High prevalence of the G101W germline mutation in the CDKN2A (P16(ink4a)) gene in 62 Italian malignant melanoma families.

CDKN2A germline mutation frequency estimates are commonly based on families with several melanoma cases. When we started counseling in a research setting on gene susceptibility analysis in northern and central Italy, however, we mostly found small families with few cases. Here we briefly characterize those kindred, estimate CDKN2A/CDK4 mutation test yields, and provide indications on the possibility of implementing formal DNA testing for melanoma-prone families in Italy. In September 1995 we started genetic counseling in a research setting at our Medical Genetics Center. Screening for CDKN2A/CDK4 mutations was performed on families with two melanoma patients, one of whom was younger than 50 years at onset, the other complying with one of the following: 1) being a first-degree relative, 2) having an additional relative with pancreatic cancer, or 3) having multiple primary melanomas. Sixty-two of 67 (80%) melanoma cases met our criteria. Four previously described CDKN2A mutations (G101W, R24P, V126D, and N71S) were found in 21 of the 62 families (34%) with a high prevalence of G101W (18/21). The percentage of families with two melanoma cases/family harboring a mutation was low (7%, 2/27), but rose to 45% (9/20) if one of the melanoma patients carried multiple melanomas or if pancreatic cancer was present in that family. In the 15 families with three melanoma cases the presence of a mutation was higher (67%, 10/15) and reached 100% in the 4 families with four or more melanoma cases. Our results suggest that CDKN2A/CDK4 counseling-based mutational analysis may be reasonably efficient also for families with two melanoma cases, if one patient carries multiple melanomas or if pancreatic cancer is present in the family.     

Prevalence of CDKN2A mutations in pancreatic cancer patients: implications for genetic counseling

Germline mutations in CDKN2A have been reported in pancreatic cancer families, but genetic counseling for pancreatic cancer risk has been limited by lack of information on CDKN2A mutation carriers outside of selected pancreatic or melanoma kindreds. Lymphocyte DNA from consecutive, unselected white non-Hispanic patients with pancreatic adenocarcinoma was used to sequence CDKN2A. Frequencies of mutations that alter the coding of p16INK4 or p14ARF were quantified overall and in subgroups. Penetrance and likelihood of carrying mutations by family history were estimated. Among 1537 cases, 9 (0.6%) carried germline mutations in CDKN2A, including three previously unreported mutations. CDKN2A mutation carriers were more likely to have a family history of pancreatic cancer (P=0.003) or melanoma (P=0.03), and a personal history of melanoma (P=0.01). Among cases who reported having a first-degree relative with pancreatic cancer or melanoma, the carrier proportions were 3.3 and 5.3%, respectively. Penetrance for mutation carriers by age 80 was calculated to be 58% for pancreatic cancer (95% confidence interval (CI) 8–86%), and 39% for melanoma (95% CI 0–80). Among cases who ever smoked cigarettes, the risk for pancreatic cancer was higher for carriers compared with non-carriers (HR 25.8, P=2.1 × 10⁻¹³), but among nonsmokers, this comparison did not reach statistical significance. Germline mutations in CDKN2A among unselected pancreatic cancer patients are uncommon, although notably penetrant, especially among smokers. Carriers of germline mutations of CDKN2A should be counseled to avoid tobacco use to decrease risk of pancreatic cancer in addition to taking measures to decrease melanoma risk.     

CDKN2A (P16(INK4a)) and CDK4 mutation analysis in 131 Australian melanoma probands: effect of family history and multiple primary melanomas.

Mutation analysis of two genes involved in melanoma susceptibility (CDKN2A/p16(INK4a) and CDK4) was undertaken in 131 probands with a family history of melanoma. Screening of all three exons of CDKN2A and exon 2 of CDK4 by single-strand conformation polymorphism (SSCP) analysis and/or direct sequencing identified a total of 10 different CDKN2A germline mutations, including 6 not previously described in the germline. All but one has been previously proven to, or is likely to, affect the structure and function of p16(INK4a). The incidence of CDKN2A mutation was 8.4% (11/131), but was significantly higher in families with three or more cases of melanoma (10/66, 15.1%) than in those in which only two relatives were affected (1/65, 1.5%). The incidence of CDKN2A mutation was also higher in families with three or more cases of melanoma and at least one member with multiple primary melanomas (6/19, 31.6%) than in similar families without multiple primary melanomas (4/47, 8.5%). One novel CDK4 variant of uncertain significance was found in a kindred that also carries a CDKN2A mutation. Genes Chromosomes Cancer 25:339-348, 1999.     

