Metagenomics for Deciphering Invasive Paediatric Infections 

Speaker: Kathryn Ann Harris, United Kingdom

Introduction:

Dr. Catherine Harris, Consultant Clinical Scientist at Royal London Hospital, presented findings from a study on the use of metagenomic sequencing (mNGS) to diagnose invasive infections in children. The study aimed to validate a novel, pan-pathogen sequencing-based test and assess its impact on patient management.

Study Design and Objectives:

This single-centre study was conducted at Royal London Hospital under Barts Health NHS Trust. The study began in March 2023 and aimed to enrol 300 children with suspected infections.

Main objectives:

• Develop and validate an mNGS-based test for identifying pathogens in pediatric infections.
• Compare mNGS results with routine diagnostics.
• Evaluate potential clinical impact via retrospective case note review.
• Secondary goals:
  1. Facilitate clinical adoption through quality frameworks.
  2. Enable use of the test in future research.

Method:

The mNGS test was developed for blood and CSF samples. A 1 mL EDTA blood or CSF sample was collected from patients after informed consent.

1. Processing steps:
   1. Plasma separated via low-speed centrifugation.
   2. RNA viruses converted to cDNA.
   3. Sequencing library prepared and run on Oxford Nanopore MinION platform.
   4. Real-time sequencing data converted to base-called reads and analyzed with the WF Metagenomics pipeline (Epi2Me).
2. Key features:
   1. Single test to detect DNA/RNA from bacteria, viruses, and fungi.
   2. No need for prior assumptions about the pathogen.
   3. Internal positive control spiked into each sample.
   4. Cost: ~£100 per sample (excluding staff time).
   5. Turnaround: 2.5 days, potentially reducible to 1.5 days.

Why mNGS for Pediatrics?

• Detects all potential pathogens from a small sample volume.
• Ideal for pediatric cases where sample volume is limited.
• Eliminates need for multiple sequential PCRs.
• Nanopore advantages: small footprint, low capital cost, real-time sequencing.

Study Recruitment and Sample Analysis:

By March 2024:

• 273 participants recruited.
• 240 samples sequenced and analyzed.
• 18 samples excluded due to technical failure or insufficient volume.
• 18 still pending analysis.

Participants were categorized into 3 groups based on routine diagnostic results.

Key Findings:

1. Group 1: All Routine Tests Negative (n=140)
   1. 61% (n=85): mNGS also found no pathogens.
   2. 15% (n=21): Non-pathogenic viruses like GB virus C or Torque teno virus detected.
   3. 24% (n=34): Potential pathogens detected by mNGS, not found on routine tests.
   4. This group highlighted the challenge of interpreting results in the absence of positive findings. The high percentage of negative results is a limitation but also supports the test’s potential as a rule-out tool.
2. Group 2: Virus Detected in Routine Tests (n=58)
   1. Mostly respiratory viruses identified in respiratory samples.
   2. mNGS detected the same virus in blood in ~10% of cases.
   3. Surprisingly, several respiratory viruses were also found in blood via mNGS.
3. Group 3: Bacteria/Fungi Detected in Routine Tests (n=48)
   1. Only 9 participants had blood cultures positive on the same day.
   2. 3 matched on mNGS.
   3. 2 showed different but likely relevant pathogens on mNGS.
   4. 4 had no pathogens detected by mNGS.
   5. Remaining 39 had culture-positive results from other samples/days.
   6. Only 2 matched mNGS findings.

These results suggest mNGS can detect organisms even when standard tests do not, though concordance varies depending on timing and sample type.

Case Highlights:

• Case 1: HSV-1 Infection
  1. 2-year-old with gingivostomatitis.
  2. mNGS confirmed HSV-1 in blood.
  3. Could have helped stop antibiotics earlier if available in real-time.
• Case 2: Pneumococcal Meningitis
  1. 5-month-old with confirmed S. pneumoniae on mNGS in both CSF and delayed blood sample.
  2. Diagnosis with mNGS could have avoided additional antigen and PCR testing.

Conclusion:

• mNGS showed promise in identifying pathogens missed by routine tests, particularly in blood and CSF.
• Good correlation with standard tests when run on the same sample type and day.
• Main limitation: large proportion of samples (61%) had no detectable pathogen by any method.
• Potential as a rule-out tool to support early antimicrobial decision-making.
• Interpretation remains critical, requiring integration with clinical and lab data.

