New Test Rapidly Distinguishes Viral, Bacterial Infections

Larry Hand

September 18, 2013

A new test that analyzes patients' immune responses, rather than the pathogens themselves, can rapidly distinguish viral infections from bacterial infections, according to an article published in the September 18 issue of Science Translational Medicine.

If carried into clinical use after further evaluation, the assay could help physicians better decide which patients need antibiotics and avoid inappropriate prescribing of antibiotics to patients who will not benefit.

Aimee K. Zaas, MD, MHS, associate professor of medicine at the Duke Institute for Genome Sciences and Policy, Durham, North Carolina, and colleagues previously reported that they could classify individuals according to respiratory viral infections based on microarray profiles of blood samples — work that has been reproduced by other researchers.

In the current study, the researchers used a Taqman low-density array (TLDA) platform to develop a reverse transcription–polymerase chain reaction (RT-PCR) assay and a predictive algorithm to test the blood of patients presenting with infection symptoms.

To develop the assay, they first tested 17 healthy volunteers who had been inoculated with influenza H3N2/Wisconsin and 24 who had been inoculated with influenza H1N1/Brisbane. They found that using expression data from a relatively small number of genes, they could classify the volunteers at 100% accuracy for H3N2 and 87% accuracy for H1N1.

"The performance of the RT-PCR assay to identify symptomatic versus asymptomatic individuals with experimentally inoculated influenza H1N1 or H3N2 is excellent...with cross-validation highly accurate between influenza subtypes," the researchers write.

They then validated the assay by testing peripheral blood RNA from 102 patients presenting with respiratory infections at Duke University Medical Center; Henry Ford Medical Center in Detroit, Michigan; or Monash General Hospital in Melbourne, Australia. The patients had either microbiologically proven viral respiratory infection or systemic bacterial infection. In this population, sensitivity of the RT-PCR assay was 89% (95% confidence interval [CI], 72% - 98%), and specificity was 94% (95% CI, 86% - 99%).

Ruling in or out

"[T]he viral infection 'score' represents a measure with potential clinical benefit for both 'ruling in' and 'ruling out' viral respiratory infection...with possibilities for improved triage, decreased utilization of inappropriate antibacterial therapy, and guiding judicious use of limited antiviral resources in a pandemic setting," the researchers write.

"It's an approach that could turn out to be very useful clinically, and certainly is likely to help us understand viral infections and how they differ from bacterial infections better," Andrew T. Pavia, MD, chief of the Division of Pediatric Infectious Diseases, University of Utah Health Sciences, Salt Lake City, and an infectious diseases advisor to the Infectious Diseases Society of America, told Medscape Medical News.

"What they've done is to turn the approach to diagnosing viral or bacterial infection around on its head," he continued. "We normally use tests to detect the germ or the pathogen that we're interested in. What they've done is to try and use the pattern of the human response to the infection to tell whether it's a bacteria or a virus. Obviously, that's terribly important, because if we can accurately detect who has a virus, we can avoid excessive antibiotic use and restrict antibiotic use to those people who have a bacterial infection."

Dr. Pavia explained that at this time, even with the most sophisticated devices available, physicians can only detect about 16 different viruses of approximately 30 or 40 that might be causing an illness. "The potential for a test like this, turned into a device, is perhaps it could be cheaper and more efficient. It wouldn't tell you, 'you have influenza B or you have corona virus HKU,' but it could tell you, 'you don't have a bacterial infection and we can spare you the antibiotic exposure.' "

In their analysis, the researchers conclude, "[W]e have established a 'proof of concept' that host expression of a relatively small set of genes, as measured by RT-PCR from blood RNA, can be used to classify viral respiratory illness in unselected individuals presenting at an emergency department for evaluation of fever."

"What they've done is pretty efficient," Dr. Pavia said. "But I think there's a lot more confirmatory work that needs to be done. You now have to take this into a more real-world setting looking at a variety of children and adults and prove that it works on other people. You would have to take what they've done, using the RT-PCR TLDA card, not suitable outside a research lab, and develop a convenient device. It's all doable."

This research was supported by the by the Defense Advanced Research Projects Agency, the National Institutes of Allergy and Infectious Diseases, and the Clinical Science Research and Development Service of the Veterans Health Administration. Six authors report filing for a provisional patent on the respiratory viral response signature. Four authors report receiving funding from Novartis. The other authors have disclosed no relevant financial relationships.

Sci Transl Med. 2013;5:203ra126. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: