Anti-angiogenesis, the inhibition of blood vessel growth in tumors, is a popular avenue of cancer therapeutics - one that has paid off for drugs such as Genentech Inc.'s Avastin, which is approved as first-line combination treatment for patients with metastatic colorectal cancer.

Targeting the vasculature, rather than the tumor itself, has several theoretical advantages. For one thing, each blood vessel supplies a lot of tumor cells. For another, the blood vessel cells divide, and thus mutate more slowly than the tumor cells, which could make resistance less of a problem.

Avastin targets the vascular endothelial growth factor, of VEGF, receptor, and works by inhibiting the growth of new vasculature. An alternative approach is to target the existing vasculature within the tumor.

The trick, of course, is to induce clotting specifically in the tumor vasculature, while avoiding blood clotting in normal vasculature. Luckily, tumor blood vessels are not the most camouflaged of structures; they have a variety of molecular features that distinguish them from normal blood vessels.

In the May 18, 2005, issue of the Journal of the National Cancer Institute, researchers from University Hospital Cologne in Germany and the University of Michigan at Ann Arbor have harnessed one such distinguishing feature to give tumors, and, only tumors, a stroke.

The scientists first constructed a fusion protein consisting of a coagulation-inducing factor, a soluble tissue factor and a vascular cell-adhesion molecule that generally is not expressed by normal blood vessels, though it is expressed in several pathological conditions.

While it is not directly toxic to the blood vessel lining or tumor cells, soluble tissue factor induces blood clotting when it is close to a membrane.

"The idea [behind the fusion protein] is to replace the portion of full-length tissue factor, which integrates in all membranes, with a specific binding agent that brings it in contact with cells/membranes of your choice, thereby inducing coagulation only there," Claudia Gottstein, laboratory head at the University Hospital Cologne, and senior author of the paper, told BioWorld Today via email. She added that "other requirements have to also be met to kick off coagulation. These factors can result in additional selectivity for the tumor, because tumor vessels support coagulation better than normal vessels."

While a tissue factor and vascular cell-adhesion molecule had been joined before by biochemical means, a fusion protein is a more precise instrument: "Biochemical conjugation is dependent on reactive chemical groups, and in every antibody there is more than one reactive group," Gottstein said. "It is not possible to exactly control how many and which groups undergo the chemical reaction, therefore the end product is a mixture of different, though similar, molecules."

Gottstein and her colleagues first confirmed the specificity and activity of the protein in vitro, and then tested its effects in a series of animal experiments. In short-term experiments, both mouse models of human Hodgkin's lymphoma and small-cell lung cancer had plugged blood vessels after a systemic injection of fusion protein; in Hodgkin's lymphoma mice, an average of 75 percent of the tumor was necrotic three days after the injection, and in small-cell lung cancer mice, an average of 25 percent. In longer-term experiments, repeated injections of fusion protein slowed tumor growth in both models.

In a final series of experiments, Gottstein and her colleagues transplanted mice with both human endothelial cells and human tumor cells; that procedure leads to mice that have a tumor fed by both mouse and human blood vessels. In those experiments, though, the scientists did not see the large areas of necrosis they had seen in the earlier studies - there were fibrin clots in the tumors and three days after treatment, tumors in treated mice had grown less than those in controls.

Gottstein and her group have developed a monoclonal antibody that they hope to use as a targeting agent; they also are investigating whether metastatic tissue expresses vascular cell-adhesion molecules. "We have so far not confirmed the exact mechanism of action; we just know that the treated tumors were significantly smaller," Gottstein said. But even from what she knows already, she said that though the approach is comparatively new, "there is a great potential to destroy solid tumors via vascular occlusion."