Abstract
Estimates of penetrance (or risk) of breast cancer among BRCA mutation carriers in published studies are heterogeneous, prohibiting direct combined estimates. Estimates of prevalence of BRCA mutations are more homogeneous and could allow combined estimates of prevalence. We propose a combined estimator of penetrance from combined estimates of the prevalence of BRCA mutations in women with and without breast cancer and from the probability of breast cancer by using Bayes' Theorem. The relative risk of having breast cancer with positive family history and the prevalence of positive family history contribute to the combined estimate of penetrance if family history is present. The combined estimate incorporates variation in estimates from different resources. The method is illustrated by using data from Ashkenazi Jewish women unselected for family history and for those with family history. Risks of breast cancer conferred by BRCA1 and BRCA2 mutations are estimated to be 8.39 per cent (6.56, 10.68 per cent) and 2.66 per cent (1.85, 3.82 per cent) by 40 years old, and 47.45 per cent (37.39, 57.72 per cent) and 31.85 per cent (23.72, 41.26 per cent) by 75 years old, respectively. For those with family history, risks of breast cancer conferred by BRCA mutations appear to be higher.
Original language | English (US) |
---|---|
Pages (from-to) | 1775-1787 |
Number of pages | 13 |
Journal | Statistics in Medicine |
Volume | 26 |
Issue number | 8 |
DOIs | |
State | Published - Apr 15 2007 |
Keywords
- BRCA mutations
- Breast cancer risk
- Family history
- Meta-analysis
- Penetrance
- Prevalence
ASJC Scopus subject areas
- Epidemiology
- Statistics and Probability