Interview: Still stealing the spotlight

At 72, you'd think angiogenesis pioneer Judah Folkman would have the grace to slow down a little. Not so, finds Susan Williamson.

A penchant for correlating clinical observations with causal relationships, along with perseverance, determination and strong ability to challenge conventional ways of thinking, helped to make Judah Folkman a pioneer in the field of angiogenesis and anti-angiogenic drugs.

It hasn't always been an easy road to travel for Folkman, a professor at the Harvard Medical School, but he has received support along the way -- not least from his friends. Such as the old school friend who became a human guinea pig in the name of Folkman's anti-angiogenic research. The friend -- also a professor of medicine -- came to Folkman about six years ago after being told nothing could be done about the multiple melanomas on his legs. Folkman immediately told him to join a clinical trial that was just beginning on some anti-angiogenic drugs. He did, and has been taking low doses of thalidomide ever since, except when he chose to stop taking the treatment to prove that it was working -- the tumours grew back, yet when he went back on the treatment three or four months later, they went away again. Taking low doses of oral thalidomide three or four months each year now keep his melanomas at bay.

Affectionately known as the father of angiogenesis, Folkman visited Australia for the first time late last year to attend the Australian Health and Medical Research Congress, where he stole the show at the plenary, dropping in-jokes and jibes without losing a beat. He's a scene-stealer outside the lecture hall, too. At the age of 72, Folkman continues to conduct research in his lab, working on the emerging class of anti-cancer or anti-angiogenic drugs that provide the possibility of treating cancer much earlier.

His vision for the future is that anti-angiogenic therapy will become a routine way of controlling or regressing tumours, and that clinicians will have a bunch of anti-angiogenic therapies to choose from to treat specific cancers. "Penicillin was the first antibiotic, but after that they were unsuccessful and initially thought not to work," he says. "But now there are over 6000 [antibiotics]. I see anti-angiogenesis drugs as comparable to this -- the initial period will be refined, and in the future specific drugs will be available for specific tumours."

'Fortuitous observation'

Folkman's foray into angiogenesis began when he was drafted into the navy as a surgeon in the 1950s to start a new program for artificial blood. It was here that Folkman, curious about whether thyroid cells could support tumour growth, began experimenting with dog and rat thyroid glands and human melanoma cells. The melanoma cells grew in the light-coloured thyroid, but only to pinhead-sized black dots. But when the same melanoma cells were put into mice they grew uncontrollably.

With this "fortuitous observation", Folkman later deduced that the tumours did not grow because they did not have a blood supply, and went on to try to publish the work. Between 1962 and 1971 he continued to submit the work and continued to receive rejections. "People had a problem with me deducing a lot from a little," he quips. He finally managed to get the paper published, thanks to an insightful editor at the New England Journal of Medicine who attended a talk by Folkman in 1971. And he hasn't stopped since.

"In 1971 there were only three papers published in the world literature that mentioned angiogenesis in the title -- two from my lab and one that criticised these papers," recalls Folkman. "Now there are 20,000 papers published per year."

Making the switch

A tumour can grow to 1-4mm in size without its own blood supply -- about the distance nutrients and oxygen from surrounding tissues can penetrate. An 'angiogenic switch' is then required to enable the tumour to grow its own blood vessels and obtain the supply of nutrients and oxygen it needs to grow larger. Most people live with microscopic or in situ tumours that remain harmless throughout their lifetimes. Too small to be detected with any current imaging techniques, these tumours do not switch to the angiogenic phenotype.

Folkman cites evidence from autopsies of car accident victims to show that these microscopic carcinomas occur in many people. For example, 39 per cent of women older than 40 had in situ tumours in the breast, and 40 per cent of men older than 50 had them in the prostate gland. Such figures are much higher than the incidence of breast cancer and prostate cancer in the general population. "Only one in a thousand [in situ tumours] will undergo the angiogenic switch in a lifetime," says Folkman.

The body's endogenous anti-angiogenic defence mechanisms prevent or block angiogenesis from occurring in these tumours, thus preventing them from growing uncontrollably. "We are beginning to think that the safest anti-angiogenesis inhibitors may be those in the body, the endogenous inhibitors," says Folkman, "especially if you want to use them to prevent recurrent cancer, say after mastectomy or after colon cancer resection, or if you want to use them to prevent new cancer in high-risk patients such as women with a breast cancer gene." Folkman's team has driven the search for these endogenous anti-angiogenic compounds since 1980. Two that they have discovered include endostatin and angiostatin, both now in clinical trials.

The discovery of endostatin came out of Folkman's hankering for explaining clinical clues. When he was a medical student, the accepted wisdom was that people with Down syndrome had a much lower incidence of cancer than the general population because they died young. But today, people with Down syndrome no longer die so young -- most live to 60 or 70, yet they still have a very low incidence of cancer. "They don't get brain, breast or prostate cancer, and only one report of pancreatic cancer exists," says Folkman, although people with Down syndrome do suffer the same incidence of testicular cancer and a form of mild leukaemia. They also have the same rate of diabetes, but suffer no diabetic retinopathy and no atherosclerosis.

In pursuing the reason for this, Folkman's team found that people with Down syndrome carry an extra copy of the gene for collagen 18 (COL18AI), which encodes for the protein collagen 18 of which the anti-angiogenic protein endostatin is an integral component. This results in people with Down syndrome having 10 times the level of endostatin of people without Down syndrome.

