Possible heterogeneity in the segregation pattern of breast cancer in families with bilateral breast cancer.

We investigated the segregation pattern of breast cancer in families with bilateral breast cancer, classifying families with respect to menopausal status (premenopausal versus postmenopausal) and the interval between diagnosis of the two primary tumors in the probands. Probands were "synchronous" if both primaries were diagnosed within 1 year; "asynchronous" if the interval was at least 2 years. Results for four complex segregation analyses are here presented; the findings support heterogeneity in the transmission of breast cancer. In the asynchronous premenopausal-cases-only analysis, a dominant Mendelian gene can explain the breast cancer pattern. A recessive gene is sufficient to describe the breast cancer distribution in the synchronous premenopausal-cases-only analysis. The synchronous all-cases and the asynchronous all-cases analyses add postmenopausal cases of breast cancer to the premenopausal ones, considering any case to be affected. In the asynchronous all-cases analysis, neither the single-locus model nor the mixed model (that is, a major locus plus other factors, genetic and/or cultural) without generation differences in heritability can be rejected by the unrestricted mixed model with generation differences in heritability. For the synchronous all-cases analysis, a mixed model with generation differences in heritability is necessary to explain the breast cancer transmission. Potential sources of error and possible interpretations are discussed.