Germline CDKN2A Variant Cascade Testing Across Four Generations Reveals Familial Melanoma–Breast Cancer Genotype–Phenotype Correlation

ABSTRACT

This study reports co‐segregation of a pathogenic CDKN2A variant with both melanoma and breast cancer in a four‐generation pedigree. Eighteen individuals were test positive (n = 10), obligate (n = 5) or assumed carriers (n = 3) of the CDKN2A variant. Eleven of these had multiple melanomas, with initial diagnoses from teens to fifties. Six of thirteen female carriers had breast cancer (n = 5 test positive, n = 1 assumed carrier), with diagnoses ranging from thirties to sixties. Additional cancer diagnoses included pancreatic, and head and neck cancers. This study illustrates a possible genotype–phenotype association between a pathogenic CDKN2A variant and the co‐occurrence of melanoma and breast cancer in a hereditary context.

CDKN2A is a well‐known melanoma susceptibility gene, which also confers an increased risk for pancreatic cancer. This three‐generation pedigree shows a CDKN2A variant co‐segregating with both melanoma (orange) and breast cancer (blue). This report, combined with previous case studies, illustrates an increased risk for breast cancer for specific CDKN2A variants.

Summary.

• To the best of our knowledge, this is the first report of both primary melanoma and breast cancer co‐segregating with a CDKN2A variant in a large pedigree.

• Validating this genotype–phenotype correlation in additional pedigrees could significantly enhance clinicians' ability to assess risk and tailor personalized screening strategies to other high‐risk individuals.

Breast cancer and primary melanoma are the second and third most common cancers in Australia, respectively, accounting for 28% and 11% of all cancer diagnoses each year (Australian Institute of Health and Welfare ²⁰²⁴). The most frequently implicated predisposition genes are BRCA1 and BRCA2 for breast cancer (Kuchenbaecker et al. 2017) and CDKN2A for familial melanoma (Cust et al. 2011). A variant in a predisposition gene significantly increases the cumulative risk of developing these cancers by 80 years of age, with an estimated lifetime risk of 69%–72% for BRCA1/BRCA2 mutation‐positive breast cancer (Kuchenbaecker et al. 2017), and 52% for CDKN2A mutation‐positive melanoma (Cust et al. 2011).

Approximately, 1%–3% of melanoma cases are classified as familial, and 20%–40% of those cases are attributed to a variant in a high‐risk melanoma gene (CDKN2A, POT1, BAP1, POLE, CDK4, TERT, TERF2IP, ACD) (Maas et al. 2022; Potrony et al. ²⁰¹⁵). CDKN2A accounts for 67%–84% of variant positive cases (Maas et al. 2025). Individuals carrying a pathogenic variant in a high‐penetrance melanoma gene are more likely to have a family history of melanoma, an earlier age of melanoma onset, and/or be diagnosed with multiple primary melanoma (Zocchi et al. 2021). While CDKN2A is primarily associated with an increased risk for familial melanoma, hereditary cancer predisposition syndrome studies have demonstrated that CDKN2A also increases the relative risk for pancreatic cancer (12.3‐fold) (Kimura et al. 2021) and respiratory cancers (4.9–15.6) (Kimura et al. 2021; Helgadottir et al. ²⁰¹⁴; Borg et al. ²⁰⁰⁰). CDKN2A produces two tumour suppressor proteins, the canonical p16INK4a and the alternate p14ARF. The majority of CDKN2A variants reported in association with melanoma are loss‐of‐function (LOF) missense variants located in p16INK4a (Goldstein 2004).

Variants in moderate‐risk melanoma genes (MITF, MC1R, GOLM1, ATM, EBF3, NEK11) have also been found to alter melanoma risk (Potrony et al. 2015; Dalmasso and Ghiorzo ²⁰²⁰; Maas et al. ²⁰²³; Wallingford et al. ²⁰²⁴). Specifically, MC1R has nine common red hair colour (RHC) variants classified as ‘r’ (V60L, V92M and R163Q) and ‘R’ (D84E, R142H, R151C, R160W, D294H and I155T), which increase melanoma risk by approximately 1.5‐ and 2.0‐fold, respectively (Cust et al. 2012). Furthermore, RHC variants have been proven to modify the penetrance of CDKN2A variants by increasing the risk of early‐onset melanoma (< 40‐years of age) and multiple primary melanoma (Box et al. 2001).

This study reports segregation of a CDKN2A variant with both hereditary melanoma and breast cancer in an extensive pedigree. Five of the six individuals in the family with a personal history of breast cancer underwent panel testing of hereditary breast cancer genes (including BRCA1, BRCA2, TP53 and PTEN) and no pathogenic, likely pathogenic, or variants of uncertain significance were identified. Prior to study commencement, the sixth individual died of metastatic cancer (diagnosis 30–40 years) having been diagnosed with both breast cancer and melanoma.

