UCSF-Harvard Team Reports Encouraging Results In Drug Study Of Mice With Cancer

The study, reported in the April 30 issue of Science, provides evidence that these drugs, so-called angiogenesis inhibitors, can act against tumors in a mouse model that more closely resembles the conditions of human tumor development than the models previously used.

It also provides the first indication that the drugs vary in their efficacy depending on the stage of cancer targeted - drugs showed encouraging results when used to treat distinct stages of tumor development, including premalignancy. The model could therefore provide a platform for exploring the potential of perhaps even more effective angiogenesis inhibitors, or combinations of inhibitors, not just for the treatment of cancer, but for prevention, the researchers said.

"We've shown that these drugs have beneficial effects when used to treat mice developing organ-specific cancer in which tumors arise from premalignant lesions in their natural tissue microenvironment, representing an important extension of the evidence for efficacy of angiogenesis inhibition in cancer," said the senior author of the study, Douglas Hanahan, Ph.D., a professor of biochemistry at UCSF.

Many angiogenesis inhibitors are currently being investigated in laboratories around the world. If the results continue to be encouraging, said the lead author of the study, Gabriele Bergers, Ph.D., a visiting postdoctoral fellow in the Department of Biochemistry and Biophysics, and the Hormone Research Institute at UC San Francisco, "the strategy will be to determine which anti-angiogenesis compounds are most effective at which stage of cancer in mouse models, and to develop dosing regimens that are fine-tuned to target these specific stages of disease progression."

The ultimate question, of course, the researchers emphasized, is whether the drugs will have similar effects in humans, and the answer is presently unknown. The researchers conducted their study in mice genetically engineered to express a cancer gene, or oncogene, in every insulin-producing islet cell of the pancreas. The result is that they develop a cancer known as islet carcinoma.

The model was created in such a way that new, spontaneous pancreatic tumors continuously emerged, providing a spectrum of disease progression that enabled the researchers to target early-stage cancer cells. (There is a three to four week period before cells expressing an oncogene develop the capability to form solid tumors with invasive, or malignant, properties.)

The premalignant cells in the model were observable, therefore, before they had undergone the "angiogenic switch," the point at which newly forming tumor masses begin emitting signals to recruit blood vessels to fuel their lethal growth. This period may serve as a paradigm for drugs aimed at "preventing" cancer development or recurrence. The classic tumor-transplant models of cancer do not afford this opportunity.

The researchers examined the efficacy of four compounds, AGM-1470, BB-94, angiostatin and endostatin, and a combination of angiostatin and endostatin. The goal of the first stage of the trial was to prevent the angiogenic switch in cells prior to the initial formation of solid tumors; in the second, to intervene in the rapid expansion of small tumors; and in the third to cause regression of large, end-stage cancers. (While each of the drugs acts differently, each in some way inhibits the molecular activity that promotes the growth of new blood vessels.)

In the precancerous lesions, BB-94, endostatin, angiostatin and the combination of angiostatin and endostatin prevented activation of the angiogenic switch in over half of the lesions that otherwise would have turned on the switch.

In the intervention trial, all four drugs were efficacious, impairing tumor growth from 60 percent to 88 percent. (The combination drug was not tested.) In the regression trial, AGM-1470 and the combination of endostatin and angiostatin each produced significant tumor regression in mice with substantial tumors, in contrast to endostatin or angiostatin alone, or BB-94. The former drugs reduced tumor volume by around 60 percent (compared to tumor volume at the starting point of the trial) and by around 72 percent (in untreated mice at end-stage disease).

None of the agents completely prevented the angiogenic switch in the precancerous stage, or blocked the growth of small tumors, or completely resolved lethal tumor burden. The fact that the treated mice still had cancer is not surprising, the researchers said, for two reasons: First, the time course of the trials was limited by design and by limited availability of the experimental drugs. Second, there was a relentless emergence and progression of new spontaneous pancreatic tumors due to the characteristics of the model, wherein tumor development is set in motion by an oncogene in virtually all of the 500,000 pancreatic islet ß cells.

"Our findings are encouraging about the potential value of angiogenesis inhibitors for treating cancer; we believe we can do even better," said Hanahan. "We've illustrated an approach to cancer therapy and prevention with this model that should be applicable to other mouse models of cancer, and in turn will hopefully guide the development of better ways to treat human cancers."

The study was conducted as part of the ongoing collaboration of the Hanahan laboratory at UCSF and that of Judah Folkman, M.D., at Children's Hospital Medical Center and Harvard Medical School. Folkman has pioneered the field of angiogenesis research and its application to cancer.

Other co-authors of the study were Kashi Javaherian, Ph.D., research associate in the Department of Surgery, Children's Hospital Medical Center and Harvard Medical School, Kin-Ming Lo, Ph.D., vice president of research of Lexigen Pharmaceuticals Corp.

The study was funded by grants from the National Cancer Institute.