Low-dose craniospinal radiation therapy for medulloblastoma

Abstract

At the University of California, San Francisco, 65 children with medulloblastoma of the posterior fossa were treated postoperatively with craniospinal irradiation; the dose to the posterior fossa was 54 Gy. The 26 children initially treated had only radiation therapy, receiving 30 to 40 Gy to the spine and 40 to 50 Gy to the brain. Subsequently, 39 children were treated with low-dose craniospinal irradiation and chemotherapy; 24 to 30 Gy was directed to the whole brain and 24 to 26 Gy to the spinal axis. Chemotherapy generally consisted of procarbazine just before, and hydroxyurea during, radiation therapy. Poor-risk and good-risk patients (defined by tumor resection < 75% or >75%, positive or negative myelogram, positive or negative cerebrospinal fluid analysis, age < or >2 years, respectively) were evenly distributed between the low-dose and high-dose craniospinal radiation therapy groups. Median follow-up was 51 months (range, 24 to 228 months). Kaplan-Meier actuarial survival for all patients was 73% at 5 years, 70% at 10 years. Freedom from disease progression was 68% at 5 years, 65% at 10 years. Whereas poor-risk patients treated with low-dose craniospinal irradiation and chemotherapy had a 5-year survival of 58% and a 5-year freedom from disease progression of 39%, those figures in the comparable good-risk patients were 83% and 77%, respectively. For both good-risk and poor-risk patients, the posterior fossa was the primary site of recurrence. Tumors recurred in the frontal region, probably under blocks, in three patients receiving low-dose irradiation and in two receiving the higher dose. Reducing the dose of whole-brain and spinal irradiation and giving chemotherapy did not result in a higher rate of recurrence in the brain or spinal cord. Intellectual and social function appeared better in patients receiving the lower dose. We did not study whether chemotherapy benefitted good-risk patients. Craniospinal axis irradiation at a lower dose than conventionally used does not compromise local control or survival in patients with medulloblastoma, and may reduce toxicity.