Whole exome sequencing and subsequent clinical panel testing were performed on a proband (II:1; Female) who had > 20 melanoma diagnoses (initial diagnosis in twenties) and breast cancer (diagnosis 50–60 years). Genetic testing identified a rare (minor allele frequency = 0.00001) CDKN2A LOF missense variant affecting both p16INK4a and p14ARF transcripts (NM_000077.5:c.159G>C (NP_000068.1:p.Met53Ile) and NM_058195.4:c.202G>C (NP_478102.2:p.Asp68His) respectively) which has been previously reported as pathogenic in melanoma (Lang et al. 2007) and pancreatic cancer (Kimura et al. 2021), and is classified as pathogenic in ClinVar as per American College of Medical Genetics (ACMG) criteria (Landrum et al. 2014). Functional studies of this variant have shown that the base change significantly reduces the CDKN2A (p16INK4a) cell‐cycle inhibitory function through downstream inhibition of Rb, increasing the risk for uncontrolled cell proliferation in not only melanocytes but also mammary epithelial cells and pancreatic epithelial cells (Kimura et al. 2021; Danishevich et al. ²⁰²³). This variant also alters the p14ARF transcript, potentially inhibiting p53, a cell cycle regulator (Sargen et al. 2022). Pathogenic variants impacting the function of both p16INK4a and p14ARF are associated with a wider spectrum of cancer risks than variants that impact either transcript alone (Sargen et al. 2022). However, it has been suggested that p14ARF has a minimal tumour suppressive role in breast cancer (Aftab et al. 2019). Cascade testing was offered to all living relatives of the proband, and ten consented to provide a saliva sample for genetic testing through the Familial Melanoma Clinic at the Princess Alexandra Hospital.

The four‐generation pedigree (Figure 1) as reported by the proband (II:1), and confirmed by individuals undergoing testing, included an extensive cancer history. Fifteen individuals had a reported personal history of melanoma alone. Six individuals had early onset (initially diagnosed from 15 to 30 years); the youngest diagnosed in her teens (IV:3). Five of these underwent genetic testing, and all were positive for the CDKN2A variant. An additional relative with multiple melanomas diagnosed in his forties was an obligate carrier (II:9). One relative (II:14) who tested negative was diagnosed with late‐onset single primary melanoma (diagnosed 60–70 years), which is consistent with sporadic melanoma presentation in Queensland Australia, thus representing a phenocopy. A further five individuals had a history of both melanoma and breast cancer (two of whom were diagnosed with early‐onset melanoma). Four of these were living and accepted testing with positive results. One relative (II:13) with breast cancer only also tested positive. A remaining unaffected relative (II:3) who consented to genetic testing received a negative result. Family history revealed a deceased relative (I:1) had a history of early onset melanoma (diagnosed 30–40 years) and pancreatic cancer (diagnosed 60–70 years), another (II:8) died of metastatic colorectal cancer (30–40 years), and an additional relative (III:8) was diagnosed with leukemia in his teens.

FIGURE 1.

Family pedigree. Four generation pedigree showing cancer diagnoses and CDKN2A and MC1R status. Arrow denotes proband; *identifies those who have had genetic testing in our clinic; OC, obligate carrier; AC, assumed carrier. Some details have been changed to ensure confidentiality.

Where possible, MC1R variants were evaluated in positive individuals, given the association with melanoma risk. Of the 10 individuals who tested positive for the CDKN2A variant, eight had a personal history of melanoma and seven of these had MC1R RHC genotyping. All except one (III:10) of these individuals carried ≥ 1 MC1R RHC variant. Four of those individuals also had a history of primary breast cancer. Of note, the remaining individual who had a history of breast cancer, but not melanoma, had a CDKN2A variant in the absence of an MC1R RHC variant (III:2). Individual characteristics and CDKN2A and MC1R RHC variant carrier status are listed in Table 1 and denoted on the pedigree (Figure 1).

TABLE 1.

Summary of characteristics and genetic status for CDKN2A positive individuals and assumed/obligate carriers. ^a

Individual on pedigree	Age	Sex	CDKN2A M53I	MC1R variants	#Melanoma(s)/Dx age	Other cancer/Dx age
I:1	d. 70s	F	Obligate carrier		Single primary Dx 30s	Pancreas/60s
I:2	d. 70s	M	Obligate carrier		Single primary Dx 30s
I:3	d. 80s	M	Obligate carrier		Multiple primaries First Dx 40s
I:4	d. 90s	F	Obligate carrier		Multiple primaries First Dx 50s	Cervix/40s
II:1	60s	F	Positive	wt/R	Multiple primaries First Dx 20s	Breast/50s
II:2	60s	F	Positive	wt/R	Multiple primaries First Dx 20s	—
II:4	60s	F	Positive	wt/r	Multiple primaries First Dx 50s	Breast/60s
II:5	60s	F	Positive	wt/r	Multiple primaries First Dx 20s	Breast/40s Head & Neck b /50s
II:9	70s	M	Obligate carrier	—	Multiple primaries First Dx 40s	—
II:10	d.30s	F	Assumed carrier	—	Single primary Dx 30s	Breast/30s
II:12	70s	F	Positive	wt/R	Multiple primaries First Dx 40s	Breast/40s
II:13	70s	F	Positive	wt/wt	—	Breast/50s
III:1	40s	M	Assumed carrier		Single primary Dx 20s	—
III:2	40s	F	Positive	wt/R	—	—
III:3	40s	F	Positive	wt/R	Multiple primaries First Dx 20s	—
III:5	40s	M	Assumed carrier		Single primary Dx 30s	—
III:10	50s	F	Positive	wt/wt	Multiple primaries First Dx 50s	—
IV:3	Teens	F	Positive		Multiple primaries First Dx teens	—

Abbreviations: Dx, diagnosis; r, MC1R variant moderately associated with red hair color; R, MC1R variant strongly associated with red hair color; wt, wildtype.

