Cutting off cancer's supply lines

Targeting the blood vessels that feed tumours is not the silver bullet once hoped for, but refinements to the strategy may suggest further ways to treat the disease. Erika Check Hayden reports.

Network of blood vessels surrounding a tumour — a potential target for therapy. Credit: CLOUDS HILL IMAGING LTD/SPl

Drugs that aim to choke off a tumour's blood supply, known as angiogenesis inhibitors, have been hailed as opening a new era in cancer therapy. But a flurry of animal studies suggests that such drugs may in certain situations actually accelerate the spread of cancer.

"We're just finding the limitations of these types of agents in the clinic," says John Ebos, a cancer researcher at the University of Toronto, Canada. "I don't think it's unique — various types of therapies, such as chemotherapy and radiation, also have limitations. It's just a question of how we can overcome it."

This is a key time in the long and controversial history of these drugs. In May, the US Food and Drug Administration (FDA) is expected to decide whether to expand use of bevacizumab, the first angiogenesis inhibitor. This monoclonal antibody, sold as Avastin by South San Francisco-based Genentech, was approved in 2004 for treating metastatic colon cancer in combination with chemotherapy. It has since been approved in the United States and elsewhere for other uses, and on 31 March an FDA advisory committee recommended the drug also be approved for glioblastoma, a deadly brain cancer for which few other treatments are available. The agency's decision is expected in May.

“I think there's been that growing feeling of why aren't they working better, and we're now uncovering some explanations. , ”

According to regulatory papers filed in January, Genentech may before then reveal results of a clinical trial to test the use of bevacizumab as an 'adjuvant' used with chemotherapy in patients whose colon tumours have been surgically removed. The 2,710-patient phase III trial investigated whether those who take bevacizumab are more likely to survive without recurrence of their disease than patients who do not take the drug.

"There are tens of thousands of patients with early colorectal cancer who don't get Avastin right now," says Geoffrey Porges, an analyst with Sanford C. Bernstein in New York City. Success as an adjuvant therapy "would open up a market at least as large as the current metastatic market", he thinks, noting that trials testing the drug as an adjuvant for other cancers are under way. Last year, the drug racked up $4.8 billion worldwide in sales; its wholesale price is about $50,000 per year of treatment.

The original idea behind bevacizumab and other angiogenesis inhibitors was championed by Judah Folkman of Harvard Medical School. In 1971, Folkman wrote in the New England Journal of Medicine1 that all tumours depend on the constant growth of new blood vessels, a process called angiogenesis, and that blocking it should eliminate the cancer. The popularity of the idea waned in the early 2000s following disappointing results in clinical trials of two anti-angiogenic compounds, angiostatin and endostatin, that were discovered in Folkman's lab. It regained ground when bevacizumab was approved.


Click for larger image.

Since 2004, two other angiogenesis inhibitors have been approved in major markets worldwide: sunitinib, sold as Sutent by Pfizer, for use in advanced kidney cancer and gastrointestinal stromal tumours, and sorafenib, sold as Nexavar by Bayer, for use in kidney and liver cancer. Both are small-molecule drugs that target kinases, in particular vascular endothelial growth factor, or VEGF, which is also targeted by bevacizumab. Many more such compounds are in late-stage clinical trials (see table).

But these drugs have not been the magic bullet that Folkman envisaged. In major cancers, such as breast and colon, they have helped patients to survive longer when given with chemotherapy, but not when given alone. The drugs seem to grant most patients slower progression and a few extra months of survival — a real benefit, but not a cure.

The tumours fight back

That has led researchers to ask whether the drugs are working as Folkman hypothesized. Some, such as Rakesh Jain of the Massachusetts General Hospital in Boston, have suggested that the drugs "normalize" blood-vessel growth around tumours. Cancer blood vessels are normally leaky and chaotic; by correcting this, angiogenesis inhibitors may turn the vessels into a more efficient pipeline for delivering chemotherapy, Jain suggests.

Some physicians who treat patients with cancer have also noticed that when the disease does return after treatment aimed at angiogenesis, it is more aggressive than in patients not treated with the drugs. Other researchers now have evidence that may validate this observation. In mouse studies2, 3, researchers have reported that the drugs can speed the spread of tumours to nearby tissues and distant organs.

Reporting recently in Cancer Cell, two teams investigated the effects of angiogenesis-inhibiting drugs and of knocking out the gene encoding VEGF. One team, led by Douglas Hanahan of the University of California, San Francisco, and Oriol Casanovas of the Catalan Institute of Oncology in Barcelona, Spain, reported that tumours spread faster and more often to both near and distant organs in mice treated with drugs or lacking the VEGF gene2. The other group, led by Robert Kerbel of the University of Toronto, reported similar effects3, and found that in some situations the treated mice actually died earlier than untreated animals.

Kerbel's group further studied how metastasis changed depending on when the drugs were given. When given either before or long after metastatic tumour cells were injected into the mice, or after primary tumours had been surgically removed, the drugs hastened metastasis. But when given while the mice were still carrying primary tumours — those that had yet to metastasize — the drugs actually helped shrink the tumours. If the data hold true in humans, they suggest that the timing of drug delivery can have a major impact on a patient's response.

