Finding Ways to Starve the Cancer Seed

Oncologists often describe cancer as a seed that grows in the body's soil. For these seeds to become tumors, the "soil" must be stocked with nutrients such as growth factors to help them proliferate. "Cancer is not a single-cell disease but involves cancer cells and how they collaborate or cooperate with surrounding cells," says Leland Chung, director of molecular urology and therapeutics at Winship Cancer Institute, Emory University School of Medicine. New approaches to starving cancer cells figured prominently at the recent meeting of the American Association for Cancer Research (AACR) in San Francisco. Blood vessels feed tumors and remove waste products, and they are necessary for tumor growth. Inhibiting this blood vessel growth could prevent tumor growth. "The whole concept is so fundamentally sound," says J. Clifford Murray, University of Nottingham Cancer Research Campaign Department of Clinical Oncology. "The endothelial cell [lining the blood vessels] is immediately attractive because of its accessibility. It's the first cell within a tumor that any reagent is going to reach." The efficacy of this approach, however, is as yet unknown.1 "We are going into a difficult time at the moment, waiting to see if these agents will bear any fruit," says Murray. Oncologists seem most impressed with an antivascular endothelial growth factor (anti-VEGF) monoclonal antibody. "The data at AACR that impressed me the most were the trials of ...[an] anti-VEGF in combination with chemotherapy. Researchers were getting quite significant improvements in survival [rates]," observes Murray. Much of the skepticism stems from conflicting results seen in past trials, and AACR presentations revealed possible biological mechanisms that could underlie the differences. For example, the popular belief that angiogenesis inhibitors (AIs) should work on all tumors (because tumors require a blood supply) is now uncertain as multiple researchers report tumor-specific angiogenic biology. Luisa Iruela-Arispe, associate professor in the department of molecular, cellular, and developmental biology, University of California, Los Angeles, reported that specific, mutated oncogenes in a tumor could influence how it responds to AIs. She showed that endostatin was more effective in mice with neu mutations than those with p53 null mutations. "You might have two different tumors that are the same size and time of incidence, but with two different initiation oncogenes. Using the same [AI] might return different results," said Iruela-Arispe. Shahin Rafii, an assistant professor in the department of medicine, Cornell University Medical College, showed that endothelial progenitor cells can originate from bone marrow, and that only certain tumors seem able to stimulate their release.2 Angiogenesis requires endothelial cells; scientists had believed that they were recruited exclusively from the existing vasculature. Rafii also reported that the majority of endothelial cells express the VEGF-1 receptor, but the bone marrow-derived endothelial cells express the VEGF-2 receptor. "This is suggesting that VEGF recruits these cells by two different mechanisms," he said. He also proposed that targeting the endothelial cells' mobility from the bone marrow could be a novel therapeutic approach. Iruela-Arispe also showed why markers of antiangiogenic activity might have been elusive. Her group studied thrombospondin, an endogenous AI, and found that it increases the active form (but not the total amount) of MMP-9, a matrix-metalloproteinase that helps endothelial cells infiltrate the basement membrane as they migrate to form new blood vessels; it also regulates the bioavailability of VEGF.3 Usually, researchers measure total protein rather than the active-to-inactive ratio; such telltale differences would have previously been missed. "Looking at the overall protein levels or mRNA levels is not necessarily telling you everything," Iruela-Arispe says in an interview. Other targets, such as pericytes, known for blood vessel stabilization and maturation, also were discussed as potential therapeutic targets. ©American Association for Cancer ResearchThe Ties that Bind: The currently known VEGFs and their receptors. VEGFR-1 and VEGFR-2 have seven extracellular immunoglobulin homology domains, but in VEGFR-3, the fifth immunoglobulin domain is cleaved on receptor processing into disulfide-linked subunits. VEGFR-1 AND VEGFR-2 mediate angiogenesis, whereas VEGFR-3 is also involved in lymphangiogenesis. Surprise Surprise Lymph vessels develop in a tumor through a process called lymphangiogenesis. Previously, researchers believed that tumors did not contain lymph vessels, but, with newer molecular and imaging techniques, scientists have identified them in some tumors. In addition, Murray demonstrated that VEGF-D expression, which is associated with lymphangiogenesis, is an independent predictor of poor survival in colorectal cancer patients.4 "At some point, [angiogenesis and lymphangiogenesis] must diverge," he says. At AACR, Judah Folkman of the Children's Hospital in Boston discussed the production of a new assay that separates the process of lymphangiogenesis from angiogenesis. Says Lynn Chang, a surgical resident and lead investigator on the assay research, "Recently what people have done is transfected VEGF-C and [VEGF]-D into tumor cells and put them into mice." Those tumors grew, developed lymphatic vessels, and metastasized at a much higher rate. "That is really what got people excited about lymphangiogenesis," he says. They devised the assay, he continues, to discover whether they could inhibit lymphangiogenesis as well as angiogenesis. The researchers put a pellet with a very low