Innovations in the Detection of Rare and Intracellular Pathogens 

Speaker: Gilbert Greub, Laussane

Introduction:

Professor Gilbert Greub presented a detailed overview of innovations in the detection and diagnosis of rare intracellular bacteria, with a particular emphasis on tick-borne infections, chlamydia-like organisms, and the utility of homemade high-throughput PCR platforms in identifying pathogens beyond the reach of commercial diagnostic panels.

Clinical Case Illustrations and Diagnostic Challenges:

Two clinical cases highlighted the limitations of standard diagnostic panels and the importance of considering rare pathogens:

1. Atypical Pneumonia in a Child Post-Alpine Exposure: A 6-year-old girl presented with pneumonia unresponsive to amoxiclav. Standard panels for Chlamydia pneumoniae and Mycoplasma pneumoniae were negative. Further investigation using lab-developed PCR detected Chlamydia abortus, likely contracted from environmental exposure to sheep, illustrating the diagnostic value of flexible, custom assays.
2. Persistent Fever in an Adult Post-Tick Bite: A case of prolonged fever, hepatosplenomegaly, and pancytopenia initially mimicked lymphoma. Diagnosis of “Candidatus Neoehrlichia mikurensis” was achieved using a specific PCR, with confirmation via 16S rRNA sequencing. Treatment with doxycycline led to clinical resolution.

Molecular Innovation: High-Throughput PCR Platform:

Prof. Greub's team developed over 100 in-house PCR assays targeting rare bacteria, fungi, and parasites. Key characteristics include:

• Use of Hamilton robotics for DNA extraction and PCR setup in 384-well formats.
• Custom primer/probe design targeting conserved genomic regions.
• Validation through plasmid-based controls and clinical sample testing.
• Pathogens targeted include Chlamydia abortus, Parachlamydia acanthamoebae, Neoehrlichia mikurensis, Rickettsia, Bartonella, and emerging agents like Rhabdochlamydia, Simaknia negevensis etc.

Zoonotic and Environmental Pathogens:

• Through extensive tick surveillance and citizen science initiatives, 62,000 ticks were screened:
• No Coxiella burnetii detected, dispelling tick-transmission concerns.
• Frequent detection of Anaplasma phagocytophilum, Rickettsia helvetica, and Neoehrlichia mikurensis.
• Neoehrlichia found to be the third most common tick-borne pathogen after Borrelia and Rickettsia.

New Pathogen Discovery and Implications:

Prof. Greub recounted the discovery of Neoehrlichia mikurensis from Rattus norvegicus on Mikura Island, Japan. Despite being uncultured (“Candidatus” status), it is phylogenetically positioned between Ehrlichia and Anaplasma. Its co-occurrence with Borrelia in ticks may explain post-Lyme syndromes. The potential pathogenic role of chlamydia-like organisms in preterm neonates and children with cystic fibrosis or asthma-like syndromes was also highlighted.

Diagnostic Approach Recommendations:

1. Panel-free, anthropological diagnostics using flexible PCR platforms.
2. Combination of PCR and serology to optimise detection across different disease phases.
3. Use of 16S rRNA gene sequencing and shotgun metagenomics for pathogen discovery, though the latter was noted as less sensitive.
4. Development of immunofluorescence assays and attempts at cell culture for further pathogen characterisation.

Conclusion:

Greub advocated for systematic search and recognition of underdiagnosed intracellular bacteria, particularly in cases of immunosuppression, zoonotic exposure, or unexplained systemic infections. He urged laboratories to invest in customisable, scalable molecular platforms to address the diagnostic gaps left by commercial panels, particularly for rare or emerging pathogens.

Key Takeaways:

