We derive ascertainment--corrected estimates of cancer penetrance for carriers of the breast cancer susceptibility genes BRCA1 and BRCA2. Obtaining population--based data on carriers is prohibitive: mutations at BRCA1/2 are rare and genotyping is expensive. Data on individuals identified, because of their family history of disease, to be at high risk of breast cancer is, however, available. Estimates of penetrance derived from such data will be biased if left uncorrected for ascertainment mechanism. We make such a correction by estimating population sampling weights from a population--based case--control study of breast cancer with extensive family history data. The Bayesian analysis we employ incorporates a modified segregation model and prior data on penetrance derived from the literature. Markov chain Monte Carlo methods are used for inference.
Return to Biostatistics Working Group