Molecular diagnostic methods have undergone explosive growth in the last decade and play an increasingly important role in clinical microbiology laboratories. These methods can rapidly detect the presence or absence of nucleic acids from organisms in a specimen without waiting for growth from culture. Real-time polymerase chain reaction (PCR), the most commonly used molecular diagnostic in clinical microbiology laboratories, amplifies pathogen-specific nucleic acids, allowing for detection and quantification of a pathogen’s genetic material in a specimen with high sensitivity and specificity. PCR-based tests have been developed further into multiplex assays that allow for simultaneous detection of several agents. However, even multiplex PCRs can only identify predefined targets, so one must have suspect organisms or targets in mind in order to detect them.

Next-generation sequencing (NGS)-based tests present the possibility of an agnostic diagnostic method capable of comprehensive detection of multiple pathogens simultaneously and directly from a patient sample. Despite the incredible promise of NGS-based tests for infectious diseases, the question remains whether we are able to overcome significant hurdles to make this testing relevant on a wider scale.

What Is NGS?

In the past three decades, many NGS platforms have emerged that enable high-throughput massively parallel sequencing of thousands to billions of DNA fragments. This is in contrast to the single DNA sequences generated with first-generation Sanger sequencing methods that are used to identify unknown microbes present in a clinical sample or resolve mutations in known genes. Sanger sequencing can be difficult to interpret when performed on complex or polymicrobial samples, so it is usually only performed on pure microbial isolates or clinical samples that are normally sterile.

A major advantage of NGS compared with PCR is that prior knowledge of the target organism(s), and thus target-specific primers, is not required. NGS can also generate sequences of numerous pathogens in one sequencing run, and reliably identify multiple organisms that may be present in one specimen. In recent years, both the instrumentation and the running costs of NGS have decreased significantly, making it more suitable for clinical usage. The differences in molecular approaches for diagnosis of infectious diseases are summarized in Table 1.

See Table 1 in CLN September PDF

In NGS, genomic material in a clinical specimen or isolate is fragmented, randomly amplified, and used to prepare a library of genomic fragments that are then sequenced. The sequencing method varies by different NGS platforms. Two commonly used sequencing platforms are Illumina and Ion Torrent. Illumina uses sequencing by synthesis during which a fluorescence signal is created when a nucleotide is incorporated, while Ion Torrent sequencers measure the changes in pH generated during incorporation of nucleotides. The signals generated from fluorescence or pH changes for each genomic fragment are independently and simultaneously recorded and translated into nucleotides (A, C, G, or T). Using bioinformatics software, the sequenced fragments are then assembled with or without (de novo assembly) the use of a reference sequence.

Nanopore also has emerged as an attractive sequencing platform. Unlike Illumina and Ion Torrent platforms, Nanopore allows sequencing of a single strand of DNA with a maximum length of up to a few hundred thousand base pairs without active DNA synthesis. The signal of ionic current changes is recorded when the single- stranded DNA passes through a protein nanopore and is translated into nucleotides. This enables direct, real-time analysis of DNA or RNA fragments and a reduction of sequencing time from days to hours. Nanopore’s MinIONs sequencer is particularly attractive because despite its small device size and low equipment cost, it maintains good performance in pathogen detection and surveillance.

Practical Applications of NGS

Major applications of NGS in clinical microbiology laboratories include: whole genome sequencing, metagenomic NGS (mNGS), and targeted NGS (tNGS). WGS is the sequencing and assembly of an entire microbial genome directly from a specimen or clinical isolate. A common application of WGS is the simultaneous identification and typing of microbial pathogens for hospital and public health epidemiological studies. It provides better resolution and more information when compared to sequencing by the Sanger method or traditional pulse field gel electrophoresis (1). In addition, WGS from clinical isolates, particularly for gram-negative bacteria and members of the Mycobacterium tuberculosis complex, is a powerful tool to detect and characterize resistance markers.

Finally, WGS is an important tool for emerging pathogen detection and characterization. In December 2019, lower respiratory specimens from a cluster of patients with pneumonia of unknown cause all linked to a seafood market in Wuhan, China, were collected and underwent WGS. Bioinformatic analysis revealed that sequences from these samples were of an unknown pathogen that matched the genome of lineage B betacoronaviruses, including SARS-CoV. This virus was later named SARS-CoV-2. Without NGS-based WGS technology, it would have taken weeks to months to culture the virus for identification. Thus, NGS-based WGS is a powerful tool for rapid discovery of novel pathogens, which has changed how we can respond to outbreaks like the COVID-19 pandemic.

