Benchmarking transcriptional host response signatures for infection diagnosis

Chawla, Daniel G.; Cappuccio, Antonio; Tamminga, Andrea; Sealfon, Stuart C.; Zaslavsky, Elena; Kleinstein, Steven H.

doi:10.1016/j.cels.2022.11.007

Cited by 6 publications

(5 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Likewise, an 81-set gene signature was used to discriminate bacterial and viral infection in immunocompromised hosts [95]. However, it is clear that while domain-level discrimination using host signatures is demonstrated widely, there is a need for significant standardisation and benchmarking at the species level as cross-reactive unintended infections and aging processes can take over in the analysis [96]. It is now becoming increasingly clear that host response signatures for bloodbased infection diagnostics [97][98][99] will complement similar approaches for direct detection of pathogens [100].…”

Section: Discussionmentioning

confidence: 99%

Using the Traditional Ex Vivo Whole Blood Model to Discriminate Bacteria by Their Inducible Host Responses

Chick,

Rees,

Lewis

et al. 2024

Biomedicines

View full text Add to dashboard Cite

Whole blood models are rapid and versatile for determining immune responses to inflammatory and infectious stimuli, but they have not been used for bacterial discrimination. Staphylococcus aureus, S. epidermidis and Escherichia coli are the most common causes of invasive disease, and rapid testing strategies utilising host responses remain elusive. Currently, immune responses can only discriminate between bacterial ‘domains’ (fungi, bacteria and viruses), and very few studies can use immune responses to discriminate bacteria at the species and strain level. Here, whole blood was used to investigate the relationship between host responses and bacterial strains. Results confirmed unique temporal profiles for the 10 parameters studied: IL-6, MIP-1α, MIP-3α, IL-10, resistin, phagocytosis, S100A8, S100A8/A9, C5a and TF3. Pairwise analysis confirmed that IL-6, resistin, phagocytosis, C5a and S100A8/A9 could be used in a discrimination scheme to identify to the strain level. Linear discriminant analysis (LDA) confirmed that (i) IL-6, MIP-3α and TF3 could predict genera with 95% accuracy; (ii) IL-6, phagocytosis, resistin and TF3 could predict species at 90% accuracy and (iii) phagocytosis, S100A8 and IL-10 predicted strain at 40% accuracy. These data are important because they confirm the proof of concept that host biomarker panels could be used to identify bacterial pathogens.

show abstract

Section: Discussionmentioning

confidence: 99%

Using the Traditional Ex Vivo Whole Blood Model to Discriminate Bacteria by Their Inducible Host Responses

Chick,

Rees,

Lewis

et al. 2024

Biomedicines

View full text Add to dashboard Cite

show abstract

“…Respiratory samples present a special challenge because they harbor microbial communities, including potential pathogens, even in states of health. Host gene expression can help distinguish bona fide infection, and several studies have demonstrated the utility of blood transcriptional profiling for this purpose ( 16 , 17 , 20 ). However, this approach precludes identification of the etiologic respiratory pathogens.…”

Section: Discussionmentioning

confidence: 99%

Integrated host/microbe metagenomics enables accurate lower respiratory tract infection diagnosis in critically ill children

