Granulocytic Airway Inflammation and Clinical Asthma Outcomes.

Prospective longitudinal studies looking at the relationship between airway inflammation and clinical asthma outcomes in real life are of great importance to better understand the disease and choose appropriate treatment. The Severe Asthma Research Program III (SARP III) is a network of U.S. investigators whose purpose is to advance the understanding of severe asthma through integration of mechanistic studies with detailed phenotypic and endotypic characterization (1).

In this issue of the Journal, Hastie and colleagues (pp. 882–892) report on the stability/instability of airway granulocytic inflammation by analyzing induced sputum cell content over a 3-year period and relate granulocytic inflammation to clinical outcomes such as exacerbation, day-to-day asthma control, and lung function decline in 206 subjects with asthma, the large majority of whom were on inhaled corticosteroids (ICS)/long-acting β2-agonists as maintenance treatment (2). Treatment change was left at the discretion of the investigators, but initiation of any biologics led to exclusion of the patients. In their article, Hastie and colleagues proposed two patient group classifications. One was based on sputum eosinophil or neutrophil count variability, identifying three groups as predominant low, highly variable, and predominant high. The other was based on the combination of low/high sputum eosinophil and neutrophil, leading to four groups called paucigranulocytic, neutrophilic, eosinophilic, or mixed granulocytic, as defined by Simpson and colleagues (3). The current study brings several key results. The first important finding is that the predominant low eosinophil group is the largest, representing 59% of the patients enrolled, whereas the highly variable and predominant high groups accounted for 15% and 25%, respectively. Therefore, most of the patients display a relative stability in their eosinophilic phenotype over the follow-up. The proportion of the predominant low eosinophil group may appear surprisingly high in this study, which mainly focused on severe asthma compared with a previous severe asthma cohort study (4), but it may be overestimated because the authors have excluded patients using biologics. However, recent real-world data have shown that anti–IL-5 and omalizumab are generally administrated in patients with significant increases in sputum eosinophils (5, 6).

Second, the predominant low eosinophil group shows the highest baseline lung function, no disease progression as reflected by change in FEV1, and the lowest exacerbation rate. The highly variable group shows reduced expiratory flow rates with greater airway obstruction, higher reversibility to albuterol, and the highest rate of exacerbation despite a higher use of ICS/long-acting β2-agonists and leukotriene receptor antagonists (Figure 1). The predominant high eosinophil group shows equally altered expiratory flow rate, airway obstruction, and reversibility as the highly variable group but with a lower rate of healthcare use, which is actually comparable to the predominant low eosinophil group. Adopting similar classification with sputum neutrophils does not reveal any relevant disease characteristics apart from older age in those in the predominant high neutrophil group, which is confirmatory of previous findings (7). The results of the current study (2) further support the role of fluctuation of sputum eosinophil count in modifying asthma control as it has been highlighted by Demarche and colleagues (8) in a real-life study in which fluctuation in Asthma Control Questionnaire six-item version was partly and independently determined by fluctuation in sputum eosinophils and FEV1. In Hastie’s study, the exacerbation rate was greater in the group with high sputum eosinophil variation than in the group with persistently high count, indicating that a temporary rise in sputum eosinophils may be a factor for asthma exacerbation; this finding is in agreement with the strategy of maintaining a low level of sputum eosinophils to reduce exacerbation (9). The current study (2) clearly demonstrates, however, that exacerbation may happen in those patients with predominant low sputum eosinophils, and other SARP III publications have shown that high plasma IL-6 concentrations, which are associated with blood neutrophilia and metabolic dysfunction and obesity (10), were predictive of exacerbation (11). It is worth noting that the group with high sputum neutrophil count variation was not at greater risk of exacerbation, as opposed to the group with high eosinophil fluctuation.

Figure 1.

Relationship between granulocytic airway pattern and asthma clinical outcomes. The arrow in the top right panel indicates that neutrophils may enhance eosinophil degranulation.

Third, in addition to asthma control and exacerbation rate, the current study also looked at pre- and post-bronchodilation FEV1 decline over 3 years. Overall, the changes in pre- and post-bronchodilation FEV1 were very limited and the predominantly low and highly variable eosinophil groups even displayed a small increase of 1–2% in pre- and post-bronchodilation FEV1% predicted after 3 years. The airway inflammatory pattern that was associated with the greater loss in lung function was characterized by persistent eosinophil-high and neutrophil high-infiltrate, which is called mixed granulocytic asthma. This phenotype was previously associated with the most altered lung function in another large cross-sectional study (7) (Figure 1). Hastie and colleagues found an average loss in post-bronchodilation FEV1 between 1% and 2% predicted after 3 years. By contrast, neither the isolated predominant high sputum eosinophil group nor the isolated predominant high sputum neutrophil group displayed such a loss in lung function. This would therefore suggest that it is the interaction between eosinophils and neutrophils that enhances the potential of airway remodeling, assuming that the latter is reflected by progressive decline in FEV1. There are in vitro studies that support a role for neutrophils directly or indirectly enhancing eosinophil function, with neutrophil elastase favoring eosinophil degranulation (12) and maturation of IL-33 (13), a potent agonist for eosinophils. Moreover, there is in vivo evidence that airway neutrophils may prevent ICS from delivering beneficial effects in asthma (14). In another longitudinal study, the group with persistent high eosinophilic airway inflammation did not show accelerated lung function decline but had lower sputum neutrophil counts than the fast decliner group with highly variable sputum eosinophils (4), further supporting the idea that the combination of eosinophils and neutrophils is important in lung function decline. On the other hand, the current study by Hastie and colleagues show that healthcare use in mixed granulocytic asthma was, however, not strikingly different from that seen in other groups, illustrating that processes leading to progressive lung function decline may be different from the ones resulting in loss of control and exacerbation. Whether biologics, which are potent at reducing exacerbation rates in subjects with T2 severe asthma, may actually prevent the loss of lung function in those with mixed granulocytic asthma remains to be clarified in long-term longitudinal studies.

Finally, it is worth noting that, whichever classification is chosen, the patients in SARP III show, over 3 years of follow-up, an improvement in asthma control together with consistent reduction in exacerbation, which may even approach 60% in the predominant eosinophil low group (2). It is even more remarkable that, over the same duration, the proportion of patients with high-dose ICS was reduced, thereby pointing to an overdosing of ICS in real life in many patients with asthma. There is no doubt that improved drug adherence linked to study participation may have contributed to this observation, but we might not exclude that a reduction in the dosage of ICS may really be beneficial to some patients, especially in the groups without high sputum eosinophil counts. Recent studies focusing on noneosinophilic or non-T2 asthma have shown maintained or improved asthma control in some patients when stepping down ICS (15, 16). As advocated by Beasley and colleagues (17), it is time to set a new target for asthma: for patients with uncontrolled asthma, we should consider underlying inflammatory phenotypes and associated comorbid traits before initiating any systematic step-up in ICS dose.