Clinical and genetic analysis of 18 pancreatic carcinoma/melanoma‐prone families

Bartsch DK, Langer P, Habbe N, Matthäi E, Chaloupka B, Sina M, Hahn SA, Slater EP. Clinical and genetic analysis of 18 pancreatic carcinoma/melanoma‐prone families. Families with both melanoma and pancreatic cancer are extremely rare and some are affected with the autosomal dominant inherited familial atypical multiple mole melanoma‐pancreatic cancer (FAMMM‐PC) syndrome. The phenotypic and genotypic expressions of such pancreatic cancer–melanoma prone families are not well defined. The National Case Collection of Familial Pancreatic Cancer of the Deutsche Krebshilfe includes 110 pancreatic cancer families, 18 of which (16%) show an association of pancreatic cancer and melanoma. These 18 families were analysed regarding their phenotype and the prevalence of germline mutations in the candidate genes CDKN2A, BRCA2, CHEK2, NOD2, ARL11 and Palladin (PALLD). There were two types of families: five families with the FAMMM‐PC phenotype and 13 PC/melanoma families without the multiple mole phenotypes (PCMS). The prevalences of PC and melanoma in the two types of families were similar. The prevalence of other tumour types, especially breast carcinoma, was higher (11%) in PCMS‐ than in FAMMM‐PC families (2.4%, p = 0.02). CDKN2A mutations were identified in 2 of 18 (11%) PCMS families. A cosegregating BRCA2 mutation was detected in one PCMS family without breast cancer. None of the reported germline mutations in the NOD2, Palladin, ARL11 or CHEK2 genes were detected in either type of family. In conclusion, families with an accumulation of PC and melanoma show a large variety of phenotypic expression, which is not always consistent with the FAMMM‐PC phenotype. More PC/melanoma‐prone families need to be analysed to clarify whether such families represent variations of the FAMMM‐PC syndrome or two distinct hereditary cancer syndromes.     

Phenocopies in melanoma-prone families with germ-line CDKN2A mutations

PurposeCarriers of CDKN2A mutations have high risks of melanoma and certain other cancers. In this study we examined the occurrence of tumors among CDKN2A wild type (wt) members of melanoma-prone families with CDKN2A mutations.MethodsSwedish and US melanoma-prone families with CDKN2A mutations were included. Data was collected on tumors diagnosed among family members. Among the CDKN2A mutated families, members with CDKN2A wt status who were diagnosed with melanoma were designated phenocopies.ResultsOf patients with melanoma in the CDKN2A mutated families (n = 266), 7.1%, were seen among members with CDKN2A wt status (phenocopy rate). Among the CDKN2A wt family members of the CDKN2A mutated families (n = 256), 7.4% were diagnosed with melanoma. The prospective relative risk for melanomas was significantly higher among the CDKN2A wt subjects compared with population-based controls (7.4 (95% confidence interval 1.7–33.2)), while no elevated risks of nonmelanoma cancers were seen and their offspring did not have significantly elevated risks of melanoma or other cancers.ConclusionMembers of CDKN2A mutation carrying families who test negative for their family’s mutation have moderately increased risk for melanoma and should, in addition to being considered for continuing dermatologic surveillance, be encouraged to follow sun safety recommendations and practice skin self-exams.     

Functional,structural, and genetic evaluation of 20 CDKN2A germ line mutations identified in melanoma‐prone families or patients

Germline mutations of the CDKN2A gene are found in melanoma‐prone families and individuals with multiple sporadic melanomas. The encoded protein, p16^INK4A, comprises four ankyrin‐type repeats, and the mutations, most of which are missense and occur throughout the entire coding region, can disrupt the conformation of these structural motifs as well as the association of p16^INK4a with its physiological targets, the cyclin‐dependent kinases (CDKs) CDK4 and CDK6. Assessing pathogenicity of nonsynonymous mutations is critical to evaluate melanoma risk in carriers. In the current study, we investigate 20 CDKN2A germline mutations whose effects on p16^INK4A structure and function have not been previously documented (Thr18_Ala19dup, Gly23Asp, Arg24Gln, Gly35Ala, Gly35Val, Ala57Val, Ala60Val, Ala60Arg, Leu65dup, Gly67Arg, Gly67_Asn71del, Glu69Gly, Asp74Tyr, Thr77Pro, Arg80Pro, Pro81Thr, Arg87Trp, Leu97Arg, Arg99Pro, and [Leu113Leu;Pro114Ser]). By considering genetic information, the predicted impact of each variant on the protein structure, its ability to interact with CDK4 and impede cell proliferation in experimental settings, we conclude that 18 of the 20 CDKN2A variants can be classed as loss of function mutations, whereas the results for two remain ambiguous. Discriminating between mutant and neutral variants of p16^INK4A not only adds to our understanding of the functionally critical residues in the protein but provides information that can be used for melanoma risk prediction. Hum Mutat 0, 1–11, 2009. © 2009 Wiley‐Liss, Inc.     

