Focal vs. Craniospinal Irradiation in Multi-Modality Therapy for Atypical Teratoid/Rhabdoid Tumor (ATRT): Results From the Pediatric Proton/Photon Consortium Registry

Abstract

CNS Atypical teratoid rhabdoid tumor (ATRT) are rare with poor outcomes. Despite known leptomeningeal spread no consensus exists regarding focal or craniospinal radiation (CSI), typically given after surgery and chemotherapy. We queried the Pediatric Proton/Photon Consortium Registry (PPCR), which prospectively collates tumor, treatment, and outcome data from all consenting pediatric patients treated with radiotherapy (RT) at 19 institutions, for impact of RT dose and volume.PPCR identified 65 evaluable ATRT patients (2012-2019). Overall survival (OS) analysis was performed using Kaplan Meier and log-rank P-values. Cox proportional hazards regression was used to determine hazard ratios among variable groups. Median follow-up was calculated using the reverse Kaplan Meier method.Median age was 6.7 years (range 2.6-21.4) and median follow-up was 23.4 months (1.3-70.7). Patients had surgery, chemotherapy and proton radiation. Thirty-three (51%) had initial gross total resection (GTR). Stage was M0 (88%), M+ (11%), or unknown (1%). Two-year OS (n = 65) was 64%; for M0 patients 65% and for M+ patients 38% (P = 0.5). For M0 patients, 48 (84%) had focal RT (median 54 Gy, 50.4-55.8 Gy) and 9 (16%) had CSI (median 36 Gy, 23.4-39.6 Gy) followed by boost (median 54 Gy, 52.2-55.8 Gy). Two-year OS (M0 patients) for focal RT was 63% and for CSI was 80% (HR 1.9, P = 0.5). Two-year EFS (M0) for focal RT was 51.9% and for CSI was 80% (HR 2.8, P = 0.3). There were 14 recurrences (4 local, 8 leptomeningeal, 2 combined) which occurred in M0 patients receiving focal RT compared to no recurrences for 9 M0 patients with CSI (P = 0.095). Local control at 2 years (n = 54, Table) with primary dose ≥ 54Gy (n = 33) was 100% compared to 84.2% for 50.4Gy (n = 21, P = 0.02). Local control with GTR (upfront or re-resection) was similar (n = 51, Table) to lesser surgery extent. In 7 M+ patients, 3 with focal RT are alive and 2 of 4 with CSI are alive. For all patients, risk of death (P = 0.8) and progression (P = 1.0) were not different for focal versus CSI. Overall risk of death was also not significant achieving GTR versus subtotal resection (STR) (HR 0.85, P = 0.7).To our knowledge, this is the largest series of ATRT treated with RT and hypothesis generating for optimal radiation field (CSI or focal) in M0 patients and cumulative dose (50.4Gy or 54Gy). Practice patterns strongly favored focal RT and 54Gy showed improved local control. CSI may reduce risk of recurrence but few patients received CSI. Analyses of chemotherapy selection are pending. Further study of radiation field in ATRT will be challenging given prohibitive toxicity of CSI in infant brain tumors. Patterns of failure analyses of ACNS 0333, using high dose chemotherapy and focal RT in M0 ATRT with encouraging results, are warranted and in progress.