1. Homemade PCR Panels Enhance Pathogen Detection: Custom-developed assays allow detection of intracellular bacteria not included in standard commercial tests, essential for rare pathogen diagnosis.
2. Tick-Borne Pathogens Extend Beyond Lyme Disease: Neoehrlichia mikurensis, Rickettsia, and Anaplasma are increasingly recognised and should be considered in patients post-tick exposure
3. Coinfection in Ticks May Explain Atypical Post-Lyme Syndromes: Frequent co-occurrence of Neoehrlichia with Borrelia may contribute to persistent or atypical symptoms.
4. Diagnostic Strategy Should Include PCR + Serology: Combining PCR (early detection) and serology (later phase) increases diagnostic yield for intracellular infections.
5. NGS Has Limited Sensitivity in Low-Burden Infections: 16S rRNA Sanger sequencing remains more sensitive than shotgun metagenomics in low-copy-number scenarios.
6. Rare Chlamydia-Like Organisms May Contribute to Neonatal Morbidity: Associations observed between Parachlamydia and respiratory illness or ventilation needs in neonates warrant further investigation.
7. Zoonotic Risk Includes Companion Animals and Environment: Children’s proximity to animals like cats, sheep, and guinea pigs increases exposure risk to pathogens like Chlamydia felis and C. abortus.
8. Need for Broader Diagnostic Access and Public Health Surveillance: Laboratories should develop in-house testing capabilities and engage in environmental sampling to detect emerging threats.
9. Citizens Can Contribute to Pathogen Mapping: Public involvement in tick collection enables broader surveillance of vector-borne diseases.
10. Chlamydia pneumoniae May Play a Role in Cystic Fibrosis and Asthma: Detected in CF patients and children with asthma-like symptoms, it may be underappreciated in chronic airway inflammation.
11. Obligate Intracellular Pathogens Remain Underdiagnosed: Due to their inability to grow in standard media, these organisms require specific molecular and cell culture approaches.
12. Pathogen Discovery is an Ongoing Endeavour: Discovery and characterisation of pathogens like Neoehrlichia underscore the importance of exploratory diagnostics in infectious disease practice.

Proteomic Profiling to Differentiate Bacterial Infections in Febrile Infants Under 90 Days in UK Emergency Departments 

Speaker: Clare Mills, United Kingdom

Introduction:

Febrile infants under 90 days of age pose a significant diagnostic challenge due to their immature immune systems and non-specific symptoms. They are at elevated risk for serious bacterial infections, including urinary tract infections, meningitis, and bacteraemia. Current diagnostic approaches rely on clinical appearance, dipstick testing for UTI, and blood biomarkers such as CRP and PCT to identify low-risk infants who might avoid invasive testing and hospitalisation.

Study Objective and Design:

Dr Claire Mills presented findings from a biomarker discovery project aiming to identify novel blood-based protein biomarkers that could outperform CRP and PCT in diagnosing bacterial infections in infants.

• Sample Source: The study used 445 plasma samples from the PHENO study—a UK-wide multicentre observational study that enrolled 1,800 febrile infants under 90 days from 35 emergency departments.
• Proteomic Approach: Biomarker discovery was conducted using the Olink proteomics platform, which is well suited for small plasma volumes and low-abundance proteins like cytokines.

Biomarker Discovery and Validation:

1. Discovery Phase: Initial biomarker discovery was conducted on 20 plasma samples (bacterial vs non-bacterial). A volcano plot demonstrated clear differential expression patterns despite infant immune immaturity.
2. Biomarkers Identified: 15 candidate proteins were selected—some upregulated in bacterial infection (e.g., TRAIL, E-selectin), others downregulated.
3. Validation Tools: ELISA and Luminex assays were used for full cohort testing. A systematic literature review also identified additional candidate proteins.

Diagnostic Accuracy and Signature Development:

• No single novel biomarker showed superior diagnostic accuracy (AUC) over CRP or PCT.
• TRAIL and E-selectin trended towards better performance than CRP but not PCT.
• A five-protein signature derived using logistic regression achieved higher AUC than CRP, but not meaningfully better than PCT.
• Subgroup analysis indicated that the five-protein signature retained better performance in detecting Gram-positive invasive bacterial infections (IBIs), where CRP and PCT showed decreased sensitivity.

Clinical Implications and Limitations:

1. The study confirmed the difficulty of surpassing the diagnostic utility of PCT in febrile infants.
2. The five-protein panel showed potential for improving diagnosis of Gram-positive IBIs, which are less inflammatory.
3. However, the clinical benefit of such marginal improvement remains uncertain due to the small Gram-positive IBI sample size (n=9).
4. Larger cohorts are needed to validate findings and assess the real-world impact on clinical decision-making.

Conclusion:

While no single novel biomarker outperformed CRP or PCT, the development of a five-protein signature offered slightly improved diagnostic performance, particularly for Gram-positive infections. These findings support the potential role of proteomic profiling but emphasise the need for larger studies to establish clinical utility.

ESPID 2025, 26-30 May, Bucharest