Characterization of individuals at high risk of developing melanoma in Latin America: bases for genetic counseling in melanoma

PurposeCDKN2A is the main high-risk melanoma-susceptibility gene, but it has been poorly assessed in Latin America. We sought to analyze CDKN2A and MC1R in patients from Latin America with familial and sporadic multiple primary melanoma (SMP) and compare the data with those for patients from Spain to establish bases for melanoma genetic counseling in Latin America.MethodsCDKN2A and MC1R were sequenced in 186 Latin American patients from Argentina, Brazil, Chile, Mexico, and Uruguay, and in 904 Spanish patients. Clinical and phenotypic data were obtained.ResultsOverall, 24 and 14% of melanoma-prone families in Latin America and Spain, respectively, had mutations in CDKN2A. Latin American families had CDKN2A mutations more frequently (P = 0.014) than Spanish ones. Of patients with SMP, 10% of those from Latin America and 8.5% of those from Spain had mutations in CDKN2A (P = 0.623). The most recurrent CDKN2A mutations were c.-34G>T and p.G101W. Latin American patients had fairer hair (P = 0.016) and skin (P < 0.001) and a higher prevalence of MC1R variants (P = 0.003) compared with Spanish patients.ConclusionThe inclusion criteria for genetic counseling of melanoma in Latin America may be the same criteria used in Spain, as suggested in areas with low to medium incidence, SMP with at least two melanomas, or families with at least two cases among first- or second-degree relatives.     

Population-based prevalence of CDKN2A and CDK4 mutations in patients with multiple primary melanomas

The presence of multiple primary cutaneous melanomas (MPM) has been advocated as guidance to identifying melanoma families. Frequencies of CDKN2A mutations in materials of sporadic MPM cases from pigmented lesion clinics vary between 8 and 15%. Patients with MPM have therefore been regarded as good candidates for CDKN2A mutational screening. We describe a population-based study where all persons in Norway diagnosed with MPM between 1953 and 2004 (n = 738 alive per April 2004) were invited to participate. Three-hundred-and-ninety patients (52.8%) responded confidentially. Mutations in CDKN2A were found in 6.9% of the respondents. Eighty-one MPM patients (20.8%) reported that they belonged to melanoma families, and 17 (21.0%) of these harboured a CDKN2A mutation, compared to 3.2% of the nonfamilial cases. The probability of finding a CDKN2A mutation increased when the patients had three or more melanomas, or a young age of onset of first melanoma. We identified five novel CDKN2A variants (Ala57Gly, Pro81Arg, Ala118Val, Leu130Val, and Arg131Pro) and four that previously have been reported in melanoma families (Glu27X, Met53Ile, Arg87Trp, and Ala127Pro). A large deletion (g.13623_23772del10150) encompassing exon 1alpha and the 5' part of exon 2 was detected in six patients with a family history of melanoma. Three patients, belonging to the same family, had the CDK4 Arg24His mutation. The frequency of CDKN2A mutations was lower than previously reported in other studies, an observation which probably is due to the population-based design of our study.     

