UCSF

Abstract Alterations in the CDK4/6 – RB signaling pathway are common causes of dysregulation of the cell cycle in many cancers, including glioblastoma. Palbociclib is an oral, highly selective, reversible inhibitor of CDK4/6, which leads to phosphorylation of RB and cell-cycle arrest. In a two-arm study, we evaluated the efficacy and safety of palbociclib in patients with recurrent glioblastoma. Notable eligibility criteria included confirmation of RB proficiency by IHC; ≤ 3 relapses; KPS ≥ 60; secondary glioblastomas were included; there were no limitation on prior treatments including bevacizumab. Patients were administered oral palbociclib 125 mg daily for 21 consecutive days followed by a 7 day break. Arm 1 planned for 15 patients to receive palbociclib for 7 days prior to indicated surgical resection for progression, followed by palbociclib. Arm 2 planned for 15 patients to receive palbociclib without additional resection. The primary objective was PFS-6, which was hypothesized to be 30% in this heavily pretreated population, and null hypothesis of 10%. Secondary objectives of toxicity, OS, and ORR. Exploratory results included biomarker assessment and pharmacodynamic effects for the surgical Arm 1 patients. A total of 22 patients were enrolled; 6 on Arm 1 and 16 on Arm 2. Median age for all arms was 47.5 years old (range 23 – 78 years old); 54% (12 patients) were male; and the median KPS was 90 (range 60 – 100). Palbociclib was started at first recurrence in 50% (11 patients – 3 in Arm 1 and 8 in Arm 2) of patients; at second recurrence in 36% (8 patients – 3 in Arm 1 and 5 in Arm 2); and at third recurrence in 14% (3 patients, all in Arm 2). Bevacizumab had previously been used in 77% (17 patients – 4 in Arm 1 and 13 in Arm 2) of cases. The trial was stopped early secondary to futility, with 95% (18 of 19) evaluable patients progressing within 6 months of intiating treatment. Four samples were available for immunohistochemistry for Rb and Ki67. There were no consistent changes in Rb expression or cell proliferation when compared to samples from diagnosis. Median progression free survival for all patients was 5.14 weeks (range 5 days – 142 weeks) and median overall survival was 15.4 weeks (range 2 – 274 weeks). Two patients (10%) had treatment related AEs that were grade ≥3. In this trial, palbociclib did not appear to have been an effective treatment for recurrent glioblastoma. However, this was a heavily pretreated patient population and targeting the CDK4/6 pathway may deserve further exploration.