New Technology Enables Differentiation of Bacterial and Viral Infections

It can help inform appropriate antibiotic prescribing, but likely would only form part of testing protocol, experts say.

With the COVID-19 pandemic still in full swing, and flu season on the way in much of the United States, clinicians need tools for differential diagnoses.

That tool may be in the offing. On September 20, the Food and Drug Administration (FDA) approved the MeMed BV test on the point-of-need platform MeMed Key for use in children and adults. The technology is designed to help healthcare providers distinguish between bacterial and viral infections and, hopefully, limit inappropriate antibiotic prescribing.

“For those of us who care for acutely ill children, we have been waiting decades for accurate, rapid diagnostics to confidently guide the care of moderately ill children without a clear focus of infection or recognizable viral illness,” Rich Bachur, MD, Professor of Pediatrics and Emergency Medicine, Harvard Medical School, and Chief, Division of Emergency Medicine, Boston Children’s Hospital, said in a press release issued by the company. “This novel test offers promise to help differentiate those children with self-limited viral illness from those with possible bacterial infection, thereby supporting the judicious use of antibiotics.”

According to the company, MeMed BV decodes the body’s immune response to infection, the “host response,” as opposed to merely detecting the presence of a microbe. This enables robust diagnosis when the infection site is inaccessible or unknown, even when the pathogen is undetectable using conventional tests, or when the cause of infection are emerging new pathogens, allowing for more informed antibiotic prescribing decisions, it said.

The technology identifies a “host-protein signature” using both viral- and bacterial-induced biomarkers: tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), interferon gamma-induced protein-10 (IP-10), and C-reactive protein (CRP). In a study using samples collected from 175 patients with viral and 139 with bacterial infections, signature sensitivity was 93.5% and specificity 94.3%. Both were significantly higher than those for CRP, procalcitonin, interleukin-6, human neutrophil lipocalin, white blood cell count, absolute neutrophil count, and prediction rules. In addition, the signature identified as viral 50 of 57 viral patients prescribed antibiotics.

More recently, in a study that has been posted on a preprint site and has not yet been peer-reviewed, the signature’s area under the receiver operating characteristic curve (AUC) in the accurate identification of patients infected with SARS-CoV-2 was 0.86. This performance was superior to IL-6 (AUC 0.77) and CRP (AUC 0.78). In addition, the signature differentiated patients who further deteriorated after meeting a severe outcome from those who improved and projected 14-day survival probabilities, according to the researchers.

As promising as the technology appears to be, clinicians would like to see how it performs in larger clinical trials before incorporating it into clinical practice.

“Does it differentiate COVID-19 specifically from other viral respiratory infections such as flu, RSV, the familiar circulating coronaviruses such as OC43, etc.? If so, then it might be useful,” an epidemiologist and infectious disease specialist told Contagion on the condition of anonymity. “If not, what do I do with the result without further testing?”