Faculty Bibliography

BRCA1 and BRCA2 mutation carriers have elevated risks of breast and ovarian cancers. The risks for cancers at other sites remain unclear. Melanoma has been associated with BRCA2 mutations in some studies, however, few surveys have included non-melanoma skin cancer. We followed 2729 women with a BRCA1 or BRCA2 mutation for an average of 5.0 years. These women were asked to report new cases of cancer diagnosed in themselves or in their family. The risks of skin cancer were compared for probands with BRCA1 and BRCA2 mutations. Of 1779 women with a BRCA1 mutation, 29 developed skin cancer in the follow-up period (1.6%). Of the 950 women with a BRCA2 mutation, 28 developed skin cancer (3.0%) (OR = 1.83 for BRCA2 versus BRCA1; 95% CI 1.08-3.10; P = 0.02). The odds ratio for basal cell carcinoma was higher (OR = 3.8; 95% CI 1.5-9.4; P = 0.002). BRCA2 mutation carriers are at increased risk for skin cancer, compared with BRCA1 carriers, in particular for basal cell carcinoma.

OBJECTIVE: To investigate whether or not fertility is reduced in women carrying mutations in the BRCA genes (BRCA1 and BRCA2), compared with noncarrier family members. DESIGN: Matched case-control study. SETTING: Academic. PATIENT(S): A total of 2,254 BRCA carriers and 764 noncarrier controls who were from the same families as the carriers, but tested negative for the BRCA mutation. INTERVENTION(S): None. MAIN OUTCOME MEASURE(S): History of fertility problems (yes/no) and prior use of fertility medications (yes/no) in cases and controls, obtained through administered questionnaire. RESULT(S): There was no difference in mean parity between carriers (1.9) and noncarriers (1.9). CONCLUSION(S): This study, which is the first epidemiologic study to specifically investigate the effect of carrying a BRCA mutation on parity and fertility, suggests that there is likely little or no effect of the BRCA gene mutation on fertility. To our knowledge previous epidemiologic studies have not investigated the effect of carrying a BRCA mutation on parity or fertility. However, we expect that if fertility differences exist, they would be toward the limits of reproductive life, thus larger studies with age-stratified analyses are needed to definitively answer this question.

Among women with a mutation in BRCA1 or BRCA2, the risk of breast cancer is high, but it may be modified by exogenous and endogenous factors. There is concern that exposure to carcinogens in cigarette smoke may increase the risk of cancer in mutation carriers. We conducted a matched case-control study of 2,538 cases of breast cancer among women with a BRCA1 (n = 1,920) or a BRCA2 (n = 618) mutation. One non-affected mutation carrier control was selected for each case, matched on mutation, country of birth, and year of birth. Odds ratios were calculated using conditional logistic regression, adjusted for oral contraceptive use and parity. Ever-smoking was not associated with an increased breast cancer risk among BRCA1 carriers (OR = 1.09; 95% CI 0.95-1.24) or among BRCA2 carriers (OR = 0.81; 95% CI 0.63-1.05). The result did not differ when cases were restricted to women who completed the questionnaire within two years of diagnosis. A modest, but significant increase in risk was seen among BRCA1 carriers with a past history of smoking (OR = 1.27; 95% CI 1.06-1.50), but not among current smokers (OR = 0.95; 0.81-1.12). There appears to be no increase in the risk of breast cancer associated with current smoking in BRCA1 or BRCA2 carriers. There is a possibility of an increased risk of breast cancer among BRCA1 carriers associated with past smoking. There may be different effects of carcinogens in BRCA mutation carriers, depending upon the timing of exposure.

Germ-line mutations in the TP53 gene are rare, but predispose women to a range of cancer types, including early-onset breast cancer. Breast cancers in women from families with the Li-Fraumeni syndrome often occur before age 30. The prevalence of deleterious TP53 mutations in unselected women with early-onset breast cancer is not precisely known. If mutations were found to be sufficiently common, it might be prudent to offer genetic testing to affected women in this age group. We screened the entire TP53 gene in the germ-line DNA from 95 women of various ethnic groups who were diagnosed with breast cancer before age 30, and who had previously been found to be negative for BRCA1 and BRCA2 mutations. No TP53 mutation was found. This study does not support a policy that TP53 testing should be offered routinely to unselected women with early-onset breast cancer in the absence of a family history of cancer.

