Increased risk of neuroectodermal tumors and stomach cancer in relatives of patients with Ewing's sarcoma family of tumors.

Background: Previous studies of the genetic epidemiology of Ewing's sarcoma have shown neither an increased incidence nor a distinct pattern of cancers in family members of Ewing's sarcoma patients.

Purpose: Because of a new biologic and cytogenetic classification of Ewing's sarcoma family of tumors, we wanted to reinvestigate the incidence and distribution of cancers in relatives of probands with Ewing's sarcoma family of tumors.

Methods: Patients treated at the Pediatric Branch and the Radiation Oncology Branch of the National Cancer Institute between 1965 and December 1992, or their next of kin, were asked to complete a questionnaire on the history of cancer in all first- and second-degree relatives. The incidence of cancer in family members was compared with Connecticut Tumor Registry rates specific for sex, age, and 5-year calendar-year intervals. Observed/expected (O/E) ratios, 95% confidence intervals (CIs), and tests of homogeneity were calculated.

Results: Four thousand six hundred seventy-eight family members with 196,640 person-years at risk entered the analysis. Overall, there was no increased risk of cancer (observed 472; O/E = 0.9; 95% CI = 0.8-1.0). However, several tumor types were found in significant excess. These tumors included stomach cancer (observed 34; O/E = 2.0; 95% CI = 1.4-2.8), melanoma (observed 23; O/E = 1.9; 95% CI = 1.2-2.8), brain tumor (observed 18; O/E = 1.9; 95% CI = 1.1-3.0), and bone cancer (observed 7; O/E = 4.2; 95% CI = 1.7-8.6). Risks of these cancers were higher among maternal than paternal relatives, but these differences were not statistically significant. There was a significant deficit of bladder cancer (observed 5; O/E = 0.2; 95% CI = 0.1-0.5) and rectal cancer (observed 0; O/E = 0.0; 95% CI = 0.0-0.1). Second-degree relatives had a significant cancer deficit (observed 389; O/E = 0.9; 95% CI = 0.8-0.95). This deficit was accounted for by the observed deficit of bladder and rectal cancer and is probably related to under-reporting or misclassification of cancer in second-degree relatives. Family members of 10 probands with second malignancies did not have an increased risk of all cancers (observed 20; O/E = 1.2; 95% CI = 0.7-1.8) but had an increased risk of both melanoma (observed 3; O/E = 7.3; 95% CI = 1.5-21.0) and breast cancer (observed 8; O/E = 3.2; 95% CI = 1.4-6.3).

Conclusion: Finding an increased risk of neuroectodermal tumors and stomach cancer in families of patients with Ewing's sarcoma family of tumors suggests that these tumors might share a common etiology. Further studies should try to confirm this hypothesis and to examine if genetic factors may have a role in these families by assessing the mode of inheritance and examining families with multiple affected members.