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

Abstract

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.

Viewing both sides of the coin for infectious disease diagnosis

“…To address these challenges, Mick and colleagues, in a recent issue of the JCI , developed a multipronged approach that distinguished pediatric patients with LRTIs from those with alternative diagnoses (including noninfectious illness and nonpulmonary infections), while simultaneously identifying the microbiological cause of the LRTI ( 4 ). Other studies combining the host immune response with microbial metagenomic next-generation sequencing (mNGS) have focused on RSV ( 5 ), adult LRTI ( 6 ), sepsis ( 7 ), and tuberculous meningitis ( 8 ).…”

“…Other studies combining the host immune response with microbial metagenomic next-generation sequencing (mNGS) have focused on RSV ( 5 ), adult LRTI ( 6 ), sepsis ( 7 ), and tuberculous meningitis ( 8 ). In this study, the authors leveraged a previously enrolled cohort of critically ill children aged 31 days to 18 years with acute respiratory failure requiring mechanical ventilation and from whom tracheal aspirate samples were available ( 4 ). Of 261 children with confirmed infection, 95% were intubated within two days of admission, timing that is consistent with community-acquired pneumonia.…”

“…Comparing this pediatric host response signature to one developed in an adult cohort ( 5 ), there was no overlap in differentially expressed genes. Nevertheless, it is possible the pediatric signature would perform well in the adult cohort or vice versa, because many combinations of differentially expressed genes can substitute for one another in a given classification task ( 9 ), though this possibility was not evaluated in Mick et al ( 4 ). Ideally, a universal, age-agnostic signature could be identified for future development.…”

“…Microbial mNGS provides a broad screen for potential pathogens that is not available by standard PCR panels or culture. To distinguish potential bacterial and fungal pathogens from commensals, Mick and colleagues applied a rules-based model that ranked high-abundance species relative to the abundance of background species ( 4 , 6 ). 70% of the samples showed concordance with culture.…”

“…Host response also provides context when microorganisms of unclear clinical importance are identified, such as through metagenomic approaches. The study presented by Mick, Tsitsiklis, and colleagues ( 4 ) goes beyond these scenarios by revealing how analysis of the host response and metagenomics can be combined synergistically. By simultaneously viewing both sides of the coin, host and pathogen, they generated a harmonious characterization of LRTI status with exceptional accuracy.…”