Breast cancer, brain tumors not caused by viruses, study finds

An estimated 1 in 5 cancers is caused by an infectious agent, either bacterial or viral. For years, scientists have traced the linkages piecemeal, scouring epidemiological data for clues about the agents behind specific cancers. In the new study, researchers screened the genomes of thousands of tumors at once to identify evidence of viral infection.

"In cancer research and treatment, there has been a lot of focus on associations that have not been proven, some of which have actually have been shown to be wrong," said Ka-Wei Tang, M.D., a doctoral candidate at University of Gothenburg, Sweden who worked on the study. "Researchers are starting to realize that we need truly unbiased methods to uncover meaningful associations. In this study, we take advantage of the opportunities offered by next-generation genetic sequencing, which has really revolutionized cancer research over the past decade."

The findings suggest treating brain tumor patients with antivirals is misguided, said Tang. Treatments that are not scientifically supported have the potential to harm patients by introducing false hope, side effects, and unnecessary medical costs.

"Patients are often devastated by a cancer diagnosis, and as a sick patient you always want to try anything that might work," said Tang. "But it doesn't help patients to undergo unnecessary treatment, and we're able to prove that the science does not support a viral cause for these cancers."

The researchers sequenced the genomes of tumor samples collected from more than 4,000 patients. After filtering out the genetic information corresponding to known human genes, they screened for known virus genes and used an additional analysis to identify any previously unknown viruses.

The results provide a map of the associations among viral infections and the 19 cancer types included in the study. "This method is very effective at detecting pathogens in an unbiased way. It's important to use an unbiased method so we do not present false associations," said Tang.

Previous studies have suggested links between breast cancer and Epstein-Barr virus and between brain tumors and cytomegalovirus. The new study disproves those associations.

In addition to these somewhat surprising findings, the study also bolsters the evidence for other previously-known virus-cancer associations. For example, the study found strong evidence linking liver cancer with hepatitis B and cervical cancer with human papilloma virus.

Understanding the viral causes behind some cancers can inform treatment options and also aid in prevention efforts. For example, Tang said the hepatitis B vaccine has prevented almost as many cancer cases as smoking cessation. Knowing the cause of a particular patient's cancer can also help doctors personalize their treatment.

"In the end, this is something that is good for patients, because if we are able to trace a cancer's cause, we are able to tailor the treatment accordingly," said Tang. "This type of information can improve patients' survival and quality of life."

The study was not able to identify so-called "hit-and-run" viruses, which may have infected a person in the past but are no longer present in the tumor, or latent infections, in which a virus becomes dormant within a person's cells. New research addressing latently-infected tumors will be presented at Experimental Biology 2014.

Ka-Wei Tang presented the findings during the Experimental Biology 2014 meeting on Tuesday, April 29.