Bronchiectasis is a heterogeneous disease in its clinical presentation, aetiology and prognosis.1–4 The pathophysiological substrate necessary for its formation is the existence of bronchial inflammation, sometimes complemented by a bronchial infection by pathogenic microorganisms that amplifies the inflammation and causes disease progression,2,5 and, in turn, clinical deterioration and exacerbations.3,4 Although in most cases the predominant inflammation in patients with bronchiectasis is neutrophilic,2,6–9 the eosinophil count may also be elevated in respiratory samples from individuals with bronchiectasis.10–14

Given that counting the relative number of eosinophils from sputum samples or bronchial biopsies can be expensive, time-consuming or invasive, in recent years blood eosinophil counts (BEC) have been used to assess the severity, prognosis or response to treatment of bronchiectasis,10–14 as in the cases of other diseases such as COPD,15–17 although the correlation between the BEC and bronchial eosinophil numbers seems to be relatively modest.16,17 Thus, a low (less than 50–100cells/μL) or high (>300cells/μL) BEC have been related to more severe forms of bronchiectasis, a greater number of exacerbations and even, in patients with peripheral eosinophilia, a better response to inhaled corticosteroid (IC) treatment (even after asthma and other eosinophilic diseases have been ruled out).10–12 Nevertheless, only baseline BEC is taken into account to assess its correlation with outcomes or when making therapeutic decisions. It is known that BEC can change over time due to a number of intrinsic and extrinsic factors16 so the situation should be reconsidered after a given time with a new determination of BEC.

In some patients, however, eosinophilia or eosinopenia may persist over time,17 but it is not known whether this persistence is associated with any specific clinical phenotype of patients or with outcomes of interest in bronchiectasis. Therefore, our objective with the present study was to assess the characteristics of patients with bronchiectasis and persistent eosinophilia or eosinopenia, defined as their presence in at least three measures, with one separation of at least one year, taken consecutively from the baseline value as well as to analyse the association of persistent eosinophilia or eosinopenia with important outcomes in bronchiectasis.

Of the 2642 patients with non-cystic fibrosis bronchiectasis and an age of at least 18 years, 582 individuals had complete baseline data and at least two additional valid yearly measures of BEC. Of these, 69 patients with asthma (6 with ABPA and 6 taking anti-eosinophil biological treatment) were excluded, along with 11 patients taking systemic corticosteroids. Therefore, 502 subjects were finally included in the analysis (Fig. 1).

Fig. 1.

Flow-chart of the study. BEC: blood eosinophils counts; ABPA: allergic bronchopulmonary aspergillosis.

(0.26MB).

Of these 502 included patients, 123 (24.5%) presented baseline eosinophilia (414.2 [117.5]cells/μL) and 83 (16.6%) eosinopenia (49.3 [22.1]cells/μL). Of those patients with baseline eosinophilia, 71 (57.7%) presented persistent eosinophilia, whereas, of those patients with baseline eosinopenia, 47 (56.6%) presented persistent eosinopenia. Table 1 shows the BEC values according to the presence of eosinophilia or eosinopenia and their persistence.

Patients with persistent eosinophilia presented more dyspnoea, comorbidities, bronchial infection by Pseudomonas aeruginosa (PA), a greater number and severity of exacerbations, higher use of ICs and greater global disease severity compared with those without persistent eosinophilia (Table 2). However, there were no differences between patients with persistent and non-persistent eosinopenia (Table 3). Moreover, those patients with persistent eosinophilia also presented greater disease severity, more PA bronchial infections and a greater number/severity of exacerbations than those with persistent eosinopenia (Table 4). However, as can be seen in Fig. 2A–C, although the highest mean value in all three multidimensional severity scores (FACED, E-FACED and BSI) was seen in the persistent eosinophilic group, both the eosinopenic groups (persistent and non-persistent) had greater severity values than the group with non-persistent eosinophilia, with very similar results regarding exacerbation (Fig. 3A) and hospitalisations (Fig. 3B).

Fig. 2.

Comparison between the FACED value (A), E-FACED value (B) and BSI (C) in the four studied groups. Data expressed as media and standard deviation.

(0.3MB).

Fig. 3.

Comparison between the number of exacerbations (A) and hospitalisations (B) in the four studied groups. Data expressed as media and standard deviation.

(0.2MB).

According to our results, 57.7% and 56.6% of the patients with bronchiectasis who presented eosinophilia and eosinopenia at baseline, respectively, persisted in the same group after three measures in a space of two years. Although no significant differences were found in terms of clinical characteristics, exacerbations or severity of patients based on persistence of eosinopenia, those patients with persistent eosinophilia presented greater severity and a higher number of exacerbations than those with non-persistent eosinophilia. Moreover, although patients with non-persistent eosinopenia presented greater severity than those with non-persistent eosinophilia at baseline, those with persistent eosinophilia presented greater severity than those with persistent eosinopenia.