Haplotype analysis of two recurrent CDKN2A mutations in 10 melanoma families: Evidence for common founders and independent mutations

Germ-line mutations in CDKN2A have been shown to predispose to cutaneous malignant melanoma. We have identified 2 new melanoma kindreds which carry a duplication of a 24bp repeat present in the 5′ region of CDKN2A previously identified in melanoma families from Australia and the United States. This mutation has now been reported in 5 melanoma families from 3 continents: Europe, North America, and Australasia. The M53I mutation in exon 2 of CDKN2A has also been documented in 5 melanoma families from Australia and North America. The aim of this study was to determine whether the occurrence of the mutations in these families from geographically diverse populations represented mutation hotspots within CDKN2A or were due to common ancestors. Haplotypes of 11 microsatellite markers flanking CDKN2A were constructed in 5 families carrying the M53I mutation and 5 families carrying the 24bp duplication. There were some differences in the segregating haplotypes due primarily to recombinations and mutations within the short tandem-repeat markers; however, the data provide evidence to indicate that there were at least 3 independent 24bp duplication events and possibly only 1 original M53I mutation. This is the first study to date which indicates common founders in melanoma families from different continents. Hum Mutat 11:424–431, 1998. © 1998 Wiley-Liss, Inc.     

Predicting functional significance of cancer‐associated p16INK4a mutations in CDKN2A

Inherited mutations affecting the INK4a/ARF locus (CDKN2A) are associated with melanoma susceptibility in 40% of multiple case melanoma families. Over 60 different germline INK4a/ARF mutations have been detected in more than 190 families worldwide. The majority of these alterations are missense mutations affecting p16^INK4a, and only 25% of these have been functionally assessed. There is therefore a need for an accurate and rapid assay to determine the functional significance of p16^INK4a mutations. We reviewed the performance of several in vivo functional assays that measure critical aspects of p16^INK4a function, including subcellular location, CDK binding and cell cycle inhibition. In this report the function of 28 p16^INK4a variants, many associated with melanoma susceptibility were compared. We show that assessment of CDK4 binding and subcellular localization can accurately and rapidly determine the functional significance of melanoma‐associated p16^INK4a mutations. p16^INK4a‐CDK6 binding affinity was unhelpful, as no disease‐associated mutation showed reduced CDK6 affinity while maintaining the ability to bind CDK4. Likewise, in silico analyses did not contribute substantially, with only 12 of 25 melanoma‐associated missense variants consistently predicted as deleterious. The ability to determine variant functional activity accurately would identify disease‐associated mutations and facilitate effective genetic counselling of individuals at high risk of melanoma. Hum Mutat 31:1–10, 2010. © 2010 Wiley‐Liss, Inc.     

Development of esophageal squamous cell cancer in patients with FAMMM syndrome: Two clinical reports

Familial atypical multiple mole melanoma (FAMMM) syndrome is a hereditary syndrome characterized by multiple dysplastic nevi and melanoma. Patients with FAMMM may have a heterozygous, inactivating, pathogenic germline variant in the CDKN2A gene, especially the NM_000077.4: c.225_243del19 (p.p75fs) variant, also known as p16-Leiden variant. Patients with this variant are at high risk for developing melanomas and pancreatic cancer due to somatic inactivation of the wild-type CDKN2A allele. The combination of an inactivating germline CDKN2A mutation and somatic inactivation of the wild-type CDKN2A allele in the same cell results in tumor formation. It has been suggested that carriers of a germline CDKN2A mutation are also at increased risk for several other cancer types, including esophageal cancer. Here, we describe two unrelated patients with the p16-Leiden variant who developed esophageal squamous cell cancer. Evidence of loss of the wild-type CDKN2A allele was obtained in the tumor tissue of both patients indicating biallelic inactivation of p16 in the tumor cells. These results suggest that these patients developed esophageal squamous cell cancer in the context of FAMMM syndrome.     

Prospective risk of cancer in CDKN2A germline mutation carriers

Background: The CDKN2A gene is the major known high-risk melanoma susceptibility gene. Susceptibility to other cancers has also been suggested. However, most studies examining the risks of other cancers classified individuals according to the family''s CDKN2A mutation rather than determining individual mutation status. For non-population-based studies, risks could also be biased because of cancer occurrence prior to family ascertainment. Methods: We examined the risk of non-melanoma cancer in 117 mutation-positive and 136 mutation-negative members from 15 families that had at least two first degree relatives with melanoma and CDKN2A mutations restricting the analysis to the period after the families were ascertained (that is, the prospective period) and using individual mutation data. The families have been followed prospectively for 4–26 years starting in the 1970s. Results: Overall, there was no significant association for mutation-negative subjects (Obs/Exp = 0.3, 95% confidence interval (CI) 0.0 to 1.2) although this group had only two observed cancers. In contrast, mutation-positive subjects had a significantly increased risk for all cancers combined (Obs/Exp = 12/5.5 = 2.2, 95% CI 1.1 to 3.8) primarily because of digestive system tumours, particularly pancreatic cancer. No other organ systems or individual tumour sites showed significantly increased risks. Conclusions: Differences in CDKN2A–non-melanoma cancer associations across studies may result from variation in genetic backgrounds, insufficient follow up, misclassification of mutation carriers, or the presence of other genetic and/or environmental risk factors in both CDKN2A mutation carriers and non-carriers. Larger sample sizes, prospective follow up, and individual mutation data will be required to understand these differences.     

