Introduction

Von Hippel-Lindau (VHL) disease is a rare autosomal dominant disorder characterized by a predisposition to haemangioblastomas of the central nervous system (cHAB) and retina (rHAB), renal cell carcinomas (RCC), pheochromocytomas and paragangliomas, endolymphatic sac tumors (ELST), pancreatic neuroendocrine tumors (PNET), papillary cystadenomas of epididymis, and adnexal papillary tumors of probable mesonephric origin (APMO)[1–3]. The disease is caused by germline mutations in the VHL tumor suppressor gene on a short arm of chromosome 3. In most cases, VHL disease is caused by single base substitutions, microdeletions or microinsertions, however approximately 30% of VHL patients carry a deletion of a part or of the entire VHL gene[4–6].

Patients with small intragenic mutations of the VHL gene have a typical disease presentation. However this may not be the truth in cases of carriers of large deletions. Large deletions may involve other genes in the proximity of the VHL gene, i.e. FANCD2, C3orf10, CNTN6, IRAK2, GHRL or MLH1 genes that may have an effect on the disease presentation. For example, there is some evidence to suggest that large deletions of VHL and of C3orf10 are associated with lower life time risk of kidney cancer compared to the deletions which do not include the C3orf10 gene[7–10]. However, the impact of the extent of large deletions encompassing VHL and other adjacent genes on phenotype has not been extensively studied. In particular, it was not investigated whether large VHL deletions that include the FANCD2 gene - an important gene in the DNA damage repair signaling pathway, have an effect on disease phenotype.

Material and methods

Patients

We studied a phenotype of a family with a large deletion of the VHL gene that involves the FANCD2 gene. This family included 11 family members (Figure 1), 4 patients had VHL-related tumors, and 2 patients had other tumors (breast cancer, prostate cancer) (Table 1). Blood samples were available from 3 family members (a 53 year old woman with bilateral breast cancer and VHL disease, her 31 year old son with VHL disease and her healthy 30 year old daughter).

Figure 1
figure 1

Pedegree of the affected family.

Full size image
Table 1 VHL related and non-VHL related lesions detected in the affected family
Full size table

Mutation analysis

DNA was extracted from whole blood using standard techniques. Peripheral blood leukocytes obtained from the patient were processed for DNA isolation using standard methods. The entire coding region of VHL gene was sequenced using direct Sanger sequencing as described previously[6]. Large genomic rearrangements were detected using MLPA analysis using the SALSA MLPA kit P016 VHL assay (MRC Holland, Amsterdam, The Netherlands). This assay includes 17 probes from distal chromosome 3p25 (from telomere to centromere): one probe for CNTN6, two probes for FANCD2, two probes for C3ORF10, four probes for VHL exon 1, three probes for VHL exon 2, two probes for VHL exon 3, IRAK2 probe, GHR probe, and MLH1 probe. The products of MLPA reactions were analyzed with automated sequencer (model 373A, ABI, USA).

Results

In two patients with VHL disease, a large germline deletion involving the exon 1 of the VHL gene, the C3ORF10 gene and a fragment of the FANCD2 gene was detected. The presence of the deletion was confirmed using MLPA in repeated analyses using DNA samples isolated from two different blood samples for each patient. In brief, in MLPA, the peak height was significantly lower for the six probes (three probes for VHL exon 1, two probes for C3ORF10 and 1 probe for FANCD2 gene) indicating the presence of a heterozygous deletion. Based of the localization of these six probes, we estimate that the length of this deletion is approximately 50 kb (between 43 kb to 54 kb). The telomeric breakpoint lies within the FANCD2 gene between exon 35 and exon 43 of the FANCD2 gene (a signal for the probe specific for exon 43 of the FANCD2 gene was abnormal, and signal for the probe for exon 35 of the FANCD2 gene was appropriate). The centromeric breakpoint is located in intron 1 the VHL gene.

The clinical presentation of the VHL disease in this family includes CNS heamangioblastoma seen in four patients, kidney cancer diagnosed in one patient and no retina heamangioblastoma or pheochromocytoma. Age of diagnosis and localization of these typical VHL-associated lesions are shown in Table 1.

