IT-30 ReACT: A PHASE II STUDY OF RINDOPEPIMUT VACCINE (CDX-110) PLUS BEVACIZUMAB IN RELAPSED GLIOBLASTOMA

BACKGROUND: EGFRvIII, a constitutively active tumorigenic EGFR deletion mutation, is linked to poor long-term survival. The investigational vaccine rindopepimut consists of the unique EGFRvIII peptide sequence conjugated to keyhole limpet hemocyanin (KLH), delivered intradermally with GM-CSF. Three phase II studies of rindopepimut in newly diagnosed, resected, EGFRvIII+ glioblastoma have shown encouraging PFS and OS. Bevacizumab (BV), an agent with activity in glioblastoma, may augment EGFRvIII-specific immune response and antitumor activity through inhibition of VEGF and its immunosuppressive properties. METHODS: "ReACT" is a Phase II study of rindopepimut plus BV in patients with 1st or 2nd relapse of EGFRvIII+ glioblastoma. BV-naive pts (Group 1; n = 70) are randomized 1:1 to BV plus double-blinded injection of either rindopepimut or control (KLH). BV-refractory patients (progression within 2 months of BV; Group 2/2C, n = 98) receive BV plus open-label rindopepimut. RESULTS: 234/700 (33%) screened patients are EGFRvIII+. 115 patients (Group 1 = 72, Group 2/2C = 43) are enrolled. Primary toxicity is Grade 1-2 injection site reaction. In group 1, for rindopepimut + BV vs. KLH + BV, objective response rate (ORR; investigator-assessed, RANO criteria) is 23% (6/26) vs 12% (3/25) [35% vs. 20% including responses observed at a single time point ("unconfirmed")]. In Group 2/2C (evaluable n = 30), one unconfirmed PR and one sustained PR (patient continues treatment at 15 months) occurred. Two additional patients had pre-study progression >2 months after BV; one maintained a CR for 11.1 months (peak anti-EGFRvIII titer = 1:3,276,800), while the second experienced an unconfirmed CR. Median peak rindopepimut-induced anti-EGFRvIII humoral response is 1:25,600 [range