Radiomic Features Extracted from Magnetic Resonance Imaging (MRI) Are Associated with Clinical Outcomes in Low-Grade Glioma

Abstract

Angeles (UCLA). Materials/Methods: 43 patients diagnosed with AM treated with RT at UCLA from 1999-2015 were identified and retrospectively analyzed. We noted that the definition of grade II AM changed in 2007. Patients were selected based on the grading system during the year they were diagnosed. Data on age, race, gender, surgical type, radiation dose and modality were investigated. Estimates of relapse-free survival (RFS) and overall survival (OS) were performed using the Kaplan-Meier method. Results: 30 patients received adjuvant radiation therapy (ART) after surgery at the initial diagnosis. Of those, 14 had gross total resection (GTR), 16 had subtotal resection (STR). 13 received salvage radiation therapy (SaRT) for recurrent AM following previous surgery. Of those, 9 received GTR at initial diagnosis, 4 underwent STR. The median age at diagnosis was 59, range 38-79. There were 31 Caucasian, 7 Asian, 3 African American, and 2 Hispanic. There were 21 females and 22 males. The median follow up was 32.5 months, range 1.5-108. In the ART group, 1 patient received stereotactic radiosurgery (SRS) and 29 patients received fractionated external beam radiotherapy (EBRT). The median dose was 55.8 Gy (range 48.6-59.4). The single SRS dose was 12 Gy, which is noted to be unusually low for SRS to AM. In the SaRT group, 4 patients received SRS and 9 patients received fractionated EBRT. The median dose was 54 Gy (range 50.4-60); the median SRS dose was 14 Gy (range 14-16). In the ART group, the RFS at 3 and 5 year was 90% and 78%, respectively; the OS at 3 and 5 year was 95%. In the SaRT group, the RFS at 3 and 5 year was 62.5%; the OS at 3 and 5 year was 100%. Compared to historic controls reporting ART for AM, the RFS and OS at 5 years range from 50-92% and 67-84%, respectively. We performed an analysis using the Surveillance, Epidemiology, and End Results Program (SEER) database evaluating patients with grade II AM who underwent GTR without ART between 2004-2010. The 3 and 5 year OS were 89% and 84%, respectively. Conclusion: ART for AM at our institution provides similar RFS and OS compared to historical controls published in the literature. With further data collection, we plan to evaluate the outcomes of these patients based on surgical resection, STR vs GTR, to determine the role of ART in AM. Patients from the SEER database with GTR had high 3 and 5 year OS. SaRT for resected AM that recurred also provided excellent OS, although RFS was not as high in the adjuvant setting. The role of ART in AM after GTR remains controversial. However, there appears to be a benefit to ART vs SaRT for local control and further studies are necessary to evaluate this potential benefit.