Metagenomic sequencing improves detection of rare pathogens in CNS infections

A groundbreaking seven-year study reveals how metagenomic sequencing offers a more accurate, comprehensive approach to diagnosing difficult CNS infections—spotting pathogens that traditional tests often overlook.

Study: Seven-year performance of a clinical metagenomic next-generation sequencing test for diagnosis of central nervous system infections. Image Credit: Kateryna Kon / Shutterstock

In a recent study published in the journal Nature Medicine, researchers analyzed the seven-year performance of metagenomic next-generation sequencing (mNGS) testing to diagnose central nervous system (CNS) infections. Encephalitis, meningitis, and myelitis associated with CNS infections can cause severe, life-threatening illnesses. Delays in their diagnosis and treatment have been linked to higher morbidity and mortality. Clinical mNGS has emerged as a comprehensive approach for diagnosing infectious diseases, allowing for the detection of diverse microbes without targeting a specific pathogen in advance. This hypothesis-free and agnostic method could be useful in cases of CNS infections for which invasive methods like brain biopsy and limited cerebrospinal fluid (CSF) samples are challenging to obtain.

The University of California, San Francisco (UCSF) clinical mNGS test was developed as a validated DNA/RNA metagenomic sequencing assay in 2016. Previous studies have shown that mNGS can increase the diagnostic yield, providing actionable information in cases of suspected CNS infections.

a,b, Distribution of tests ordered by state (a) and internationally (b). A total of 4,075 mNGS tests were performed from CSF samples collected from the United States, California being the most frequent state of origin (n = 2,420 samples). Reference laboratories such as Associated Regional and University Pathologists, Inc., Labcorp and Mayo Clinic (n = 722) receive tests from multiple states, so the location of individual samples cannot be tracked and thus are excluded from the figure. 14.8% (n = 715) of samples were sent from pediatric hospitals. c, Number of tests performed by year and number of positive results, excluding results that were reported possible or likely contaminants. *Data shown are samples analyzed up to April 2023.

Study Overview and Key Findings

In the present study, researchers assessed the clinical applicability of mNGS testing over seven years. In total, the clinical microbiology laboratory at UCSF performed 4,828 mNGS tests between June 2016 and April 2023. Over 84% of tests were performed for patients from 46 states in the United States. Around 56% of patients were males, and 24.2% were children under 18.

The median turnaround times for non-UCSF and UCSF patients were 11.4 and 8.2 days from sample collection to result and 3.8 and 3.6 days from sample processing to result, respectively. Approximately 10.6% of samples contained at least one environmental or commensal organism and were reported as contaminants. Contaminants were more frequent in non-UCSF samples, possibly due to variations in sample handling, acquisition, and transport methods.

After excluding samples with contaminants, 14.4% of samples were positive for a pathogen. The average annual positivity rate was estimated at 16%, with UCSF samples showing a higher rate (16.2%) than non-UCSF samples (13.9%), which may be attributed to the more frequent reporting of subthreshold results in UCSF samples. Pathogens detected at subthreshold levels included West Nile virus, Powassan virus, Balamuthia mandrillaris, Mycobacterium tuberculosis, Coccidioides sp., and Histoplasma capsulatum. Notably, most subthreshold detections were later confirmed by orthogonal testing. In total, 797 organisms were detected, with DNA viruses being the most commonly identified, followed by RNA viruses, bacteria, fungi, and parasites.

The most frequently detected RNA viruses were human immunodeficiency virus, arboviruses, and enteroviruses. Uncommon arboviruses, such as La Crosse virus, Potosi virus, St. Louis encephalitis virus, and Cache Valley virus, were also detected. Detected parasites included Naegleria fowleri, Angiostrongylus cantonensis, B. mandrillaris, and Toxoplasma gondii. Fungal pathogens included Coccidioides sp., Cryptococcus sp., Fusarium sp., and Histoplasma capsulatum.

Analysis of Clinical Data

Next, the researchers analyzed laboratory and clinical metadata from a subset of UCSF patient samples. This cohort included 1,164 samples from 1,053 individuals; 55% were males, 15.2% were children, and 35.8% were immunocompromised. Further, 87.7% of the cohort were hospitalized for a median of 12 days; 38.7% required intensive care, and 10.2% died within two months. This cohort included 18% of cases with CNS infection, 37.1% with autoimmune diseases or other non-infectious conditions, 0.1% with prion disease, and 44.8% with unknown etiology.

Immunocompromised patients had a higher mNGS positivity rate (16.7%); it was also higher in patients with meningoencephalitis and meningitis. Overall, 180 cases were positive for at least one microorganism detected by mNGS testing; among these, 227 organisms were detected. Of these detections, 135 were true positives, 85 were incidental detections, and four were false positives. Thirty-five detections were classified as subthreshold.

Notably, mNGS testing showed lower sensitivity in detecting fungal pathogens, as 10 fungal infections missed by mNGS testing were identified by antigen testing, culture, or both. The performance metrics for mNGS testing in diagnosing CNS infections were 63.1% sensitivity, 92.9% accuracy, 99.6% specificity, 92.3% negative predictive value, and 97.1% positive predictive value.

Moreover, the diagnostic yield of mNGS was higher (63.1%) than other methods, including CSF direct detection (CSF-DD), non-CSF-DD, and indirect serologic testing. Notably, mNGS testing identified 60 infections missed by CSF-DD methods, while CSF-DD methods detected 26 infections that mNGS missed.

Conclusions

In sum, mNGS testing demonstrated high sensitivity for detecting pathogens in a cohort with severe and diagnostically challenging CNS infections. It was the primary or sole diagnostic tool for 30.4% of infections, capable of identifying diverse pathogens, including emerging and unexpected microbes, which are often hard to detect using conventional methods. While the sensitivity of 63.1% indicates that mNGS should not replace microbiological testing, this study suggests several use cases for mNGS testing: 1) detection of unculturable organisms, 2) identification of rare or unexpected infections, 3) broad viral diagnostics, and 4) outbreak investigations.

Further studies are needed to assess the cost-effectiveness and overall clinical impact of mNGS testing.