The efficacy of angiogenesis inhibitors in cancer is limited by resistance mechanisms that are poorly understood. Notably, instead of inducing angiogenesis, some cancers vascularize by the non-angiogenic mechanism of vessel co-option. Here we show that vessel co-option is associated with a poor response to the anti-angiogenic agent bevacizumab in patients with colorectal cancer liver metastases. Moreover, we find that vessel co-option prevails in human breast cancer liver metastases, a setting where results with anti-angiogenic therapy have been disappointing. In our preclinical mechanistic studies, we show that cancer cell motility mediated by the Arp2/3 complex is required for vessel co-option in liver metastases in vivo and that combined inhibition of angiogenesis and vessel co-option is more effective than inhibiting angiogenesis alone in this setting. Vessel co-option is therefore a clinically relevant mechanism of resistance to anti-angiogenic therapy and combined inhibition of angiogenesis and vessel co-option may be a warranted therapeutic strategy.

Background:Liver metastases present with distinct histopathological growth patterns (HGPs), including the desmoplastic, pushing and replacement HGPs and two rarer HGPs. The HGPs are defined owing to the distinct interface between the cancer cells and the adjacent normal liver parenchyma that is present in each pattern and can be scored from standard haematoxylin-and-eosin-stained (H&E) tissue sections. The current study provides consensus guidelines for scoring these HGPs.Methods:Guidelines for defining the HGPs were established by a large international team. To assess the validity of these guidelines, 12 independent observers scored a set of 159 liver metastases and interobserver variability was measured. In an independent cohort of 374 patients with colorectal liver metastases (CRCLM), the impact of HGPs on overall survival after hepatectomy was determined.Results:Good-to-excellent correlations (intraclass correlation coefficient >0.5) with the gold standard were obtained for the assessment of the replacement HGP and desmoplastic HGP. Overall survival was significantly superior in the desmoplastic HGP subgroup compared with the replacement or pushing HGP subgroup (P=0.006).Conclusions:The current guidelines allow for reproducible determination of liver metastasis HGPs. As HGPs impact overall survival after surgery for CRCLM, they may serve as a novel biomarker for individualised therapies.