Recent studies have shown that both high (at least 300eosinophils/μL) and low (less than 50–100eosinophils/μL) basal values of BEC were associated with a more pronounced clinical picture, greater severity of bronchiectasis (according to the multidimensional scales used), a higher number of exacerbations and a poor prognosis.10–12 We have used in our study the same cutoff points of eosinophils in order to make the studies comparable. However, these observations were made by assessing only the baseline value of BEG, without taking into account that these values may vary over time as a result of various circumstances.16,17 In fact, a recent study showed how the degree of correlations between the baseline BEG values and subsequent measures did not remain robust after six months, forcing us to reassess, since this variation could imply changes in the classification of the severity of the disease and even have therapeutic implications.27 Therefore, it is not known whether the persistence of elevated or decreased eosinophil counts in bronchiectasis patients may have different associations with significant disease outcomes beyond any elevated or decreased of baseline BEC measurement. This analysis has, however, been carried out in patients with COPD. Casanova et al observed in patients with COPD from two large series (424 patients from the CHAIN series and 308 patients from the BODE series) that the persistence of eosinophilia defined as at least 300cells/μL in at least three separate determinations within a period of two years was not related to the number of exacerbations, although it was related to a higher mortality (15.8% vs 33.7%; p=0.026) than that of the baseline measure.17 In contrast, there was no analysis of patients with eosinopenia (also important, since in COPD it has been observed that a decreased BEC is associated with a lack of response to ICs and more adverse effects28 from this treatment, while in bronchiectasis it has been associated with a greater severity and also lack of response to ICs).10,12

Our group recently reported a U-shaped relationship between the baseline BEC and the severity and number of exacerbations in bronchiectasis. In fact, both eosinophilic (at least 300cells/μL) and, more clearly, eosinopenic (less than 50cells/μL) groups were associated with more severe presentation of bronchiectasis.10

The present study shows some novel results: 1. Persistent eosinophilic bronchiectasis patients had the greatest severity of bronchiectasis compared with non-persistent eosinophilia and even with eosinopenia (both persistent and non-persistent); 2. The severity in the eosinopenic group did not increase over time; 3. In spite of the lack of the change in the severity of the eosinopenic group, this group presented more severity than the non-persistent eosinophilic group (but not more severity than the eosinopenic group).

Therefore, according to our results, it would be important to monitor the BEC of those bronchiectasis patients with baseline eosinophilia in order to detect those moving towards persistent eosinophilia, since this group of patients will present a more severe form of bronchiectasis. However, after two years the severity of bronchiectasis patients with baseline eosinophilia remains constant. In other words, those patients with persistent eosinophilia and baseline eosinopenia are the patient with more severe forms of bronchiectasis and a greater number and severity of exacerbations.

One of the prominent aspects of the present study that could be most controversial is the exclusion of patients with asthma (with a higher probability of persistent eosinophilia) and those with systemic corticosteroids or biological treatments with a higher probability of persistent eosinopenia. Our objective was to assess whether bronchiectasis per se could present a persistent eosinophilic or eosinopenic endotype associated with a particular phenotype that was not influenced by other diseases. Patients being treated with ICs were not excluded as these have not been shown to influence peripheral eosinophil levels.

Among the strengths of this study is that it is the first to be carried out on the value of the persistence of eosinophilia and the persistence of eosinopenia in patients with bronchiectasis and that it has been carried out using the RIBRON Registry, which brings together more than 40 centres throughout Spain.18,19 Among the limitations is the lack of an external validation of the results, since eosinophil values can be influenced by various geographical factors. Furthermore, not all our patients presented repeated eosinophil measurements, so we cannot rule out the existence of some selection bias and a subsequent reduction in statistical power. Finally, we do not have information on some important outcomes, such as mortality, so it is important for future studies to address this issue by trying to consolidate “persistent eosinophilia” as a special endotype related to a characteristic phenotype. Similarly, it is important to assess whether this condition could be a “treatable trait” the role of inhaled corticosteroids,29–31 biologic treatment32 and the impact on chronic bronchial infection and its treatment in bronchiectasis.5

In conclusion, taking into account only the baseline BEC, patients with eosinopenia showed more severe bronchiectasis and a greater number/severity of exacerbations than those with eosinophilia, although patients with persistent eosinophilia showed more severity and a greater number/severity of exacerbations than those with persistent eosinopenia. Therefore, it seems necessary to monitor the values of peripheral eosinophils in steady-state bronchiectasis, especially in those patients with baseline eosinophilia.

Study design: MAMG; Data acquisition: MAMG, CO, AB, RG, MGC, LM, OS, RG, RM, JRL, EB, CP, JLR, GO and DR; Data analysis: GL and MAMG; Data interpretation and writing of the manuscript: All authors. All authors critically reviewed the manuscript and approved its final submitted version.

None declared.

None declared.

The authors state that they have no conflict of interests.