Metagenomic NGS (mNGS) allows for sequencing all the nucleic acids directly from patient specimens including pathogen and human DNA and RNA without culture. This method provides an unbiased detection of all microbial groups, resistance markers, and virulence factors, as well as host biomarkers associated with different disease states. Prior knowledge of a potential pathogen is not required for this type of diagnosis. Clinical tests have been developed to detect the nucleic acids of microbes from various specimen types such as blood, joint fluid, and cerebrospinal fluid (CSF) to aid the diagnosis of various infections (2,3). A significant limitation of mNGS is that most of the nucleic acids in clinical samples are from the host, so the host genome dominates sequence reads. This can result in decreased analytical sensitivity for detection of pathogens present at relatively low burden.

TNGS uses a process to enrich for microbial sequences of interest before library preparation to improve analytical sensitivity. The most common enrichment method for clinical applications is amplification of a highly conserved region of bacteria or fungi before sequencing. For example, in tNGS for bacteria, the primers are designed to detect and amplify the 16S ribosomal RNA gene, present in all bacteria. Another example of this enrichment method is the use of PCR to first enrich for SARS-CoV-2 RNA in clinical samples before performing NGS to detect mutations of the viral genome. This technique also can be used to detect mutations associated with resistance in viruses such as HIV, hepatitis B, and cytomegalovirus directly from clinical samples with high sensitivity. The enrichment step amplifies the nucleic acids of the target to millions of copies, significantly increasing the number of target-specific sequencing reads when compared with mNGS, where the majority of the sequence reads are from the host genome.

What NGS-Based Tests Are Currently Available?

To date, no NGS-based tests for diagnosis of infectious diseases have received premarket approval or 510(k) clearance from the U.S. Food and Drug Administration (FDA). However, several laboratory-developed, NGS-based tests for pathogen detection directly from patient samples are available under CLIA certificates at select commercial and reference laboratories. A selection of these tests is listed in Table 2. While not an exhaustive accounting of all NGS tests currently available or in development, the most popular NGS-based tests for infectious disease diagnoses are discussed below.

See Table 2 in CLN September PDF

TNGS tests for pathogen detection directly from clinical samples are available from several reference labs. These tests, usually called “universal” or “broad range” PCR, begin with amplification of genes such as the 16S rRNA region for bacteria or the internal transcribed spacer (ITS) region for fungi using universal primers. The amplified gene is sequenced, and results are compared to known sequences in curated databases for organism identification. This testing is available for samples from normally sterile sites (CSF, sterile body fluids, tissues, etc.), and has been useful in patients with high suspicion or evidence of an infectious process on histopathology, but with negative cultures or conventional tests. Universal PCR testing can be particularly useful for detection of fastidious or uncultivable organisms, such as Bartonella or Mycoplasma, in deep-seated infections such as endocarditis, osteomyelitis, and native/prosthetic joint infections (4).

The mNGS Pathogen Dx test from the Department of Clinical Microbiology at University of California, San Francisco was the first described clinical mNGS test for unbiased pathogen detection directly from patient samples. This test detects bacterial, fungal, parasitic, and RNA and DNA viruses from CSF samples, and is available to external clients. In a landmark case, first reported in 2014, mNGS was used to diagnose a case of neuroleptospirosis in a 14-year-old immunocompromised boy with meningoencephalitis after four months of illness and negative conventional test results (5). In clinical validation studies, this test was reported to have clinical sensitivity and specificity of 73% and 99%, respectively, for pathogen diagnosis compared to conventional clinical testing (6).

The Karius Test, which detects microbial cell-free DNA from blood plasma samples, is among the most popular commercially available NGS tests. This test has two major advantages: sample type and turnaround time. First, the Karius Test is performed on plasma, an abundant, easy-to-collect, and noninvasive sample, unlike CSF or surgically collected samples. The use of plasma is based on the premise that during sepsis or serious infections, fragments of nucleic acids from the offending pathogen are shed into the bloodstream. Thus, microbial cell-free DNA in plasma can be a marker, not only for bloodstream infection or sepsis, but also other serious infections such as pneumonia, deep-seated abscesses, endocarditis, etc.

Secondly, the Karius Test has a stated turnaround time of 2 working days from sample receipt. While there may be delays due to shipping, this is generally shorter than most culture-based or reference/send-out laboratory tests. The Karius Test can detect more than 1,000 bacteria, fungi, parasites, and select DNA viruses. Detected microorganisms are reported quantitatively as DNA molecules per microliter of plasma (MPM), and are compared to reference MPM ranges established in healthy, asymptomatic individuals. In an initial clinical validation study of 350 emergency department patients meeting sepsis alert criteria, this test had a sensitivity of 92.9% when compared with a composite reference standard including microbiological testing and clinical adjudication (7).

