Aveo's cancer drug research rests on finding the right mouse

Company tries to make tumors in rodents match human tissue
By Jeffrey Krasner, Globe Staff | July 2, 2007

Want a better cancer drug?

That's the thinking behind Aveo Pharmaceuticals Inc. of Cambridge, a five-year-old company that says it has a better way to induce cancer in laboratory mice. As a result, research conducted on those mice is a better indicator of how humans will respond to experimental drugs.

"Our models replicate more faithfully what actually happens in a human tumor," said Tuan Ha-Ngoc , chief executive of Aveo.

As a result, the company says, drugs developed using its unique cancerous mice will have a greater chance of being approved.

Aveo's approach takes aim at a fundamental problem facing biotechnology companies. Many drug candidates show promise in animal tests, so they're moved into clinical trials on humans, which often take years and cost hundreds of millions of dollars.

But only one in five drugs that enter human clinical trials ultimately are approved, according to the Tufts Center for the Study of Drug Development. Murray O. Robinson, Aveo's senior vice president for research, said the statistics are even bleaker for experimental cancer treatments.

"Twenty drugs go into the clinic, and only one comes out," said Robinson. "That's a 95 percent false positive rate."

A recent example is Telcyta , an experimental drug developed by Telik Inc. of Palo Alto, Calif. The drug produced promising results in early clinical tests when used to treat ovarian and lung cancer. But in December, the company said Telcyta failed three pivotal tests. The firm's stock lost about 70 percent of its value overnight.
Robinson said many clinical trials fail because experimental drugs are first tested in mice on what are called xenograft tumors. Scientists extract tumor cells from a human, and cultivate them in lab dishes. Those cells replicate, and they can then be injected under the skin of laboratory mice, where they develop into tumors.

Robinson described these tumors with the disdain of a French chef examining a hothouse tomato.

Because the cells are grown in the lab, they don't closely resemble the original tumor. They don't develop naturally, in an organ like the breast or colon. The tumor cells interact with the plastic containers in which they are cultivated.

"They've changed and mutated," said Robinson. "They're not the same cells that were originally extracted."

It's a problem for many cancer researchers.

"Developing mouse models that more closely resemble human tissue would be a great breakthrough," said Dr. Sandra Orsulic , assistant professor of pathology at Harvard University and a specialist in creating mice with ovarian cancer for research at Massachusetts General Hospital. "Any new mouse model would be better than what we have at the moment."

Aveo makes its better tumor by starting with mouse stem cells. In the lab, scientists are able to give the stem cells genetic instructions so they develop into a particular form of cancer, with a propensity to grow in a particular organ within the mouse's body. They also introduce a genetic on/off switch, so the cancer doesn't develop until a special agent is delivered through the mouse's drinking water. As the tumors grow in the mice, they develop small variations, the same way that tumors within humans differ slightly from one patient to another.

Finally, the cells are programmed to produce luciferase -- the same enzyme that makes fireflies glow. When tumors develop, scientists can measure their growth -- or shrinkage -- by examining the mice under special light that makes the enzyme glow. That makes it possible to chart the progress of the disease without sacrificing a mouse and cutting it open to view the tumor directly.

Aveo was originally known as GenPath Inc. The company has raised $118 million in four rounds of venture capital, including $53 million in May, and has cash to last through the end of next year.

Aveo calls its better mouse model the Human Response Prediction platform. Ha-Ngoc said biotech firms, big pharmaceutical companies, and contract research outfits have sought to purchase the technology. He's not interested in selling .

Instead, the firm is using the platform to develop its own drugs in-house. The first candidates are drugs that it licensed from large Japanese pharmaceutical firms . Another is an antibody it discovered in-house and is developing with drug giant Schering-Plough Corp., which invested $10 million in Aveo.

Many biotech companies have promised better results developing drugs with specialized genetic expertise. Most have failed. Millennium Pharmaceuticals Inc. of Cambridge touted its gene ingenuity for years, but the only drug it sells is one it bought from another company.

Should Aveo's method succeed, it will face another challenge. The variations within the mouse tumors mean that any drugs will be targeted to a small subgroup of say, lung cancer patients. Such drugs could be successful, but they wouldn't likely be blockbusters generating $1 billion or more in annual sales.

Tuan said he isn't discouraged.

"I can live with a drug that generated $500 million," he said.

Jeffrey Krasner can be reached at krasner@globe.com.

» back to top