting carrier status. MELPREDICT performed well in 2,116 GenoMEL familial cases (AUC 0.752, 0.730–0.775); adding a family history of pancreatic cancer improved discrimination (AUC 0.772, 0.750–0.793; NRI 0.40, P < .0001) whereas adding phenotypic risk factors did not (PMID 30731170).

BAP1 tumour predisposition syndrome

Collation of 181 BAP1 germline-variant-positive families worldwide (106 published, 75 unpublished) carrying 140 unique variants confirmed the core spectrum — uveal melanoma, mesothelioma, cutaneous melanoma, renal cell carcinoma and BAP1-inactivated melanocytic tumours — and showed that some missense-variant families exhibit the full phenotype. Median age of onset was lower in null than missense carriers for all tumours combined (P < .001), for mesothelioma, cutaneous melanoma and non-melanoma skin cancer (all P < .001) (Walpole 2018, PMID 30517737). BAP1 is the one predisposition gene that spans both intraocular and cutaneous melanoma, and the somatic side of the same gene defines the high-metastatic-risk class of sporadic uveal melanoma — see uveal melanoma.

Telomere-maintenance genes

POT1 was found by sequencing 184 melanoma cases from 105 pedigrees in the UK, Netherlands and Australia that were negative for known predisposition genes. Loss-of-function variants either disrupted POT1 mRNA splicing or altered conserved OB-domain residues, in both cases disrupting protein–telomere binding and increasing telomere length; carrier families showed early onset and multiple primaries (Robles-Espinoza 2014, PMID 24686849). Longer telomeres predisposing to melanoma inverts the usual telomere-shortening-drives-cancer intuition and links POT1, ACD, TERF2IP and TERT into one mechanistic class (PMID 26337759). Population-based and national testing series continue to refine which POT1 variants are pathogenic (Papadakis 2025, PMID 40851494; Simonin-Wilmer 2023, PMID 36539277).

MITF p.E318K

The p.E318K substitution sits in a SUMO consensus site; it impairs MITF SUMOylation, enhances binding to the HIF1A promoter and increases transcriptional activity, with globally increased locus occupancy and gain-of-function effects on melanocytic and renal-cell clonogenicity, migration and invasion. Carriers had more than fivefold increased risk of melanoma, renal cell carcinoma or both (Bertolotto 2011, PMID 22012259). Whether risk extends further is contested: a systematic review plus whole-exome analysis across TCGA and enriched cohorts found minimal evidence for non-melanoma cancer risk among people with low inherited melanoma risk (OR 1.168, 0.78–1.74, P = .454), with an unexpected association for uterine carcinosarcoma (OR 9.24, 2.08–37.17, P = .024) (Guhan 2020, PMID 33051548). The genome-first cohorts replicated the melanoma and kidney associations and added cervix, nasal cavity/middle ear and non-melanoma skin (PMID 42201696).

Common-variant architecture

GWAS meta-analysis of 36,760 melanoma cases and 375,188 controls identified 54 genome-wide significant loci with 68 independent SNPs; combining with GWAS of nevus count and hair colour and with transcriptome-association methods raised the total to 85 susceptibility loci, reinforcing nevogenesis, pigmentation and telomere maintenance as the three axes. Acral melanoma was uniquely unrelated to pigmentation in the regional and host-factor analyses (Landi 2020, PMID 32341527) — the germline counterpart of the somatic divergence described in acral and mucosal melanoma.

A separate GWAS meta-analysis across UK Biobank, FinnGen, QSkin and Q-MEGA compared in situ with invasive melanoma: 6 significant loci for in situ, 18 for invasive, genetic correlation r = 0.96 (0.76–1.15), and SNP heritability 6.7% (4.1–9.3) versus 4.9% (2.8–7.2). Loci near IRF4, KLF4 and HULC had larger effects for in situ disease and MC1R a larger effect for invasive disease; a case-case polygenic risk score was higher in invasive cases (OR per SD 1.43, 1.16–1.77) (Ingold 2024, PMID 39141363). The authors' framing is important: a PRS that separates invasive from in situ risk is exactly the instrument that could stratify screening without aggravating overdiagnosis — the open problem in screening and overdiagnosis.

