The Health Outcome
In western countries, the lifetime risk of breast cancer is about 8-12%. Mutations in two genes, BRCA1 and BRCA2, substantially increase this risk. However, these mutations are rare, accounting for fewer than 5% of breast cancer cases. While there may be other rare, highly penetrant disease-causing mutations that remain to be discovered, the majority of breast cancer is likely to involve interactions between common, weakly-penetrant genetic polymorphisms and environmental factors.
The Finding
Ziv et al (1) have reported that a single nucleotide polymorphism in the gene encoding the transforming growth factor beta 1 protein (TGFB1) is associated with breast cancer in a cohort of white, elderly women in the United States. A HuGE Net e-journal club abstract of this article is available (2). The women were recruited between 1986 and 1988 for a prospective study on osteoporotic fractures and were aged 65 or over at the time of enrolment. After a follow-up period (mean duration 9.3 years), they were invited to participate in a study that involved recording whether they had developed breast cancer since enrolment, and genotyping a blood sample for the T29->C polymorphism in the TGFB1 gene. This polymorphism has been reported to affect the circulating levels of the TGFB1 protein. Of 6795 women who provided samples, 3112 consented to genotyping, which was successful in 3075 cases. 146 of these women had developed breast cancer since enrolment in the study. (The incidence of breast cancer among non-participants was not significantly different.) The researchers found that homozygosity, but not heterozygosity, for the C allele of the TGFB1 gene was associated with a decreased risk of breast cancer in the study population. They report hazard ratios, relative to the T/T genotype, of 0.43 for the C/C genotype (95% CI 0.22-0.84, p=0.01) and 1.06 for the T/C genotype (CI 0.75=1.49, p=0.74). Adjusting for factors such as age, age at menarche, estrogen use, body mass index and bone mineral density did not significantly affect this result. The authors calculate that, if the association is a true one, it could account for as many as 60% of breast cancer cases in older women.
Public Health Implications
This is a well conducted cohort study with no obvious sources of bias, but the number of cases is small and the confidence intervals are wide. Furthermore, the possibility of Type 1 statistical error must be seriously considered in all genetic association studies. It would be premature to attempt to assess the public health implications of the findings until they have been confirmed by other studies. An association between TGFB1 polymorphisms and breast cancer risk is biologically plausible, but although the C/C genotype has been shown in other studies to be associated with higher serum TGFB1 levels, there is little evidence that this variant itself has a functional effect. There are other polymorphisms in the TGFB1 gene that also need to be investigated but were not genotyped in this study. If the association between the T allele and breast cancer proves to be a true one, it can be calculated that this allele would account for 2.7% of the excess familial risk of breast cancer
References
- Ziv E JAMA, 2001;285:2859-2863
- Electronic journal club abstraction form
