Preradiation chemotherapy for newly diagnosed childhood brain tumors. A modified Phase II trial

A poor-risk population of children with primary malignant central nervous system (CNS) tumors, other than gliomas, can be identified by their young age, by the presence of disease dissemination at diagnosis, and possibly by subtotal resection of the primary tumor. These children require at least neuraxis radiation therapy and possibly chemotherapy for disease control. Unfortunately, once neuraxis radiation is administered, tolerance of subsequent chemotherapy is limited. The authors have explored a multimodal treatment approach in 14 poor-risk patients initially consisting of a modified Phase II chemotherapy trial followed by neuraxis radiation. The diagnoses were medulloblastoma (5), pineoblastoma (3), cerebral primitive neuroectodermal tumor (3), germinoma (2), and choroid plexus carcinoma (1). Eleven patients had disseminated CNS disease, and two had bone marrow involvement at diagnosis. Nine patients received 2 courses of intravenous cyclophosphamide (80 mg/kg) alone over 8 weeks, and five others received three daily doses of intrathecal Ara-C (50 mg/m2) and oral hydroxyurea (40 mg/kg) with each course of cyclophosphamide. There were four complete responses (two dysgerminomas, one pineoblastoma, and one primitive neuroectodermal tumor), one partial response (medulloblastoma), and three mixed responses (two medulloblastomas, one pineoblastoma) to chemotherapy alone, for a response rate of 57%. Twelve patients subsequently tolerated the planned dose of neuraxis radiation. The median survival of all patients was 11 months, and seven of eight deaths were related to recurrent disease. The hematologic toxicity was appreciable, and one death resulted from gram-negative septicemia. Through the use of this type of Phase II trial, valuable information can be obtained on the response rates to specific chemotherapy agents administered prior to radiation. Although cyclophosphamide alone was an active agent in this context, these treatment regimens did not have an important affect on survival.