Syndromes, phenotypes and negative findings

  • FAMMM. The familial atypical multiple mole melanoma phenotype is the clinical face of much CDKN2A carriage: roughly 5–10% of cutaneous melanoma occurs in kindreds with hereditary predisposition, and CDKN2A mutations are found in approximately 20–40% of those kindreds. Lynch and Fusaro described the melanoma–pancreatic cancer association in 1991, and the review notes bluntly that adequate pancreatic screening did not exist at the time of writing (Lynch 2016, PMID 26892865). The dysplastic-nevus phenotype that defines FAMMM clinically has itself been contested since the late 1970s, with unresolved nomenclature and diagnostic uncertainty (Drozdowski 2023, PMID 36038073).
  • Xeroderma pigmentosum. Autosomal-recessive nucleotide-excision-repair deficiency produces extreme UV sensitivity and early skin cancer including melanoma (Kraemer 1993/updated, PMID 20301571). At the common-variant level, genotyping 94 SNPs across seven XP genes in 714 unselected melanoma patients and 1,841 controls found only XPC and XPD associated with melanoma susceptibility, and for XPC rs2228000 the direction was protective (CT OR 0.15, TT OR 0.11, both P < .001) (Paszkowska-Szczur 2013, PMID 23436679).
  • A negative result worth recording. Whole-genome or exome sequencing of 160 cutaneous and/or uveal melanoma families from four countries (307 individuals) found two truncating BRCA1 mutations co-segregating with cutaneous melanoma, but 33 further rare missense variants were all benign or of unknown consequence, and in a sporadic series of 763 melanomas the one deleterious BRCA1 allele was lost in the tumour. The authors conclude the evidence is insufficient to attribute familial melanoma susceptibility to BRCA1, and that the previously described BRCA2–uveal melanoma association is a rare source of increased risk (Johansson 2019, PMID 31464824). Absence claims of this kind are re-checked at each sweep.
  • Panel testing in uveal melanoma. Among 70 uveal melanoma patients meeting NCCN testing criteria, 13% (8/69) of unrelated individuals carried a pathogenic or likely pathogenic variant (BAP1, BRCA1, BRCA2, MBD4, MUTYH, POT1, XRCC2); eight would have been missed by BAP1-only testing as recommended by ASCO in 2024, and no association was seen with tumour size or stage (Byrne 2026, PMID 42415513).

The retinoblastoma border

Survivors of heritable retinoblastoma carry increased melanoma risk; survivors of non-heritable retinoblastoma do not. The distinction is load-bearing and "retinoblastoma survivors" must never be written as a single risk group.

Measure Heritable Non-heritable Source
Any subsequent malignant neoplasm, SIR 11.9 (95% CI 10.4–13.5), n = 239 0.8 (0.5–1.2), n = 25 Schonfeld 2021, PMID 33473166
50-year cumulative incidence, first SMN 33.1% (29.0–37.2) — PMID 33473166
SIRs for melanoma, CNS, oral cavity and breast 3.1–17 as a range across those four cancer types, not a melanoma-specific point estimate not increased PMID 33473166
Melanoma cases observed (1,851 White survivors, 1914–2006, followed to 2016) 33 of 1,020 7 of 831 Kleinerman 2021, PMID 34153328
50-year cumulative incidence, melanoma 4.5% 0.7% PMID 34153328
50-year cumulative incidence, non-melanoma skin cancer 3.7% 1.5% PMID 34153328
Median age at skin-cancer diagnosis ~20 years younger than non-heritable reference PMID 34153328

The Kleinerman cohort supplies the melanoma-specific figures that Schonfeld's abstract does not: 4.5% versus 0.7% cumulative incidence at 50 years, with melanomas distributed in patterns resembling melanoma-prone families rather than the head-and-neck predominance of the non-melanoma skin cancers, and with sun sensitivity and phenotype not varying by skin-cancer status — pointing to genetic rather than exposure mechanism (PMID 34153328). RB1 is not among the eight familial melanoma genes tested in the genome-first analysis (PMID 42201696), so RB1 carriers fall outside standard melanoma germline panels by construction. Retinoblastoma's own evidence lives at retinoblastoma second cancers and survivorship and is not restated here.

