EPIDEMIOLOGY OF OSTEOSARCOMA AND EWINGS-SARCOMA IN CHILDHOOD - A STUDY OF 305 CASES BY THE CHILDRENS-CANCER-GROUP

BACKGROUND. The Children's Cancer Group conducted a case-control study to determine the role of a broad range of environmental and familial factors in the etiology of Ewing's sarcoma and osteosarcoma in childre n. These factors included radiation exposure and, for children with os teosarcoma, parental exposure to beryllium. METHODS. The parents of 15 2 children with osteosarcoma and 153 children with Ewing's sarcoma wer e interviewed by telephone. Controls were obtained by random digit dia ling and were matched to cases by age and race. RESULTS. Female osteos arcoma patients had earlier onset of breast development (age 11.4 vs. 11.8 years, P = 0.03) and menarche (age 12.1 vs. 12.5 years, P = 0.002 ) but no significant differences in growth, whereas male osteosarcoma patients were similar in age at the onset of secondary sexual characte ristics but reported significantly less weight gain during their growt h spurt (6.6 vs. 11.7 kg, P = 0.003). For children with Ewing's sarcom a, the growth spurt began earlier (age 12.1 vs. 12.7 years, P = 0.12) and resulted in less weight and height gain (5.2 vs. 9.7 kg, P = 0.002 , and 10.2 vs. 12.7 cm, P=0.02, respectively) for males, but no differ ences were observed among females. For factors not related to growth a nd development (including a wide range of occupational, medical, and h ousehold exposures), there was little evidence of an etiologic role wi th respect to either tumor type. CONCLUSIONS. Differences between case s and controls with respect to growth and development showed no consis tent pattern. This study did not identify any important risk factors f or either type of childhood bone tumor. Cancer 1998;83: 1440-8. (C) 19 98 American Cancer Society.