Metagenomics in the diagnosis of pneumonia: a systematic review

Pneumonia is a major cause of morbidity and mortality, yet causative organisms are rarely identified. Metagenomic next-generation sequencing (mNGS) may improve this, however the impact on outcomes is uncertain and the relative utility in pneumonia subtypes or different clinical samples is unclear.

This systematic review compared mNGS with conventional microbiology, assessing pathogen identification, turnaround time and clinical impact. It also reviewed difference related to the pneumonia subtype, or sample used. We searched EMBASE, MEDLINE, Scopus, and the Cochrane Central Register of Controlled Trials (CENTRAL) for studies comparing mNGS with conventional microbiology, in patients with pneumonia. All study types were included (excluding case reports of <10 people). Eligibility assessment and data extraction was performed by two independent reviewers, and risk of bias assessed using the QUADAS-2.

We identified 74 studies including 10,566 people, with all studies having high risk of bias or applicability concerns in at least one domain. Conventional tests identified a pathogen in 43.2% of cases (3559/8229) compared to 78.9% for mNGS (6242/7980). mNGS had faster turnaround times than conventional microbiology (MD -73.6h, CI-111.36 - -35.73 hrs) but meta-analysis was not possible for other outcomes. Most studies did not specify the pneumonia subtype. 92%of studies used exclusively or predominantly bronchoalveolar lavage fluid (BALF), and the utility of other sample types could not be established.

mNGS may identify possible pathogens with greater frequency than standard techniques although bias in study design limits definite conclusions, and the clinical relevance of this remains unresolved. There is a need for studies assessing clinical outcomes, and the differential impact in pneumonia subtypes and in samples other than BALF.