Prognostic value of microbial and host determinants for Legionnaires' disease outcome – ProgLegio

Legionnaires’ disease (LD) due to Legionella pneumophila serogroup 1 (Lp1) is an important cause of community- and hospital-acquired pneumonia. Development of man-made water systems created aerosols allowing the direct access of this opportunistic bacterium to human lungs. Thus LD became an emerging disease in the 1970’s. About 1300 LD cases are notified annually in France and despite prevention, its prevalence did not significantly decrease in the last decade. Among LD patients, 98% are hospitalized and 40% require intensive care unit (ICU) admission. The mortality rate remains high (10% to 30% in ICUs and for hospital-acquired LD) despite improved diagnostic and therapeutic management of patients. However, LD is characterized by clinical polymorphism and different severity; disease progression is highly variable from one patient to another, even in immunocompetent patients, rendering the clinical course of LD difficult to predict.

Our hypothesis is that the most severe clinical outcome of LD may be due to specific bacterial or host determinants, or both combined, leading to a higher capacity to infect the lung and/or to an inadequate immune response of the host. Thus, our proposal aims to study, the influence of these bacterial and human factors on the severity and evolution of infection by investigating clinical data combined with pulmonary and blood sample analyses taken during infection, on the same national cohort of 300 patients. The monitoring of such a cohort has never been done at the international level except for rare outbreak investigations; however, this is possible in France thanks to a large number of cases diagnosed per year and the existence of a national monitoring network. The specific objectives of this project are: (1) to investigate the human genetic factors that predispose to severe Legionella infections using a whole exome sequencing (WES) approach; (2) to investigate the emergence of an antibiotic resistant population of Legionella during treatment only detectable by high-throughput sequencing as suggested by not yet published data; (3) to assess the prognostic value of pulmonary bacterial load during infection; (4) to assess, using multiple phenotypic tests, whether LD severity is correlated with the capacities of Legionella to better infect alveolar macrophages and lungs and/or to trigger an inadequate immune response by the exploration of both NF-kB and inflammasome pathways; (5) to identify bacterial genes or SNPs associated with LD severity using a large-scale comparative genomic analyses. These analyses should also allow us to learn what is the genomic diversity of the infecting L. pneumophila population and whether the L. pneumophila strains evolve /adapt during infection to the human; (6) to analyze the diversity of lung microbiota using metagenomic and NGS approaches in order to identify specific microflora that could be associated to severity of infection and the changes in microbiota associated with disease progression. By using this comprehensive approach, our global aims are to provide predictive markers of LD severity, and/or objective criteria of increased virulence of Legionella strains, in order to improve environmental surveillance, clinical prognosis of the patients and potentially to reduce mortality.

As our scientific questions are at the interface between fundamental and clinical research, we created an interdisciplinary consortium including teams of clinical microbiologists expert in diagnostic and epidemiology of LD and in the genotypic and phenotypic characterization of Legionella strains, clinical epidemiologists specialized in data management and statistical analysis on large cohort of patients, physicians specialized in molecular immunology and human genetics, as well as microbiologists expert in molecular analysis of Legionella pathogenesis and in genome-wide approaches. This project draws on the expertise and networks of the National Reference Center of Legionella.