Metagenomics: a new frontier for routine pathology testing of gastrointestinal pathogens

Abstract

Background: Accurate and comprehensive identification of enteropathogens, causing infectious gastroenteritis, is essential for optimal patient treatment and effective isolation processes in health care systems. Traditional diagnostic techniques are well established and optimised in low-cost formats. However, thorough testing for a wider range of causal agents is time consuming and remains limited to a subset of pathogenic organisms. Metagenomic next-generation sequencing (mNGS) allows the identification of all pathogens in a sample in a single test, without a reliance on culture or introduction of target selection bias. This study aims to determine the ability to routinely apply mNGS testing, in comparison to traditional culture or polymerase chain reaction (PCR) based tests, for the identification of causal pathogens for gastrointestinal infections.

Results: The performance of mNGS, PCR and microscopy, culture and sensitivity (MCS) assays was established using 2,619 prospectively collected faecal samples from patients with symptomology indicative of infectious gastroenteritiss. Commonly experienced pathogens including Aeromonas spp, Campylobacter spp, Salmonella spp and Giardia spp, in single and co-infected patients, were used to establish test outcomes. When testing for these organisms, using the combined result from either or both PCR and MCS testing as the comparator, the mNGS assay had clinically acceptable sensitivity (89.2-100%). Further, the mNGS assay detected 14 additional enteropathogens, that were either not detected or not tested, by initial PCR/MCS testing.

Conclusions: The advantage of mNGS compared to other syndromic testing systems is the broad range of detectable targets and the ability to interrogate samples without clinician informed or assay specific bias. With the development of newer sequencing assays, it is now feasible to test for a wide range of target organisms in a sample using a single mNGS test. Overall, the mNGS based approach enabled pathogen detection that was comparable to conventional diagnostics and was shown to have the potential to be extended for the detection of many pathogens and genes of clinical interest. In conclusion, the mNGS assay offers an easy, sample to answer workflow with rapid detection of enteropathogens and has the potential to improve diagnosis, therapy and infection control precautions.

Keywords: Enteropathogens; Faecal; Gastrointestinal; Infection; Metagenomics; Pathogen.

Fig. 1

Outline of Retrospective Study to Assess mNGS Assay Performance

Fig. 2

Outline of Workflow for Clinical Metagenomics

Fig. 3

Outline of required controls for sample processing to proceed to report generation

Fig. 4

Established Limit of Detection for Representative Species used to Demonstrate Clinical Performance of Metagenomic Assays. The limit of detection was established using known isolates as a spike in to pre-established screened negative faecal samples at known numbers of pathogenic organisms per gram of faeces. A Each organism was graphed to map the number of target organisms to the resulting diagnostic reads recorded. An example of one species within each of the genus evaluated is included in section A. B Tabulated summary of the analysis of the spike in samples over the range of 0 to 5 x 10⁹ target organisms per gram of screened negative faeces, tested in 3 replicates with both mNGS and PCR and reported as Detected (+) or Not Detected (-)

Fig. 5

Test Outcomes for MHC Clinical Sample Set Tested by PCR, MCS, mNGS and Discrepant PCR assay. A clinical sample set (MHC) was established from samples submitted for routine pathology testing, being symptomatically indicative of the presence of a gastrointestinal pathogen. The outcome from testing of these samples was determined for the initial pathology service (diagnostic PCR testing and MCS testing), and for the subsequent mNGS testing, and discrepant PCR testing, as required. The results were collated at a genus level for Aeromonas, Camplyobacter, Salmonella and Giardia species, given that PCR testing lacked resolution to the species level

Fig. 6

Clinical Performance Metrics for mNGS when Applied as a Routine Pathology Test. Standard performance metrics for the mNGS assay were established and are outlined for diagnostic sensitivity, specificity, accuracy, limit of detection, negative predictive value and positive predictive value

Fig. 7

Correlation of Diagnostic Read Identification with mNGS testing to Ct Value with PCR Testing of Clinical Samples. Comparison was made from samples within the MHC clinical sample set of the number of diagnostic reads (NGS read count) and PCR Ct value, displayed as the value below the upper Ct value cut off established by the manufacturer (CT units from max)

Fig. 8

Detection of Adenovirus with mNGS and PCR Testing. A Subset of the MHC Clinical Sample Set was used to apply analysis for less prevalent organisms including Adenovirus. Comparison was made between the initial laboratory PCR test, the mNGS test, and the confirmatory PCR result to come a consensus outcome