2020. Concordance of Direct vs. Indirect Pathogen Detection Using the BioFire(®) System

BACKGROUND: Polymerase chain reaction (PCR) is a highly sensitive and specific method for pathogen detection. While direct methods enable rapid identification, they are limited by pathogen titer, available assays, or sample matrix. Transcriptomic analysis addresses these limitations by measuring systemic host gene expression changes to infections. The BioFire System uses sample-to-answer multiplex PCR that was adapted to detect 42 transcripts differentially expressed during viral and bacterial infections. Here we report concordance between indirect detection of viral respiratory pathogens and the FDA-cleared BioFire(®) Respiratory Panel 2 (RP2). METHODS: Paired nasal pharyngeal swabs and blood samples were obtained by informed consent from patients with suspected acute respiratory illness. Swabs (COPAN FLOQSwab(™)) were collected and stored in viral transport media (BD) for BioFire RP2 testing and peripheral blood samples were collected in PAXgene tubes (Qiagen) for testing with the research use only human response (HR) panel. A logistic regression model was developed to classify viral and nonviral positive samples using normalized quantification cycles for each assay. Probabilities of viral infection for each sample were calculated via cross-validation and an optimal threshold of positivity was identified. RESULTS: Overall accuracy of the HR panel relative to BioFire RP2 was 86% (CI(95%) 80%, 91%) for all viral infections with an area under the receiver operating characteristic curve of 0.87 (CI(95%) 0.76, 0.95). Accuracy varied by infection etiology: Influenza virus 100% (CI(95%) 88%, 100%) Human Rhinovirus/Enterovirus 63% (CI(95%) 35%, 85%), other viral infections 50% (CI(95%) 25%, 75%). While most of the BioFire RP2 negative results exhibited low viral probabilities, strong viral probabilities were measured in a few samples; this may be indicative of an infection at a low titer or the presence of a viral pathogen not on the BioFire RP2 Panel. CONCLUSION: These results demonstrate that indirect transcriptomic analysis resulted in similar accuracy to direct viral pathogen detection in acute respiratory patients; however, additional research is needed to elucidate the relationship between transcriptional results and infectious etiologies. DISCLOSURES: C. Gritzen, BioFire Diagnostics, LLC.: Employee, Salary. T. Wilson, BioFire Diagnostics, LLC.: Employee, Salary. J. Nawrocki, BioFire Diagnostics, LLC.: Employee, Salary. M. Deneris, BioFire Diagnostics, LLC.: Employee, Salary. C. Baird, BioFire Diagnostics, LLC.: Employee, Salary. E. Ott, BioFire Diagnostics, LLC.: Employee, Salary. J. Jones, BioFire Diagnostics, LLC.: Employee, Salary. J. Bastar, BioFire Diagnostics, LLC.: Employee, Salary. H. Kim, BioFire Diagnostics, LLC.: Employee, Salary. S. House, Washington University School of Medicine: Research Contractor, Research support. D. Cohen, Nationwide Children’s Hospital: Research Contractor, Research support. A. Leber, Nationwide Children’s Hospital: Research Contractor, Research support. R. Crisp, BioFire Diagnostics, LLC.: Employee, Salary. A. Hemmert, BioFire Diagnostics, LLC.: Employee and Investigator, Salary.