Abstract P2-16-04: Plasma Biomarker Analyses in the AVADO Phase III Randomized Study of First-Line Bevacizumab + Docetaxel in Patients with Human Epidermal Growth Factor Receptor (HER) 2-Negative Metastatic Breast Cancer

Abstract

Abstract Background: Inhibition of vascular endothelial growth factor (VEGF) signaling by bevacizumab (Bv; Avastin) combined with chemotherapy (chemo) is associated with improved clinical benefit in several malignancies, including metastatic breast cancer (mBC). Bv+chemo was evaluated in HER2 negative mBC in AVADO where baseline biomarker (BM) sampling was included as an optional element to identify prognostic and predictive biomarkers. Different angiogenic markers have been evaluated in trials containing Bv, with most trials having focused on circulating VEGF. Thus far, VEGF seems to have more a prognostic than predictive value as demonstrated across different indications (lung, renal, and colorectal cancer [Bernaards et al. ASCO 2010]). Trends for the predictive value of intracellular adhesion molecule-1 (ICAM-1) have been noted in two NSCLC studies (E4599 and AVAiL) and for basic fibroblast growth factor in AVAiL. The value of angiogenic plasma markers has not yet been established in mBC. Material and Methods: In AVADO, 736 mBC patients (pts) were randomized to docetaxel 100mg/m2 plus either Bv7.5mg/kg, Bv15mg/kg or placebo. Primary endpoint was progression-free survival (PFS). Baseline plasma samples were analyzed for VEGF, ICAM-1, VEGF Receptor 1 and 2 (VEGFR1, 2), E-selectin and vascular cell adhesion molecule-1 using a novel ELISA based multiplex assay. Samples for BM analysis were available for 396 pts, equally distributed across treatment arms. Median levels of these markers were pre-specified as a cut point to categorize pts as low or high. Pre-treatment values were correlated with PFS (before start of subsequent antineoplastic therapy) using simple and multiple regression approaches as well as subgroup analyses. Results: All subgroups had HR point estimates in favor of the Bv arms, with the exception of pts with low VEGFR2 levels at a Bv dose of 7.5mg/kg. In general, high plasma levels of VEGF and VEGFR2 were associated with better treatment effect in this study. Low ICAM-1 levels did not show the trend towards a larger treatment effect previously seen in E4599 and AVAiL. Further analyses will be presented. Discussion: We report the first plasma BM data in a phase III placebo-controlled randomized study in mBC with Bv. In this subset analysis, using a novel ELISA assay, VEGF and VEGFR2 were identified as promising candidates for predictive value for PFS. A similar observation has been made in an independent BM analysis of pancreatic cancer (AVITA; data on file). Plasma data have so far not shown these trends in mCRC and mNSCLC. Further prospective evaluation of these markers is warranted. Authors D.W.M. and S.L.d.H. contributed equally. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-16-04.