O10.07 RANDOMIZED PHASE II STUDY OF AXITINIB VS. STANDARD OF CARE IN PATIENTS WITH RECURRENT GLIOBLASTOMA

INTRODUCTION: Glioblastoma (GB) associated neo-angiogenesis is mediated by vascular endothelial growth factor receptor (VEGFR) signal transduction. Axitinib is an oral small molecule tyrosine kinase inhibitor with high affinity and specificity for the VEGF-receptors, approved for the treatment of metastatic renal cell carcinoma. METHODS: Axitinib (5 mg BID starting dose) vs. best alternative choice of therapy was studied in an open label, randomized, phase II clinical trial in patients (pts) with recurrent GB. Six-month progression-free survival (6mPFS) was the primary endpoint using Fleming's single-stage procedure. RESULTS: Between Sep 2011 and Oct 2013, 44 pts who failed surgery, RT and temozolomide were randomized 1:1 at 3 sites. Median age 54y (range 20-79), 33M/11F, WHO-PS 0/1/2/3: 6, 26, 11, and 1 pt. In the control arm 20 pts received bevacizumab and 2 pts received lomustine at standard doses. Axitinib (n = 22) was generally well tolerated. Most common axitinib related AEs consisted of dysphonia (3x G2), fatigue (6x G2, 2x G3), hypertension (5x G2), oral hyperesthesia (7x G2, 1x G3), diarrhea (2x G2, 2x G3), and hypothyroidism (3x G2). In 4 pts axitinib dosing was interrupted and subsequently dose reduced because of toxicity; in 4 other pts axitinib was dose escalated to 7 or 10 mg BID (2 pts each). The confirmed best overall tumor response rate by RANO criteria was 28% in the axitinib arm (2 CR / 4 PR) vs. 23% in the control arm (1 CR / 4 PR). Five additional pts in the axitinib-arm and 1 additional patient in the control-arm had an unconfirmed respons, and respectively 5 and 7 patients had a SD; 1pt in each arm was not evaluable for tumor response. Corticosteroids could be stopped in 4/12 and tapered in an additional 5/12 pts in the axitinib arm. At baseline all pts had an increased uptake on 18F-FET PET and a magnetic resonance spectroscopy (MRS) signal pattern showing an increased level of choline (Cho) and a decreased level of N-acetylaspartate (NAA) at the site of rGB. A decrease (-26 to -100%) of SUVmax/background on 18F-FET PET was documented in 6/7 pts on axitinib at the time of response on MRI. MRS-spectra obtained at the time of response and progression are currently being analysed. After a median follow-up of 10.5 mths, the 6mPFS for axitinib was 22% (95% CI 5-40) vs. 19% (95% CI 2-36) for the control arm. Median PFS and OS were respectively 2.9 vs. 2.6 mths, and 10.3 vs. 7.4 mths for pts treated in the axitinib vs. control arm. CONCLUS