7th joint conference "Joint Microbiology & Infection Conference 2024" of the DGHM and VAAM

Chair

PD. Dr. med.

Metagenomic analysis of cell-free DNA from whole blood specimens of patients with suspected infections: performance and therapeutic impact in clinical routine

Diagnostic and Clinical Microbiology

Background: Sensitivity of blood culture analysis is limited especially in cases of existing antibiotic therapy and some pathogens, like Aspergillus spp., do not grow in blood cultures at all. Metagenomic analysis of cell-free DNA (cfDNA) from whole blood has the potential to compensate for the disadvantages of blood culture diagnostics.Methods: The study was conducted over 3 months at the University Hospital Erlangen. Adult and pediatric patients with suspected infections were included. cfDNA from whole blood was analysed by metagenomic next-generation sequencing with an Illumina NextSeq instrument. Sequences were analysed with the DISQVER® platform (Noscendo GmbH, Germany), comprising a CE-IVD-labelled software algorithm and a curated database. In parallel, a minimum of two blood cultures were analysed.Results: 203 samples of 156 patients (25 pediatric, 131 adult) were analysed. 12 samples (5.9%) were excluded from metagenomic analysis due to technical reasons. Mean time from sampling to result of metagenomic analysis (including logistics) was 3 days. A total of 160 pathogens were detected (1-10 pathogens per sample, mean 2). The positivity rate was 42.4%. Detected pathogens were 103 bacteria, 51 viruses, 4 fungi and 2 parasites. The most common bacteria were Enterobacteriaceae (n=31), Enterococci (n=16) and Staphylococci (n=11). The most common viruses were Epstein-Barr-virus (n=15), Herpesvirus-6B (n=10) and Cytomegalovirus (n=9). Mycobacterium avium, Legionella pneumophila, Tropheryma whipplei, Rhizomucor pusillus, and Leishmania infantum were detected in one sample each. Parallel blood cultures were positive in 11.2%. However, 62% of patients received antibiotic therapy at the time of blood culture sampling. The bacterial pathogens detected by metagenomic analysis were confirmed by classical microbiological diagnostics in 47% of cases. The analysis of the clinical data and the impact of the metagenomic analysis on anti-infective therapy are currently being investigated.Conclusions: Metagenomic analysis of cfDNA with the DISQVER® platform found a pathogen in a significant number of samples in which classical microbiological diagnostics were negative.

Oral Presentation

Prof. Dr.

Full-length 16S rRNA gene nanopore sequencing for bacterial identification in routine diagnostics

Healthcare-associated infections and pathogens: Prevention, surveillance, outbreaks und antibiotic stewardship

Introduction: Culture-negative bacterial infections can result from various factors, including prior antibiotic therapy, slow-growing bacteria, or extended transport times. In such cases, 16S rRNA gene Sanger sequencing can be employed to identify uncultured bacteria. However, Sanger sequencing is limited to monobacterial infections and may require significant time for results. Advances in sequencing technologies, such as nanopore sequencing, offer potential benefits for species identification in polybacterial infections, expanded species identification, and in-house sequencing capabilities.Goals: The primary objectives of this study were to assess the accuracy and efficacy of full-length 16S rRNA gene nanopore sequencing in bacterial species identification, comparing it to the established Sanger sequencing method. Additionally, we aimed to evaluate whether nanopore sequencing could provide enhanced resolution for specific microbial species.Materials &amp; Methods: Forty samples, cultured and sequenced with Sanger in routine diagnostics, were selected for this comparative study. Subsequently, these samples underwent nanopore sequencing, and the results were compared among the three methods.Results: Our findings demonstrated a high level of concordance between Sanger and nanopore sequencing for species identification, including the resolution of mixed sequences in polybacterial infections. Nanopore sequencing successfully identified certain species that would have remained undetected using Sanger sequencing.Summary: In conclusion, this study highlights nanopore sequencing as a potentially valuable tool, especially in cases of culture-negative infections and critically ill patients where urgent bacteriological diagnosis is crucial. Further exploration and validation of nanopore sequencing in larger cohorts should be pursued to comprehensively assess its utility and integration into routine diagnostic workflows.

