A CASE-CONTROL STUDY OF THE ETIOLOGY OF EWING SARCOMA

An interview case-control study was undertaken to search for risk factors for Ewing's sarcoma. The 208 cases, aged 5 months to 22 years at diagnosis and all white but one, were identified from hospitals participating in the intergroup Ewing's Sarcoma Study therapeutic trials. Two controls were sought for each case: a sibling control and an age-matched regional population control identified through random-digit dialing telephone procedures. A questionnaire was administered to the parents of cases and controls. Parents were more likely to have smoked during the pregnancy with the case than during the pregnancy with the unaffected sibling. Risks rose with the number of cigarettes the mother smoked per day during the pregnancy. Concepti exposed to less than 1 pack/day were at 3.2 times the risk, and those exposed to 1 pack or more were at 6.7 times the risk of the nonexposed. However, risks associated with smoking were lower and not statistically significant in analyses using the region-matched controls. Hernias, mostly umbilical and inguinal, were diagnosed six times more frequently among the cases compared to region-matched controls. However, hernias occurred in just 10% of cases, and the matched siblings had hernias diagnosed with the same frequency as the cases. An apparent excess of heart disorders among cases versus siblings seems likely to be an artifact of increased medical surveillance of cases. Fathers of Ewing's sarcoma cases were 2 times more likely to work as farmers at the time of conception than fathers of region-matched controls and 3 times more likely to have farming as their usual occupation, but the findings did not reach statistical significance. Ewing's sarcoma was not linked to other factors which were examined: antecedent bone trauma-cancer among relatives; demographic characteristics including religion, education, and income; and anthropomorphic measures (child's birth height or weight or parents' height or weight). Although a wide range of demographic, maternal reproductive, child health, occupational, and family history data were obtained, Ewing's sarcoma remains a cancer of unknown cause.