Radiosurgery for Primary and Metastatic CNS Malignancies in the Pediatric Population

Abstract

<h3>Purpose/Objective(s)</h3> Safety and efficacy of stereotactic radiosurgery (SRS) are not well described for pediatric patients. This study presents disease control and toxicity outcomes after SRS for malignant CNS lesions in pediatric patients. <h3>Materials/Methods</h3> Retrospective chart review was performed for 31 pediatric patients treated at a single institution with a non-invasive stereotactic radiosurgery instrument for primary or metastatic CNS malignancies between 2000-2020. Analyses were performed with descriptive statistics, Kaplan-Meier, and proportional hazards modeling. <h3>Results</h3> Twenty-five patients were treated with SRS for focal recurrences of primary CNS malignancies, 1 patient was treated in the early adjuvant setting after initial resection, and 5 patients were treated for brain metastases. Primary CNS histologies included ependymoma (n=14), glioma (n=4), medulloblastoma (n=2), and meningioma (n=2). 65% were WHO grade 3 or 4 at the time of diagnosis. Brain metastases were from non-germinomatous germ cell tumors (n=3), papillary thyroid cancer (n=1), and neuroblastoma (n=1). 71% of patients had prior involved-field external beam radiation to the brain (median dose 59.4 Gy in 33 fractions). Median age at SRS was 14 years (range 4-21). SRS was predominantly performed in a single fraction (median dose 17 Gy, range 15-20) to a total of 42 targets among 29 patients, with a median total gross target volume (GTV) of 1.3 cc. Two patients underwent fractionated radiosurgery to 30 Gy in 5 fractions for larger brain metastases (GTVs 8.5 cc & 23.2 cc). Median follow up after SRS was 44 months (interquartile range 20-72). The 3-year actuarial rates of local control, intracranial progression-free survival, and overall survival after SRS were 77.6%, 33.3%, and 74.6%, respectively. 9 patients (29%) had no evidence of intracranial disease after SRS at a median follow up of 39 months. 6 patients (19%) developed local recurrence at a median of 13 months after SRS. 16 patients (52%) developed distant brain recurrence without local failure at a median of 9 months after SRS. Among primary CNS tumors, there was no significant association of histology, grade, or dose with disease control outcomes on univariate analysis. All patients with subsequent CNS recurrence after SRS underwent salvage therapy, including additional radiosurgery (36%), systemic therapy (23%), resection (9%), or combined modality treatment (32%). In total, 4 patients developed CTCAE grade 2-3 toxicity from SRS, all of which were related to radiation treatment effect and occurred within 1 year of SRS. 2 patients received steroids alone, and 2 underwent invasive procedures (1 shunt revision, 1 resection of nonviable tumor). <h3>Conclusion</h3> SRS for malignant CNS lesions in the pediatric population provides effective local control and is generally well-tolerated. Although there remains a substantial risk of distant CNS failures, all patients were able to receive additional salvage therapy after SRS.