dose of the fibroblast growth factor (FGF) in a murine cornea. At high doses FGF stimulates both angiogenesis and lymphangiogenesis, but at low doses it stimulates lymphangiogenesis at a dramatically higher rate. The researchers also implanted a tumor under the mouse's skin and found that tumors can secrete circulating factors that inhibit lymphangiogenesis in the eye. "A remote fibrosarcoma can suppress lymphangiogenesis largely, but a distant melanoma does not suppress lymphangiogenesis," said Folkman. "It could be a known angiogenesis inhibitor, or it could turn out to be a pure lymphangiogenesis inhibitor." Folkman reported that, to date, they know that the COX-2 inhibitors rofecoxib (Vioxx) and celecoxib (Celebrex) inhibit lymphangiogenesis, but thalidomide, which has antiangiogenetic properties, does not. Inhibiting lymphatics could have contradictory effects, researchers caution. Although the lymphatics disseminate tumor cells throughout the body, they also deliver helpful inflammatory cells to tumors. Murray says it is known that inflammatory infiltrates are a very important positive prognosticator in colon cancer patients. "We want to have inflammatory cells get into the tumor. [But] if we start targeting lymphatics you might alter the traffic of the inflammatory cells within the tumor." A Vicious Cycle Tumors frequently metastasize to bone because it is a rich source of growth factors. "Bone provides fertile soil," says Theresa Guise, associate professor of molecular medicine and medicine/endocrinology at University of Texas, San Antonio. Tumors that reach bone can cause cells there to secrete even more growth factors. "This can initiate a vicious cycle," Guise says. Prostate cancer is particularly prone to metastasize to bone. "They really need the stromal cells to be able to grow." In fact, says Emory's Chung, when prostate cancer cells metastasize, they often express bone-like proteins. "For some reason, the bone matrix protein probably allows them to survive in the bone environment," he says. Stromal cells often act as intermediaries between tumors and the bone microenvironment. "There are hundreds of interactions where tumor cells communicate with stromal cells, and then the stromal cells make factors that act on bone itself or the tumor cells," Guise says. Chung developed a gene therapy virus that targets both prostate tumor cells and a subset of the surrounding stromal cells. At AACR, he presented data showing that, in mice and in vitro, the virus' destruction of both prostate cancer cells and bone stromal cells effectively inhibited prostate tumor cell growth. "Forty percent of the mice with preestablished cancer in the bone are cured," says Chung.5 In the early stages, though, research has identified agents that might fight cancer by targeting the bone microenvironment. For example, bisphosponates that are currently available to treat cancer-related bone pain could also kill cells that destroy bone. "It's hard to tell it if lengthens life," says Kent Osborne, breast center director at Baylor College of Medicine in Houston. "There are a couple of hints [indicating] that, by using it even before the bone metastases are obvious, ... it might." Combination Therapy In general, oncologists do not believe that agents targeting the cancer-support network will be highly effective by themselves. Rather, they predict that such agents will increase the efficacy of tumor-directed agents when given in combination. "The microvascular endothelial cell has become an important second ... target in cancer treatment," said Folkman. "If you cotarget the tumor and [bone] stroma you can hopefully acquire a much stronger response," concurred Chung. A number of presentations at AACR showed positive results when AIs were delivered along with other treatments. For example, biochemist Douglas Hanahan, University of California, San Francisco, reported striking efficacy in mouse models by combining two angiogenesis inhibitors, SU5416 and SU6668, which act on complementary targets. Neither of these agents alone showed strong efficacy against established tumors; together, however, they produced tumor regression. Hanahan's group also tested imatinib mesylate (Gleevec), a marketed agent that inhibits the platelet-derived growth factor (PDGF) receptor (one of the targets of SU6668) with another antiangiogenic agent, and found similar efficacy. Nottingham's Murray suggests, "I don't think targeting the tumor will ever go away.... But we will see it used in combination with strategies targeting the support cells as well." Mignon Fogarty (mignon@welltopia.com)is a freelance writer in Santa Cruz, Calif. References 1. M. Fogarty, "Learning from angiogenesis trial failures," The Scientist, 16[6]:33, March 18, 2002. 2. S. Rafii et al., "Contribution of marrow-derived progenitors to vascular and cardiac regeneration," Seminars in Cell & Developmental Biology, 13:61-7, February 2002. 3. J.C. Rodriguez-Manzaneque et al., "Thrombospondin-1 suppresses spontaneous tumor growth and inhibits activation of matrix metalloproteinase-9 and mobilization of vascular endothelial growth factor," Proceedings of the National Academy of Sciences, 98:12485-90, Oct. 23, 2001. 4. J.D. White et al., "Vascular endothelial growth factor-D expression is an independent prognostic marker for survival in colorectal carcinoma," Cancer Research, 62:1669-75, March 15, 2002. 5. S. Matsubara et al., "A conditional replication-competent adenoviral vector, Ad-OC-E1a, to cotarget prostate cancer and bone stroma in an experimental model of androgen-independent prostate cancer bone metastasis," Cancer Research, 61:6012-9, Aug. 15, 2001.

Finding Ways to Starve the Cancer Seed

Interested in reading more?

Become a Member of