Germline mutations of the CDKN2 gene in UK melanoma families

Germline mutations in CDKN2 on chromosome 9p21, which codes for the cyclin D kinase inhibitor p16, and more rarely, mutations in the gene coding for CDK4, the protein to which p16 binds, underlie susceptibility in some melanoma families. We have sequenced all exons of CDKN2 and analysed the CDK4 gene for mutations in 27 UK families showing evidence of predisposition to melanoma. Five different germline mutations in CDKN2 were found in six families. Three of the mutations (Met53Ile, Arg24Pro and 23ins24) have been reported previously. We have identified two novel CDKN2 mutations (88delG and Ala118Thr) which are likely to be associated with the development of melanoma, because of their co-segregation with the disease and their likely functional effect on the CDKN2 protein. In binding assays the protein expressed from the previously described mutation, Met53Ile, did not bind to CDK4/CDK6, confirming its role as a causal mutation in the development of melanoma. Ala118Thr appeared to be functional in this assay. Arg24Pro appeared to bind to CDK6, but not to CDK4. No mutations were detected in exon 2 of CDK4, suggesting that causal mutations in this gene are uncommon. The penetrance of these mutant CDKN2 genes is not yet established, nor is the risk of non-melanoma cancer to gene carriers.      

Features associated with germline CDKN2A mutations: a GenoMEL study of melanoma-prone families from three continents

Background

The major factors individually reported to be associated with an increased frequency of CDKN2A mutations are increased number of patients with melanoma in a family, early age at melanoma diagnosis, and family members with multiple primary melanomas (MPM) or pancreatic cancer.

Methods

These four features were examined in 385 families with ⩾3 patients with melanoma pooled by 17 GenoMEL groups, and these attributes were compared across continents.

Results

Overall, 39% of families had CDKN2A mutations ranging from 20% (32/162) in Australia to 45% (29/65) in North America to 57% (89/157) in Europe. All four features in each group, except pancreatic cancer in Australia (p = 0.38), individually showed significant associations with CDKN2A mutations, but the effects varied widely across continents. Multivariate examination also showed different predictors of mutation risk across continents. In Australian families, ⩾2 patients with MPM, median age at melanoma diagnosis ⩽40 years and ⩾6 patients with melanoma in a family jointly predicted the mutation risk. In European families, all four factors concurrently predicted the risk, but with less stringent criteria than in Australia. In North American families, only ⩾1 patient with MPM and age at diagnosis ⩽40 years simultaneously predicted the mutation risk.

Conclusions

The variation in CDKN2A mutations for the four features across continents is consistent with the lower melanoma incidence rates in Europe and higher rates of sporadic melanoma in Australia. The lack of a pancreatic cancer–CDKN2A mutation relationship in Australia probably reflects the divergent spectrum of mutations in families from Australia versus those from North America and Europe. GenoMEL is exploring candidate host, genetic and/or environmental risk factors to better understand the variation observed.     

Somatic alterations in the melanoma genome: A high‐resolution array‐based comparative genomic hybridization study

We performed DNA microarray‐based comparative genomic hybridization to identify somatic alterations specific to melanoma genome in 60 human cell lines from metastasized melanoma and from 44 corresponding peripheral blood mononuclear cells. Our data showed gross but nonrandom somatic changes specific to the tumor genome. Although the CDKN2A (78%) and PTEN (70%) loci were the major targets of mono‐allelic and bi‐allelic deletions, amplifications affected loci with BRAF (53%) and NRAS (12%) as well as EGFR (52%), MITF (40%), NOTCH2 (35%), CCND1 (18%), MDM2 (18%), CCNE1 (10%), and CDK4 (8%). The amplified loci carried additional genes, many of which could potentially play a role in melanoma. Distinct patterns of copy number changes showed that alterations in CDKN2A tended to be more clustered in cell lines with mutations in the BRAF and NRAS genes; the PTEN locus was targeted mainly in conjunction with BRAF mutations. Amplification of CCND1, CDK4, and other loci was significantly increased in cell lines without BRAF‐NRAS mutations and so was the loss of chromosome arms 13q and 16q. Our data suggest involvement of distinct genetic pathways that are driven either through oncogenic BRAF and NRAS mutations complemented by aberrations in the CDKN2A and PTEN genes or involve amplification of oncogenic genomic loci and loss of 13q and 16q. It also emerges that each tumor besides being affected by major and most common somatic genetic alterations also acquires additional genetic alterations that could be crucial in determining response to small molecular inhibitors that are being currently pursued. © 2010 Wiley‐Liss, Inc.     

