Abstract
BACKGROUND: The underlying cause of fever of unknown origin (FUO) remains unidentified in up to 51% of cases despite systematic evaluation. Microbial cell-free DNA next-generation sequencing (mcfDNA-NGS) offers an agnostic, noninvasive approach to pathogen identification, but the utility and clinical impact of this assay in FUO remain unknown. METHODS: This retrospective cohort study evaluated adult patients referred for FUO evaluation at a tertiary medical center between November 2019 and November 2023. Patients underwent both standard microbiologic testing (ST) and mcfDNA-NGS. Diagnostic impact was assessed in 4 domains: new diagnoses, earlier time to diagnosis, avoidance of invasive procedures, and non-hypothesis-driven diagnoses. Logistic regression was used to identify predictors of positive mcfDNA-NGS testing. RESULTS: Among 176 patients, mcfDNA-NGS was positive in 44.3%, with 49% of these cases considered clinically significant. Infectious cause of FUO was identified in 39% of patients, noninfectious in 35%, and unknown in 26%. mcfDNA-NGS contributed to a positive diagnostic impact in 30% of cases, mainly by earlier diagnosis (16%) and potential for avoidance of invasive procedures (10%). Positive mcfDNA-NGS was significantly associated with higher Charlson comorbidity index score (odds ratio [OR], 1.22; P < .001) and white blood cell (WBC) count ≤4.5 × 10(9) cells/L (OR, 8.61; P < .001). Conversely, FUO without localization was associated with a decreased likelihood of positive mcfNDA testing (OR, 0.18; P < .001). CONCLUSIONS: mcfDNA-NGS effectively complements ST in diagnosing FUO, providing earlier detection and minimizing invasive testing. Clinical predictors such as high comorbidity and low WBC count may guide the optimal use of mcfDNA-NGS in FUO. Prospective evaluation of optimal timing and use of mcfDNA-NGS and cost-benefit analysis in FUO is needed.