Although a bacterial culture can identify a huge swath of known pathogens, a viral culture is not nearly as representative. “Viruses are usually hunted one by one using serology and PCR,” said Don Ganem, MD, FIDSA, who gave the John F. Enders Lecture at the 2010 IDSA Annual Meeting in Vancouver. “But when this approach is exhausted, a diagnosis of viral infection is often made by exclusion, with the clinician saying, ‘I think you have a virus.’”

This problem also bedevils the discovery of new viruses. “Can we do better?” asked Dr. Ganem, a professor in the departments of Microbiology/Immunology and Medicine at the University of California at San Francisco (UCSF), who described a new method using genomics. “The dream is to take everything we know about virology and do one test that could be used for diagnosis and discovery.”

There are opportunities for discovering viruses in acute infections, Dr. Ganem noted, such as upper respiratory infections (URIs), encephalitis, myocarditis, and hepatitis, and in chronic diseases such as cancer, autoimmune disorders, and inflammatory diseases.

Working with Joseph DeRisi, PhD, and a team at UCSF, Dr. Ganem uses a genomics-based method to detect viral pathogens. The team has created a microarray of viral genomes with the viruses of humans, animals, invertebrates, plants, fungi, and bacteria. The array has 28,000 individual DNA sequences used to detect homologies with other viruses in a process called “virochip sample analysis.”

Once the laboratory receives a sample from a patient, the idea is to extract RNA from the sample and hybridize it to the array, then look for hybridization patterns that identify what kind of virus is present in the sample. This can be done through a visual inspection of the clustered array or through computational analyses.

In a real-world use of the virochip, Dr. Ganem’s team studied cohorts of URIs in children and in asthmatics and when compared to PCR, the virochip had a sensitivity of 85 to 90 percent and a specificity of 99 percent. But interesting problems arose along the way: “Sometimes the microarray detects plant viruses that come from food the patient ate,” he said. “This background noise is one of the drawbacks of the method.”

The virochip is currently not a realistic clinical test because of its technical sophistication and high cost, Dr. Ganem noted. “It might be better used as an instrument of discovery research.”

To date, the team has discovered a new clade of noncultivatable, very divergent rhinoviruses (HRV-C), a novel human cardiovirus (HTCV), and enterovirus 109. Besides the array, the UCSF team is also looking at high-throughput cDNA sequencing as the next frontier, which provides 20-40 million reads per sample. This technology, however, requires sophisticated bioinformatics and can cost up to  $3,000 per sample, Dr. Ganem said.

Audio and synchronized speaker slides from the 2010 IDSA Annual Meeting are available for purchase online.

How useful is this article?

< Previous Article | Next Article >