One of the two X chromosomes in female mammalian cells is inactivated in early embryonic life. Females are therefore mosaics for two different cell types, cells with the paternal X chromosome as the active X, and cells with the maternal X chromosome as the active X.¹ In young females, the distribution of the two cell lines is close to normal, with a mode corresponding to the 50:50 ratio. Skewed X inactivation is a marked deviation from the 50:50 ratio and may be due to chance, to genetic factors involved in the X inactivation process, or to a selection process.^2–4 Elderly females have an increased frequency of skewed X inactivation in peripheral blood cells, most probably also reflecting a selection process.^5,6

Breast cancer results from genetic and environmental factors causing accumulation of mutations and eventually cancer. Germline mutations in BRCA1 and BRCA2 account for 15–20% of familial breast cancer, but less than 5% of all breast cancers.^7,8 In a study of patients with ovarian cancer, a higher frequency of skewed X inactivation was found in patients with invasive cancer compared with those with borderline cancer and healthy controls.⁹ We have reported an association between skewed X inactivation and breast cancer in young patients.¹⁰ The patients were part of a consecutive series with no information on family history or mutation analysis. Of this consecutive series approximately 30% are expected to have a positive family history.¹¹ It was therefore of interest to analyse X inactivation in subsets of patients with breast cancer. In this report we describe the results of X inactivation analysis of 272 patients with familial breast cancer, 35 with BRCA1/BRCA2 germline mutations, and 292 with sporadic breast cancer.

METHODS

Breast cancer cases

Blood samples from the patients had been collected after informed consent at Karolinska Hospital (material 1) and Huddinge Hospital and Söder Hospital (material 2) in Stockholm between 1989 and 2000. Material 1 was collected at a hospital where patients are referred for genetic counselling because of an increased risk of breast cancer. These patients were screened for both BRCA1 and BRCA2 mutations.^12–15 Material 2 was a consecutive series of breast cancer patients collected at surgery. These patients were screened for mutations in BRCA1 exon 11 only, as the previous studies from the Stockholm region showed that 70% of the BRCA1 mutations were detected in exon 11.^12,14,15 Furthermore, the estimated BRCA1 mutation frequency in an unselected breast cancer population from this region is less than 1%¹⁶ and mutations in BRCA2 are even less frequent.^13,15 Blood samples were drawn on average five years after onset. The study was approved by the ethics committee at the Karolinska Institutet.

Familial cases (from materials 1 and 2)

The patients were 306 women with non-BRCA1/BRCA2 breast cancer. Of these, 183 were low risk patients with one close (first or second degree) relative with breast cancer, and 123 were high risk patients with two or more close relatives with breast cancer.

BRCA1/BRCA2 mutation cases (from material 1)

The patients were 31 women with a BRCA1 mutation and eight with a BRCA2 mutation.

Sporadic cases (from material 2)

The patients were 322 women without a family history of breast cancer.

Controls

The controls were previously analysed women. As skewing of X inactivation increases after the age of about 55 years,¹⁷ it was necessary to have controls of different age groups. Young controls were 259 women from two populations of blood donors aged 20 to 54.^10,18 Middle aged controls were a population of 91 women aged 55 to 72 years, who were part of a routine mammography screening programme where blood samples were collected after two negative screenings.¹⁰ Elderly controls were 202 dizygotic twins aged 73 to 93 years.¹⁰

X chromosome inactivation analysis

DNA was extracted from peripheral blood cells according to standard procedure. The X inactivation pattern was determined by polymerase chain reaction (PCR) analysis of a polymorphic CAG repeat in the first exon of the androgen receptor (AR) gene.¹⁹ Methylation of HpaII sites in close proximity to this repeat correlates with X chromosome inactivation. After digestion with the methylation sensitive enzyme HpaII, a PCR product is obtained from the inactive X chromosome only. Additionally, PCR amplification of a segment in exon 1 of the pseudoautosomal MIC2 gene was carried out on both undigested and digested samples to control for completeness of digestion.²⁰MIC2 is unmethylated on both the active and inactive X chromosome, and no PCR product is expected from completely digested samples.²¹ PCR products from undigested and digested DNA were separated on an ABI 3100 automated sequencer, and analysed by GeneScan software (Applied Biosystems, Foster City, California, USA).

The X inactivation pattern was recorded as the relative amount of the PCR product of the smallest allele. The pattern was classified as skewed when 90% or more of the cells preferentially expressed one X chromosome.

Statistical methods

The Pearson χ² test was used for testing categorical variables. The Fisher two tailed exact test was used where appropriate. As we divided the patients into three groups according to their ages, the probability (p) value to reach statistical significance was set at 0.017 (Bonferroni correction).

RESULTS

Patients with a family history of breast cancer

Of the 306 patients with familial non-BRCA1/BRCA2 breast cancer, 272 were heterozygous for the CAG repeat in the AR gene and were therefore informative for the X inactivation assay. When the total material of familial cases, including both low risk and high risk cases, was compared with controls in the respective age groups, a significantly higher frequency of skewed X inactivation was found in young patients (27 to 54 years) compared with young controls (11.2% and 2.7% respectively; χ² = 11.15, p = 0.001) (table 1). The frequency of skewed X inactivation in the total material of young low risk patients (13.2%) was significantly higher than in controls of similar age (2.7%; χ² = 11.25, p = 0.001). There was some discrepancy between the two materials of young low risk patients with respect to skewing of X inactivation, patients from material 1 being more skewed than those from material 2. However, both materials showed a distinct difference in X inactivation skewing compared with controls. Young high risk patients were also more skewed than young controls, but the difference was not significant (6.5% and 2.7%, respectively; Fisher’s exact test, p = 0.25).