Clinical Utility of Metagenomic NGS Tests From Patient Samples

There is no prospective controlled clinical trial data evaluating the utility of NGS-based tests for agnostic pathogen detection directly from clinical specimens. Most available publications are case reports or retrospective studies comparing the results of diagnostic NGS tests to standard of care testing. Patient outcome data are often not included, making it difficult to ascertain the clinical impact of this testing. Theoretically, agnostic mNGS may offer a significant advantage over conventional testing in specific patient populations, such as the immunocompromised, in whom obscure or rare pathogens may be disease causing or in specimens from patients previously treated with antimicrobials, where culture-based testing may be falsely negative.

While numerous case reports have described diagnoses made from mNGS that otherwise would have been missed using conventional testing, when evaluated systemically, the clinical utility of mNGS tests remains questionable. Several independent retrospective studies have now reported limited utility of both CSF and plasma mNGS assays for unbiased pathogen detection.

Recently, a large, single-center retrospective study of CSF samples submitted for mNGS reported an overall positivity rate of 15% (12/80). Of the 12 positives, only five were deemed potential pathogen detections. Patient outcomes were only available for three out of five patients with potential pathogens, two of which had changes in management due to mNGS results (8). The results of this study are in contrast to a multicenter prospective clinical study evaluating the clinical impact of CSF mNGS in 204 pediatric and adult patients, where mNGS detected 55% of infections, with 22% being solely detected with mNGS (9). In this multicenter study, 54% of solely mNGS diagnosed infections directly impacted management and treatment decisions of the treating physicians (9). These contrasting results suggest that the clinical impact of mNGS from CSF can be quite variable across institutions and patient populations. In patients with CNS disease and high suspicion of infection despite negative microbiologic tests, mNGS from CSF may be beneficial, but additional studies are needed to identify optimal utilization criteria.

More data evaluating the real-world clinical impact of plasma mNGS is now available. In a multi-center retrospective study of a cohort of 82 patients, including children and immunocompromised patients, the Karius Test was only positive in 61% of cases, and only affected patient management in 11% of cases (10). This was due to a majority of cases positive for microorganisms being deemed non-contributory to the infectious process or confirmation of diagnoses from conventional test results. Similar studies from single-center pediatric institutions also found low overall performance that rarely resulted in changes to antimicrobial management with plasma mNGS when compared to conventional testing (11,12).

While the studies mentioned above included both immune competent and immunocompromised patients, there does appear to be a trend toward greater clinical utility of plasma mNGS in highly immunocompromised patients, particularly those at high risk for invasive fungal infection (IFI). This may be due, in part, to the lack of sensitive diagnostics capable of timely detection of the opportunistic pathogens, to which this patient population is uniquely susceptible (13).

Importantly, all major studies evaluating plasma mNGS report missed detections compared to conventional testing. While some missed detections are inherent to the test limitations (e.g., detection of RNA viruses by DNA sequencing-based tests), others have occurred with “claimed’’ organisms capable of being detected by the test, such as Staphylococcus aureus, Candida spp., Mycobacterium tuberculosis, nontuberculous mycobacteria, and DNA viruses (HSV, adenovirus) among others (10,11). This underscores that mNGS is not sufficient to replace traditional testing methods, and should always be performed in addition to conventional testing.

At more than $2,000 per test billed directly to the patient, the cost associated with mNGS should not be overlooked, especially since this testing is additive to current standard of care diagnostics. As such, indiscriminate use of mNGS coupled with low rates of clinically significant or actionable results can lead to a relatively low return on investment.

Another limitation of mNGS-based diagnostics for infectious diseases is the decreased clinical specificity due to the detection of commensal microorganisms. Studies with both the CSF and plasma mNGS tests have reported clinical false-positives deemed to be unlikely causes of the disease presentation by the clinical care team or reviewing specialists (8,10,11). In rare instances, the results of mNGS tests can lead to unnecessary treatment or additional diagnostic investigations (10).

Finally, molecular detection alone does not yield antimicrobial susceptibility information. This may limit the ability to target antimicrobial therapy and inadvertently extend utilization of broad-spectrum antimicrobials. All of these concerns highlight the importance of performing these tests in consultation with specialists in infectious diseases, clinical microbiology, and pathology to appropriately adjudicate the clinical significance of findings and determine their effect on patient management.

