Lisavanbulin in patients with recurrent glioblastoma: Phase 2a results and a consolidated analysis of response-predictive biomarkers

2044 Background: Lisavanbulin (BAL101553, prodrug of BAL27862) destabilizes microtubules and promotes tumor cell death by modulating the spindle assembly checkpoint. BAL27862 is a lipophilic small molecule shown in rodents to penetrate the brain, with antitumor activity in orthotopic glioblastoma (GBM) models. In the Phase 1 part of this study 1 (NCT02490800), 2 of 5 patients with recurrent IDH-mutated G4 astrocytoma treated at active dose levels (15–30 mg/day) showed long-lasting objective responses and strong end-binding protein 1 (EB1) expression in GBM tissue by IHC. EB1 is a regulator protein on microtubules. Methods: The objective of the Phase 2 study was to investigate prospectively the response-predictive value of EB1, and to identify RNA-based response signatures in patients with recurrent GBM. A Simon’s two-stage design was used with an objective response rate (ORR) ≥ 2/9 required in Stage 1 to enable a final ORR ≥ 6/19. A prescreening program identified patients with EB1-positive archival GBM tissue. All patients received 25 mg oral lisavanbulin once daily. RNA-seq was performed on archival GBM tissues. Results: GBM tissue samples from 64 of 629 patients (10.2%) obtained from 13 sites in four countries were EB1-positive, and 18 of these patients received lisavanbulin. Of 9 patients with measurable disease evaluable for response in Stage 1, one patient had a partial response, and another had a 44% target lesion area reduction. Despite sustained activity in these patients, formal stage transition criteria were not met, and the study was closed. IHC testing for EB1 did not show sufficient enrichment for response, but it is thought that IDH status may play a role. In addition, RNA-seq analyses identified a five-gene signature that is distinct from expression patterns observed with EB1-IHC positivity, and predicts current responses irrespective of IDH status. This signature is characterized by homeobox gene downregulation, which may be implicated in the control of microtubule dynamics. Conclusions: This Phase 2a study supports previous study results that lisavanbulin is associated with durable responses and clinical benefit in a subset of patients with GBM. RNA-seq analyses of GBM samples suggest further evaluation of the lisavanbulin predictive response signature. 1 Lopez et al, JCO 2019;37,15 suppl, 2025. Clinical trial information: NCT02490800 . [Table: see text]