Infectious Notion.

Lessons from gene therapy promote viruses as cancer fighters

Poisons. Radiation. Scalpels. Cancer treatments can sound as dangerous as the disease itself. Standard therapies tend to swipe at cells almost indiscriminately, damaging healthy tissue while killing tumors. The lack of specificity can mean even more misery to an already suffering patient.

As candidates for the anticancer toolkit, viruses may seem to belong to the same therapeutic category as poison and radiation. Researchers, however, have new strategies to make viruses into a more compliant instrument.

Whereas traditional cancer treatments blast the disease with a swath of killing power, viruses can provide a fine, deadly beam--potentially no wider than a gene--and perhaps a substantial improvement in care.

The idea of using viruses as anticancer agents isn't new. The story reaches back nearly a century to doctors who noticed tumor regressions in cancer patients after they contracted viral diseases such as pneumonia. Until recently, however, technical limitations made research into viral treatments sporadic at best.

In one 1956 study, for example, doctors introduced adenovirus into 30 patients with cervical carcinoma. The virus, responsible for the common cold, had been isolated 3 years earlier and was known to kill epithelial cell types that are the first manifestation of most tumors. The experiment succeeded in initially shrinking many tumors, says Steven Linke, a molecular biologist at the National Cancer Institute in Bethesda, Md. However, the tumors returned, and the patients died within less than 3 months.

With the ascendancy of chemotherapy in the 1960s, anticancer viral research retreated. Now, however, as scientists are unlocking the mechanisms that regulate cell division and unraveling the puzzles of infection and immunity, interest in viral treatments for cancer is reemerging.

The latest direction of research has grown out of gene therapy techniques, which use viruses to deliver genes to living cells, says Edmund C. Lattime, a researcher at the Cancer Institute of New Jersey in New Brunswick. In gene therapy, scientists typically modify viruses so they can't replicate in the patient's body and cause infection.

Early clinical trials of gene therapy to combat genetic disorders weren't particularly impressive, he says. "A lot of us were naive when we started this a number of years ago, thinking we were just going to take viruses ... and that they were automatically going to do what we wanted them to do," says Lattime.

After struggling with the problems of how to insert genes into huge quantities of cells, and, especially, how to evade immune problems when repeated injections of cells were required, success appeared to be at hand in early 1999. That year, for example, scientists used a modified adenovirus to deliver normal versions of a gene called p53 to the lung tumors of 25 people. In many cases of cancer, mutations in p53 stop malignant cells from going into a natural cell-death phase called apoptosis. In 18 of the patients, the injections either reduced or arrested tumor growth (SN: 5/15/99, p. 310).

Scientists have also had success with a related technique called suicide gene therapy. The strategy is to introduce to a tumor site a gene encoding a toxin that can kill cancer cells. The technique also can deliver genes to block the formation of blood vessels feeding the tumor or to activate chemotherapy agents, says Antonio Chiocca, a neurosurgeon at Massachusetts General Hospital in Boston.

However, gene therapy has also had its much-publicized pitfalls.

In 1999, Jesse Gelsinger, an 18-year-old volunteer with a non-life-threatening disease that prevented proper processing of nitrogen, received gene therapy at the University of Pennsylvania in Philadelphia. The goal was to introduce the gene for an enzyme that would control his condition. Instead, he died within 15 days of receiving a high dose of the adenovirus carrying the gene. His death sparked congressional hearings, and the Food and Drug Administration temporarily halted similar gene-therapy trials at other research sites.

Sparked by the limitations of gene therapy, some cancer researchers are now attempting to capitalize on what viruses do best: replicate.

In both traditional and suicide therapy techniques, the virus is an inert vehicle that delivers its payload to a tumor. In contrast, cancer researchers are now developing therapies in which injected viruses reproduce and spread throughout the body.

As risky as traditional gene therapy can be, introducing a replicating virus into a patient is fraught with even more danger. Cancer specialists must strike a delicate balance. For safety, the virus must be purified, weakened, and modified. Ideally, it should be nonlethal in its original form, just in case any safeguard fails. It must be innocuous enough to slip past a patient's immune system, but infectious enough to reach all the tumor sites.

What lures the researchers is the possibility of finding or creating a replicating virus that confines its activities to the boundaries of its target tumor and doesn't infect healthy tissue.

Chemotherapy drugs have traditionally had a therapeutic index of no better than 6 to 1, meaning that they destroy up to 6 cancer cells for every healthy cell that dies. In contrast, one adenovirus currently being developed to fight prostate cancer by Calydon Pharmaceuticals in Sunnyvale, Calif., has demonstrated in animal experiments a therapeutic index of 10,000 to 1.

Work with replicating viruses has had a twofold goal: making agents that boost a patient's immune system and others that directly attack the tumor. Often, the same modified viral genome can be the staging ground for both efforts.

In one immune strategy, researchers use a traditional vaccine approach. Tumors themselves elicit an immune response from the body, but often it's not strong enough to overcome the malignant growth. Viruses manipulated to carry the gene for a tumor protein can boost the immune response. Researchers hope that such strategies will not only destroy an existing tumor but also trigger an immune-system memory that can ward off cancer recurrence.

In other instances, researchers load a virus with genes for proteins called cytokines, chemicals that normally activate elements of the immune system. Once inside the malignant cells, the genes turn the tumor into a cytokine-production mill and make it the agent of its own destruction.

Some of the latest work to exploit the relationship between viruses and the immune system offers the possibility of sidestepping a traditional problem in cancer therapies: getting the viral vehicle to all the sites of malignancy.

