However, only a fraction of treated patients typically derive clinical benefit. Predictive biomarkers identifying patients most likely to respond would allow for a more targeted approach to treatment and, therefore, would be of significant clinical value. To date, no validated BU 4061T biomarker has been identified for any angiogenesis inhibitor despite extensive investigation. Lambrechts et al recently described efforts to identify predictive biomarkers for bevacizumab: although potential markers have been identified in certain tumor types, as yet none have proven robust. A recent prospective study found an association between low VEGF-A levels and both progression-free survival and overall survival in patients with nonsquamous NSCLC. However, because the study did not include a control arm it was not possible to differentiate between prognostic and predictive value of the biomarker. Motesanib is a potent small-molecule inhibitor of VEGF receptors 1, 2, and 3; platelet-derived growth factor receptor; and Kit, with demonstrated antitumor activity as monotherapy and in combination with chemotherapy. In the motesanib first-in-human study, analysis of potential biomarker candidates showed a strong pharmacodynamic response of placental growth factor and further suggested that increased levels of PLGF from baseline were associated with increased motesanib exposure and possibly correlated with tumor shrinkage. PLGF is a VEGF-A homolog and a VEGFR1 ligand that is up-regulated during hypoxia, and may be involved in pathologic angiogenesis, possibly by increasing the responsiveness of endothelial cells to VEGF-A. The increase in PLGF following motesanib treatment possibly represents a compensatory upregulation in response to VEGF pathway blockade. Subsequent phase 2 studies with motesanib showed a consistent association between increased levels from baseline in PLGF and outcomes across different tumor types, including thyroid cancer, breast cancer, and non�Csmall-cell lung cancer. LDK378 Furthermore, other inhibitors of the VEGF pathway have been known to induce pharmacodynamic changes in PLGF, which, in some cases, have been associated with outcomes including objective response and OS. Taken together, the data suggested that PLGF may serve as a biomarker for the biologic effect of VEGF receptor inhibitors, and as such, it may be a potential biomarker identifying a population most likely to benefit from continued treatment with these agents. The PLGF data collected in motesanib phase 2 studies formed a strong body of evidence that supported further prospective testing of PLGF as a potential biomarker in the large international phase 3 Motesanib NSCLC Efficacy and Tolerability study of motesanib plus carboplatin/paclitaxel versus placebo plus carboplatin/paclitaxel in patients with nonsquamous NSCLC. However, the study did not 
meet its primary endpoint, and PLGF analysis with a validated assay developed specifically as a companion diagnostic test did not reveal an association between change from baseline in PLGF and OS. To date, MONET1 remains the only large, prospective study of a biomarker candidate for an angiogenesis inhibitor. Considering the body of evidence for PLGF as a biomarker for motesanib and the rigorous analysis of data that formed the basis of the PLGF hypothesis for MONET1, the study’s negative biomarker results demonstrate the challenges in the development of a valid predictive biomarker. Here we describe the processes we undertook in an effort to develop PLGF as a pharmacodynamic biomarker for motesanib using an ongoing phase 3 study of motesanib in patients with NSCLC and supporting data from the preceding phase 2 study of motesanib in NSCLC.