Information on the absence of breast cancer after mammography screening was available in the controls aged 55 to 72 years only. When this group was compared with the total material of familial cases in the same age group, a difference in the frequency of skewed X inactivation was also found (4.4% and 9.7%, respectively), but this was not significant (χ² = 2.11, p = 0.15). Middle aged low risk and high risk patients were also more skewed than controls, but the differences were not significant.

The elderly patients had a lower, and close to significant, frequency of skewed X inactivation than elderly controls (χ² = 5.07, p = 0.02). A lower frequency of skewed X inactivation was also observed for elderly low risk and high risk patients, but neither was significantly different from controls.

Patients with BRCA germline mutations

Twenty seven patients with a BRCA1 mutation and eight with a BRCA2 mutation were informative for X inactivation analysis. Only one young and one middle aged patient had a skewed X inactivation pattern, which was not significantly different from the frequency in the controls of the same age groups (table 1).

Patients with sporadic breast cancer

Of 322 patients with sporadic breast cancer, 292 were informative for X inactivation analysis. There was no significant difference between young patients and young controls with respect to X inactivation skewing (6.2% and 2.7%, respectively) (table 1). There was, however, a strong tendency for middle aged patients (55 to 72 years) to be more skewed than middle aged controls (13.6% v 4.4%, respectively) (χ² = 5.17, p = 0.02).

DISCUSSION

We found a higher frequency of skewed X inactivation in young patients (<55 years) with familial non-BRCA1/BRCA2 breast cancer than in young controls (p = 0.001) (table 1). Young low risk patients were significantly more skewed in X inactivation than young controls. The same association was not found for the high risk patients. However, the number of young high risk patients was only 31, and the low frequency of patients with skewed X inactivation may therefore reflect statistical fluctuation and requires further study.

Middle aged patients (55 to 72 years) with sporadic breast cancer also had a higher frequency of skewed X inactivation than middle aged controls, though this did not reach significance (p = 0.02; table 1). In a previous study we found an increased frequency of skewed X inactivation in a consecutive series of young breast cancer patients with no available information on family history or mutation analysis.¹⁰ Our current findings indicate that skewed X inactivation may be a risk factor in both sporadic and familial breast cancer. One explanation could be that some of the familial cases are carriers of a mutation in an X linked tumour suppressor gene. An additional preferential inactivation of the normal X chromosome could lead to early onset of breast cancer. In the haematopoietic stem cells of the sporadic cases, a mutation during early years in an X linked tumour suppressor gene on the active X chromosome could lead to a selective growth advantage for the mutation carrying blood cells. As a result, skewed X inactivation may become apparent at middle age or older. Thus a mutated X linked factor that confers a growth advantage to blood cells may also be involved in sporadic breast carcinogenesis—for example, as an angiogenetic factor. Another explanation could be that in some females who are born with skewed X inactivation cells are expressing a factor that is harmless to young females but which confers damage to breast epithelia owing to the hormonal changes after the menopause.

Elderly healthy women have an increased frequency of skewed X inactivation in their peripheral blood cells. This increase in frequency seems to occur after the age of 55 years and continues throughout life.^5,17 In this study we found an unexpected lack of increased frequency of skewed X inactivation with age in the familial breast cancer cases, as 7.7% of the elderly patients had skewed X inactivation compared with 21% of elderly controls (table 1). A lower frequency of skewed X inactivation (14%) was also observed in elderly subjects in our previous study of a consecutive series of breast cancer patients.¹⁰ This is in contrast to our finding in a series of patients with cervical cancer, where the same increase in skewing was found both in elderly patients and controls (23% and 21%, respectively).²² Thus skewed X inactivation may be a risk factor for the development of breast cancer in young women with a family history of breast cancer, while the older patient group may be protected by the lack of X inactivation skewing. A discrepancy between early and late onset of breast cancer in this respect may also reflect different genetic backgrounds or tumour pathways between these two kinds of breast cancer.

Buller and coworkers⁹ found a skewed X inactivation (⩾75:25) in nine of 11 patients with ovarian cancer where a BRCA1 mutation was identified. It is suggested that the BRCA1 protein is involved in maintaining the inactive X chromosome in the inactive state through XIST RNA localisation.²³ It was therefore of interest to look for an association between patients with BRCA mutations and X inactivation skewing. We did not find a significant increase in frequency of skewed X inactivation in the BRCA1/BRCA2 breast cancer patients. However, the number of patients was very small, and an association between skewed X inactivation and cancer in the BRCA1/BRCA2 patients cannot be excluded.

Information on the timing of blood sampling in relation to treatment was not available for the patients in this study. There is a possibility that chemotherapy may cause neutropenia and lymphopenia which could lead to skewed X inactivation. However, no change in X inactivation pattern was seen in patients who had received chemotherapy a year after treatment in a study by Buller and coworkers.⁹

In summary, we have found a higher frequency of skewed X inactivation in young patients with familial non-BRCA1/BRCA2 breast cancer and in middle aged patients with sporadic breast cancer. This indicates that skewed X inactivation may be a risk factor for the development of breast cancer and could indicate an effect of X linked genes.