Prevalence of variations in melanoma susceptibility genes among Slovenian melanoma families

Background  

Two high-risk genes have been implicated in the development of CM (cutaneous melanoma). Germline mutations of the CDKN2A gene are found in < 25% of melanoma-prone families and there are only seven families with mutation of the CDK4 gene reported to date. Beside those high penetrance genes, certain allelic variants of the MC1R gene modify the risk of developing the disease.     

A large Norwegian family with inherited malignant melanoma, multiple atypical nevi, and CDK4 mutation

Mutations in two loci encoding cell-cycle-regulatory proteins have been shown to cause familial malignant melanoma. About 20% of melanoma-prone families bear a mutation in the CDKN2A locus, which encodes two unrelated proteins, p16INK4A and p14ARF. Mutations in the other locus, CDK4, are much rarer and have been linked to the disease in only three families worldwide. In the 1960s, a large Norwegian pedigree with multiple atypical nevi and malignant melanomas was identified. Subsequently, six generations and more than 100 family members were traced and 20 cases of melanoma verified. In this article, we report that CDK4 codon 24 is mutated from CGT to CAT (Arg24His) in this unusually large melanoma kindred. Intriguingly, one of the family members had ocular melanoma, but the CDK4 mutation could not be detected in archival tissue samples from this subject. Thus, the case of ocular melanoma in this family was sporadic, suggesting an etiology different from that of the skin tumors. The CDK4 mutation in the Norwegian family was identical to that in melanoma families in France, Australia, and England. Haplotype analysis using microsatellite markers flanking the CDK4 gene and single-nucleotide polymorphisms within the gene did not support the possibility that there was a common founder, but rather indicated at least two independent mutational events. All CDK4 melanoma families known to date have a substitution of amino acid 24. In addition to resulting from selection pressure, this observation may be explained by the CG dinucleotide of codon 24 representing a mutational hot spot in the CDK4 gene.     

Deletion at chromosome arm 9p in relation to BRAF/NRAS mutations and prognostic significance for primary melanoma

We report an investigation of gene dosage at 9p21.3 and mutations in BRAF and NRAS, as predictors of relapse and histological markers of poor melanoma prognosis. Formalin‐fixed primary melanomas from 74 relapsed and 42 nonrelapsed patients were sequenced for common BRAF and NRAS mutations (N = 71 results) and gene dosage at 9p21.3 including the genes CDKN2A (which encodes CDKN2A and P14ARF), CDKN2B (CDKN2B), and MTAP was measured using multiplexed ligation‐dependant probe amplification (MLPA), (N = 75 results). BRAF/NRAS mutations were detected in 77% of relapsers and 58% of nonrelapsers (Fisher's exact P = 0.17), and did not predict ulceration or mitotic rate. There was no relationship between BRAF/NRAS mutations and gene dosage at 9p21.3. Reduced gene dosage at MTAP showed a borderline association with BRAF mutation (P = 0.04) and reduced gene dosage at the interferon gene cluster was borderline associated with wild type NRAS (P = 0.05). Reduced gene dosage in the CDKN2A regions coding for CDKN2A was associated with an increased risk of relapse (P = 0.03). Reduced gene dosage across 9p21.3 was associated with increased tumor thickness, mitotic rate, and ulceration (P = 0.02, 0.02, and 0.002, respectively), specifically in coding regions impacting on CDKN2B and P14ARF and CDKN2A. Loss at MTAP (P = 0.05) and the interferon gene cluster (P = 0.03) on 9p21 was also associated with tumor ulceration. There was no association between reduced gene dosage at 9p21.3 and subtype or site of tumor. This study presents supportive evidence that CDKN2B, P14ARF, and CDKN2A may all play a tumor suppressor role in melanoma progression. © 2010 Wiley‐Liss, Inc.