Abstract IA01: Heterogeneity of the tumor vasculature: Why doesn't anti-VEGF/VEGF receptor therapy work better?

Abstract

Abstract The microvasculature that supplies human and mouse cancers, and for that matter healing wounds, is both similar and similarly heterogeneous (1-3). It includes at least six distinct blood vessel types (4,5). Four of these arise from preexisting venules or capillaries and are the product of angiogenesis: mother vessels, glomeruloid microvascular proliferations, capillaries, and vascular malformations. The remaining two blood vessel types, feeding arteries and draining veins, arise from preexisting arteries and veins and so result from arterio-venogenesis. Most of these vessel types are present in all human cancers thus far examined. Further, an adenovirus expressing VEGF-A164 (Ad-VEGF-A164) can replicate each of these vessel types in immunodeficient nude mice. Studies with the Ad-VEGF-A164 model have allowed us to demonstrate the temporal sequence and kinetics of new tumor blood vessel formation and to explore the mechanisms of anti-VEGF/VEGFR therapy. Anti-VEGF/VEGFR drugs have added significantly to the cancer therapy armamentarium, but, unfortunately, are not the wonder drugs that had been hoped for. There are at least three reasons for their limited success: 1. Some tumors, e.g., some lung metastases, do not require new blood vessels for tumor survival and growth. 2. Other tumors are able to survive in extreme hypoxia and so require minimal new blood vessel support. And 3. Anti-VEGF/VEGFR does not target all tumor blood vessels equally; in fact, only a subset is affected, particularly mother vessels (6,7). Also, the mechanisms by which anti-VEGF/VEGFR drugs act are not well defined. The generally accepted view is that they kill the endothelial cells of affected blood vessels. We now propose an alternate and/or additional explanation, namely, that vessel killing can be an indirect effect. Anti-VEGF/VEGFR drugs interfere with VEGF signaling; as a result they inhibit eNOS expression and so reduce expression of its product, NO. NO is a potent arterial dilator (8) that serves to maintain normal arterial tone. The reduced NO resulting from anti-VEGF/VEGFR therapy is expected to cause arterial contraction, and is likely responsible for the hypertension that is commonly observed in patients receiving these drugs (9). We have now found that these drugs act potently on the feeding arteries that supply downstream mother vessels. Mother vessels are lined only by endothelial cells, without smooth muscle coverage, and are therefore uniquely susceptible to damage when immediately upstream feeding arteries contract. These findings could have therapeutic significance. Treating patients receiving anti-VEGF/VEGFR drugs with anti-hypertensive therapy could interfere with their beneficial effects, which may be mediated through the selective contraction of tumor feeding arteries. Finally, anti-angiogenesis therapy to date has almost exclusively targeted VEGF and its receptors. This focus may be too narrow. There is a need for new vascular targets apart from the VEGF/VEGFR axis, if anti-vascular therapy is to have a more important role in cancer therapy. One such target that we have been studying is the tetraspanin-like plasma membrane protein, TM4SF1 (10-12).