VIENNA – As metagenomic next-generation sequencing (mNGS) becomes cheaper and easier to use, presenters on Saturday at the European Society of Clinical Microbiology and Infectious Diseases global congress in Vienna made their case for implementing the method in clinical settings to diagnose infectious diseases.
In one presentation, Hans Linde Nielsen, a clinical associate professor at Aalborg University Hospital in Denmark, reported data on the use of cell-free mNGS to detect pathogens in hospitalized adult patients with suspected bloodstream infections. The study, which was described last month in Microbiology Spectrum, included 181 patients who qualified for analysis but excluded 141 from the final analysis because they had a negative blood culture and low-to-medium suspicion of a true bloodstream infection. Patients also had to be able to provide informed consent to be included, Nielsen noted.
Nanopore sequencing was conducted with Oxford Nanopore Technologies' PromethIon instrument, while simultaneous blood culture was incubated in Becton Dickinson's Bactec FX Top instrument. Species identification was performed using conventional biochemical tests and mass spectrometry, according to the paper.
Eleven patients in the final study had a confirmed bloodstream infection with a positive blood culture, 25 patients had negative blood cultures but a high suspicion of bloodstream infection, and four had a negative blood culture and no suspicion of infection. Twelve healthy blood donors were also included to establish the background for the metagenomic assay.
After nanopore sequencing was conducted because it is "much faster" than other sequencing options, all 11 patients with positive blood cultures had confirmed bloodstream infections with mNGS, Nielsen said. Twelve patients were positive only with mNGS, and 17 had negative blood culture and mNGS results. Nanopore sequencing also identified all clinically relevant pathogens related to the acute infection.
Overall, the researchers' nanopore sequencing assay had 100 percent sensitivity and 59 percent specificity in their study.
Nielsen cautioned that the study had some limitations, including a low number of cases and selection bias. The researchers also restricted inclusion criteria by requiring informed consent, which excluded severely ill patients. However, Nielsen said that the proof-of-concept study showed that rapid nanopore sequencing may be useful in culture-negative cases where bloodstream infections are suspected.
During the same presentation, two researchers from England provided additional prospective data to make the case for clinical usage of mNGS. Luke Snell, a clinical research fellow at St. Thomas' Hospital in London, said that his hospital has used same-day mNGS to complement routine testing, with preliminary reporting in two hours and total turnaround time of less than seven hours. The nanopore sequencing method detects both DNA and RNA in one test and one tube, and every run includes an internal control, an external positive control, and a mock negative swab to judge contamination, he said. The method, which was described in a paper published in Communications Medicine last year, also includes an automated bioinformatic pipeline with validated thresholds for detection.
In a study of 114 samples mostly from intubated patients in the intensive care unit that underwent nanopore sequencing, 94 percent were able to report out on the same day, although seven samples were failures and did not report out.
Sensitivity was between 89 percent and 98 percent, and 24 percent of the pathogens discovered through mNGS were not identified by routine clinical testing. In about 28 percent of cases, clinicians used the results to change antimicrobial prescriptions, allowing clinicians to adapt their treatment to the patient and escalate or de-escalate treatment as necessary.
"People assume that with metagenomics you'll be detecting more, therefore you'll be treating more," Snell said. "But what we've found actually is that metagenomics gives you confidence in the diagnosis and therefore you can tailor your antibiotics."
In addition, 14 percent of the organisms identified were important for infection control and public health, so embedded mNGS testing may also have benefits when used as a sentinel diagnostic for public health purposes, Snell said.
Since this data readout, Snell said that his hospital has improved its laboratory protocol to gather three times more viral reads and improve the limit of detection, although it has not used the new protocol in routine service yet.
As a result of the successful implementation of mNGS at St. Thomas' Hospital, the UK's National Health Service launched a pilot service at 10 sites across the country to implement the protocol for mNGS for diagnostic use in respiratory cases.
Robbie Hammond, a research fellow at Great Ormond Street Hospital for Children in London and one of the participants in the pilot service, also reported data at the conference. Hammond's team used a unified metagenomic method that relied on untargeted nanopore sequencing with instrumentation from Oxford Nanopore Technologies as well as differential cell lysis. The method had a six-hour turnaround time for the preliminary report, making it a "true same-day result" and includes an automated bioinformatic pipeline with validated thresholds.
The original local validation study included 65 pediatric samples and used a different extraction protocol than Snell's and had a higher volume of upper airway samples, rather than samples from intubated patients. The sensitivity of the method was 50 percent for fungal pathogens, 80 percent for bacterial pathogens, and 81 percent for viral pathogens, and the specificity was between 93 percent and 100 percent.
The lower sensitivity for fungal results was partially the result of polymicrobial samples causing the researchers to miss the lower-abundance organism and resulting in a false negative, Hammond said.
In a follow-up study with 177 samples from 130 patients, sensitivity was between 84 percent and 100 percent with specificity totaling between 99 percent and 100 percent. When looking retrospectively at the clinical impact, Hammond said that about 30 percent of the results had some impact on care with the "vast majority" related to changes in antimicrobial therapy.
Hammond's team also gathered feedback from ICU staff, who largely thought mNGS was a positive addition and liked the speed, sensitivity, and breadth, although they noted some difficulty with interpreting the test results. Hammond said that implementing diagnostic nanopore sequencing would require additional work to lower costs and use fewer staff members but that it was "feasible to implement this protocol" and well received overall.
Also, Ekkehard Siegel, a professor at the University Medical Center of Johannes Gutenberg-University Mainz, presented data on his hospital's implementation of mNGS. Since 2020, his team has analyzed 2,177 infectious disease samples in a clinical routine setting, and in 50 percent of samples at least one pathogen was detected. Within his hospital, clinicians often utilize routine diagnostic testing first and then move to mNGS for difficult cases — there is a "general acceptance of the method," and it is "increasingly used as a second-line diagnostic approach for complicated differential diagnostic scenarios," he noted. However, he added that it is also used as a first-line diagnostic approach in certain hematology cases.
When looking at 1,249 patients with mNGS samples taken within one week of blood culture, Siegel said that 130 samples had matched findings with blood culture. While 235 samples were only positive with blood culture, 144 of those findings were likely contaminants and 70 were typical clinical pathogens. However, he noted that almost all those 70 samples had a second culture collected on the fifth, sixth, or seventh day after the original blood culture and without a second sequencing test, so they were mostly samples from complicated patients who had a secondary infection.
Clinicians "didn't repeat the expensive NGS diagnostic, they just repeated the blood culture," he added.
When the researchers used mNGS and collected blood culture on the same day, there were only 59 findings detected with blood culture alone and most of those were likely contaminants, rather than relevant clinical pathogens.
In contrast, there were 632 samples with findings that were detected by mNGS and not by blood culture. More than 300 of those findings were typical clinical pathogens, and 95 were pathogens not cultivable in blood culture or using standard methods, Siegel said.
Industry movement
Industry has also joined the move toward using nanopore sequencing for infectious disease diagnostics. In 2023, BioMérieux and Oxford Nanopore Technologies announced that they would work together to explore new applications for NGS-based infectious disease diagnostics. The partnership was announced a year after ONT established its subsidiary Oxford Nanopore Diagnostics to develop validated clinical diagnostic tools.
Earlier this month, Oxford Nanopore Technologies and Danaher subsidiary Cepheid said that they would collaborate to develop and commercialize an automated infectious disease sequencing workflow. That workflow is intended for research use only, but the companies are also developing a workflow for positive blood cultures that could progress to regulatory-approved clinical diagnostics.