Phase 3 Randomized Trial of Postirradiation Chemotherapy in Patients with Newly Diagnosed Ependymoma: A Report from the Children's Oncology Group.

BACKGROUND:The utility of chemotherapy for childhood ependymoma is uncertain without a prior randomized trial. METHODS:ACNS0831, a multicenter phase 3 randomized study conducted through the Children's Oncology Group (COG), included patients 1-21 years with newly diagnosed intracranial ependymoma. Patients with complete/near-total resections (GTR/NTR) or complete response (CR) to induction therapy were randomized to radiation (RT) alone or RT followed by chemotherapy (RT-CHEMO). Primary outcomes were event-free survival (EFS) and overall survival (OS). Due to anticipated noncompliance with chemotherapy, an "as treated" analysis was planned. Patients with subtotal resection (STR) were non-randomly assigned to RT-CHEMO. Grade 2 supratentorial tumors with GTR or CR to induction therapy were observed. RESULTS:Of 449 eligible patients, 325 with GTR/NTR or CR were randomized. Five-year EFS was 63.7% (95% CI: 55.1%-71.1%) for RT only (n = 161) versus 69.2% (60.8%-76.3%) for RT-CHEMO (n = 164) (1-sided log-rank P = .299, HR = 0.866). Five-year OS was 86.9% (79.8%-91.6%) for RT only versus 88.3% (81.8%-92.6%) for RT-CHEMO (1-sided log-rank P-value = .172, HR = 0.757). The "as treated" and "as randomized" analysis results were similar. Sixty-three subjects with STR were assigned to RT-CHEMO; 5-year EFS was 33.6% (22.1%-45.5%) and OS 74.0% (60.5%-83.5%). Supratentorial grade 2 tumors with GTR or CR were observed with a 5-year EFS of 66.9% (49.0%-79.7%) and OS of 100%. Molecular classification was provided for 94% (n = 422) of all subjects. CONCLUSIONS:Primary analysis showed no benefit for maintenance chemotherapy. Further follow-up is important to assess its effect on late relapses. This is the largest cohort of molecularly classified ependymomas treated on a Phase 3 randomized trial.