June 3, 2026

mNGS Validation for Sterile Body Fluid Infections: ICCMg Poster Results

Infections in sterile body fluids — cerebrospinal fluid, synovial fluid, vitreous fluid, peritoneal fluid — remain among the most difficult diagnostic challenges in clinical microbiology. Traditional culture and PCR methods have well-documented sensitivity limitations, particularly when patients have received prior antimicrobial treatment or when the causative organism is fastidious. A validation poster presented at the International Conference on Clinical Metagenomics (ICCMg) in November 2024 by the Tel Aviv Sourasky Medical Center (Ichilov) team demonstrates that metagenomic next-generation sequencing (mNGS) can close this diagnostic gap with high accuracy across multiple sterile body fluid types.

Download the poster: Tel Aviv Sourasky Medical Center — ICCMg Poster (PDF)

Key Takeaways

mNGS achieved 92.1% sensitivity and 100% specificity across 51 sterile body fluid samples, validated against culture and PCR reference standards.
The study spanned five sample types — cerebrospinal fluid, synovial fluid, vitreous fluid, peritoneal fluid, and brain abscess — detecting over 20 distinct bacterial species plus Herpes simplex virus type I and Cryptococcus neoformans.
Host depletion using the Devin Microbial Enrichment Kit achieved over 99% human DNA removal with greater than 10- to 1000-fold microbial enrichment, enabling sufficient microbial signal from low-biomass specimens.
The RPM-ratio method (RPM in sample divided by RPM in the no-template control) provided systematic separation of true pathogens from reagent background noise.
Three false-negative results, all Staphylococcus aureus from joint fluid, were attributed to DNA degradation rather than a platform limitation.

Study Design and Sample Workflow

The poster, titled “Validation of a Novel Metagenomics Sequencing Method for the Diagnosis of Infectious Diseases from Sterile Body Fluids,” describes a 51-sample study. Forty samples were confirmed positive by conventional cultures or PCR, while 11 negative controls consisted of CSF collected during therapeutic lumbar puncture for non-infectious indications — a rigorous control design that strengthens the specificity calculation.

The sample preparation workflow followed a defined sequence. Sterile body fluid specimens underwent host depletion using the Devin Microbial Enrichment Kit, which employs ZISC Technology to remove over 99% of human DNA while preserving microbial DNA. This host depletion step is critical: without it, human-origin reads overwhelm the sequencing run and obscure microbial signal. The Devin Host Depletion Filter achieves microbial enrichment of 10- to 1000-fold, making downstream detection feasible even from low-biomass specimens.

Following host depletion, extracted DNA was processed with the Unison Ultralow Library Prep Kit, designed for the extremely low DNA inputs typical of host-depleted sterile fluid samples. Libraries were sequenced on an Illumina NextSeq500 at a depth of 5 to 10 million reads per sample. Taxonomic classification was performed using KrakenUniq, with RPM-ratio calculations applied to distinguish true pathogen signal from background contamination.

Performance Results and RPM-Ratio Analysis

Across the 40 positive samples, the mNGS workflow detected a broad range of pathogens. Bacteria identified included Streptococcus pneumoniae, Neisseria meningitidis, Listeria monocytogenes, Klebsiella pneumoniae, Escherichia coli, Staphylococcus aureus, Haemophilus influenzae, Neisseria gonorrhoeae, Cutibacterium acnes, and multiple other species spanning both common and less frequently encountered organisms such as Streptococcus constellatus, Staphylococcus lugdunensis, and Citrobacter freundii. Viral detection included Herpes simplex virus type I, and yeast detection included Cryptococcus neoformans.

The RPM-ratio approach — dividing the reads per million for a given taxon in the clinical sample by the RPM for that same taxon in the no-template control (NTC) — proved effective at separating true positives from background noise. This metric adjusts pathogen reads relative to the negative control, enhancing the accuracy of pathogen identification and providing a reproducible framework for clinical reporting.

Overall performance reached 92.1% sensitivity and 100% specificity. Notably, the study identified one case where an HSV-2 PCR-positive result was determined to be a false positive, and one Cryptococcus neoformans case had been diagnosed solely by antigen testing — demonstrating that mNGS can both confirm and challenge conventional diagnostic results.

Clinical Significance and Limitations

The three false-negative cases — all Staphylococcus aureus from joint fluid samples — are an important finding. The investigators attributed these misses to DNA degradation, a known challenge in joint fluid specimens that may be subject to intense inflammatory conditions. This represents a pre-analytical limitation rather than a failure of the sequencing or bioinformatic pipeline itself.

The breadth of detection demonstrated in this study underscores the clinical value of mNGS for sterile body fluid infections. A single mNGS assay covered bacteria, viruses, and fungi simultaneously across five specimen types, replacing the need for multiple targeted tests that would otherwise exhaust limited sample volumes. This is particularly relevant for specimens such as vitreous fluid taps and CSF collections, where repeat sampling is often not feasible and the differential diagnosis spans multiple pathogen classes.

The authors concluded that this validation study highlights the potential of metagenomics to enhance detection of a broad spectrum of pathogens from sterile body fluids, with the caveat that pre-analytical sample quality — particularly in joint fluid — remains a factor requiring further investigation.

About the Authors

This poster was presented by Katya Levytskyi, Matan Slutzkin, Avi Gadoth, Zohar Habot-Wilner, Ronen Ben-Ami, Yael Paran, Hanoch Goldshmidt, and Amos Adler from the NGS team at Tel Aviv Sourasky Medical Center and the Faculty of Medicine at Tel Aviv University. Dr. Amos Adler is the director of clinical microbiology at Tel Aviv Sourasky Medical Center, with research focused on antimicrobial resistance mechanisms and clinical application of microbial genomics.