Bloodstream infections (BSI) are common, and identifying the causative organism is crucial for effective patient management. Shotgun metagenomics (SMg) has emerged as a promising diagnostic tool; however, standardized protocols are lacking. This study aimed to evaluate the use of SMg for diagnosing BSI in patients with confirmed or suspected infections, using stored samples collected at the time of blood culture (BC). DNA extraction was performed with Add-on 10 complement and SelectNA Blood Pathogen kit (Molzym) and SMg sequencing was performed on an Illumina MiSeq instrument (Illumina). The outputs from five taxonomic classification tools were compared with routine blood culture. Of the initial 51 samples (36 BC-positive and 15 BC-negative), 36 (71 %) were included in the taxonomic classification analysis. Fifteen samples were excluded due to a low DNA library yield (n = 8) or low sequencing output (n = 7). In two cases, SMg results matched BC findings involving one Cutibacterium acnes and one Staphylococcus aureus. These organisms could be clearly distinguished from the background level of bacterial DNA. Aside from these, SMg identified additional bacterial findings that overlapped with BC results but at low abundance making interpretation more difficult. Most SMg reads were suspected to represent contaminations, originating either from the patient or the laboratory. The output from the different taxonomic classification tools were overall similar but displayed notable differences related to their strategies for identifying bacterial findings. Based on these results, we discuss the challenges associated with SMg-based diagnosis of BSI and highlight key areas requiring further research to improve its clinical utility.
Keywords: Bacteremia; Bloodstream infection; High-throughput nucleotide sequencing; Metagenomics; Sepsis.
Copyright © 2025 The Authors. Published by Elsevier B.V. All rights reserved.