In a November 1999 paper in HUMAN GENE THERAPY, a team led by Robert L. Martuza at Georgetown University Medical Center in Washington, D.C., reported that cancer-ridden mice injected with a strain of herpesvirus experienced tumor regression both at the injection site and in remote growths. Further, they found, remote tumors regressed even at sites that the herpesvirus hadn't reached. The virus appeared to act like a red cape waved before a bull: It directed the immune system's attention to cancer cells, even distant ones, previously ignored.

Other strategies focus on finding viruses that will infect and kill cancer cells while leaving healthy cells alone. At least two families of viruses, parvovirus and reovirus, appear to naturally have such selectivity. Scientists also can engineer the property into other viruses, such as poliovirus, herpesvirus, adenovirus, and a candidate currently being tested, a lentivirus derived from HIV.

One way that researchers change viruses into cancer fighters is by taking out the genes that encode enzymes needed for replication. The virus will then grow and kill only in actively dividing cancer cells and others that have a rich supply of those enzymes. Slowly dividing healthy cells are a poor source of enzymes and so aren't affected by the virus.

Some researchers suspect that certain viruses are naturally attracted to cells in which a specific molecular pathway has gone awry. This so-called ras signaling pathway, which controls cell proliferation, differentiation, and death, is abnormally active in most cancer cells, reports Patrick Lee, a virologist at the University of Calgary in Alberta.

"We found that the correlation is amazing," says Lee. "We believe that 80 percent, maybe even more, of all [cancerous] cell lines have an activated ras signaling pathway."

In a paper scheduled for publication in the PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES, Lee and his colleagues give evidence that some laboratory-modified herpesviruses infect only cells with ras activation. Though the natural herpesvirus can cause meningitis and encephalitis, researchers are able to delete the genes that make the virus dangerous.

James M. Markert, a neurosurgeon at the University of Alabama at Birmingham, works with the HSV-1 strain of herpes-virus. He tries to exploit the virus' natural affinity for the central nervous system by pitting it against neural gliomas. Inaccessible, inoperable, and protected against chemotherapy agents by the blood-brain barrier, these cancers are nearly always fatal. With currently available treatments, half the patients die within a year of diagnosis, and the 5-year survival rate is less than 5.5 percent, says Markert.

An advantage to herpesvirus, says Markert, is that it possesses a large genome, into which researchers can pack more foreign material than other viruses will accommodate. Thus, researchers can try a double-whammy strategy, modifying a single glioma-seeking herpesvirus to carry genes that encode both immune-enhancing cytokines and that deliver suicide instructions to the tumor.

Markert and his colleagues have tried this herpes-plus-immunotherapy technique against gliomas in mice. In a study described in the Feb. 29 PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES, they demonstrated that treated mice survive more than twice as long as untreated animals.

In the May GENE THERAPY, researchers working with two strains of defanged, modified herpesvirus in the United States and in Scotland reported reassuring results of the first human safety trials in patients with malignant brain tumors. David H. Kirn of the Viral and Genetic Therapy Programme at Hammersmith Hospital in London, hails the safety results as "remarkable" in an editorial accompanying the report.

Researchers weren't studying efficacy in these trials, but their reports of "anecdotal cases of tumor shrinkage or prolonged progression-free intervals were encouraging," says Kirn.

Although Lee helped discover that the herpesvirus targets the ras pathway, it s not the virus he champions as an anticancer weapon. Lee's enthusiasm is reserved for the reovirus, which targets the ras pathway even without modification

The reo part of the name is an acronym for respiratory enteric orphan. The orphan designation, says Lee, means that the virus hasn't been linked to any known human disease, very unlike herpesvirus. Lee favors the reovirus because, also unlike herpesvirus, it is extremely easy to grow in large quantities. "What's the advantage of reovirus over herpes?" Lee asks, rhetorically. "Given a choice, which would you pick?"

The choice, however, isn't limited to the herpes and reovirus families. In fact, the workhorse of the field so far, says Lattime, is adenovirus, the same pathogen that yielded such dismal results in the 1950s and the tragedy last year.

The adenovirus has been studied the most and development of treatments that use it are furthest along, Lattime says. Its mechanism for seeking cancer cells is known. A research team led by Frank McCormick, a molecular biologist at the University of California, San Francisco, reported in 1996 that adenovirus targets cancer by identifying cells in which p53 can no longer prevent indiscriminate growth.

The efficacy of adenovirus, particularly when used in conjunction with more traditional chemotherapy, is impressive. In the August NATURE MEDICINE, a team headed by Fadlo R. Khuri of the Texas Medical Center in Houston reports that 25 of 30 patients with advanced head and neck cancer responded favorably to a combination of chemotherapy and an adenovirus called ONYX-015. The tumors in eight patients disappeared entirely.

Even with the progress being made with adenovirus, it's likely that several viruses will emerge as cancer therapies, each with its own strengths and weaknesses. Reovirus isn't the disease threat that herpesvirus is, but researchers can't modify reovirus to evade the immune response, as they have herpesvirus, Chiocca says. In fact, Lattime expresses doubt that cancer patients, scheduled to receive reovirus in the first safety trials this year, will be able to receive multiple injections without developing an immune response against the treatment.

Immunity and the other broad challenges that the scientists face with all viral families can be overcome, Chiocca predicts. Immunosuppressive drugs, carefully administered, may become necessary elements of viral therapy.

The researchers agree that the trend in their field is to try to find the perfect niche for each virus. "I think, based on past experience, that there really isn't any magic bullet for all cancers," says Linke.

Instead, the scientists want to appreciate the idiosyncrasies of each viral strain and each cancer type and find matches between disease and therapy. In the future, Chiocca predicts, "there will be a variety of different viruses with engineered mutations ... so that each virus will be like a different drug."

Still, one team of geneticists at the University of Alabama at Birmingham cautions in the July NATURE BIOTECHNOLOGY, "Claims of selective magic bullets need to be modest, though, because much remains to be known about the regulation of viral replication and how to harness it."