FAMILIAL CLUSTERING OF COLON, BREAST, UTERINE, AND OVARIAN CANCERS ASASSESSED BY FAMILY HISTORY

The aggregation of colon, endometrial, ovarian, and possibly breast ca ncers in families has been described as a ''cancer family syndrome'' ( now called Lynch syndrome II). To determine if the familial clustering of these malignancies was more common in women with cancer than witho ut, we analyzed data from the Iowa Women's Health Study (IWHS), a popu lation-based sample of 41,837 women aged 55-69 years. Self-reported in formation was collected on history of colon, uterine, ovarian, and bre ast cancers in female first-degree relatives. A family history of canc er of the breast (odds ratio [OR] = 1.4), colon (OR = 1.3), and uterus (OR = 1.3), but not ovary (OR = 1.2), was significantly more common a mong women with a personal history of any of these four cancers (all P < 0.05); the pattern of the ORs suggested strongly that the clusterin g tended to be site-specific. Age-adjusted relative risks (RR) of inci dent colon cancer over 5 years of follow-up (N = 237) were calculated with regard to family history. Colon cancer incidence was increased am ong women with a family history of breast (RR = 1.3), uterine (RR = 1. 4), colon (RR = 1.5), and ovarian (RR = 1.3) cancers, although none of the risk estimates achieved statistical significance. RR was, however , significantly related to the number of different cancer sites report ed among family members (P(trend) = 0.008). These data on a representa tive sample of postmenopausal women suggest that family histories of c olon, breast, uterine, and ovarian cancers are associated with an incr eased risk of cancer at the same site, but provide little support for the hypothesis that Lynch syndrome II is a non-random occurrence. (C) 1993 Wiley-Liss, Inc.