What testing changes

Domain Evidence that testing changes outcome
Pancreatic surveillance in CDKN2A carriers 347 carriers, median 5.6 years follow-up: 36 PDACs in 31 people (8.9%), cumulative incidence 20.7% by age 70; 83.3% resectable at imaging, 33.3% stage I; median survival 26.8 months, 5-year survival 32.4% (19.1–54.8), and 44.1% (27.2–71.3) among the 71.0% resected. Nine of 347 (2.6%) had surgery for a lesion that proved not to be PDAC (Klatte 2022, PMID 35658523)
Dermatologic surveillance in carriers 62 relatives of index cases (48 CDKN2A, 14 MITF p.E318K) followed a mean 60.9 months: 5 of 39 CDKN2A carriers with follow-up developed a new cancer; none of 12 cancer-free MITF carriers developed melanoma over a mean 24.6 months (Gironi 2025, PMID 40623218)
Melanoma mortality No randomised evidence that germline-directed skin surveillance reduces melanoma mortality

The asymmetry is the point: the strongest outcome evidence for melanoma germline testing concerns a non-melanoma cancer. Pancreatic surveillance in CDKN2A carriers has 20 years of prospective data with a stage shift and a survival figure far above unselected PDAC; melanoma surveillance in the same carriers has cohort description only. National recommendation documents for genetic testing in melanoma exist (Avril 2015, PMID 25600792) and behavioural and surveillance considerations for carriers have been reviewed (Pauley 2022, PMID 35372037), but the guideline synthesis lives in guidelines.

Does surveillance of carriers change outcomes?

Sweden has the longest-running answer. Melanoma-prone families identified through a nationwide preventive programme beginning in 1987 were followed through the Swedish Cancer Registry and Cause of Death Registry, and melanoma-specific survival compared across four cohorts: CDKN2A mutation carriers with a first invasive melanoma before (n = 53) or after (n = 43) programme inclusion, and CDKN2A wild-type cases before (n = 255) or after (n = 122). Carriers were diagnosed with their first invasive melanoma at significantly younger ages (38 and 42 years) than wild-type cases (48 and 57 years) (Pissa 2023, PMID 36156317). This is the only dataset located in this build that compares melanoma-specific survival before and after entry into a familial surveillance programme.

Prospective dermatological surveillance of healthy CDKN2A and MITF p.E318K carriers has also been described: among 62 relatives of index cases (48 CDKN2A, 14 MITF) followed a mean 60.9 months, 5 of 39 CDKN2A carriers with follow-up developed a new cancer, additional cutaneous melanoma and pancreatic cancer occurred in 43.75% and 21.87% of CDKN2A families respectively, and none of 12 cancer-free MITF carriers developed melanoma over a mean 24.6 months (Gironi 2025, PMID 40623218).

On MITF p.E318K prevalence, a hospital-based case-control study genotyped 531 patients — 271 with multiple primary melanoma and wild-type p16INK4A, 191 probands from melanoma-prone families with a single melanoma and wild-type p16INK4A, and 69 probands from families carrying p16INK4A-affecting CDKN2A mutations — against 499 age- and sex-matched cancer-free controls, recruited 1992–2014 (Potrony 2016, PMID 26650189). The variant is found independently of CDKN2A status, which is why panel rather than single-gene testing is the practical approach (see also PMID 42415513).

