Clinical Utility of Unbiased Metagenomic Next-Generation Sequencing in Diagnosis of Acute Infectious Diseases: A Prospective Case Series

Abstract

Background. Unbiased metagenomic next-generation sequencing (mNGS) allows the identification of any pathogen based on nucleic acid sequence. Pan-pathogen detection may constitute an ideal diagnostic test for acute infectious diseases (ID), often caused by a wide variety of pathogens. Methods. To evaluate mNGS utility for ID diagnosis, acutely ill patients were enrolled in a prospective, institutional review board-approved research study comparing mNGS with conventional testing. Patients were referred for mNGS testing by ID specialists. Inclusion criteria included (1) high clinical suspicion of infection but no diagnosis to date, (2) extensive negative workup (“last resort” testing), (3) confirmation of infection by an agent suspected based on symptoms and travel history for which testing was not readily available or (4) for which speciation or genomic characterization might be clinically useful. Clinical samples were sequenced by mNGS, and data were analyzed using the SURPI bioinformatics pathogen detection pipeline. Results were communicated to referring providers. Results. We enrolled 28 patients from 4 hospitals with clinical syndromes including neurological symptoms (n = 15), pneumonia (n = 2), influenza-like illness (4) or other (n = 7). (1) In patients with high suspicion of infection, credible infectious etiology was found in 4 of 9 patients by mNGS, including neuroleptospirosis and astrovirus encephalitis diagnoses impacting management. (2) “Last resort” mNGS testing was performed in 10 cases; mNGS as well as conventional testing revealed no infectious etiology in all 10. (3) mNGS confirmed suspected etiology in 4 of 4 patients, including Angiostrongylus cantonensis in a patient with eosinophilic meningitis. (4) In 4 cases, mNGS was used to confirm infection by genomic characterization. Leishmania infantum was speciated in an imported case of visceral leishmaniasis, Mycobacterium tuberculosis in a lung transplant patient, and influenza A(H1N1)pdm2009 in 2 previously vaccinated individuals. Conclusion. mNGS for pan-pathogen ID adds diagnostic value where infectious etiologies are suspected. In all 15 cases in which mNGS testing failed to detect a credible infectious etiology, conventional microbiological testing was also negative, suggesting that mNGS testing may be potentially useful as a “rule-out” assay to exclude infection. Disclosures. All authors: No reported disclosures.