NOVA Online (see text links below)
Cancer Warrior
Designing Clinical Trials
by Robert Cooke

High expectations are hard to live up to, especially if they're someone else's. Judah Folkman found himself in that position as his two big antiangiogenesis drugs, endostatin and angiostatin, finally headed toward clinical trials near the end of 1999. Folkman was expected to cure cancer—it said so in The New York Times. Now all he needed was for the drugs to work in people as unequivocally as they had in mice. But it was not so simple, of course, and the results would not be known quickly.

Previewing a candidate Doctors previewing a candidate for a clinical trial of endostatin.
The first phase of the trials, as in tests of all drugs overseen by the Food and Drug Administration, was designed to test for safety, not efficacy. So those receiving the first test doses would be the patients with the most advanced cancers, patients for whom everything else had failed, and they would receive very small doses to start with. Moreover, the experience with mice as well as with patients like Jennifer LaChance and Tonya Kalesnik, who had been treated with interferon, indicated that antiangiogenesis drugs have a slow, cumulative effect, taking up to a year to eliminate tumors. It stretched the limits of reasonable expectation to think that the critically ill patients in the first trials would take a few injections of low-dosage, slow-acting drugs and waltz out of their hospital wards, cancer-free.

Obviously, such waltzing wasn't likely. A study that had recently been done at the Johns Hopkins Medical Institutions, in Baltimore, showed starkly that very little about a drug's potential is discernible from the results of Phase I trials. Because the patients admitted to such trials are desperately ill, and because the doses of drugs given are so small, half of the patients fail—they drop out within 1.8 months, even if the drug being tested ultimately turns out to be effective. In contrast, in trials of drugs that turn out to be flops—never making it to the market—half of the patients drop out by 1.6 months. Statistically that's almost dead even, and the take-home lesson is that Phase I toxicity trials generally don't say much about a candidate drug's ultimate success. Thus, by any measure, the expectations for Folkman's antiangiogenic agents were far beyond reason.

O'Reilly Michael O'Reilly discovered endostatin and angiostatin.

Endostatin on trial

Endostatin, the second of the antiangiogenic agents discovered by Michael O'Reilly less than half a decade earlier, was the first to be put to the test, in trials involving 15 patients at each of three sites: the Dana-Farber Cancer Institute in Boston, the M.D. Anderson Cancer Center in Houston, and the University of Wisconsin Medical Center in Madison. The clinical researchers would start with very small doses—only 15 milligrams per kilogram of body weight daily, measures so insignificant that they were unlikely to have any effect—and then gradually escalate them to see if there came a point when the drugs became toxic and should be stopped. The idea was to see if, and how much, endostatin could be given safely. If the drug passed that test, proving itself nontoxic, then the dosages would be increased to levels where they could begin to be tested for efficacy—to see if they actually worked against tumors.

Even then, it could be yet another year before the results were in. Because antiangiogenesis works by shutting down the growth and migration of blood-vessel cells, it's a far slower process than poisoning or blasting tumor cells with radiation or chemotherapy. The tests in mice had made it clear that the antiangiogenic drugs had to be continued for a very long time, and that stopping too early would allow the blood vessels to regrow, reigniting tumor growth. Slow and steady was the rule. So the first patients selected for the trials had to wonder how lucky they were: The protocols of the studies made it unlikely any of the first patients would be saved by the new drugs.

Mice Studies in mice have shown that, to be effective, antiangiogenic drugs must be administered for an extended period.
Still patients lined up, brave and desperate people who knew their chances were worse if they didn't try at all. The first endostatin trials were scheduled for October 1999 at Dana-Farber, and as news of that first opportunity got around, thousands of patients called, wrote, and e-mailed, frantically hoping to qualify for one of the 30 slots in the small toxicity test. The same rush happened a few months later in Wisconsin, and again in Texas, when the two other Phase I endostatin trials were opened under auspices of the National Cancer Institute. Most of the applicants were disappointed. Some patients arrived with complications beyond cancer, such as high blood pressure or diabetes, or even the wrong kind of cancer. People with brain tumors, for example, were excluded because the doctors could not know whether endostatin might weaken the blood vessels feeding the brain tumor so much that hemorrhaging into the brain might occur.

Unlike most clinical trials, in which all the treatment centers do exactly the same things, the doctors running the three endostatin trials were allowed to set some of their own rules, select the kinds of tumors they would treat, and decide how to enroll their patients. This was because antiangiogenesis was so new and untested that no one really knew how to use endostatin. "Since no one knows," explained Dr. Mark Kieran, director of pediatric neurooncology at Dana-Farber, "having one person write the protocol would be very arbitrary."

Surgery Patients chosen for the endostatin trial had to have tumors that defied all other treatments, including surgery.

In Boston, patients were selected almost on a first-come, first-served basis, while in Houston and Madison they were chosen from among patients who were already under care at the hospitals. There were only two things patients in all three trials had to have in common: Their malignancies had to have defied all previous treatments, and the patients couldn't be so near death that they might succumb in just a week or two, before they could be evaluated for signs of drug toxicity. Such an immediate death would be seen as a waste of the precious drug while raising serious questions about why the patient died. Was it the cancer or the drug?

Continue: Watching and Waiting

Dr. Folkman Speaks | Cancer Caught on Video
Designing Clinical Trials | Accidental Discoveries | How Cancer Grows
Help/Resources | Transcript | Site Map | Cancer Warrior Home

Editor's Picks | Previous Sites | Join Us/E-mail | TV/Web Schedule
About NOVA | Teachers | Site Map | Shop | Jobs | Search | To print
PBS Online | NOVA Online | WGBH

© | Updated February 2001
Feedback Shop Site Map Search Home Text Version Cancer Warrior Home