Circulating angiogenic factors as predictors of benefit from bevacizumab (bev) beyond progression in metastatic colorectal cancer (mCRC): Translational analyses from the phase III BEBYP trial.

Abstract

382 Background: TML and BEBYP trials demonstrated that the strategy of prosecuting bev beyond progression is effective in mCRC. Previous analyses from phase II studies showed that a dynamic modulation of plasma angiogenic factors occurs during first-line treatment with chemotherapy (CT) plus bev and a wide variability in soluble Vascular Endothelial Growth Factor Receptor 2 (sVEGFR2) levels was observed at the time of progression. Moving from preliminary analyses in murine models we selected a pool of candidate ligands to be tested in the clinical setting. Methods: sVEGFR2, Placental Growth Factor (PlGF), Platelet-Derived Growth Factor-C, basic Fibroblast Growth Factor, Angiopoietin-2 and sTie-2 were assessed by ELISA on plasma samples collected at baseline in a cohort of 59 patients enrolled in phase III BEBYP trial of 2nd-line CT ± bev beyond progression to a bev-containing first-line regimen. Plasma levels were defined high or low adopting the median as cut-off. Results: A significant interaction between treatment arm and baseline sVEGFR2 levels was observed (p=0.036). Among 30 patients with high sVEGFR2 levels, the prosecution of bev was associated with a significant benefit in terms of PFS (median: 10.4 vs 3.4 months, HR 0.37 [95%CI:0.10-0.58], p=0.0015), that was not evident among 29 patients with low sVEGFR2 levels (5.4 vs 5.0 months, HR 0.98 [95%CI:0.45-2.11], p=0.956). Despite a trend towards a greater benefit from bev among 30 patients with high PlGF levels (HR 0.45 [95%CI: 0.13-0.86]), no interaction between treatment arm and baseline PlGF levels was observed (p=0.210). Combined analysis of sVEGFR2 and PlGF showed that prosecuting bev provided a substantial benefit in PFS in the subgroup with high levels of both ligands (10.5 vs 2.3 months, HR 0.25 [95%CI:0.01-0.45], p=0.043). Conclusions: sVEGFR2 levels at the time of first progression may predict benefit from the prosecution of bev. Impressive results deriving from concomitant testing of sVEGFR2 and PlGF may be affected by excessive subgrouping and should be considered cautiously. Given their potential clinical value these data need prospective confirmation. Clinical trial information: NCT00720512.