^a As reported by family members who completed genetic testing.

^b Non‐smoker.

The identified CDKN2A p.Met53Ile has been reported in familial melanoma cases globally in families of White‐European ancestry, including families from Australia (Walker et al. 1995), North America (FitzGerald et al. ¹⁹⁹⁶) and the United Kingdom (Lang et al. 2007). More specifically, a study investigated 18 families of Scottish ancestry which revealed data supporting p.Met53Ile being a founder mutation that arose now around 90 generations ago (Lang et al. 2007). Breast cancer was reported in three of those families; however, no further information on the number of affected individuals or age‐of‐onset was provided (Lang et al. 2007). An increased prevalence of breast cancer has been reported in some CDKN2A positive families, including nine families with a duplication of three bases (NM_000077.5:c.335_337dup (NP_000068.1:p.Arg112dup)) in which breast cancer incidence was collectively increased (Borg et al. 2000; Potrony et al. ²⁰¹⁴). However, a statistically significant risk of breast cancer in CDKN2A carriers has not been not been consistently reported (Kimura et al. 2021; Helgadottir et al. ²⁰¹⁴; Borg et al. ²⁰⁰⁰). As pancreatic cancer is elevated in CDKN2A carriers, all individuals over 50 years of age were referred to a pancreatic cancer screening study (Broseghini et al. 2025).

A limitation of our study is that not all family members consented to cascade testing and thus segregation could not be confirmed in five living individuals with a personal history of melanoma. Furthermore, while no individuals had pathogenic variants in the four breast cancer predisposition genes tested, not all genes implicated in familial breast cancer were tested. It is possible that the increased incidence of breast cancer in this family is due to a variant in a different gene. Melanoma diagnoses were confirmed through pathology reports, where available. All breast cancers were self‐reported, but this has been associated with a high degree of accuracy (97%) in the literature (Abraham et al. 2009). Additionally, sun damage was not assessed, though prior research has shown that UV exposure does not predict the risk of developing melanoma in CDKN2A positive individuals (Cust et al. 2011).

This study reports CDKN2A genetic testing results and affection status of twelve relatives in a four‐generation pedigree. CDKN2A carriers had a history of early‐onset and/or multiple melanoma (youngest in their teens), and the majority carried an MC1R RHC variant. Six of thirteen women (46%) in this pedigree assumed or confirmed to have a CDKN2A p.Met53Ile variant had a personal history of breast cancer. Diagnosis ages ranged from the thirties to sixties. Four of the seven women without breast cancer were under 50 years of age, after which age breast cancer risk in the general population is most frequently diagnosed (Rojas and Stuckey 2016). To our knowledge, this is the first extended pedigree with both primary melanoma and breast cancer co‐segregating with a pathogenic CDKN2A variant. Confirming this genotype–phenotype correlation in other pedigrees could assist clinicians in assessing risk and personalizing screening.

Author Contributions

Conceptualization: A.M.‐L., J.B., and H.P.S. Data curation and analysis: J.B., E.J.M., and A.M.‐L. Writing: J.B., E.J.M., and A.M.‐L. Review and editing: J.B., E.J.M., E.D., C.A.P., H.P.S., and A.M.‐L.

Ethics Statement

This study was approved by the Human Research Ethics Committees of the Princess Alexandra Hospital (HREC/2019/QMS/55944) and the University of Queensland (UQ2020000771) and conducted in accordance with the Declaration of Helsinki.

Consent

Consent for the publication of de‐identified patient data was obtained by the authors prior to article submission stating that the participant gave consent with the understanding that this information may be publicly available.

Conflicts of Interest

H.P.S. is a shareholder of MoleMap NZ Limited and e‐derm consult GmbH and undertakes regular teledermatological reporting for both companies. H.P.S. is a Medical Consultant for Canfield Scientific Inc., Blaze Bioscience Inc., MoleMap Australia Pty Limited, and a Medical Advisor for First Derm. Other authors declare no conflicts of interest.

Berkman, J. , Maas E. J., DeBortoli E., Primiero C. A., Soyer H. P., and McInerney‐Leo A.. 2025. “Germline CDKN2A Variant Cascade Testing Across Four Generations Reveals Familial Melanoma–Breast Cancer Genotype–Phenotype Correlation.” Pigment Cell & Melanoma Research 38, no. 5: e70055. 10.1111/pcmr.70055.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.