AI-09 * PATTERNS OF SALVAGE TREATMENT FOR HIGH GRADE GLIOMA IN THE TEMOZOLOMIDE ERA

Abstract

PURPOSE: High grade gliomas (HGGs) are aggressive malignancies that recur despite trimodality treatment with surgery, radiation therapy (RT), and temozolomide (TMZ). We analyzed patients treated at our institution for factors associated with mortality in recurrent disease following trimodality therapy. METHODS: From 1/2000 to 9/2012, 437 consecutive patients with HGG (92 Grade III and 345 Grade IV) received trimodality therapy. Bevacizumab (Bev) was given in 11.7%. Initial RT schema was full-course to 56-60 Gy in 76%, whole-brain in 3%, short-course in 12%, and high-dose hypofractionated in 9%. Patient, tumor, upfront and salvage treatment-related factors were assessed. Survival rates were calculated by Kaplan-Meier methods, and Cox-regression modeling was utilized to calculate hazard ratios (HR); all confidence intervals (CI) are 95%. RESULTS: The median follow-up was 14 months; 17 months for surviving patients. The 1-year local control was 36.8% (31.8-41.8%), 59% (50.2-69.8%), and 30.7% (25.3-35.4%) for the entire cohort, Grade III, and Grade IV patients, respectively (P < .001). Salvage surgery was performed in 27.6%, salvage RT in 11.3%, and salvage chemotherapy (CT) in 83.1% of patients. Salvage CT included Bev in 41.6% and TMZ in 52.3% of patients. Factors associated with overall survival (OS) after LF included age at diagnosis (P < .001), KPS (P < .001), initial surgery (P < .001), initial uni- vs. bilateral disease (P < .001), grade (P < .001), initial RT schema (P < .001), salvage RT (P = .024), salvage Bev (P = .006), and salvage TMZ (P < .001). On multivariable analysis, the only treatment-related factor that retained significance for reduced mortality in patients with LF was use of salvage Bev [HR 0.65 (43-.97), P = .036]. CONCLUSIONS: In this large retrospective series of patients with HGG treated with trimodality therapy, local failure was common. Bevacizumab was the only physician-controlled factor associated with improved survival after failure. However, only 11.3% of our cohort received salvage RT. Further controlled studies are necessary to determine optimal salvage therapy.