PHASE II TRIAL OF THE PHOSPHATIDYINOSITOL-3 KINASE (PI3K) INHIBITOR BUPARLISIB (BKM120) IN RECURRENT GLIOBLASTOMA CONDUCTED BY THE IVY FOUNDATION EARLY PHASE CLINICAL TRIALS CONSORTIUM

BACKGROUND: The PI3K pathway is activated in most GBMs and represents a potential therapeutic target. Buparlisib is an oral, pan-Class I PI3K inhibitor that enters the brain at therapeutic concentrations and inhibits the growth of U87 and human glioma tumor sphere models. METHODS: The Ivy Consortium conducted a phase II study of buparlisib in recurrent GBM patients with activation of the PI3K pathway (mutation, homozygous deletion or loss of IHC of PTEN, PIK3CA or PIK3RI mutations, or detectable pAKT). Additional eligibility criteria included radiologic progression, 1st or 2nd relapse, >18 yrs, KPS >60, adequate bone marrow and organ function, no prior bevacizumab or enzyme-inducing antiepileptic drugs. Patients received buparlisib 100mg daily. The study consisted of 2 concurrent cohorts. In Cohort 1 (up to 15 patients) buparlisib was given for 8-12 days prior to surgery for recurrent disease. Patients underwent FDG PET, pharmacokinetic studies, and tumor was obtained for drug concentrations and pharmacodynamic effects. In Cohort 2, 50 patients with unresectable GBM received buparlisib with a primary endpoint of PFS6. RESULTS: 13 patients were enrolled into cohort 1, 50 into cohort 2. Treatment was fairly well-tolerated with no grade 4 toxicities. Grade 3 toxicities were asymptomatic lipase elevation (6), rash (4), hyperglycemia (3), fatigue (4), elevated ALT/AST (2), and 1 each with depression, anxiety, hypophosphatemia, thrombocytopenia and lymphopenia. Analysis of tumor specimens from Cohort 1 showed reduction of pAKT by IHC in 4/6 (67%) of evaluable patients. The combined cohorts showed minimal efficacy with median PFS of 1.8 months and median OS of 10.9 months. Best response was stable disease. Only 1 patient in each cohort achieved PFS6. Of the first 40 patients who underwent exome sequencing, there were 4 PIK3CA (10%), 2 PIK3R1 (5%) and 13 PTEN (33%) mutations. Patients with PTEN loss and or PI3K mutations paradoxically had a worse outcome. CONCLUSIONS: Buparlisib is well-tolerated in patients with recurrent GBM and achieves adequate tumor concentration to inhibit pAKTS473. However, single agent efficacy was minimal. Final correlation of tumor genotype with outcome will be presented. SECONDARY CATEGORY: n/a.