Risk characterization of familial cancer using the Swedish Family-Cancer database with a special reference to breast cancer

This thesis reports on epidemiological studies on family related cancer in Sweden. The Family-Cancer Database was constructed in 1996 from several national registries. Family relations were available from the Generation register, cancer cases from the Cancer registry and socioeconomic information from Censuses. The sources were linked together using personal ID-numbers and then unidentified. A particular advantage of the database is that the contribution of both parental lineages on cancer risk can be examined. The main method was to calculate relative incidence measures using birth cohorts to compare the study groups selected by index cases, probands. Cancer risk in young and middle aged offspring was increased approximately 1.1 times from the usual risk when one of the parents had cancer, if both parents had cancer, the risk was 1.3-1.4 times higher. When offspring cancer risk was analyzed by index cancer, same as in parent, the risk elevated to 2-3 times for many cancer sites like colorectum, skin, melanoma and endocrine glands, or even to 4-8 times for uterus, testis and thyroid. Analysis of familial cancer risks between non-index sites provides etiologic understanding on genetic and environmental risk factors. Novel findings associated parental-offspring sites of pancreas-breast, breast-testis and uterus-nervous system. For these, the familial relative risks were modest, but increased in those whose parents were diagnosed before ages 50. Mutations in known cancer-related genes may explain some of these findings. For melanoma, pancreatic and liver cancer, environmental factors are important etiologic factors and may contribute to the observed familial effects. The molecular genetic explanation may be that rare dominant single genes increase susceptibility at many sites, or that overlapping sets of genes control susceptibility at multiple sites. The proportion of familial breast cancer among all breast cancers diagnosed before age 54 was 8.7%. The familial relative risk was about 1.8, but is likely to decrease to about 1.5 or less in the aging population. The higher familial relative risks were evident in young women, being 4.0 when both the mothers and their daughters were diagnosed at ages before 40 years. In mothers and daughters, ovarian cancer risk was increased in combination with breast cancer. Breast cancer was studied in three of the papers. Cancer cases can be divided into familial cases and sporadic cases, individuals in these groups are at high risk, medium risk or low risk. The high risk cases in the familial group may have hereditary causes. If new cancer genes are found they could affect both the familial and sporadic groups, especially with low penetrance or gene-environment interaction as expected features. Family studies are useful to distinguish both hereditary and environmental risk factors and their possible action together