Mick¹,

Tsitsiklis²,

Kamm³

et al. 2023

Journal of Clinical Investigation

View full text Add to dashboard Cite

BACKGROUND Lower respiratory tract infection (LRTI) is a leading cause of death in children worldwide. LRTI diagnosis is challenging because noninfectious respiratory illnesses appear clinically similar and because existing microbiologic tests are often falsely negative or detect incidentally carried microbes, resulting in antimicrobial overuse and adverse outcomes. Lower airway metagenomics has the potential to detect host and microbial signatures of LRTI. Whether it can be applied at scale and in a pediatric population to enable improved diagnosis and treatment remains unclear. METHODS We used tracheal aspirate RNA-Seq to profile host gene expression and respiratory microbiota in 261 children with acute respiratory failure. We developed a gene expression classifier for LRTI by training on patients with an established diagnosis of LRTI ( n = 117) or of noninfectious respiratory failure ( n = 50). We then developed a classifier that integrates the host LRTI probability, abundance of respiratory viruses, and dominance in the lung microbiome of bacteria/fungi considered pathogenic by a rules-based algorithm. RESULTS The host classifier achieved a median AUC of 0.967 by cross-validation, driven by activation markers of T cells, alveolar macrophages, and the interferon response. The integrated classifier achieved a median AUC of 0.986 and increased the confidence of patient classifications. When applied to patients with an uncertain diagnosis ( n = 94), the integrated classifier indicated LRTI in 52% of cases and nominated likely causal pathogens in 98% of those. CONCLUSION Lower airway metagenomics enables accurate LRTI diagnosis and pathogen identification in a heterogeneous cohort of critically ill children through integration of host, pathogen, and microbiome features. FUNDING Support for this study was provided by the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Heart, Lung, and Blood Institute (UG1HD083171, 1R01HL124103, UG1HD049983, UG01HD049934, UG1HD083170, UG1HD050096, UG1HD63108, UG1HD083116, UG1HD083166, UG1HD049981, K23HL138461, and 5R01HL155418) as well as by the Chan Zuckerberg Biohub.

show abstract

“…Respiratory samples present a special challenge since they harbor microbial communities, including potential pathogens, even in states of health. Host gene expression can help distinguish bona fide infection, and several studies have demonstrated the utility of blood transcriptional profiling for this purpose(16, 17, 20). However, this approach precludes identification of the etiologic respiratory pathogens.…”

Section: Discussionmentioning

confidence: 99%

“…However, this approach has not been well studied in the diagnostically challenging critically ill pediatric population. Moreover, while blood gene expression can in some cases distinguish between the response to viral and bacterial infection (16)(17)(18)(19)(20)(21), it cannot pinpoint the specific pathogens active in the respiratory tract, which is critical for optimal antimicrobial therapy.…”

Section: Introductionmentioning

confidence: 99%

Integrated Host/Microbe Metagenomics Enables Accurate Lower Respiratory Tract Infection Diagnosis in Critically Ill Children

Mick

Tsitsiklis

Kamm

et al. 2022

Preprint

View full text Add to dashboard Cite

BACKGROUND. Lower respiratory tract infection (LRTI) is a leading cause of death in children worldwide. LRTI diagnosis is challenging since non-infectious respiratory illnesses appear clinically similar and existing microbiologic tests are often falsely negative or detect incidentally-carried microbes. These challenges result in antimicrobial overuse and adverse outcomes. Lower airway metagenomics has the potential to detect host and microbial signatures of LRTI. Whether it can be applied at scale and in a pediatric population to enable improved diagnosis and precision treatment remains unclear. METHODS. We used tracheal aspirate RNA-sequencing to profile host gene expression and respiratory microbiota in 261 children with acute respiratory failure. We developed a random forest gene expression classifier for LRTI by training on patients with an established diagnosis of LRTI (n=117) or of non-infectious respiratory failure (n=50). We then developed a classifier that integrates the: i) host LRTI probability, ii) abundance of respiratory viruses, and iii) dominance in the lung microbiome of bacteria/fungi considered pathogenic by a rules-based algorithm. RESULTS. The host classifier achieved a median AUC of 0.967, driven by activation markers of T cells, alveolar macrophages and the interferon response. The integrated classifier achieved a median AUC of 0.986 and significantly increased the confidence of patient classifications. When applied to patients with an uncertain diagnosis (n=94), the integrated classifier indicated LRTI in 52% of cases and nominated likely causal pathogens in 98% of those. CONCLUSIONS. Lower airway metagenomics enables accurate LRTI diagnosis and pathogen identification in a heterogeneous cohort of critically ill children through integration of host, pathogen, and microbiome features.

show abstract

Benchmarking transcriptional host response signatures for infection diagnosis

Cited by 6 publications

References 47 publications

Using the Traditional Ex Vivo Whole Blood Model to Discriminate Bacteria by Their Inducible Host Responses

Using the Traditional Ex Vivo Whole Blood Model to Discriminate Bacteria by Their Inducible Host Responses

Integrated host/microbe metagenomics enables accurate lower respiratory tract infection diagnosis in critically ill children

Integrated Host/Microbe Metagenomics Enables Accurate Lower Respiratory Tract Infection Diagnosis in Critically Ill Children

Contact Info

Product

Resources

About