In this review, we propose that age-related changes in mammographic density and breast tissue involution are closely related phenomena, and consider their potential relevance to the aetiology of breast cancer. We propose that the reduction in mammographic density that occurs with increasing age, parity and menopause reflects the involution of breast tissue. We further propose that age-related changes in both mammographic density and breast tissue composition are observable and measurable phenomena that resemble Pike's theoretical construct of 'breast tissue ageing'. Extensive mammographic density and delayed breast involution are both associated with an increased risk of breast cancer and are consistent with the hypothesis of the Pike model that cumulative exposure of breast tissue to hormones and growth factors that stimulate cell division, as well as the accumulation of genetic damage in breast cells, are major determinants of breast cancer incidence.

Breast and colon cancer are among the most common cancers in the developed world. Several epidemiological studies suggest that the occurrence of one of these two cancers in a woman may predispose to the development of the other. The occurrence of both forms of cancer in the same woman may be because of chance or common susceptibility. In order to determine how frequently double primary cancers have a hereditary basis, we conducted a registry based study at a single Montreal hospital. Cancer rates in first degree relatives of patients with multiple primaries were compared with provincial age standardised incidence rates and relative risks (RRs) were estimated. In first degree relatives under 45 there was a total of 15 cancers observed, compared with 3.70 expected, giving an RR of 4.05 (95% CI: 2.27-6.68). The RR for colon cancer was significantly increased among male relatives. For relatives less than 45 years old at diagnosis, the RR for colon cancer was 66.7 (95% CI: 13.8-195) (three cases observed, 0.045 expected). For all ages the RR was 5.02 (95% CI: 2.04-10.5). The RR for breast cancer was 5.92 (95% CI: 1.91-13.8) for female relatives under 45 (five cases observed, 0.845 expected) and 2.14 (95% CI: 1.07-3.83) for breast cancer at any age. These results suggest that there may be genes that predispose to both breast and colon cancer in certain people.

The association between sebaceous neoplasms of the skin and visceral cancers, known as Muir-Torre syndrome, is described in three patients, including one with an extensive history of cancer in his family. The first patient, a 54-year-old man, developed multiple sebaceous adenomas, epitheliomas, and carcinomas in association with a colonic carcinoma 6 years after cardiac transplantation. Family history in this patient disclosed colon cancer in 17 relatives. The second patient was a 51-year-old man who had recurrent adenocarcinoma of the sigmoid colon, adenocarcinoma arising in Barrett's esophagus, and sebaceous epithelioma during a period of 15 years. The third patient was a 90-year-old man with a sebaceous adenoma followed 5 months later by adenocarcinoma of the sigmoid colon with liver metastases. Muir-Torre syndrome in 129 other patients published in the literature is reviewed. Although it is a rare disease, Muir-Torre syndrome requires recognition because skin lesions may be the first sign of the syndrome and this may lead to early diagnosis of associated visceral cancers. Moreover, because this syndrome appears to be inherited, family members should be screened for visceral cancer, especially colorectal adenocarcinoma.

Hereditary nonpolyposis colon cancer (HNPCC) is an autosomal dominant trait responsible for approximately 6% of colorectal cancers. Linkage of the HNPCC trait to the D2S123 locus on 2p15-16 has previously been reported in two families. This HNPCC locus is now designated "COCA1." We have tested seven Canadian HNPCC families, who have a variety of clinical presentations, for linkage to a panel of microsatellite polymorphisms in the vicinity of D2S123. One family was clearly linked to the COCA1 locus (LOD = 4.21), and a second family is likely to be linked (LOD = 0.92). In three families linkage was excluded. In the remaining two families the data were inconclusive. In the linked family, individuals with cancer of the endometrium or ureter share a common haplotype with 12 family members with colorectal cancer. This supports the suspected association between these extracolonic neoplasms and the HNPCC syndrome. In addition, five of the six individuals with adenomatous polyps (but no colorectal cancer) have the same haplotype as the affected individuals, while the sixth carries a recombination. One individual with colorectal cancer carries a recombination that places the COCA1 locus telomeric to D2S123. This study localizes the COCA1 gene to an 8-cM region that is consistent with the location of the hMSH2 gene. We also confirm that families presently classified as HNPCC are genetically heterogeneous.