Pencil beam scanning proton therapy for pediatric intracranial ependymoma

To assess the clinical outcome and late side effect profile of pencil beam scanning proton therapy (PT) delivered to children with intracranial ependymoma. Between July-2004 and March-2013, 50 patients with intracranial ependymoma (n = 46, grade 3) received involved-field PT at Paul Scherrer Institute (PSI). Median age at time of PT was 2.6 years (range 1.1-15.2). Thirty-six patients had infratentorial and 14 supratentorial ependymomas. Seventeen patients presented with macroscopic residual disease after subtotal resection before starting PT (8 with ≤1.5 cc and 9 with >1.5 cc residual tumor respectively). Forty-three (86 %) patients received post-operative chemotherapy before PT according to protocols; 44 (88 %) patients younger than 5 years required general anesthesia. Median prescribed dose was 59.4 Gy (RBE) (range 54-60) delivered in 1.8-2 Gy (RBE) per fraction. Late toxicity was assessed according to CTCAE v4.0. With a mean follow-up time of 43.4 months (range 8.5-113.7) seven patients experienced local failure (6 with infratentorial tumors and 1 with supratentorial tumor); four of the local failures were in patients with residual disease ≥1.5 cc at the time of PT and 3 without residual macroscopic disease. Five patients died from tumor progression. Actuarial 5-year Local Control rates were 78 ± 7.5 % and 5-year OS rates were 84 ± 6.8 %. Three patients developed grade ≥3 toxicity: 2 developed unilateral deafness (infratentorial tumors infiltrating into the internal acoustic canal), one patient developed a fatal brainstem necrosis. Repeated general anesthesia in children younger than 5 years was delivered without complications. Our data indicate the safety and the effectiveness of PT for pediatric ependymomas. Local control and survival rates are encouraging considering the high grade histology in 92 % of the patients and the number of patients with residual tumor ≥1.5 cc. The rates of late effects compare favorably with published photon-treated cohorts.