Editor—Germline mutations in theBRCA1 and BRCA2genes cause predisposition to breast and ovarian cancer.1Epidemiological evidence and linkage studies suggested that the likelihood that a woman with breast cancer has a genetic susceptibility to the condition is greater the younger she was at diagnosis and with increasing extent of family history of the disease. Studies of the prevalence of germline mutations in BRCA1and BRCA2 in women with breast cancer has enabled the frequency of mutations to be determined in women with different ages at diagnosis and extent of family history of breast cancer.2 3 The CASH study into the attributable risk of breast and ovarian cancer estimated that 33% of all breast cancers diagnosed by the age of 29 years, and 22% diagnosed by the age of 30-39 years, are the result of an inherited mutation.4However, the proportion of breast cancer cases diagnosed by 40 years resulting from a BRCA1 mutation was predicted to be 5.3%.5 Previous population based studies of the prevalence of BRCA1 mutations in early onset breast cancer have been in cases unselected for family history, and the majority of mutation carriers detected did have some degree of family history of either breast or ovarian cancer.6-10 The aim of this study was to establish the prevalence of BRCA1 mutations in a large series of British patients with a young age of onset and no known family history of the disease, since such patients are referred relatively frequently for genetic counselling. The presence ofBRCA1 mutations in a significant proportion of these patients would have important implications for the planning of a mutation screening strategy in diagnostic services.

Patients were ascertained from the Imperial Cancer Research Fund Clinical Breast Oncology Department at Guy's Hospital, and the family cancer clinics at the Genetics Departments of Guy's Hospital and the Royal Free Hospital, both in London, and from St Mary's Hospital, Manchester. The Manchester cases were initially ascertained as part of a population based series for other studies. Patients without a known family history of breast or ovarian cancer at referral were recruited into this study, and these patients were interviewed at home by a genetic nurse counsellor. A family history was taken by the clinical geneticist or genetic nurse counsellor, and the diagnosis confirmed in all probands with either the referring breast unit or the patient's GP or oncologist. Patients with a known family history of breast or ovarian cancer at referral were excluded from the study. The range of age of diagnosis was 22-35 years and the median age of diagnosis was 31 years. Blood samples were taken after the purpose of the study had been explained and informed consent obtained. Full genetic counselling was provided, following agreed protocols, and results were made available to those women wishing to be informed.

The 22 BRCA1 coding exons were amplified using 24 pairs of primers (exon 11 was amplified in three overlapping fragments). The fluorescent chemical cleavage of mismatch (FCCM) protocol was adapted from the method of Rowley et al. 11 PCR products were labelled by incorporation of dUTP analogues which were labelled with either R110 (blue), R6G (green), or Tamra (yellow) fluorescent dyes (PE Applied Biosystems Inc), and heteroduplex molecules then subjected to hydroxylamine modification and piperidine cleavage. Using the three different dyes to label the 24 PCR products that cover the completeBRCA1 coding sequence, three patients could be analysed for one fragment in one lane of the gel; thus three patients were completely screened in 24 lanes and six patients were screened on a 50 well gel. The FCCM technique has been reported to detect over 95% of mutations in a blind study of haemophilia A patients.11 We evaluated the sensitivity of ourBRCA1 protocol by examination of the eight polymorphisms that have been reported in theBRCA1 gene with a frequency of at least 5% (1186A/G, 2201C/T, 2430T/C, 2731C/T, 3232A/G, 3667A/G, 4427C/T, 4956A/G).12 These polymorphisms were all detected reproducibly at the expected frequency in a panel of over 400 patients who were tested, as part of our continuing studies. In addition, FCCM detected the mutations in three known BRCA1positive samples (188del 11bp, 5242 C/A, and 5382insC) that were previously found in our laboratory by SSCP analysis.13 The fluorescent chemical cleavage assay which we have developed therefore allows a rapid and sensitive mutation screen ofBRCA1.

Four mutations that were likely to be pathogenic were detected in 110 patients (3.6%) and are listed in table 1. These included three sequence variants that would be predicted to result in premature termination of translation. The 185delAG frameshift mutation, which is prevalent in the Ashkenazi Jewish population, was identified in a British patient (91032) with Jewish ancestry. The 4693-4694delAA mutation was detected in a British patient (78750) who was diagnosed with breast cancer at the age of 26 years. The truncating mutation 3875-3878delGTCT was seen in a patient of Afro-Caribbean origin (94641), who was diagnosed with breast cancer aged 33 years and with a second primary cancer at 38 years. Testing for this mutation in the parents of the patient indicated that it was inherited from her father. A novel in frame deletion, 1965-1967delTCT, was detected in a patient of West African origin (103727) who was diagnosed with breast cancer at the age of 27 years. This mutation would be predicted to result in the deletion of a single amino acid, serine 616, but would not lead to premature termination of translation. This sequence change was absent in over 350 control chromosomes in our study and fulfils all the other criteria for pathogenic status.13 The pedigrees of these four women are shown in fig 1. Although all four patients originally reported no family history of breast cancer, further investigation showed that the maternal great grandmother of patient 103727 was diagnosed with the disease at about 60 years of age. The lack of a history of breast or ovarian cancer in these families is likely to result from a combination of factors including paternal transmission of the mutation, chance, and reduced penetrance.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Pedigrees of cases in whom a pathogenic BRCA1 mutation was identified.

Eight other rare DNA sequence variants were identified (table 2). Three of these would not be predicted to alter the expression ofBRCA1 or the sequence of its encoded protein (Q1604Q, IVS22+8T/C, UGA+36T/C) and R1347G was present in a patient with a frameshift mutation.12 The pathogenic status of the other four (Q804H, S1140G, P1637L, and M1652I) remains inconclusive in the absence of a functional assay for the BRCA1 protein. At present, screening panels of ethnically matched controls is a useful means of excluding missense mutations as pathogenic mutations of high penetrance, and it would be helpful if this information was provided in the Breast Cancer database (www.nhgri.nih.gov/Intramural_research/Lab_transfer/Bic/index.html).

Our detection of pathogenic BRCA1 mutations in 3.6% of young breast cancer patients without a family history of breast or ovarian cancer is consistent with the estimate of Fordet al 5 that the proportion of breast cancer cases in the general population resulting fromBRCA1 is 5.3% below the age of 40 years. A recent population based study of young British breast cancer patients unselected for family history found BRCA1mutations in nine of 254 (3.5%) women diagnosed by the age of 36 years.8 In North American studies ofBRCA1 mutations in women unselected for family history and diagnosed below the age of 35 years, Langstonet al 7 detected mutations in 7.5%, Malone et al 9 in 6.2%, and Struewing et al 10 in 5.7%. Fitzgerald et al 6 detectedBRCA1 mutations in 13% of women diagnosed before the age of 30 years, but this included Ashkenazi Jewish patients who have founder mutations and some patients with a family history. Since we did not screen the promoter region ofBRCA1 or for deletions of entire exons,14 we cannot exclude the possibility that some mutations were missed, and the pathogenic status of several sequence variants remains unresolved. However, the important practical implication of our study is that, given the time and expense involved, it would be reasonable to attach a low priority toBRCA1 mutation screening of young isolated cases of breast cancer in the context of the provision of a publicly funded and cost effective diagnostic service. A screen of this cohort of patients for BRCA2 mutations is in progress.