Interpretation rules for this page

  • Penetrance figures are ascertainment-dependent. The 40% CDKN2A share comes from high-density melanoma pedigrees (PMID 24686849); the 0.5–0.9% prevalence comes from unselected genomic cohorts (PMID 42201696). Neither number transports to the other setting.
  • Name the variant, not just the gene. Within CDKN2A, p16INK4a-affecting and p14ARF-only variants differ in pancreatic risk to the point of clinical divergence (PMID 32482799), and within MC1R, variants that raise melanoma risk are not the same set that produce red hair (PMID 18366057).
  • Penetrance is modified, not fixed. MC1R genotype, hair colour and nevus count jointly shift melanoma risk within CDKN2A carriers (PMID 20876876), and smoking shifts extracutaneous risk (PMID 24935963).
  • Distinguish "gene associated with cancer X" from "carriers should be surveilled for X". Only pancreatic surveillance in CDKN2A carriers has prospective outcome data (PMID 35658523).
  • Cohorts here are overwhelmingly of European ancestry (>93% in both genome-first cohorts, PMID 42201696; the retinoblastoma skin-cancer cohort is explicitly White, PMID 34153328). Risk estimates for other ancestries are not established.
  • Heritable and non-heritable retinoblastoma are different exposures. Writing "retinoblastoma survivors" as one group misstates a 4.5%-versus-0.7% difference in 50-year melanoma incidence (PMID 34153328).

Open questions

  • What explains the ~50% of high-density melanoma families with no identified high-penetrance variant — additional rare genes, or polygenic burden (PMID 26337759; PMID 32341527)?
  • Should the 2.5% testing threshold be revised now that genome-first prevalence is known to be 0.5–0.9% overall but above threshold in multiple-primary and early-onset melanoma (PMID 42201696)?
  • Are the newly observed associations (BAP1–prostate, CDKN2A–biliary/breast/small intestine, POT1–myeloma) real, and do they change surveillance (PMID 42201696)?
  • Does a case-case polygenic score separating invasive from in situ melanoma risk improve screening yield without increasing overdiagnosis (PMID 39141363)?
  • Are RB1-carrier retinoblastoma survivors reached by melanoma surveillance pathways built around CDKN2A, MC1R and BAP1 (PMID 34153328; PMID 42201696)?
  • Does the CDKN2A–tobacco interaction generalise beyond the Swedish founder variant, and should smoking cessation be a formal counselling endpoint for carriers (PMID 24935963)?
  • Why is acral melanoma genetically unrelated to pigmentation loci while sharing the melanoma diagnosis (PMID 32341527)?

