This article has been updated with additional information from the University of California, San Francisco regarding the researchers' future plans.
NEW YORK – Metagenomic sequencing can quickly identify pathogens in body fluid samples of people who are ill, a new study has found.
Being able to rapidly determine the cause of patients' infections can inform clinicians' treatment approaches and guide the selection of antibiotics. However, as researchers led by the University of California, San Francisco's Charles Chiu noted, fast and accurate pathogen identification is not always possible in the clinic, as cultures take time to grow and as PCR testing requires an idea of what microbe might be the source of the infection.
Instead, Chiu and his colleagues developed a metagenomic sequencing-based approach, which can use either Illumina or nanopore sequencing to identify pathogens from cell-free DNA from different types of body fluid samples. When they tested their metagenomic next-generation sequencing (mNGS) method, they found that it had high sensitivity and specificity to detect bacteria using Illumina or Oxford Nanopore sequencing technologies, as the researchers reported on Monday in Nature Medicine. In addition, results could be delivered in as little as six hours.
The study builds on earlier work by the same group, published last year, that focused on diagnosing infections using cerebrospinal fluid.
"Rapid mNGS testing is a promising tool for diagnosis of unknown infections from body fluids," Chiu and his colleagues wrote in their paper.
The researchers collected a range of body fluid samples — such as abscess, pleural, and cerebrospinal fluids — from 158 patients, most of whom were hospitalized. Of these, 127 samples were positive for a pathogen by culture, nine were culture negative but PCR positive, and 34 were negative controls.
They then applied the mNGS testing protocol they developed to those samples. This protocol, they noted, is cross-compatible with both Oxford Nanopore and Illumina sequencing platforms, can analyze all body fluid types, and can be automated in clinical microbiology labs. The reads generated are analyzed via the sequence-based ultra-rapid pathogen identification (SURPI) software to determine which, if any, pathogen is present.
Based on these samples, the researchers determined that their approach could detect bacteria with 79.2 percent sensitivity and 90.6 percent specificity using Illumina sequencing, and with 75 percent sensitivity and 81.4 percent specificity using a nanopore sequencing approach. The performance of the test varied slightly by sample type, with the highest accuracy stemming from CSF samples.
Further, the researchers noted that nanopore sequencing began to return results in as little as 50 minutes and had an average sample-to-answer turnaround time of about six hours. Illumina sequencing, meanwhile, had an average turnaround time of about 24 hours.
Only one of the negative controls was a false positive for a bacterial pathogen by mNGS, and of the false-negative cases, Staphylococcus aureus was the most commonly missed bacterium.
The researchers suggested the lower sensitivity in detecting S. aureus, especially by nanopore sequencing, could be due to higher levels of background human host DNA.
The approach could also detect fungal pathogens, with 91 percent sensitivity and 89 percent specificity using Illumina sequencing and with 91 sensitivity and 100 specificity using nanopore sequencing.
Further, in a case series of a dozen patients whose samples were culture- and PCR-negative, but ultimately determined to have an infection, the researchers found that seven tested positive by mNGS.
"In conclusion, here we demonstrate the utility of mNGS in expanding the scope of conventional diagnostic testing to multiple body fluid types," Chiu and his colleagues wrote. "The achievable [less than six hour] turnaround time using nanopore sequencing may also be essential for infections such as sepsis and pneumonia that demand a rapid response and timely diagnosis."
In a statement, Chiu said that PCR tests that are currently being used to detect difficult-to-culture organisms could potentially all be replaced by this one metagenomic test. He and his colleagues further plan to seek regulatory approval from the US Food and Drug Administration to offer the test as part of patient care.