ONCOLOGY

A COLD FOR CANCER

Infection with a mutant virus makes some sick patients better

This summer, if all goes as planned, several dozen patients dying of head or neck cancer will each be injected with 50 billion living virus particles. Their doctors hope the infection will take hold and even spread. Radiation treatments and chemotherapy no longer help these people; there is good reason to suspect that the infection might.

This past May researchers reported that head and neck tumors shrank by half or more in six of 19 patients given a lower dose of the virus. Tumors stopped growing in five others. The results are only preliminary, but they support a bold new strategy of attacking cancer with a living drug: in this case, a mutant adenovirus.

In the wild, adenoviruses are common and cause no more harm than mild colds. The organism spreads by invading a cell, commandeering its genetic machinery and forcing the host to crank out viral clones until the cell membrane explodes. To succeed in its coup, adenoviruses must in most cells overcome a formidable defense, a protein called p53.

Like a genetic sentry, p53 monitors a cell's DNA for mutations caused by injury or viral attack. If it spots any, p53 halts the cell's reproductive cycle--preventing a virus from replicating, a mutation from propagating--and then signals for genetic repairs. Sometimes p53 goes a step further, activating a self-destruct mechanism to sacrifice the cell for the good of the body. Radiation treatment and chemotherapy injure tumor cells (as well as a lot of healthy ones) in the hope that p53 will then dispose of them or at least stop their growth.

When those treatments fail, it is often because p53 inside the tumor cells has been genetically disarmed or blocked by other proteins. Adenovirus contains genes that do the same, and that point of commonality led scientists at Onyx Pharmaceuticals in Richmond, Calif., to a clever idea.

They opened up the genome of an adenovirus and knocked out one of its anti-p53 genes. This new mutant strain, called O15, can still infect and kill defenseless cells that lack p53, so it dispatches many kinds of cancer cells handily. But it is nearly powerless against cells with normal p53--that is, most of the healthy parts of the body.

Whether O15 acts as a "smart bomb" against cancer, as some have called it, will depend on how patients' immune systems respond to the virus. In the first trial, subjects reported only flulike side effects. Many produced antibodies to O15, but Allan Balmain, head of laboratory research at Onyx, does not know whether the immune system will mop up the viral particles before they can kill the cancer or whether the body might actually go after infected cancer cells with new vigor.

Human safety trials, now under way for pancreatic, ovarian and colon cancer, do show that O15 is not perfect. After injection into the tumor, the virus does replicate, but it does not spread throughout the malignancy as hoped. David Kirn, Onyx's director of clinical research, thinks that is because his colleagues knocked out useful virulence genes that reside near the anti-p53 code on adenovirus DNA.

So Onyx is busily preparing new strains with a smaller disabled region. They appear to attack cancer more aggressively in animal studies, Kirn reports. Balmain is testing a version that makes infected cells vulnerable to the antiviral drug ganciclovir. "We can also add genes into the virus to make it kill cells better," he says. Not the usual goal in medical research, but oncology is a field accustomed to drastic measures.

--W. Wayt Gibbs in San Francisco