Abstract
Of the thousands of viruses infecting humans, only seven cause cancer in the general population. Tumor sequencing is now a common cancer medicine procedure, and so it seems likely that more human cancer viruses already would have been found if they exist. Here, we review cancer characteristics that can inform a dedicated search for new cancer viruses, focusing on Kaposi sarcoma herpesvirus and Merkel cell polyomavirus as the most recent examples of successful genomic and transcriptomic searches. We emphasize the importance of epidemiology in determining which cancers to examine and describe approaches to virus discovery. Barriers to virus discovery, such as novel genomes and viral suppression of messenger RNA expression, may exist that prevent virus discovery using existing approaches. Optimally virus hunting should be performed in such a way that if no virus is found, the tumor can be reasonably excluded from having an infectious etiology and new information about the biology of the tumor can be found.