Abstract BACKGROUND Bevacizumab, a monoclonal antibody targeting vascular endothelial growth factor A (VEGF-A), has demonstrated activity in the treatment of recurrent malignant glioma. High response rates have been observed, but particularly in WHO grade 3 gliomas, efforts to identify predictors of clinical response have been limited. Predictive markers and prognostic models are required in order to individualize treatment for this patient population. The primary endpoint of this study was to identify predictive biomarkers associated with response to bevacizumab therapy. The secondary endpoint was to identify prognostic factors associated with progression-free survival (PFS) and overall survival (OS). METHODS A total of 62 consecutive, recurrent grade 3 glioma patients were retrospectively evaluated. Eligible patients from our center had a WHO performance status of 0-2 and were administered bevacizumab and irinotecan between December 2005 and November 2014 according to a previously published clinical protocol. Baseline factors screened for potential prognostic and predictive value included: Age, gender, PS, WHO grade 3 diagnosis, tumor size and location, multifocal disease, extent of resection, number of prior chemotherapy regiments, response to prior chemotherapy, first-line treatment, number of previous recurrences, neurological deficit, corticosteroid use, necrosis, vascular proliferation, neutrophil-to-lymphocyte ratio, and expression of p53, EGFR, MIB-1, MGMT, IDH1 and ATRX. Candidate factors with p-values below 5% were considered for multivariate analysis. Prognostic models were generated by logistic regression and Cox regression, modelling response and survival endpoints. RESULTS Twenty-two patients (35.5%) demonstrated a response according to the RANO criteria. Responders had significantly prolonged OS (p = 0.007) and trended toward longer PFS (p = 0.067) as compared to non-responders (OS: 12.4 vs 4.3 months, PFS: 5.6 vs 3.2 months). Presence of necrosis (OR: 0.17, CI: 0.04-0.68, p = 0.012) and a WHO performance status (PS) of more than 1 (OR: 0.04, CI: 0.002-0.89, p = 0.042) were more common in non-responders than responders. Female gender (HR: 0.48, CI: 0.28-0.82, p = 0.008) and a PS of 0-1 (HR: 0.20, CI: 0.10-0.41, p < 0.0001) were identified as positive prognostic factors for PFS. Immunohistochemical p53 negativity (HR: 0.47, CI: 0.26-0.87, p = 0.016) and low PS (HR: 0.17, CI: 0.07-0.39, p < 0.0001) correlated with extended OS and these factors were included in a prognostic model. CONCLUSIONS A favorable baseline PS and absence of necrosis were positively associated with response to bevacizumab treatment in recurrent grade 3 glioma patients. Low PS, female gender and p53 negativity are prognostic of improved outcome in this patient group. Citation Format: Anders Toft, Thomas Urup, Ib J. Christensen, Signe R. Michaelsen, Babloo S. Lukram, Kirsten Grunnet, Michael Kosteljanetz, Vibeke A. Larsen, Ulrik Lassen, Helle Broholm, Hans S. Poulsen. Prognostic and predictive biomarkers in recurrent WHO grade 3 malignant glioma patients treated with bevacizumab and irinotecan. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr B3.
Read full abstract