Non-tuberculous mycobacterium pulmonary disease: challenges and strategies for the preclinical modeling of M. abscessus and M. avium complex

Abstract

Mycobacterium avium complex (MAC) and Mycobacterium abscessus (Mab) each present significant clinical challenges. Both mycobacterial complexes are notorious for their ability to cause chronic and severe pulmonary infections, resistance to standard antibiotics, and intricate host-pathogen interactions that complicate disease management. Although in vitro characterization and preclinical mouse models are important for developing novel therapies, they often fail to replicate the full complexity of human disease. This dissertation presents work evaluating the efficacy of inhaled antibiotics delivered via liquid aerosol in treating MAC pulmonary infections in mice and explores a cystic fibrosis-like mouse strain as a potential preclinical model for Mab pulmonary infection. Building on the Mab studies, whole-body plethysmography (WBP) was established as a robust tool for longitudinally studying the effects of Mab pulmonary infection. This technique enabled monitoring of respiratory parameters and provided a detailed assessment of mouse respiratory function over time. Subsequently, a machine learning (ML) pipeline was developed to classify infection status based on WBP data, which demonstrated the potential of WBP-coupled infection studies to monitor disease progression. By identifying the respiratory parameters most predictive of infection, this work showed the potential for WBP modelling to not only track disease progression, but also better align preclinical mouse models with clinically relevant patient-reported outcomes.