Tellevik, Marit Gjerde

Abstract Background New insights into community-acquired pleural infection microbiology may enable development of more rapid and accessible diagnostics that can streamline treatment. This study systematically compared molecular approaches for pathogen detection in pleural fluid from patients with such infections. Considering the need for rapid results with minimum hands-on time and test complexity, off-label use of the cassette-based, random-access BIOFIRE Joint Infection Panel (BF-JIP, bioMérieux) with and without supplementary in-house polymerase chain reaction (PCR) assays was evaluated. Methods Pleural fluids from 162 subjects from the Mayo Clinic (Rochester, Minnesota) and Haukeland University Hospital (Bergen, Norway), where molecular tests are routinely used for analyzing pleural fluid, were studied. The performance of the BF-JIP was compared to 16S rRNA gene PCR followed by Sanger and/or next-generation sequencing (16S rRNA gene PCR/sequencing), an in-house PCR panel specifically designed for pleural space infections, and culture. Results The BF-JIP supplemented with 3 in-house PCR assays targeting Fusobacterium nucleatum group, Streptococcus intermedius/constellatus, and Aggregatibacter aphrophilus detected 96.8% of pleural infections, providing complete microbial profiles in 70.1%. In comparison, the detection rate of 16S rRNA gene PCR/sequencing was 92.2%, providing complete microbial profiles in 87.0%. Culture detected bacteria in only 31.9% of pleural fluids, providing complete microbial profiles in 15.3%. Conclusions These results underscore the importance of molecular approaches in diagnosing pleural space infections and demonstrate the feasibility of panel-based PCR-based diagnostics for these sometimes-complex infections. The BF-JIP, designed for testing synovial fluid, could be modified to provide a useful diagnostic for testing pleural fluid.

ABSTRACT The rpoB gene has been proposed as a promising phylogenetic marker for bacterial identification, providing theoretically improved species-level resolution compared to the 16S rRNA gene for a range of clinically important taxa. However, its utility in diagnostic microbiology has been limited by the lack of broad-range primers allowing for its amplification from most species with a single PCR assay. Here, we present an assay for broad-range partial amplification and Sanger sequencing of the rpoB gene. To reduce cross-reactivity and allow for rpoB amplification directly from patient samples, primers were based on the dual priming oligonucleotide principle. The resulting amplicon is ~550 base pairs in length and appropriate for species-level identification. Systematic in silico evaluation of a wide selection of taxa demonstrated improved resolution within multiple important genera, including Enterococcus, Fusobacterium , Mycobacterium , Streptococcus , and Staphylococcus species and several genera within the Enterobacteriaceae family. Broad-range rpoB amplification and Sanger sequencing of 115 bacterial isolates provided unambiguous species-level identification for 97 (84%) isolates, as compared to 57 (50%) using a clinical 16S rRNA gene assay. Several unresolved taxonomic matters disguised by the low resolution of the 16S rRNA gene were revealed using the rpoB gene. Using a collection of 33 clinical specimens harboring bacteria and assumed to contain high concentrations of human DNA, the rpoB assay identified the pathogen in 29 specimens (88%). Broad-range rpoB amplification and sequencing provides a promising tool for bacterial identification, improving discrimination between closely related species and making it amenable for use in culture-based and culture-independent diagnostic approaches.