A multiplex polymerase chain reaction assay identified bloodstream pathogens in 94% of positive blood cultures and showed high agreement with standard microbiological methods while delivering results within 1 hour in adult patients with bloodstream infection treated at an oncology hospital.
Results from the multiplex panel were available within 1 hour following a positive blood culture signal, compared with 24 to 48 hours for conventional identification and susceptibility testing, potentially enabling earlier targeted antimicrobial therapy in high-risk patients.
In a prospective single-center study, researchers evaluated 93 positive blood culture bottles from patients with bloodstream infection using the BioFire FilmArray Blood Culture Identification 2 panel alongside routine microbiological methods, including matrix-assisted laser desorption ionization–time-of-flight mass spectrometry for identification and automated susceptibility testing.
The FilmArray Blood Culture Identification 2 panel identified microorganisms in 87 of 93 blood culture bottles (94%). Six culture-positive bottles were not detected by the panel; the researchers attributed these missed detections largely to organisms not included in the assay, although some discordant identifications were observed.
Agreement between the multiplex panel and standard methods was complete at the genus level, with five discrepancies at the species level. Across samples, the panel identified 96 microorganisms compared with 92 identified by standard microbiological methods. In one case, Streptococcus pneumoniae was detected by the multiplex assay but did not grow in subculture.
Most cultures were monomicrobial (85 of 93), while eight were polymicrobial. The multiplex panel detected multiple organisms in all polymicrobial samples, whereas standard microbiological methods identified polymicrobial growth in four.
Gram-negative bacteria predominated, particularly Klebsiella pneumoniae (43 isolates) and Escherichia coli (24 isolates). Gram-positive bacteria were identified in 11 cultures, and one culture yielded Candida auris by both methods.
The multiplex panel detected at least one antimicrobial resistance gene in 59 bacteria, with 107 total genes identified. Twenty-eight isolates carried one resistance gene, 15 carried two, 15 carried three, and one carried four. The most common genes were CTX-M (49), OXA-48–like (23), and New Delhi metallo-beta-lactamase (22), most frequently in K pneumoniae and E coli.
Phenotypic susceptibility testing was consistent with resistance gene detection in all but one isolate, a K pneumoniae strain carrying OXA-48 that remained susceptible to meropenem.
The panel covered approximately 82% of organisms identified in culture. Its main limitation was the inability to detect organisms not included in the assay. The study was also limited by its single-center design and relatively small sample size.
“FilmArray BCID2 panel results not only correlate well with conventional blood culture identification and susceptibility results but also identify antimicrobial resistance genes during a shorter period of time,” wrote lead study author İpek Mumcuoğlu of the University of Health Sciences Dr. Abdurrahman Yurtaslan Ankara Oncology Training and Research Hospital and colleagues.
The researchers reported no conflicts of interest.
