Comparison of the Panther Fusion and Allplex assays for the detection of respiratory viruses in clinical samples.

Abstract

Respiratory viral infections are the most frequent clinical syndrome affecting both children and adults, and early detection is fundamental to avoid infection-related risks and reduce the healthcare costs incurred by unnecessary antibiotic treatments. In this study, performance characteristics of two commercial methods, the Panther Fusion® assay (Hologic Inc., San Diego, CA, USA) were compared to Allplex™ respiratory panels (Seegene, Seoul, South Korea) for the detection of influenza A (Flu A), influenza B (Flu B), respiratory syncytial virus (RSV), parainfluenza virus (PIV), human metapneumovirus (hMPV), rhinovirus (RV) and adenovirus (AdV) targets. A total of 865 specimens collected prospectively and retrospectively were included, and discordant results were further examined using another commercial product, R-GENE™ respiratory kits (bioMérieux, Marcy l'Etoile, France). There was high agreement between both methods, with 98.6% concordance and a kappa (k) value of 0.9 (95% CI: 0.89-0.92). A specific analysis of both methods for each viral agent demonstrated comparable sensitivity and specificity, both ranging from 0.83 to 1 with good predictive values for the prospective part of the study. Good agreement between both methods was also found for the κ values obtained (ranging from 97.55% to 98.9%), with the lowest for hMPV (k = 0.83, 95% CI: 0.75-0.91) and RV (k = 0.73, 95% CI: 0.65-0.81). Amplification efficiency, measured according to the value of the cycle threshold (Ct) obtained in each of the amplifications in both tests, was significantly better with Panther Fusion for Flu A, Flu B, hMPV and RV. Regarding discordant results, R-GENE showed higher agreement with Panther Fusion-positive specimens (negative for Allplex; n = 28/71, 34.9%) than with Allplex-positive samples (negative for Panther Fusion; n = 7/49, 14.3%). In summary, Panther Fusion proved to be a more efficient fully-automated methodology, requiring shorter hands-on and turnaround times than Allplex.