Fever triage is challenging because bacterial and viral infections are often clinically indistinguishable. Routine assays aimed at detecting the pathogen can assist in determining the underlying infectious process. Yet, several diagnostic challenges remain: 1) in many cases the infection source is inaccessible or unknown; 2) identification of a virus does not preclude the possibility that an undetected bacteria may be the cause of the underlying illness; and 3) carriage of bacteria and viruses in healthy subjects often cause false alarms. An alternative approach to discriminating between bacterial and viral etiologies that is not limited by such challenges relies on monitoring the host's immune response to infection.
In line with this rationale, recent studies examining host-RNA signatures in response to different infections show promising results.1 However quantitative and rapid measurement (within minutes) of multiple host-RNAs remains a technical challenge, especially in resource-limited settings. In contrast, circulating host-proteins are readily amenable to rapid measurements using well-established technologies. Proteins routinely used today to support diagnosis of infection include procalcitonin and C-reactive protein (CRP), but these markers are subject to inter-patient variability. For example, elevated CRP levels are suggestive of bacterial infection, but similar levels may be observed in patients infected with adenovirus or influenza.2 To improve clinical performance, various combinations of several proteins into a single predictive score have been proposed. Such combinations have provided only limited-to-moderate diagnostic improvement to-date, likely due to the predominant focus on bacterial-induced proteins that therefore convey partially redundant information.
We reasoned that a combination of unrelated host immune proteins participating in different pathways may improve diagnostic accuracy. In particular, the inclusion of new host-proteins that are up-regulated in viral infections may be an innovative complement to bacterially-induced proteins in current clinical use. To discover novel host-biomarkers, we executed one of the largest quantitative screens to-date of host-proteins in patients with an acute infection (1002 patients and 600 proteins). Then, we screened possible protein combinations using feature selection algorithms to find the best performing signature for distinguishing bacterial and viral infections. The resulting host-signature comprised TRAIL (a novel virally-induced protein and member of the tumor necrosis factor family implicated in programmed cell death), IP-10 (small cytokine implicated in multiple cellular processes including chemotaxis and cell growth inhibition), and CRP (routinely used acute phase protein with diverse roles in tissue injury, infection and other inflammation processes). In line with our reasoning, TRAIL, IP-10 and CRP participate in different signaling pathways and exhibit complementary dynamics in response to bacterial and viral infections.
This host-signature is the basis for an inpatient diagnostic kit called ImmunoXpert™ (approved for use in Europe, Israel and Switzerland) and a point-of-care (POC) rapid diagnostic kit called ImmunoPoCTM that is currently under development. The host-signature has demonstrated clinical accuracy at differentiating bacterial and viral infections (AUC=0.94±0.02) when validated using an independent cohort of febrile patients with a wide range of clinical syndromes, ages, pathogens, time from symptom onset and co-morbidities3, as well as in follow up studies enrolling thousands of patients.
In summary, a host-signature incorporating both viral- and bacterial-induced proteins is amenable to rapid measurement, robust to patient and disease variability, and insensitive to microbial carriage. Together, these features enable superior diagnostic performance relative to standards-of-care for triaging fever3 and open the way to deliver timely actionable information (bacterial or viral/other etiology) at the point-of-care.
- Mejias, A., Suarez, N. M. & Ramilo, O. Detecting specific infections in children through host responses: a paradigm shift. Curr. Opin. Infect. Dis. 27, 228–235 (2014).
- Kunze, W., Beier, D. & Groeger, K. Adenovirus Respiratory Infections In Children. Do They Mimic Bacterial Infections? (2010). at <http://www.webmedcentral.com/article_view/1098
- Oved, K. et al. A Novel Host-Proteome Signature for Distinguishing between Acute Bacterial and Viral Infections. PLoS ONE 10, e0120012 (2015).