Central nervous system involvement in Ewing's sarcoma

Advances in the treatment of patients with Ewing's sarcoma have spurred interest in the treatment of 'sanctuary' areas much in the fashion of prophylactic CNS treatments for children with acute leukemia. The Registry contained records of 134 patients with Ewing's sarcoma treated at military hospitals between 1950 to 1977. There were 102 male and 32 female patients. Age at diagnosis ranged from 1.5 to 54 yr (mean, 17.4 yr). Primary treatments were diverse. Most patients were treated with radiation or surgery directed to the primary tumor. Median survival was 15 mth. Twenty-two patients (16%) had suspected or proven CNS metastases. Nine unautopsied patients had clinically suspected CNS lesions based on symptoms, signs, and noninvasive tests. Autopsy reports were available on 44 patients. Thirteen were found to have CNS involvement. Nine of these had CNS involvement as a result of direct extension from bone metastases. One patient had primary dural involvement. Only three patients had discrete parenchymal brain metastases. Tumor was widespread at the time of first suspicion of neurologic involvement in all 22 cases. Although the authors found an overall prevalence of suspected and proven CNS involvement of 16%, our data indicate that primary parenchymal brain metastases are considerably less frequent. Central nervous system involvement most frequently occurred as a result of direct extension from skull or vertebral lesions. The low prevalence of CNS metastases in the subgroup of patients with an extended disease-free survival and the absence of CNS relapse strongly suggest that isolated brain and leptomeningeal involvement is distinctly uncommon. Furthermore, clinical, radiologic, or autopsy evidence of discrete CNS lesions was found in only three of the entire group of 134 patients. The data support the contentions of Rosen and others that CNS prophylaxis is not necessary in this disease.