Molecular profiling of OCCC tumor tissues has revealed several potential prognostic biomarkers, predictors of treatment response, and therapeutic targets 9–14. For instance, studies have reported the hyper-activation of several signaling pathways including hypoxia-inducible factor 1α (HIF-1α)/VEGF and IL-6/STAT3 pathways in OCCC 15–17. Mabuchi et al. reported that OCCC cells under intratumoral hypoxia strongly expressed VEGF and that Bev demonstrated anti-tumor efficacy against OCCC both in vitro and in vivo 18. However, they also suggested that VEGF may not be a reliable biomarker for predicting Bev sensitivity. Therefore, although anti-angiogenic treatment by Bev appears promising based on pathogenesis, there is still no widely recognized biomarker predictive of its clinical efficacy. Here we demonstrate that high IL-6 or low Ang1 may be such predictors.
Upregulation of IL-6 and related pathway mediators as well as the anti-tumor efficacy of IL-6 pathway inhibition have been reported in OCCC 5,6,19,20. In addition, several studies have documented significant associations between poor OCCC prognosis and high IL-6 in tumor or serum samples 5,20,21. Anglesio et al. reported upregulation of the IL-6/STAT3/HIF pathway and therapeutic responses to the anti-angiogenic agent sunitinib in two chemotherapy-resistant OCCC cases 5. However, the phase II GOG-254 trial evaluating this multi-receptor tyrosine kinase inhibitor for the treatment of persistent or recurrent OCCC found minimal clinical efficacy 22. Recently, a retrospective biomarker analysis revealed longer survival by Bev-treated patients with high plasma IL-6 level compared to those with lower plasma IL-6 8. Based on these findings, we evaluated the influence of IL-6 on Bev anti-angiogenic and anti-tumor efficacies using an in vitro OCCC model to reveal potential molecular mechanisms, and further examined the association of IL-6 expression with that of the pro-angiogenic factor Ang1 among OCCC patients to help identify good candidates for Bev treatment.
In general, IL-6 has been shown to influence angiogenesis as pro-angiogenic factor 23. For instance, transgenic mice engineered to overexpress IL-6 exhibited hypervascularization of the cerebellum 24, while mice deficient in IL-6 exhibited reduced angiogenic responses to wound injury 25. Further, IL-6 induced VEGF production by tumor cells and ensuing angiogenesis 26,27. In the current study, IL-6 signal blockade reduced the anti-angiogenic activity and associated anti-tumor activity of Bev in vitro co-culture system with OCCC tumors and HUVEC. However, the moderate alteration of VEGF production by OCCC cells under IL-6 blockade led us to speculate that some other angiogenic factor(s) were responsible for the observed attenuation of Bev anti-angiogenic activity. Several VEGF-independent angiogenic mechanisms of IL-6 have been reported, including Ang1 modulation, and Kayakabe et al. reported that IL-6 could destabilize angiogenesis through inhibition of Ang1 signaling in a co-culture model of rheumatoid arthritis 28. Additional analyses in our model suggested that IL-6 promoted the anti-angiogenic activity of Bev by suppressing Ang1 release from OCCC cells. Furthermore, the association of high IL-6 expression with low Ang1 expression was found in a series of OCCC patient specimens. According to these results, Ang1 may interfere with Bev function in OCCC (Fig. 5).
Angiopoeitin-1 is thought to facilitate vessel stabilization and its signaling is thought to support the induction and growth of tumor vasculature even under VEGF blockade, resulting in improved tumor perfusion 29. Huang et al. also reported that Ang1 protected the tumor vasculature from regression, increased vessel caliber, and induced the recruitment of mural cells under anti-VEGF treatment 30. Moreover, Casanovas et al. reported that tumors showing continued progression during anti-VEGF receptor-2 antibody treatment maintained substantial Ang1 expression 31. According to these evidence, Ang1 suppression may be important for Bev to exert its efficacy. In the current study, we found that IL-6 blockade enhanced Ang1 production by tumor cells, while siRNA-induced Ang1 knockdown restored the impaired anti-angiogenic function of Bev under IL-6 blockade. Collectively, IL-6 signaling could reduce the tumor angiogenic function of Ang1 and relatively increases that of VEGF, which in turn enhances Bev function in OCCC cells.
The utility of IL-6 as a potential predictor of anti-angiogenic drug response is still debated. Earlier reports suggested that high IL-6 levels could be predictive of survival benefit from Bev treatment in patients with EOC 8, metastatic renal cancer 32, and metastatic colorectal cancer 10. In contrast, however, others reported that low IL-6 levels were associated with better Bev treatment response against hepatocellular carcinoma 33, pancreatic cancer 34, and metastatic colorectal cancer 35. In addition, a recent clinical study reported that aflibercept, a recombinant fusion protein that blocks the VEGF pathway in advanced EOC, was more effective in patients with low IL-6 levels 36. Differences in cellular IL-6 response among tumor types, study design, or clinical stage may explain these discordant results. Further research efforts are warranted to elucidate how IL-6 modulates Ang1 and the prognostic utility of IL-6 as a biomarker for anti-angiogenic drug response.
In conclusion, this study demonstrated that IL-6 enhanced the anti-angiogenic efficacy of Bev by suppressing Ang1 in addition to increased VEGF production in OCCC cells. Furthermore, the current study provides a strong rationale for prospective clinical trials of anti-angiogenic therapy for EOC including OCCC considering about IL-6 as well as Ang1 measurements to evaluate its prognostic utility.