19 December 1998 CapCure5\Folkman.MS-- 1,150 words

Angiogenisis inhibitors stop tumors permanently;
many twists in scaleup from mice to man

Lake Tahoe, NV--When an extravagant story hit the front page of the New York Times several months ago, saying that Judah Folkman, M.D., had a drug which would cure cancer, an oncologist got a phone call from a man in New York who said, "How much would it take?" How much money would it take for him to get endostatin?

He couldn't get endostatin for all the money in the world. Dr. Folkman's lab at Harvard University Children's Hospital was having a problem, typical of drug development. They couldn't scale up production.

"You can make 100 milligrams of endostatin every week in the lab without problems," Dr. Folkman said in a lecture at CapCURE'S 5th Annual Scientific Retreat here this September. That's enough for mice. "But as you begin to scale up to hundreds of grams," which would be needed for human therapy, "you lose activity and it doesn't work."

Zn active site

After they worked out the 3-dimensional structure at 1.5 angstroms resolution, they figured out why. "The important thing we've learned is that there's a zinc in the N terminus," said Dr. Folkman, who is professor of pediatric surgery and cell biology.

During scaleup, the growth medium was becoming zinc deficient, they discovered. Without zinc, the proteases chew on the end of the molecule and it lost activity, explained Dr. Folkman. Even with zinc-enriched medium, if the pH falls below 3.3, the zinc "pops out."

More significantly, the zinc, which is surrounded by 3 histidines and an aspartic acid, seems to be involved in the angiogenic activity, they reported in PNAS (Sept. 1, 1998;95(18):10443-8). Human endostatin is a 20-kilodalton internal fragment of collagen XVIII. Other angiogenesis inhibitors have been found as internal fragments of larger proteins, which is "evidence of a an evolutionary relationship between regulation of angiogenesis and regulation of the coagulation system," said Dr. Folkman.

Dr. Folkman, who has a reputation for diligently crediting his co-workers, noted the contributions of Don Wiley and Thomas Boehm.

Dormancy, resistance, toxicity

"The hallmark of these specific angiogenesis inhibitors continues to be lack of acquired drug resistance, and lack of detectable toxicity, despite long-term chronic administration," said Dr. Folkman. So you can always resume therapy.

A "key finding," said Dr. Folkman, is that angiogenesis inhibitors produce a "self-sustained dormancy that is indefinite." Tumors don't recur. Angiogenesis is blocked. They first observed this dormancy after cycles of treatment, but "now it turns out you don't need to cycle," he said. The number of cycles needed for LNCAP was zero.

Endostatin is a third-generation angiogenesis inhibitor, said Dr. Folkman. The first generation slows animal tumors but doesn't regress them. The second generation, including inhibitors of integrin, VEGF, and metalloproteinases, produces actual regression. The third generation "can regress every tumor tested so far in animals, including transplantable transgenic spontaneous tumors or orthotopic tumors," he said. "And these are very specific to endothelial cells."

Tumor dormancy was originally observed after drug cycling, said Dr. Folkman. But now it turns out that dormancy occurs after a sufficiently long time. "

In a series of mouse studies, "The cycling was done to demonstrate lack of drug resistance," said Dr. Folkman. "It took 6 cycles for Lewis lung carcinoma, it took 4 cycles for fibrosarcoma, it took 2 cycles for melanoma, but only 1 cycle for human prostate cancer."

How LNCAP responds

LNCAP cells were planted orthotopically in the mouse prostate, "and we can't see the tumor except in the paired animals who are sacrificed periodically," said Dr. Folkman. PSA was a good way to follow the progress of cancer, and correlates with tumor mass. When the tumor is "just beginning to come up out of the pelvis," the PSA is about 40 ng/ml, he said. After treatment, "it's simply eradicated." In untreated mice, the PSA went over 100. "The tumors go up and fill the pelvis, and push the liver up," he said. One set of mice has been followed for 293 days. "That's the equivalent of 27 human years."

One series of mice was deliberately under-treated, to allow the tumor to come back, so they could retreat it. The PSA went down to 4, then came up to 25, then went down again as they treated it with angiogenesis inhibitor. "You can simply retreat it," said Dr. Folkman. Androgen-independent tumors also became dormant.

"You cannot eradicate Lewis lung tumor with 1 angiogenesis inhibitor, but you can with 2 angiogenesis inhibitors," said Dr. Folkman. Human tumors have less angiogenic activity than mouse cells, and are much more easily treated than mouse tumor, "mainly because of the long passage over time of mouse tumors that have made them very highly selective for angiogenic activity."

Untreated mice become cathectic and lose weight, said Dr. Folkman. When treated, they gain weight normally. "This is very interesting," he said, because of a study in Science (Sept. 11, 1998; 281(5383):1683-6), which found that leptin, which is made in fat cells and shuts off appetite in the brain, can also drive creation of capillaries in fat. So the one molecule serves 2 functions, of matching the appetite to fat, and matching the vasculature as well. Dr. Folkman's group has found some tumor cells making leptin.

Dormant cancers

Dr. Folkman was curious about dormant prostate cancer, since it fit into a larger pattern that he had been following. In autopsies of people who have died of trauma, pathologists find microscopic dormant primary carcinomas."Not dysplasias, not pre-malignancies, they're called 'cancers' by the pathologists." He reviewed some of the studies. 39% of women 40-50 years old have small breast carcinomas, but only 1/100 are diagnosed clinically. 46% of men 60-70 have prostate cancer, but only 1/100 are diagnosed clinically. In people 50-70, "virtually 99% have thyroid cancer,"' but only 1/1,000 are diagnosed clinically. They are "competing with the lifespan."

Histologically, they are avascular carcinoma. When Dr. Folkman was in Helsinki, he visited the researchers and examined those samples. They ranged from 0.24 mm to 0.4 mm. "That is the gate which absence of angiogenesis blocks," he said.

"Could it be," said Dr. Folkman, "that normally in the human population there is a significant burden of in situ carcinomas that have not yet switched on angiogenic activity, and may never do so in a person's lifetime? Could it be similar for prostate cancer, because the molecular marker of PSA is so sensitive that we are seeing a rise in incidence?"

"One goal of angiogenesis therapy is to switch those lesions back to their pre-angiogenic state," said Dr. Folkman. "To molecular biologists I would ask, could you think of a future technology that would detect the presence of these lesions, and distinguish between those that might reach a dead end, versus those that are potentially capable of switching on angiogenesis, and thus progressing to potentially lethal cancer?"