Event-free Survival (EFS) in Children Who Have Completely Resected Ependymoma or Achieved CR and Are Treated With Post-radiation Maintenance Chemotherapy or Post-radiation Observation Only

Arm II: Randomized to Radiation With Maintenance Chemotherapy

Patients with supratentorial ependymoma (Grade II without GTR1 or Grade III) or any infratentorial ependymoma who have undergone gross or near total resection (GTR or NTR) or achieved CR either after first or second resection or after post-operative induction chemotherapy are randomized to undergo conformal radiotherapy over 6-7 weeks followed by maintenance chemotherapy. Maintenance chemotherapy comprised of vincristine IV on days 1, 8, and 15 of cycles 1-3 only, etoposide IV over 60-120 minutes on days 1-3, cisplatin IV over 1-8 hours on day 1, and cyclophosphamide IV over 30-60 minutes on days 2-3. Treatment repeats every 21 days for 4 courses in the absence of disease progression or unacceptable toxicity. Patients without GTR or NTR at enrollment require induction chemotherapy (see Arm I) and possibly second surgery before randomization.

Arm III: Randomized to Radiation Only

Patients with supratentorial ependymoma (Grade II without GTR1 or ST Grade III) or any infratentorial ependymoma (Grade II or III) who have undergone gross or near total resection or achieved CR either after first or second resection or after post-operative induction chemotherapy are randomized to undergo conformal radiotherapy over 6-7 weeks and then undergo observation.

Analysis denominator · Participants

1. ci Lower Limit

   0.627

   ci Num Sides

   TWO_SIDED

   ci Pct Value

   90.46

   ci Upper Limit

   1.197

   estimate Comment

   Hazard ratio is based on hazard of events for Arm II over hazard of events for Arm III. Confidence interval is stage-wise adjusted for multiple interim analyses.

   group Description

   The study was designed to test the superiority of RT+maintenance vs. RT alone as measured by EFS using the hazard ratio via a 1-sided log-rank test. The planned sample size was 160 subjects per arm. 85 EFS events were needed, which were expected after 8 years of accrual and 2 years of follow-up. At 5% type 1 error, the study had 93% power to detect an increase in 2-year EFS from 75% (RT alone) to 87% (RT+maintenance). The design also incorporated interim analyses for futility and efficacy.

   group Ids

   1. OG000
   2. OG001

   non Inferiority Type

   SUPERIORITY

   p Value

   0.229

   p Value Comment

   P-value is one-sided. After adjusting for interim analyses, the a priori threshold for significance for the final analysis is one-sided alpha=0.031.

   param Type

   Hazard Ratio (HR)

   param Value

   0.866

   statistical Method

   Log Rank