Current Limits to Implementation

Overall, use of this technology as a clinical diagnostic is still in its infancy. While it remains potentially powerful, more studies are needed to determine best use and interpretation of results. To date, mNGS tests are limited to select reference laboratories, as the instrumentation and technical expertise are not yet available to most clinical labs. Furthermore, guidelines for method validation, interpretation, and evaluation/proficiency testing have been proposed but are not yet standardized across the discipline, limiting widespread implementation.

Improvements to NGS technology that further reduce costs and the availability of commercial bioinformatics tools also will timulate more widespread test development. Such development can lead to future applications, including mNGS tests for other sample types. Already mNGS research tests for diagnosis of pneumonia using lower respiratory samples and prosthetic joint infection from joint fluids are in development.

It is clear that NGS has great potential to revolutionize diagnostic testing for infectious diseases. However, in its current form, it cannot replace current standard of care testing. Additionally, evidence does not yet support indiscriminate or screening-based use to rule in/out infection based on NGS test results alone. To date, best practice for use of these types of tests appears to be in patient populations where infection is suspected but conventional testing is negative, and in consultation with treating physicians, ID specialists, and clinical microbiologists to determine appropriate use and interpretation.

Huanyu Wang, PhD, D(ABMM) is the assistant director of clinical microbiology and immunoserology in the department of pathology and laboratory medicine at Nationwide Children's Hospital in Columbus, Ohio. +EMAILl: [email protected]

Sophonie Jean, PhD, D(ABMM) is the assistant director of clinical microbiology and immunoserology laboratories in the department of pathology and laboratory medicine at Nationwide Children's Hospital in Columbus, Ohio. +EMAIL: [email protected]

References

  1. Oakeson KF, Wagner JM, Rohrwasser A, et al. Whole-genome sequencing and bioinformatic analysis of isolates from foodborne illness outbreaks of campylobacter jejuni and salmonella enterica. J Clin Microbiol 2018;56.
  2. Simner PJ, Miller HB, Breitwieser FP, et al. Development and optimization of metagenomic next-generation sequencing methods for cerebrospinal fluid diagnostics. J Clin Microbiol 2018;56.
  3. Ivy MI, Thoendel MJ, Jeraldo PR, et al. Direct detection and identification of prosthetic joint infection pathogens in synovial fluid by metagenomic shotgun sequencing. J Clin Microbiol 2018;56.
  4. Kerkhoff AD, Rutishauser RL, Miller S, et al. Clinical utility of universal broad-range polymerase chain reaction amplicon sequencing for pathogen identification: A retrospective cohort study. Clin Infect Dis 2020;71:1554-7.
  5. Wilson MR, Naccache SN, Samayoa E, et al. Actionable diagnosis of neuroleptospirosis by next-generation sequencing. N Engl J Med 2014;370:2408-17.
  6. Miller S, Naccache SN, Samayoa E, et al. Laboratory validation of a clinical metagenomic sequencing assay for pathogen detection in cerebrospinal fluid. Genome Res 2019;29:831-42.
  7. Blauwkamp TA, Thair S, Rosen MJ, et al. Analytical and clinical validation of a microbial cell-free DNA sequencing test for infectious disease. Nat Microbiol 2019;4:663-74.
  8. Rodino KG, Toledano M, Norgan AP, et al. Retrospective review of clinical utility of shotgun metagenomic sequencing testing of cerebrospinal fluid at a U.S. tertiary care medical center. J Clin Microbiol 2020;58.
  9. Wilson MR, Sample HA, Zorn KC, et al. Clinical metagenomic sequencing for diagnosis of meningitis and encephalitis. N Engl J Med 2019;380:2327-40.
  10. Hogan CA, Yang S, Garner OB, et al. Clinical impact of metagenomic next-generation sequencing of plasma cell-free DNA for the diagnosis of infectious diseases: A multicenter retrospective cohort study. Clin Infect Dis 2021;72:239-45.
  11. Niles DT, Wijetunge DSS, Palazzi DL, et al. Plasma metagenomic next generation sequencing assay for identifying pathogens: A retrospective review of test utilization in a large Children’s hospital. J Clin Microbiol 2020;58.
  12. Lee RA, Dhaheri FA, Pollock NR, et al. Assessment of the clinical utility of plasma metagenomic next-generation sequencing in a pediatric hospital population. J Clin Microbiol 2020;58.
  13. Armstrong AE, Rossoff J, Hollemon D, et al. Cell-free DNA next-generation sequencing successfully detects infectious pathogens in pediatric oncology and hematopoietic stem cell transplant patients at risk for invasive fungal disease. Pediatr Blood Cancer 2019;66.