Article Text

other Versions

A germline deletion of 9p21.3 presenting as familial melanoma, astrocytoma and breast cancer: clinical and genetic counselling challenges
  1. Jaime Vengoechea,
  2. Christine Tallo
  1. Human Genetics, Emory University, Atlanta, Georgia, USA
  1. Correspondence to Dr. Jaime Vengoechea, Human Genetics, Emory University, 1365 Clifton Rd NE, Building A, Suite 1500, Atlanta, Georgia 30322, USA; jvengoe{at}

Statistics from

Baker and collaborators describe a family with melanomas, nerve sheath tumours, gliomas and haematological malignancies caused by a large, 25-gene deletion resulting from an unbalanced translocation in 9p21.3.1 The 9p21.3 region contains at least three tumour suppressor genes: CDKN2A, CDKN2B and MTAP. Somatic deletions involving this region have been identified in tumours such as breast, prostate, melanoma, leukaemia, among many. Germline mutations and deletions involving the genes in this region are causative of inherited cancer predisposition. We report the case of a family with a germline 9p21.3 deletion encompassing nine genes. The family presented with melanomas, astrocytomas, neurofibromas and breast cancer.

The CDKN2A gene encodes two proteins: p16 and p14, by means of alternative splicing. Protein p16 inhibits cyclin-dependent kinase 4 (CDK4) which controls the progression through the G1 phase of the cell cycle by phosphorylating Rb.2 In turn, p14 is a stabiliser of p53—a key protein in inducing cell apoptosis.3 CDKN2B encodes p15, which similarly inhibits CDK4 and CDK6.4 MTAP encodes methylthioadenosine phosphorylase—an enzyme that appears to also exhibit features of a tumour suppressor gene.

Mutations affecting the p16 transcript of CDKN2A can cause familial atypical multiple mole melanoma syndrome5 also known as pancreatic cancer–melanoma syndrome.6 Melanoma–astrocytoma syndrome (MAS) presents with melanomas, astrocytoma, neurofibromas, schwannomas and meningiomas but not pancreatic cancer.7 MAS is caused by mutations that affect both the p16 and p14 transcripts of CDKN2A. Most families with MAS have deletions including CDKN2A plus other genes in 9p21.3, as in the original French family described with this syndrome.7 Mutations affecting alternative splicing of CDKN2A can also present as MAS.8 Germline mutations in CDKN2B cause inherited renal cell carcinoma,9 but this malignancy has not been described as part of MAS. Germline mutations in MTAP have been identified in families with skeletal dysplasia and osteosarcoma.10

In the family reported here, the proband is a 46-year-old man who presented with no personal history of cancer. His mother was diagnosed with breast cancer in her 60s and had a germline multigene cancer panel by next-generation sequencing complemented with exon array deletion/duplication testing which identified a full deletion of the CDKN2A gene. The test did not identify pathogenic variants in genes causative of inherited breast cancer. The maternal family history is significant for several individuals with melanoma, astrocytoma and neurofibromas (see figure 1), leading to a diagnosis of neurofibromatosis type I in several relatives without cafe-au-lait spots or Lisch nodules. The proband's brother had melanoma at the age of 16. The maternal grandfather had melanoma in his 50s and died at the age of 57. A maternal aunt, aged 77, has neurofibromas. This aunt had three sons, one of which died at the age of 34 from an astrocytoma and another who died at the age of 9 of leukaemia. Another maternal aunt had breast cancer diagnosed in her 60s and neurofibromas. She had two daughters with cancer, one with breast cancer diagnosed in her late 40s and neurofibromas who also carries the CDKN2A deletion, and another with astrocytoma.

Figure 1

Pedigree, unaffected proband indicated with an arrow.

Before presenting to our clinic, the proband had CDKN2A deletion testing by exon array which identified the same deletion seen in his mother. As we did not know the extent of the deletion, which as reported involved the full gene, we ordered comprehensive genomic hybridisation microarray. This test identified a deletion in 9p21.3 involving base pairs (21,430,743-22,212,612) (see figure 2). The deletion encompasses nine genes, including CDKN2A, CDKN2B and MTAP.

Figure 2

9p21.3 region (chr9: 20500000–22500000, human genome assembly GRCh37/hg19), comparing the deletion reported by Baker et al, 1 to the one found in our family, and showing the genes predicted to be deleted in each family. Image generated using the University of Cailifornia Santa Cruz (UCSC Genome browser).

Deletion/duplication testing has been implemented into multigene platforms by relying on the next-generation sequencing readout dose coupled with spilt-end analysis, or by adding an exon array platform. Neither of these are designed to identify larger-scale microdeletions. In this case, the size of the microdeletion required adding microarray to recognise the full extent. The deletion is smaller than that reported by Baker et al1 but seems to have the full phenotypic spectrum of nerve sheath tumours, gliomas and melanomas reported by them.

Germline missense mutations of CDKN2B can cause renal cell carcinoma and germline mutations in MTAP leading to aberrant splicing and exon skipping have been related to osteosarcoma and a skeletal dysplasia phenotype. These mechanisms may have a different impact than a multigene deletion and we cannot affirm that our patient's family is at risk for these manifestations. We should also be cautious when inferring potential risks from single case reports which may not be replicated. For example, we have identified a high frequency of breast cancer cosegregating with the astrocytomas and neurofibromas in this family. However, it is too early to establish early-onset breast cancer as a part of the phenotype.

Germline deletions involving multiple tumour suppressor genes pose additional management challenges, beyond those associated with single-gene loss of function variants.Many types of cancer cells have somatic deletions involving key tumour suppressor genes but the effects of a germline deletion may be different. The interaction between the deleted genes and other loci in the genome is often completely unknown.

There are very few cases in the medical literature describing the outcomes of MAS. This poses a clinical challenge in recommending heightened screening. The lack of knowledge regarding tumour risks associated with 9p21.3 deletions must be addressed when offering testing to family members. Currently, no other relatives have sought clinical testing for the 9p21.3 deletion. The limited availability of medical management guidelines available for his familial deletion was of great concern for our proband as he prepares to share this information with his family. These issues demonstrate the need for proper genetic counselling prior to undergoing genetic testing, particularly for genes for which there is limited knowledge of the associated cancer risks. The era of next-generation sequencing has greatly expanded the availability and use of clinical genetic testing. In cancer genetics, this has led to the development of multigene panels. As these tests are relatively novel, healthcare specialists not trained in genetics may not be familiar with their limitations. This highlights the growing need for genetics healthcare providers who can order appropriate genetic testing and accurately interpret results.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
View Abstract


  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval Emory IRB.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.