Professor

Can machine learning algorithms predict negative urine cultures using flow cytometry routine data?

IntroductionUrine samples are among the most frequent samples analysed in microbiology laboratories, while a large proportion are culture-negative. The aim of this study was to test whether culture-negative samples can be predicted from routine flow cytometric data and how reliable prediction algorithms are over time.Materials &amp; MethodsIn 2023, 1325 urine samples from the University Hospital Münster were used for a training dataset (n=1032) and three independent test datasets (n=93-100 samples) that were collected three months apart. Predictors from flow cytometry were total counts per µl of bacteria, erythrocytes, yeast-like-cells, hyaline casts, crystals, leukocytes, squamous epithelial cells, non-hyaline casts and non-squamous epithelial cells in addition to type of urine sample as well as age and sex of the patient. Two different labels were applied (positive culture defined as any bacterial growth [yes/no] and detection of a clinically relevant uropathogenic species [yes, no]). Three classifiers (decision tree, random forest classifier, CatBoost) were cross-validated on the training dataset with 5 folds in order to select an optimized model with at least 95% sensitivity and maximum negative predictive value (NPV). Finally, the optimized model was trained on the complete training dataset and then evaluated on the three independent test sets.ResultsIn total, 72.5% (960/1325) samples were culture positive with a predominance of Escherichia coli (n=295). In all three test sets the classifier predicted a negative culture with a moderate &quot;balanced accuracy&quot; (63-69%), &quot;sensitivity&quot; (94-97%) and &quot;specificity&quot; (31-44%). The test performance of the prediction algorithms was stable over a period of six months. Depending on the tested samples, the negative predictive value was 73-85%.SummaryThe machine learning algorithms in our study population had a moderate NPV to rule out culture negative isolates. Additional clinical parameters (e.g. pH) might improve the test performance.

Prospective evaluation of the SeptiCyte® Rapid assay for the identification of systemic infectious diseases in patients with suspected sepsis

Question: Sepsis is a pathophysiologically complex condition with severe implications for patients and health care systems. In patients with suspected sepsis, the differentation between an infectious and non-infectious aetiology is challenging. Against this background, the SeptiCyte® Rapid assay was developed: This test detects two specific mRNA transcripts (PLA2G7 and PLAC8) that are part of the host&quot;s immune response and that are differentially expressed during the early phase of systemic infection. While this test is commercially available, real-life clinical performance data are scarce.Methods: Since November 2022, we prospectively enrolled patients with the clinical suspicion of sepsis (defined as an increase in sequential organ failure assessment (SOFA) score of ≥2 within 24 hours) and with antiinfective therapy for ≤ 48 hours on one ICU. Blood samples from the patients were collected using PAXgene RNA tubes and subjected to the SeptiCyte® Rapid assay. The test result is provided as a so-called SeptiScore, ranging from 0-15, with values ≥ 7.4 indicating a high suspicion of systemic infection. In parallel, samples from non-septic control patients from the same ICU were also analysed. Clinicians were blinded to the results.Results: Until 30 January 2024, a total 50 patients were recruited, of which 36 belonged to the study group. 56% of the patients were male and the mean age was 65.1 years. In the preliminary dataset, symptomatic patients had an average SOFA score of 9.6. In 21/36 of symptomatic patients, a probable causative pathogen was detected, most frequently Escherichia coli. The average SeptiScore among clinically septic patients was considerably higher than in control patients (8.7 vs. 4.8, see also Figure 1). In patients with suspected sepsis, those with an identified causative agent had slightly higher scores than those without (9.0 vs. 8.2).Conclusions: The SeptiScore Rapid assay was higher in patients with suspected sepsis than in control patients. Patient recruitment and further analysis are ongoing to evaluate whether the SeptiScore accurately differentiates those individuals with an infectious aetiology from those with other causes.

