NCI Logo
NCI Logo

COX-2 – A Novel Target for Reducing Tumor Angiogenesis and Metastasis

Dual therapy with axitinib and celecoxib prevented the outgrowth of spontaneous breast cancer metastasis in the lung. Metastasis in the lung appeared as dense tumor cell clusters (arrows) following hematoxylin and eosin (H&E) staining of formalin-fixed paraffin-embedded tissue sections.
Dual therapy with axitinib and celecoxib prevented the outgrowth of spontaneous breast cancer metastasis in the lung. Metastasis in the lung appeared as dense tumor cell clusters (arrows) following hematoxylin and eosin (H&E) staining of formalin-fixed paraffin-embedded tissue sections.

Angiogenesis is essential for tumor growth and metastasis, by supplying a steady stream of nutrients, removing waste, and providing tumor cells access to other sites in the body. The vascular endothelial growth factor (VEGF) and its receptors (VEGFRs) play a key role in tumor-mediated angiogenesis, and this pathway is the target of monoclonal antibodies and tyrosine kinase inhibitors (TKIs) that have been approved to treat patients with cancer. Unfortunately, tumors can use alternative angiogenesis mechanisms to escape VEGF pathway blockade, but these alternate pathways are not well understood. Brad St. Croix, Ph.D., of CCR’s Mouse Cancer Genetics Program, along with Lihong Xu, Ph.D., a Postdoctoral Fellow in the St. Croix laboratory, and colleagues set out to identify VEGF-independent mediators of tumor angiogenesis.

The researchers began their studies with the CT26 mouse colon cancer cell line, which is resistant to VEGFR2 inhibition in vivo. Using limiting dilution, they generated several clones with medium and low tumorigenic potential compared to the high tumor-forming activity of the parent line. The investigators found no correlation between the level of tumorigenesis by the cell lines and VEGF expression, suggesting that VEGF is not the major angiogenic factor used by these cells. Since immune cells can contribute to tumor angiogenesis, which might indirectly involve VEGF, they implanted small tumors into the mouse cornea, a tissue that lacks vasculature and, thus, immune cell access. Still the cells induced rapid angiogenesis that was not affected by VEGF blockade, indicating that the cells secrete a unique angiogenic factor. By purifying conditioned media from the cells, the scientists identified prostaglandin E2 (PGE2) and demonstrated that PGE2 expression correlated with the tumorigenic properties of the CT26 parent and clone cells.

COX-2 is the enzyme that performs the rate-limiting step in PGE2 synthesis, and the researchers found that high COX-2 levels associated with high PGE2 expression and tumor formation. Treatment with celecoxib, a COX-2 inhibitor, decreased tumor growth without altering VEGF expression, supporting the importance of COX-2 in this system. Likewise, forced expression of COX-2 in the low tumorigenic clones increased PGE2 expression and tumorigenesis, both of which were reversed with celecoxib, again with no change in VEGF. Using in vivo and in vitro assays, the investigators showed that overexpression of COX-2 or treatment with synthetic PGE2 or conditioned media from COX-2-expressing cells enhanced angiogenesis, suggesting the COX-2/PGE2 pathway directly acts on endothelial cells. They also found that activation of this pathway increased the number of myeloid-derived suppressor cells, which can promote further angiogenesis, demonstrating that the COX-2/PGE2 pathway promotes angiogenesis via multiple mechanisms.

The scientists next examined the interaction between the COX-2 and VEGF pathways using a VEGF knockout (VKO) of the human colon cancer cell line HCT116. VKO cells formed slow-growing, poorly angiogenic tumors compared to parental cells, but expressing either COX-2 or VEGF in the cells independently restored tumor formation and angiogenesis as well as increased secretion of PGE2 or VEGF, respectively. Because previous studies suggested crosstalk between these pathways, the researchers asked whether either could work while the other was blocked. Treating with axitinib, a VEGFR1-3 TKI, potently reduced the growth of the VKO VEGF-expressing tumors by 80 percent compared to 34 percent in VKO COX-2-expressing tumors. Celecoxib treatment, however, strongly reduced VKO-COX-2 tumors by 60 percent compared to 48 percent in VKO-VEGF tumors. In the cornea, treating with a VEGFR2 antibody only reduced VKO-COX-2 tumor growth by 20 percent. Together these studies show that the COX-2 and VEGF pathways can independently regulate angiogenesis and tumor growth.

The results of the previous studies suggest that the COX-2/PGE2 pathway may help tumors escape VEGF pathway blockade. To test this idea, the investigators treated mice bearing CT26 tumors with celecoxib, VEGFR2 antibody, or both. As expected, the CT26 tumors resisted VEGF pathway inhibition. Treatment with celecoxib slowed tumor growth, but the combination synergistically enhanced tumor growth reduction. The scientists saw similar results with axitinib and celecoxib. They then tested celecoxib and the human VEGF antibody, bevacizumab, on mice bearing HCT116 tumors. This time, both drugs individually displayed antitumor activity, but the combination failed to significantly increase efficacy. In contrast, all three combinations dramatically reduced blood vessel densities compared to the monotherapies, indicating that the two pathways collaborate to enhance tumor angiogenesis.

Because angiogenesis is also important for tumor metastasis, the researchers examined the ability of VEGF/COX-2 dual pathway inhibition to alter the formation of metastases. In the HCT116 model, neither axitinib nor celecoxib had a significant effect on liver metastases, but the combination dramatically reduced them. In the CT26 model, however, celecoxib alone significantly inhibited liver metastases, and its activity was not enhanced by axitinib. To see whether progression of other tumor types was susceptible to dual pathway blockade, the investigators tested a breast cancer model. Individually, celecoxib and axitinib decreased primary tumor growth and blood vessel density, but there was no significant efficacy increase with the combination. Similar to the HCT116 model, neither drug alone affected metastases, which formed primarily in the lung, but the combination significantly reduced them. Long-term treatment with the combination also improved survival at six months.

Finally, the scientists examined a model of microscopic metastatic disease, which is often the case in patients diagnosed with cancer. They injected mice with breast cancer cells, removed the primary tumors three weeks later, and then began therapy. After four weeks, there were widespread metastases in the untreated mice and in mice that received celecoxib or axitinib. In contrast, there were no detectable metastases in 10 of 11 mice on combination therapy. The researchers detected no significant change in weight or eating habits, suggesting the mono- and combination therapies were relatively nontoxic. By 100 days, survival was only improved in the combination therapy group, demonstrating that dual pathway blockade prolonged survival even in the presence of microscopic metastatic disease.

Taken together, these results suggest that combining COX-2 inhibitors with anti-angiogenic drugs may reduce tumor burden and enhance survival in patients with metastatic disease, but the combination must be tested in clinical trials first since these data were generated in preclinical models.

Xu L, Stevens J, Hilton MB, Seaman S, Conrads TP, Veenstra TD, Logsdon D, Morris H, Swing DA, Patel NL, Kalen J, Haines DC, Zudaire E., St. Croix B. COX-2 Inhibition Potentiates Antiangiogenic Cancer Therapy and Prevents Metastasis in Preclinical Models. Sci Transl Med. June 25, 2014. PubMed Link