Cancer: Is There a Brain Tumor Virus?

In 2002, UCSF neurosurgeon Charles Cobbs published a novel finding in a prominent cancer journal: nearly all of the two-dozen brain tumors he had analyzed were teeming with a common herpes virus called cytomegalovirus, or CMV. Normally, CMV is harmless—it lies dormant in roughly 80 percent of the population—but in Cobbs's tumor samples, the virus appeared to be actively replicating, even as it remained dormant in nearby healthy tissue. "When I first saw the data, I couldn't sleep for a week," says Cobbs. "I kept asking myself, 'can this be?'" If his findings were correct, they might shed light on the causes of brain cancer, or better yet, provide a new target for battling—maybe even preventing—the disease.

But by 2004, at least two labs had tried and failed to replicate Cobbs's results. That might have been the end of the story, were it not for the young neurosurgeon's audacity. Convinced that his methodology was better than his colleagues, he offered to show both research teams his technique. One group, led by Duke University neuro-oncologist Duane Mitchell, accepted. Last year they published the first peer-reviewed confirmation of Cobbs's work. "We have enough evidence now to say that this merits serious attention," says Mitchell. As the journal Science wrote last week, a flurry of papers exploring a possible link between CMV and brain cancer have caught the attention of at least some experts, spurring the first conference on the subject last October and touching off a handful of clinical trials.

The findings have opened a new avenue of inquiry for one of the most intractable cancers—Glioblastoma Multiforme, an aggressive brain tumor, diagnosed in 10,000 new patients every year and fatal in virtually all cases. (Sen. Ted Kennedy was stricken with the disease last year). The alleged link between CMV and brain cancer may also represent the latest reversal of a decades-old consensus that generally speaking, viruses don't cause cancer. While some scientists are urging caution in interpreting this growing body of evidence, others say that a bias against "cancer-virus" research highlights a major flaw in the way science works. Ideas that challenge the conventional wisdom are often shunned in favor of "safer" hypotheses that stand a better chance of gaining acceptance and securing research dollars. "The powers that be are really opposed to funding this kind of research," says Cobbs who is now at California Pacific Medical Center. "They would rather put their money on more discreet projects where the outcomes are clear."

To be fair, the history of cancer-virus research is littered with false starts and embarrassing missteps. In 1926, a Danish scientist scored a Nobel Prize for showing that parasitic worms cause stomach cancer; it was later discovered that the "tumors" were actually lesions, triggered by vitamin deficiency. In the early 1970s scientists still believed that many if not most human cancers were triggered by some sort of infection. Famed HIV scientist Robert Gallo spent years at the National Cancer Institute trolling for the viral culprit, but most of his studies were never replicated and by the end of the decade, the hypothesis had been abandoned. "You have to tread carefully with findings like these," says Robert Weinberg, a biology professor and cancer researcher at MIT. "The majority of these claims tend to go up in smoke."

Today we know that at least three cancers are virus-induced: cervical cancer (Human Papiloma Virus, or HPV), liver cancer (Hepatitis B), and lymphoma (Epstein-Barr virus). But many questions still need answers before scientists can add brain cancer and CMV to that list. Chief among them is whether the virus actually triggers tumor growth. Cobbs thinks this may be the case, but he says the virus's influence is probably indirect. "It's not like a typical virus—disease relationship," he says. "The cancer may stem from chronic inflammation that is triggered by the virus and persists for years and years." If he's right, scientists may one day be able to develop a vaccine that prevents brain cancer by targeting CMV, much like Merck's Gardasil protects against cervical cancer by inoculating against certain strains of HPV.

Other researchers hypothesize that rather than induce tumor formation, CMV might simply abet their growth. Studies have shown that the virus promotes angiogenesis—the creation of an extra blood supply that tumors need to survive and grow.

Even if CMV doesn't cause brain tumors, the virus promises to be a useful target for future glioblastoma therapies. Mitchell's team is testing a vaccine made from immune cells that have been trained to attack CMV proteins. While the trial is too small to be conclusive, the vaccine, which enhanced immune responses against CMV, did extend patients' median survival time from 15 to more than 20 months. Meanwhile, Swedish researchers have just completed a clinical trial of the anti-CMV medication Valcyte to see if it can prevent the recurrence of brain tumors that have been surgically removed. The data is being tabulated now; if it looks good, Roche may eventually launch a large-scale study.

To make real progress, however, scientists will need funding. Cobbs says that he has had a dozen or so grant proposals on CMV-glioblastoma research rejected by the National Cancer Institute and other funding agencies. "People from NCI have said that we need to prove CMV causes brain cancer before they fund us," says Cobbs. "It's putting the cart before the horse."