There are now at least 17 endogenous angiogenesis inhibitors known and, according to Folkman, there are probably many more. In mice that have had the genes for these endogenous angiogenesis inhibitors knocked out, such as tumstatin, thrombospondin 1, or endostatin, tumours grow much faster compared with wild-type mice. However, if the angiogenic inhibitor is given to the knockout mice in physiologic doses, the tumours decrease in size, supporting that the normal physiological function of these compounds is to inhibit angiogenesis. In addition, if the compounds are given in pharmacological doses the tumours regress significantly.

"We have a genetically engineered mouse in our lab that over-expresses endostatin plus thrombospondin and tumours grow up to about half a millimetre and don't grow beyond that," says Folkman. "[The tumours] are not angiogenic."

Bring on the biomarkers

Folkman believes it is possible to detect tumours that undergo the angiogenic switch much earlier, although he admits that his belief does challenge current ways of thinking about cancer treatment. "If molecular 'forecasting' or biomarkers were developed to detect these cancers before they develop into large tumours that become symptomatic, we could begin to move therapy earlier and earlier because you may not need to know the location [of a tumour]," Folkman says.

"We are heavily tied to anatomical localisation when it comes to treating cancer," he says of the current chemotherapy, radiotherapy and surgical techniques that are employed in cancer treatment. "But how do you treat a tumour when you have no location for it?"

Surgery and radiotherapy cannot usually be considered to treat a cancer without accurately locating the tumour mass, so treatment begins long after the tumour has become angiogenic and gained the potential to spread to other parts of the body. Similarly, cytotoxic chemotherapy is not usually administered to an asymptomatic patient whose malignant tumour cannot be seen or biopsied.

Anti-angiogenic therapy, however, may feasibly be used years before any of these other three treatment modes. These drugs are less toxic than conventional chemotherapy, have a much lower risk of drug resistance and produce no side-effects.

Folkman is also enthusiastic about combination therapies for treating tumours -- it appears as though some drugs have a better effect when used in combination with others. "Changing the schedule and using low dose or metronomic chemotherapy and anti-angiogenic drugs together has a good effect on reducing the size of tumours," he says.

For example, research on Bristol-Myers Squibb's chemotherapy drug Taxol showed that it has anti-angiogenic properties at a dose a thousand fold less than the dose given for chemotherapy. Clinical trials in women with breast cancer who had failed all drugs and had ongoing metastases and had no other options were given Genentech and Roche's anti-angiogenic drug Avastin, plus the chemotherapy drug cyclophosphamine, orally at very low doses, and their tumours reduced.

Like all pharmaceuticals, anti-angiogenic drugs are expensive to produce. As a researcher, Folkman is keen to work out how to increase endogenous inhibitors within the body and in turn make treatments cheaper. The anti-angiogenic molecule thrombospondin 1 can be raised by the antibiotic doxycycline, although not tetracycline, despite doxycycline being a slightly chemically modified form of tetracycline. "Doxycycline increases thrombospondin in ras-transfected tumour cells," says Folkman. "And ras drives down thrombospondin, so somehow doxycycline will overcome ras and increase thrombospondin."

Also, doxycycline is cheap. But how the levels of these endogenous inhibitors are raised remains to be elucidated. "There's got to be some rule about how these small molecules increase the levels of these circulating proteins, and nobody has a rule yet," says Folkman. "If anyone can figure one out they can call me and be first author!"

How does it work?

Angiogenic inhibitors are many and varied, and act in different ways. Because angiogenesis involves the formation of new blood vessels, the main targets of these drugs are the cells and factors involved in this process, such as epithelial or endothelial cells and vascular endothelial growth factor (VEGF).

AstraZeneca's Iressa (ZD1839), for example, is an oral epithelial growth factor-tyrosine kinase (EGRF-TK) inhibitor. In animals bearing human tumours, such as colon cancer, it selectively inhibits angiogenesis by inhibiting the production of transforming growth factor (TGF)-alpha, basic fibroblast growth factor (bFGF) and VEGF, proteins that all function to stimulate angiogenesis.

Another anti-angiogenic drug, Avastin, neutralises VEGF, whereas others block the receptor for VEGF along with other angiogenic stimulators at the level of the endothelial cell. The endogenous anti-angiogenic compounds that Folkman is focusing on, such as endostatin and angiostatin, block a broad spectrum of angiogenesis regulators.

More than 30 anti-angiogenic drugs are currently in trials in the US, and more than 50 are in trials worldwide.

"Trials in the US have stalled at Phase II," says Folkman, who is a little frustrated with the lack of progress in the US. "Small biotech companies don't have the money, and the big companies won't take the risk or invest long-term to take things forward."

In contrast, Folkman says colleagues in China have completed phase three trials on a slightly modified form of endostatin and are well ahead in using it to treat cancer patients.

But other drugs on the market that are available for the treatment of other diseases have high anti-antigenic activity, such as the osteoporosis drug bisphosphonate that inhibits human endothelial cell proliferation.

Thalidomide inhibits the angiogenic potential of tumours by decreasing the number of endothelial cells that are recruited from the bone marrow to a tumour, and the breast cancer drug, Tamoxifen, also has anti-angiogenic actions.

Another agent, Pfizer's Celebrex, which is widely used for arthritis, is a potent angiogenesis inhibitor which inhibits the growth of blood vessels into arthritic joints. Celebrex raises levels of endostatin to those found in Down syndrome patients.