“What they're telling us is that there are other targets that need to be considered if you're going to mess with the blood supply. , ”

A third paper4, published online in Nature Medicine on 22 March, suggested that such effects might also occur for a class of drugs called integrin inhibitors. These block the activity of integrins, which are proteins that trigger angiogenesis, among other things. Researchers led by Kairbaan Hodivala-Dilke at Queen Mary, University of London, studied the effects of two integrin inhibitors, one of which, cilengitide, is in phase III clinical trials to treat brain cancer. They found that, when given in low doses in mice, the drugs paradoxically seemed to promote angiogenesis and tumour growth.

Ebos, the first author on the Kerbel-group paper, says that, taken together, these studies don't mean anti-angiogenesis drugs are a disappointment, simply that they have limitations. "I think there's been that growing feeling of why aren't they working better, and I think we're now uncovering some of the explanations," he says.

For instance, Hanahan and Casanovas's team studied whether choking off some blood vessels, and thereby inducing oxygen deprivation (hypoxia), might be driving the tumours to search elsewhere for sustenance. When they stained cancer cells in mice with a marker for hypoxia, it showed up in the cancer cells of animals treated with angiogenesis inhibitors.

The oxygen connection

To Casanovas, this provides a link to earlier studies indicating that hypoxia can boost tumour invasiveness, and shows that the lack of oxygen caused by angiogenesis inhibitors actually induces cancerous cells to leave the tumour site in search of it. Casanovas says that his and Kerbel's teams have together found enough evidence to suggest that such effects probably occur with many angiogenesis inhibitors in many different tumour types. "Between us, we have tested five different compounds, including small molecules and antibodies, so we think the effect could be more general than strictly what we've seen in these two papers, and the same thing applies for different types of tumours," Casanovas says.

Donald McDonald, a vascular biologist at the University of California, San Francisco, suggests that the new studies are delivering a broader message. "What they're telling us is that there are other targets [in addition to VEGF] that need to be considered if you're going to mess with the blood supply." Drug companies are already developing compounds that may address some of the problems raised by the papers.

McDonald notes that researchers six years ago reported5 that hypoxic cancer cells ramped up production of a protein called Met, which binds and activates another protein called hepatocyte growth factor. This growth factor also goes by the name of scatter factor because it triggers cells to move — precisely the effect seen in the Cancer Cell papers. Administered with anti-angiogenesis drugs, Met inhibitors — the focus of development by drug and biotechnology companies — might offset the effects reported in the Cancer Cell papers, McDonald suggests.

And Hodivala-Dilke's team noted that simply using some drugs differently might improve their effectiveness. Clinical trials of cilengitide have delivered mixed results so far, and the team says its study points to a possible reason for this. They found that integrin inhibitors slowed the degradation of proteins that promote angiogenesis. These longer-lived proteins then recruited cells that make up blood-vessel walls to the tumour site, where they built new blood pipelines to the tumour. Because the effect occurred only at very low doses, the team suggests that the drugs should be delivered continuously, rather than at a high dose followed by drug-free days, as was done in the clinical trials. That way, levels would stay above those that seemed to promote angiogenesis in the study.

Porges says that the new studies have not dampened enthusiasm among drug companies who are developing angiogenesis inhibitors, integrin inhibitors and inhibitors of other growth factors thought to be involved in cancer. They expect that eventually patients will be given combinations of inhibitors, with or without chemotherapy.

Because of this, McDonald says that the angiogenesis-inhibitor story is far from over. "We have learned that what these drugs do is more complicated than the original idea, which was that they would stop tumour growth by stopping blood-vessel growth," he says. "Avastin is chapter one of angiogenesis inhibition, and we're going to move on to chapter two and chapter three. And with each chapter there will be more clinical benefit as we get a better understanding of the underlying biology."



Network of blood vessels surrounding a tumour — a potential target for therapy.



Network of blood vessels surrounding a tumour — a potential target for therapy.


  1. 1

    Folkman, J. N. Engl. J. Med. 285, 1182–1186 (1971).

  2. 2

    Pàez-Ribes, M. et al. Cancer Cell 15, 220–231 (2009).

  3. 3

    Ebos, J. M. L. et al. Cancer Cell 15, 232–239 (2009).

  4. 4

    Reynolds, A. R. et al. Nature Med. doi:10.1038/nm.1941 (2009).

  5. 5

    Pennacchietti, S. et al. Cancer Cell 3, 347–361 (2003).

Download references

Additional information

See Editorial, page 679 .

Related links

Related links

Related links in Nature Research

Nature Reviews Cancer: When good drugs do bad things

Nature Reviews Cancer Focus on Targeting Angiogenesis

Related external links

Interview with John Ebos

US National Cancer Institute fact sheet on angiogenesis inhibitors

Cancer Research UK fact sheet on angiogenesis inhibitors

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hayden, E. Cutting off cancer's supply lines. Nature 458, 686–687 (2009).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.