12:00 AM
Aug 04, 2016
 |  BC Innovations  |  Product R&D

Antibiotic wisdom

A new infection diagnostic could smarten antibiotic use

While one of the biggest problems in infectious diseases is the overuse of antibiotics, few point-of-care tools are available to help physicians decide when and what to prescribe. Focusing on host genes rather than microbial ones, a group at Stanford University has developed an assay based on an 18-gene signature that distinguishes viral from bacterial infections and doesn't depend on the species present.

"We are not trying to identify what bugs you have or if the bug is even present. We are simply saying from the host's response it looks like you have bacterial infection," said lead investigator Purvesh Khatri, who is an assistant professor of medicine at Stanford's School of Medicine.

The lion's share of attention on antibiotic resistance has gone to new antibiotic discovery, but there is little dispute that better diagnostic agents are needed. In its 2015 National Action Plan for Combating Antibiotic Resistant Bacteria (CARB), the White House called for activities to "ensure that each patient receives the right dose at the right time for the right duration" within its goal to slow the emergence of resistant bacteria and prevent the spread of resistant infections.

Khatri told BioCentury that although some fast PCR-based assays have been developed, they are primarily for viruses and are based on the organisms' antigens, which means tests need to be run separately for each type of infection. He said that according to the CDC about 30-50% of antibiotic prescriptions are given for viral infections, but there is still "no diagnostic that definitively tells us if you have a bacterial or viral infection."

Last year, Khatri's group published a study in Science Translational Medicine showing it could differentiate between the presence or absence of infection based on a set of 11 host genes - the "Sepsis Meta Score" or SMS test - identified by multicohort analysis of patients with either sepsis or a sterile systemic inflammatory response syndrome (SIRS) that can present similarly in the clinic.

Now, in a second Science Translational Medicinepaper published in July, the team has found a seven-gene host response signature that distinguishes between viral and bacterial infections, dubbed the bacterial/viral metascore. Together with the first 11 genes, the findings provide the 18-gene signature that can help physicians determine if antibiotics are needed.

To validate the signature and ensure it would stand up to variability across testing centers, the group build a data analysis program called COCONUT (COmat COnormalization Using conTrols) that allowed it to compare data from to compare data from dozens of cohorts with disparate inclusion criteria.

Read the full 2110 word article

User Sign in

Trial Subscription

Get a 4-week free trial subscription to BioCentury Innovations

Article Purchase

$100 USD
More Info >