Interestingly, one carrier of the deletion developed bilateral breast cancer at age 46 and 49. First cancer was diagnosed at the age of 46 in her right breast. The patient was treated by mastectomy with postoperative chemotherapy (CMF). Histopathological examination revealed invasive ductal breast carcinoma of 1.8 cm size. The tumor was ER- negative, and PR-negative (HER2 was not analyzed). A metastasis in 1 axillary lymph node was found. There was no evidence of distant metastases. The tumor was classified as T1c pN1a M0, stage IIA; grade G2. Three years later, at the age of 49, she was diagnosed with cancer of her left breast. It was an invasive ductal carcinoma of 2.0 cm size. The tumor was ER-positive, PR-positive, and HER2-negative. There was no evidence for lymph-node or distant metastases. The tumor was classified as T1c pN 0 M0, stage IA, grade G2 and treated with mastectomy and tamoxifen. Ki67 expression was not tested. At present, 9 and 6 years after treatment of the breast cancers there is no evidence of breast cancer recurrence. However at age of 54, lesions typical for VHL disease including cerebellar haemangioblastoma and renal cell carcinoma were diagnosed in the patient.

There was also a history of prostate cancer in one affected male family member at the age of 65 years although his DNA sample was not available for MLPA analysis.

Discussion

To our knowledge this is the first study to suggest an association between a large deletion involving the FANCD2 gene and bilateral breast cancer. FANCD2 is one of eight genes known to cause the autosomal recessive disorder Fanconi anaemia (FA), which is characterized by spontaneous chromosomal instability, immunodeficiency, and a predisposition to cancer[11–14]. FANCD2 plays an important role in the recombination DNA repair pathways[15]. The activated FANCD2 protein co-localises with other DNA repair proteins such as BRCA1, BRCA2, ATM, NBS1 and RAD51[16]. It has been suggested that FANCD2 and BRCA1 interact directly in this process[11]. In response to ionising radiation, FANCD2 is also phosphorylated by ATM, which leads to the activation of an S-phase checkpoint of the cell cycle[16]. Therefore, FANCD2 is good candidate for breast cancer susceptibility gene.

It is interesting that biallelic mutations of BRCA2, the major gene for hereditary breast cancer, were also shown to cause Fanconi anemia. However, it has been reported that FA gene mutations, other than in BRCA2, are unlikely to be a frequent cause of highly penetrant breast cancer predisposition. Analysis of the FA genes (FANCA, B, C, D1, D2, E, F, G) in 88 non-BRCA1, non-BRCA2 breast cancer families failed to identify any highly penetrant mutations for breast cancer[17]. In another study, 399 women, from 356 non-BRCA1/2 breast cancer families (some had more than one index case because multiple women were affected at the same age), were screened for FANCD2 mutations by DHPLC/sequencing and no pathogenic mutations were identified[18].

In our study, breast cancers seen in carrier of a deletion of FANCD2 gene were invasive, grade 2 ductal carcinomas. Both were tumors of ~2 cm in diameter. One of these had lymph node metastases. One tumor was ER- and PR-negative (HER2 status unknown) and the other was ER- and PR-positive, but HER2-negative. This may suggest that a mutation in FANCD2 predispose to invasive breast cancer of ductal type, however further studies are needed in this regard.

Given that both FANCD2 and BRCA2 mutations (in homozygous state) cause the same chromosomal instability syndrome called Fanconi anemia, it is likely that breast cancers that arise in carriers of a mutation in these genes may have similar clinical characteristics. In this context it is important to mention that BRCA2-associated breast cancers exhibit higher grade than sporadic tumors[19–22]. They are similar to sporadic cancers with respect to ER receptor status[20, 23–25]. However, one study reported that BRCA2 associated tumors are more likely to be ER-positive[22]. In addition, it has been reported that BRCA2-related breast cancers are less likely to over express HER2 receptor compared to sporadic cancers[20, 25]. Further studies are needed to investigate whether FANCD2- related breast cancers are similar to those seen BRCA2 mutation carriers.

In conclusion, our study suggests that a deletion of the VHL and FANCD2 gene may be associated with coexistence of VHL disease and hereditary breast cancer, but further studies are needed in this regard.

Consent

Informed consent has been obtained from the patients for publication.