Abstract :
[en] Asthma is a heterogeneous respiratory disease. The emergence of sputum analysis has allowed to categorize asthmatic patients into eosinophilic and non-eosinophilic phenotypes, suggested to reflect type 2-high and type 2-low inflammation, respectively. These phenotypes have a clinical interest because they were reported to differently respond to inhaled corticosteroids (ICS). The presence of eosinophilic airway inflammation seems indeed necessary to have a good clinical response to ICS, while non-eosinophilic asthmatics are thought to be poorly sensitive to ICS. The major aim of this thesis was to further investigate the relationship between eosinophilic phenotypes and responsiveness to ICS in real-life patients from the University Asthma Clinic of Liege.In the first part of this work, we analyzed in detail sputum and systemic inflammation in asthmatic patients classified according to their inflammatory phenotype, and we investigated how they compared with healthy subjects. Our main finding was that a raised sputum eosinophilia was present across all phenotypes as compared with healthy subjects, even in patients defined as having non-eosinophilic asthma who were characterized by a low-grade eosinophilic inflammation.The second and main part of this work focused on corticosteroid responsiveness according to the eosinophilic phenotypes. In eosinophilic asthmatics, we confirmed in a retrospective analysis that initiating or increasing the dose of ICS resulted in an improvement in asthma control, quality of life, lung function, bronchial hyperresponsiveness and severe exacerbation rate, concomitant with decreased levels of sputum eosinophils. In keeping with these data, we showed, in a retrospective longitudinal study, that asthma control was associated with individual fluctuations in sputum eosinophils over time, with an increase in sputum eosinophils being linked to a worsening of asthma control, and a decrease in sputum eosinophils being linked to an improvement in asthma control. These results also suggested that ICS brought asthma control essentially by modulating eosinophilic inflammation. Finally, in a prospective randomized controlled trial in patients who were prescribed ICS, we showed that a pharmacist-delivered educational intervention (including information about asthma and education relating to inhalation technique, adherence and non-pharmacological treatments) was more effective in decreasing the fractional exhaled nitric oxide (FENO) levels, a marker of type 2-high inflammation, than usual care. In non-eosinophilic asthmatics, we showed in a retrospective analysis that an initiation or increased dose of ICS did not result in any overall clinical benefit. These results were confirmed in a prospective study assessing the impact of an increase in the ICS dose in uncontrolled non-eosinophilic asthmatics. However, at the individual level, it is noteworthy that those patients with higher baseline FENO levels had a clinically significant improvement in asthma control with this raised ICS dose. On the other hand, our retrospective results suggested that withdrawing or decreasing the dose of ICS in non-eosinophilic asthmatics might result in an overall improvement in asthma control, quality of life and severe exacerbation rate. In a prospective study of ICS step-down in non-eosinophilic asthmatics, we showed that it was possible to completely withdraw ICS in about one third of the patients and to step-down ICS to a reduced dose in one further third, without any deterioration of asthma control and severe exacerbation rate. In those patients who were successfully weaned off ICS, it is worth noting that an improvement in asthma control and a reduction in severe exacerbation rate was observed, thereby confirming our retrospective data. Interestingly, patients in whom withdrawing ICS was unsuccessful had higher blood eosinophil levels at baseline and during the follow-up than patients who were successfully weaned off ICS.In the third part of this work, we assessed the ability of biomarkers to rule in eosinophilic or non-eosinophilic asthma. Sputum analysis is indeed not widely available and several user-friendly surrogate markers for sputum eosinophils have been proposed, of which the 3 most studied are FENO, blood eosinophils and total serum IgE. In a retrospective study, we showed that those biomarkers taken alone were not sufficiently good predictors of the airway eosinophilic status. However, combining these biomarkers could accurately predict the eosinophilic or non-eosinophilic phenotype in about 60% of patients, leaving the remaining 40% with uncertain eosinophilic status, which should be investigated by a sputum analysis.ConclusionTo conclude, our results emphasize the importance to measure the level of airway eosinophilic inflammation in clinical practice. In eosinophilic asthmatics, our data confirm in real life the effectiveness and clinical utility of ICS, which should be adjusted to reduce eosinophilic inflammation as much as possible. In non-eosinophilic asthmatics, our data cast doubt on the clinical utility of stepping-up ICS and rather support a try for stepping-down ICS, which seems feasible in two thirds of those patients irrespective of their baseline asthma control.
