Novel diagnostic tools for managing pediatric chronic cough: Assessing the feasibility of breath collection and analysis for microbial endotyping

Abstract

Rationale: Chronic cough in children often starts as protracted bacterial bronchitis and can progress to bronchiectasis if not treated. Diagnosing protracted bacterial bronchitis typically involves bronchoalveolar lavage (BAL), where multiple bacterial species are usually detected. Misdiagnosis and the invasiveness of BAL pose challenges that we believe can ultimately be solved noninvasively by microbial endotyping using a patient's exhaled breath. Here, we: 1. Assess the feasibility and acceptability of breath collection in children (<1-12 years) 2. Correlate microbial endotypes of chronic cough to BAL as a first step toward microbial endotyping using breath Methods: We recruited two cohorts of children with chronic cough: 1: 150 children for paired BAL and breath collection to identify pathogen-specific biomarkers, with 54 assessed for breath collection feasibility (success: >100mL breath collected). Children were offered a mask, mouthpiece, or straw for collection; 2: 125 children for BAL collection, analyzed for pathogens and volatile molecules, with a 24-month clinical follow-up. Volatilome analysis of BAL and breath samples was performed using two-dimensional gas chromatography time-of-flight mass spectrometry. Statistical tools will be used to correlate the volatile molecules in sample matrices and cohorts to microbial endotypes. Results: The mask proved optimal for children <2 years, while mouthpieces were effective for those >2 years and familiar with blowing. Age-appropriate motivational activities improved cooperation. However, breath collection remained challenging for children <2 (n = 16), with a 50% success rate. For older children (n = 38), we achieved a 95% success rate. We identified an average of 992 molecules in breath samples and 354 in BAL samples, surpassing previous studies and confirming the accuracy of our analysis. Hydrocarbons were the dominant identified class. Analysis of culture, volatilome, and clinical data for cohort 2 is underway. Conclusions: Breath collection from children is feasible, and almost three times more volatile molecules were identified in breath than BAL. Once the results of microbial endotyping using BAL from cohort 2 are known, we will assess whether we can endotype microbes using the more abundant volatile molecules from breath. Identifying the exact cause of protracted bacterial bronchitis will help ensure proper treatment and reduce the risk of progression to bronchiectasis.