Assessment of Multiplex Molecular Tests for Detecting Viral Infections in Lower Respiratory Tract Specimens.

Abstract

Lower respiratory tract (LRT) infections represent a major cause of morbidity and mortality, particularly among critically ill patients. Rapid molecular diagnostic tests have improved the detection of respiratory pathogens; however, most commercial assays are validated exclusively in upper RT (URT) specimens, limiting their applicability in LRT samples, which are often more representative of disease in advanced or severe cases. This study evaluated the diagnostic performance of two commercially available assays, the BioFire Respiratory Panel 2.1 Plus and the Panther Fusion SARS-CoV-2/Flu A/B/RSV assay, on bronchoalveolar lavage (BAL) specimens, using the Allplex Respiratory Panel 1/2/3 and the Allplex SARS-CoV-2 assay as reference methods validated both for URT and LRT matrices. A total of 132 BAL samples were analyzed. BioFire identified more positives than did Allplex, particularly for human rhinovirus/enterovirus (HRV/EV), human parainfluenza virus (HPIV), and non-severe acute respiratory syndrome coronavirus (SARS-CoV)-2 coronaviruses. The overall agreement between BioFire and Allplex was fair (κ = 0.237), and pathogen-specific concordance was almost perfect for SARS-CoV-2 (κ = 0.841), influenza A/B (κ = 0.808), and HPIV (κ = 0.884). The Panther assay showed substantial agreement with Allplex (κ = 0.719) and near-perfect concordance for SARS-CoV-2 and influenza viruses, while BioFire and Panther exhibited almost perfect interassay agreement (κ = 0.903). These findings demonstrate that assays validated for URT specimens can perform reliably on BAL samples, underlining the diagnostic potential of LRT matrices and the need for expanded validation of molecular respiratory panels across specimen types.