PD. Dr.

Performance evaluation of machine-assisted interpretation of Gram stains from positive blood cultures

Background: Bloodstream infections are an important cause of severe morbidity and mortality. Manual microscopy of Gram stains from positive blood cultures (PBC) is pivotal in diagnosing bloodstream infections but remains labor-intensive, time-consuming, and subjective. The objective of the present study was to evaluate the potential of a scan and analysis system that combines fully automated digital microscopy with deep convolutional neural networks (CNN) to assist Gram stain interpretation from PBCs for routine laboratory use.Methods: The CNN was trained to categorize images of Gram stains based on staining and morphology into seven different classes: negative/false-positive, gram-positive cocci in pairs (GPCP), gram-positive cocci in clusters (GPCCL), gram-positive cocci in chains (GPCC), rod-shaped bacilli (RSB), yeasts, and polymicrobial specimens (FIG1 exemplary image of the graphical user interface). 1555 Gram-stained slides of PBCs were scanned, pre-classified, and reviewed by medical professionals. The assisted Gram stain results were compared to manual microscopy as well as cultural species identification by MALDI-TOF MS as the corresponding reference standard. Furthermore, site-to-site reproducibility between three sites, repeatability, and limit of detection (LOD) were determined.Results: Comparison of assisted Gram stain interpretation to manual microscopy yielded sensitivity and specificity of 95.8% / 98.0% (GPCCL), 87.6% / 99.3% (GPCP/GPCC), 97.4% / 97.8% (RSB), 83.3% / 99.3% (yeasts) and 87.0% / 98.5% (negative/false-positive), respectively. The comparison of assisted Gram stain interpretation to MALDI-TOF MS yielded similar results as the previous comparison. In the analytical performance study, assisted interpretation showed excellent reproducibility and repeatability. With the determined LOD of 105 CFU/ml, any microorganism in PBCs should be detectable.Conclusion: The application has demonstrated its ability to classify microorganisms based on their Gram reaction, arrangement, and morphology. It has shown potential for implementation in clinical laboratories in the future. Furthermore, digital review allows off-site image interpretation.

Combination of molecular imaging with microbiome analysis: infection landscapes in clinical samples

QuestionNext-generation sequencing (NGS) is a powerful tool, which is broadly used for microbiome analysis. Its use for diagnostic applications, however, is hampered by technical challenges, such as lack of sensitivity and risk of contamination. Therefore, data might be difficult to interpret in the clinical setting. Whereas NGS allows the overall assessment of species involved, Fluorescence in situ hybridization (FISH) shows the abundance and localization of specific microbial populations. This way, key microbial species, which are directly associated with the host epithelial surfaces or invasive within the host tissue, can be identified. However, for proper FISH probe selection, prior knowledge on potential microbial candidates is required.MethodsWe developed MG-FISH (Microbiome-guided fluorescence in situ hybridization), which synergistically combines FISH with NGS out of consecutive tissue sections. Consequently, microorganisms (bacteria and fungi) are both identified and localized. We used MG-FISH to identify the bacterial species involved in a case of an aspirated, endoscopically removed peanut from a 78-year-old patient.ResultsWe amplified and sequenced microbial DNA out of methacrylate-embedded tissue sections and identified Staphylococcus, Streptococcus, Veillonella dispar and Actinomyces as most abundant microorganisms. Using specific FISH probes for the respective target species, we visualized, quantified, and localized the microorganisms.ConclusionsMG-FISH gives microbiome data a spatial dimension in situ: It allows quantification and distribution of the microorganisms involved, localization at and in the tissue, and identification of key pathogens. This will further the use of NGS for diagnostic applications.

Oral Session
Molecular Diagnostic Methods and Machine Learning

Appointment

Chair

Topic

Description

Programme

Oral SessionMolecular Diagnostic Methods and Machine Learning

Appointment

Chair

Topic

Description

Programme

Oral Session
Molecular Diagnostic Methods and Machine Learning