GENETIC-ANALYSIS OF MAMMOGRAPHIC BREAST DENSITY IN ADULT WOMEN - EVIDENCE OF A GENE EFFECT

Background: The appearance of the female breast viewed by mammography varies considerably from one individual to another because of underlyi ng differences in the relative proportions of fat, connective tissue, and glandular epithelium that combine to create a characteristic patte rn of breast density. An association between mammographic patterns and family history of breast cancer has previously been reported. However , this association has not been found in all studies, and few data are available on possible genetic components contributing to mammographic breast density. Purpose: Our purpose was to estimate familial correla tions and perform complex genetic segregation analyses to test the hyp othesis that the transmission of a major gene influences mammographic breast density. Methods: As part of a cohort study (initiated in 1944) of families with a history of breast cancer, the probands' female rel atives who were older than 40 years were asked to obtain a routine mam mogram. The mammograms of 1370 women from 258 independent families wer e analyzed. The fraction of the breast volume occupied by radiographic ally dense tissue was estimated visually from video displays of left o r right mediolateral oblique views by one radiologist experienced in m ammography who had no knowledge of individual relationships to the pro bands. Data on breast cancer risk factors were obtained through teleph one interviews and mailed questionnaires. Unadjusted and adjusted fami lial correlations in breast density were calculated, and complex genet ic segregation analyses were performed. Results: Sister-sister correla tions in breast density (unadjusted and adjusted for age and either bo dy mass index, menopausal status, hormone replacement therapy, waist-t o-hip ratio, number of live births, alcohol consumption, or cigarette smoking status) were all statistically significant (r =.16-.27; all P< .05 [two-sided]). Estimated mother-daughter correlations were smaller in magnitude (r =.01-.17) and not statistically significant. Segregati on analyses indicate that a major autosomal gene influences breast den sity. The mendelian transmission of a dominant gene provided the best fit to the data; however, hypotheses involving the inheritance of eith er a recessive gene or a codominant gene could not be ruled out. The m endelian dominant hypothesis, accounting for 29% of the variability in breast density, suggests that approximately 12% of the population wou ld be expected to carry at least one variant allele of this putative g ene. Women who inherit the variant allele would have a mean breast den sity about twice that of the rest of the population. Conclusions: Our preliminary findings suggest that, in this cohort of women at risk of breast cancer, mammographic breast density may be genetically influenc ed.