Bacteriophage combined with mNGS enhances the specificity of bacterial infection diagnosis

Abstract

Introduction

Metagenomic next-generation sequencing (mNGS) is an important tool for enhancing pathogen detection in infected patients. However, distinguishing between specimens that are infected or colonized is still a major challenge.

Objectives

To explore the composition of bacteriophages in the blood and respiratory tract of the human body, the association between bacteriophage detection and bacterial infections, and whether bacteriophages can assist in differentiating infectious pathogens according to mNGS results.

Methods

Clinical samples from hospitalized patients were collected between January 2023 and February 2024. DNA and cell-free DNA were extracted from BALF and plasma retrospectively to identify the pathogens present, and bacteriophage annotations were conducted.

Results

A total of 299 samples, comprising 136 blood samples and 163 BALF samples, were obtained from 218 patients. Compared with the samples negative for bacteria, both blood and bronchoalveolar lavage fluid (BALF) samples infected with Acinetobacter baumannii, Klebsiella pneumoniae, Pseudomonas aeruginosa, and S. aureus showed a corresponding increase in the proportions of phages related to these pathogens. In BALF samples with Acinetobacter baumannii infection, the proportions of Autographiviridae, Siphoviridae, and Myoviridae were significantly greater than those in the Acinetobacter baumannii colonization group. The sensitivity of Myoviridae for differentiating between infection and colonization was 86.36%, and the specificity was 52.94%.

Conclusion

In sepsis, compared with conventional mNGS methods alone, the use of bacteriophages combined with mNGS was more effective in identifying causative pathogens and had higher specificity. These findings may provide new ideas and tools for improving clinical infection diagnosis.