References

  1. Read J, et al. Melanoma genetics. Journal of medical genetics. 2016;53:1-14. PMID 26337759
  2. Robles-Espinoza CD, et al. POT1 loss-of-function variants predispose to familial melanoma. Nature genetics. 2014;46:478-481. PMID 24686849
  3. Helgadottir H, et al. High risk of tobacco-related cancers in CDKN2A mutation-positive melanoma families. Journal of medical genetics. 2014;51:545-52. PMID 24935963
  4. Goldstein AM, et al. Prevalence of Familial Melanoma Genes and Cancer Risk Among Genomically Ascertained Individuals. JAMA dermatology. 2026;162:692-700. PMID 42201696
  5. Klatte DCF, et al. Pancreatic Cancer Surveillance in Carriers of a Germline CDKN2A Pathogenic Variant: Yield and Outcomes of a 20-Year Prospective Follow-Up. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2022;40:3267-3277. PMID 35658523
  6. Walpole S, et al. Comprehensive Study of the Clinical Phenotype of Germline BAP1 Variant-Carrying Families Worldwide. Journal of the National Cancer Institute. 2018;110:1328-1341. PMID 30517737
  7. Bertolotto C, et al. A SUMOylation-defective MITF germline mutation predisposes to melanoma and renal carcinoma. Nature. 2011;480:94-8. PMID 22012259
  8. Raimondi S, et al. MC1R variants, melanoma and red hair color phenotype: a meta-analysis. International journal of cancer. 2008;122:2753-60. PMID 18366057
  9. Taylor NJ, et al. Estimating CDKN2A mutation carrier probability among global familial melanoma cases using GenoMELPREDICT. Journal of the American Academy of Dermatology. 2019;81:386-394. PMID 30731170
  10. Goldstein AM, et al. Prospective risk of cancer in CDKN2A germline mutation carriers. Journal of medical genetics. 2004;41:421-4. PMID 15173226
  11. Overbeek KA, et al. Genotype-phenotype correlations for pancreatic cancer risk in Dutch melanoma families with pathogenic CDKN2A variants. Journal of medical genetics. 2021;58:264-269. PMID 32482799
  12. Demenais F, et al. Association of MC1R variants and host phenotypes with melanoma risk in CDKN2A mutation carriers: a GenoMEL study. Journal of the National Cancer Institute. 2010;102:1568-83. PMID 20876876
  13. Papadakis K, et al. POT1 genetic testing in melanoma-prone families in Sweden: germline variant prevalence and tumor spectrum in identified carriers. Acta oncologica (Stockholm, Sweden). 2025;64:1102-1108. PMID 40851494
  14. Simonin-Wilmer I, et al. Population-based analysis of POT1 variants in a cutaneous melanoma case-control cohort. Journal of medical genetics. 2023;60:692-696. PMID 36539277
  15. Guhan SM, et al. Cancer risks associated with the germline MITF(E318K) variant. Scientific reports. 2020;10:17051. PMID 33051548
  16. Landi MT, et al. Genome-wide association meta-analyses combining multiple risk phenotypes provide insights into the genetic architecture of cutaneous melanoma susceptibility. Nature genetics. 2020;52:494-504. PMID 32341527
  17. Ingold N, et al. Exploring the Germline Genetics of In Situ and Invasive Cutaneous Melanoma: A Genome-Wide Association Study Meta-Analysis. JAMA dermatology. 2024;160:964-971. PMID 39141363
  18. Lynch HT, et al. Familial atypical multiple mole melanoma (FAMMM) syndrome: history, genetics, and heterogeneity. Familial cancer. 2016;15:487-91. PMID 26892865
  19. Drozdowski R, et al. Dysplastic nevus part I: Historical perspective, classification, and epidemiology. Journal of the American Academy of Dermatology. 2023;88:1-10. PMID 36038073
  20. Kraemer KH, et al. Xeroderma Pigmentosum. GeneReviews®. 2003 Jun 20 [updated 2022 Mar 24]. PMID 20301571
  21. Paszkowska-Szczur K, et al. Xeroderma pigmentosum genes and melanoma risk. International journal of cancer. 2013;133:1094-100. PMID 23436679
  22. Johansson PA, et al. Evaluation of the contribution of germline variants in BRCA1 and BRCA2 to uveal and cutaneous melanoma. Melanoma research. 2019;29:483-490. PMID 31464824
  23. Byrne L, et al. Impact of Multigene Panel Testing in High-Risk Uveal Melanoma Patients. Pigment cell & melanoma research. 2026;39:e70108. PMID 42415513
  24. Schonfeld SJ, et al. Long-term risk of subsequent cancer incidence among hereditary and nonhereditary retinoblastoma survivors. British journal of cancer. 2021;124:1312-1319. PMID 33473166
  25. Kleinerman RA, et al. Increased Risk of Skin Cancer in 1,851 Long-Term Retinoblastoma Survivors. The Journal of investigative dermatology. 2021;141:2849-2857.e3. PMID 34153328
  26. Gironi LC, et al. Dermatologic surveillance in healthy carriers of CDKN2A and p.E318K MITF germline variants from melanoma-prone families: a 14 years hospital-based experience. Melanoma research. 2025;35:424-428. PMID 40623218
  27. Avril MF, et al. [Recommendations for genetic testing and management of individuals genetically at-risk of cutaneous melanoma]. Annales de dermatologie et de venereologie. 2015;142:26-36. PMID 25600792
  28. Pauley K, et al. Considerations for Germline Testing in Melanoma: Updates in Behavioral Change and Pancreatic Surveillance for Carriers of CDKN2A Pathogenic Variants. Frontiers in oncology. 2022;12:837057. PMID 35372037
  29. Pissa M, et al. Melanoma-specific survival before and after inclusion in a familial melanoma dermatologic surveillance program in CDKN2A mutation carriers and non-carriers. Journal of the European Academy of Dermatology and Venereology : JEADV. 2023;37:284-292. PMID 36156317
  30. Potrony M, et al. Prevalence of MITF p.E318K in Patients With Melanoma Independent of the Presence of CDKN2A Causative Mutations. JAMA dermatology. 2016;152:405-12. PMID 26650189