Puumala, Emily

Abstract Background Identification of the microorganisms causing infection in children with central nervous system infection is important. Conventional cerebrospinal fluid (CSF) tests may provide limited pathogen detection. CSF metagenomic next-generation sequencing (mNGS) may provide broader pathogen detection in a single assay. Methods Results of CSF mNGS testing performed at Mayo Clinic on patients and nonpatients between January 2024 and June 2025 were retrospectively analyzed. The subcohort of children treated at Mayo Clinic was analyzed to assess test performance as compared with conventional CSF diagnostics and potential clinical impact. Results In total 134 patients (138 CSF mNGS tests) from 22 US states and 2 international sites were included. Overall positivity was 16% with viruses representing most (59%) detections. In the Mayo Clinic subcohort (n = 15), test positivity was 27%: organisms included single detections of Streptococcus mitis group and HIV type 1, as well as 2 detections of Epstein-Barr virus, all confirmed by conventional tests. Positive results were associated with CSF pleocytosis (P = .026). Higher yield was observed in participants whose care was overseen by pediatric infectious diseases (PID) specialists (P = .026) and when a specimen hold strategy was applied (P = .033). Most results (87%) did not alter antimicrobial therapy, with 2 negative results contributing to antimicrobial de-escalation. Conclusions CSF mNGS demonstrated higher diagnostic yield in patients whose specimens showed CSF pleocytosis and in those who had PID consultation. Diagnostic stewardship strategies were applied, including specimen hold approaches and PID consultation.

ABSTRACT Bloodstream infections (BSIs) caused by opportunistic yeasts, namely those of the genus Candida , demonstrate strikingly high mortality rates and require early therapeutic intervention to optimize patient outcomes. Advancements in microbiological diagnosis of BSIs have introduced rapid molecular methods, including multiplex PCR panels, to detect common organisms directly from blood culture media. The BIOFIRE FilmArray Blood Culture Identification 2 (BCID2) assay is one such method and detects seven of the most common yeast species responsible for fungemia, enabling timelier therapeutic optimization by clinicians. Here, we evaluated the clinical performance of BCID2 for yeast identification from positive blood cultures between 2021 and 2025 at a tertiary medical center, compared to the reference standard method, which couples subculturing and diagnostic mass spectrometry. A total of 252 individual yeast identification events were confirmed using the reference method during our study period, 33 (13%) of which represent organisms that lack BCID2 PCR-based targets and 219 (87%) represent organisms targeted by this assay. Of the 219 identifications compatible with the BCID2 panel targets, 209 (95%) were correctly identified. Discordance between BCID2 and the reference method commonly stemmed from the inability of BCID2 to target all yeasts encountered clinically in blood culture, as well as its reduced accuracy in cultures where >1 yeast is present. Overall, these data suggest that the BCID2 assay is a rapid and accurate method to identify the most common organisms causing yeast BSI from blood culture; however, subculture-based reference techniques are important complementary tools in the clinical microbiology laboratory. IMPORTANCE Invasive fungal diseases, including bloodstream infections (BSIs) caused by yeasts, are increasing in prevalence alongside the growing global population of immunocompromised individuals. Prompt and targeted treatment is necessary for effective management of yeast BSI, and microbiological diagnostics play an important role in tailoring patient treatment. Rapid molecular assays that simultaneously detect multiple pathogens associated with a specific syndrome are attractive strategies to expedite pathogen identification. This includes the BIOFIRE FilmArray Blood Culture Identification 2 (BCID2) assay, which detects 33 organisms including 7 yeasts commonly associated with fungemia, directly from blood culture media. Its performance has been well established for the detection of bacteria and bacterial resistance markers from blood cultures. Here, we provide a yeast-specific snapshot into the microbiological performance of BCID2, highlighting its accuracy as well as limitations in detecting medically important yeasts compared to standard culture-based identification from a four-year specimen cohort collected in a tertiary care hospital.