Clinical and inflammatory characteristics of patients with asthma in the Spanish MEGA project cohort

Abstract Introduction The MEGA (MEchanism underlying the Genesis and evolution of Asthma) project is a multicenter cohort study carried out in eight Spanish hospitals, gathering clinical, physiological, and molecular data from patients with asthma and multimorbidities in order to gain insight into the different physiopathological mechanisms involved in this disorder. Material and Methods We report the baseline clinical and physiological characteristics and biomarker measures of adult participants in the project with the aim of better understanding the natural history and underlying mechanisms of asthma as well as the associated multimorbidities across different levels of severity. We carried out a detailed clinical examination, pulmonary function testing, measurement of fractional exhaled nitric oxide (FeNO), blood counts, induced sputum, skin prick tests, chest computed tomography scan, asthma questionnaires, and multimorbidity assessment in 512 asthmatic patients. Results When compared to patients with milder disease, severe asthmatic patients showed greater presence of symptoms, more exacerbations, lower asthma control, increased airflow obstruction, and higher frequency of chronic rhinosinusitis with nasal polyps, severe rhinitis, anxiety and depression, gastroesophageal reflux, and bronchiectasis. Conclusion The MEGA project succeeded in recruiting a high number of asthma patients, especially those with severe disease, who showed lower control and higher frequency of multimorbidities.


| INTRODUCTION
According to GEMA 5.0, 1 asthma is a chronic inflammatory disease of the airways in which the pathogenesis involves various cell types and inflammatory mediators. In Spain, the prevalence of asthma ranges from 1.5% to 16.7% in the adult population and is approximately 10% among children. 1 The current treatment approach consists of stratifying patients by phenotype (clinical, inflammatory, and molecular) 2 but also by endotype (allergic asthma, nonsteroidal anti-inflammatory drug [NSAID]-exacerbated respiratory disease, eosinophilic asthma, or late-onset asthma), a strategy known as "phenoendotyping." 3 Current medications used to treat asthma reduce inflammation of the airways and relieve bronchospasm, but symptoms reappear with cessation of treatment. Recent studies show that over 50% of patients with asthma are not controlled, 1,4 indicating a need for alternative therapies. Although novel biological treatments directed against type 2 cytokines hold promise, 5,6 selection of patients most likely to respond to these treatments continues to be hindered by an inadequate understanding of the heterogeneous underlying pathophysiological and molecular mechanisms involved. 7,8 The present and future research projects should attempt to identify asthma phenotypes using data collected from multiple cohort studies with sufficiently large sample sizes, integrating data from several domains of the disease repeatedly measured over time. The MEchanism underlying the Genesis and evolution of Asthma (MEGA) project 9 is an ongoing multicenter study in Spain carried out within the framework of the CIBER of Respiratory Diseases. This consortium will perform a number of imaging studies, determinations of lung function, inflammation, and bronchial hyperresponsiveness, and conduct measurements of associated multimorbidities to establish the characteristics that shape this asthma population. In addition, the project will study the stability of different parameters at long term to determine changes in patient condition, exacerbations, control, biomarkers, as well as treatments that can influence the progression of the disease.
The aim of the present study is to describe the baseline characteristics of the adult asthmatic participants that make up the project cohort in terms of their clinical features, frequency of multimorbidities, functional features, and inflammatory biomarkers, so as to better understand the natural history of asthma in patients with different levels of disease severity.

| Design
We conducted a multicenter prospective cohort study of 512 adult patients with asthma. Consecutive (unselected) sampling was used to recruit patients from eight university hospitals in Spain.

| Data collection
Standard data collection methods were used in all participating research centers. A electronic database and case report form (CRF) were designed to collect study data. 9 2 of 10 -RIAL ET AL.

| Patient selection and sampling
Asthma severity has been made according to the classification of the Global Initiative for Asthma (GINA). 10 Asthma diagnosis (based on GINA guidelines) preceded the inclusion of patients by at least 1 year.
A standardized clinical history will be completed for each patient and validated Spanish versions of the following questionnaires will be administered: Asthma Control Test (ACT), 11 Test of Adherence to Inhalers (TAI), 12  Salsola kali, Blatella orientalis, and epithelia (cat and dog). SPTs will be considered positive at wheal diameters of at least 3 mm compared to the negative control (saline); histamine (10 mg/ml) will be used as a positive control.
Atopy is defined as the presence of at least one positive SPT or aeroallergen-specific immunoglobulin E (IgE) in serum.
The DNA, serum, exhaled breath condensate, and sputum supernatants have been stored at −80°C in each of the recruiting centers for further analysis. The detailed protocol has been published elsewhere. 9

| Statistical analyses
Comparisons between groups were performed using the unpaired, two-tailed Student's t-test for Gaussian samples and Mann-Whitney U-test for non-Gaussian samples. Normality was analyzed using the Kolmogorov-Smirnov test. Analysis of variance with Bonferroni post hoc test was performed for comparisons between more than two groups of Gaussian samples, and Kruskal-Wallis with Dunn post hoc test was applied for non-Gaussian distributions. A p value less than 0.05 was considered significant. In addition, Spearman's correlation was used to measure the association between clinical parameters.
Statistical calculations and graphs were performed with Graph-Pad Prism 6 (GraphPad Software Inc.).
In order to determine the contribution of a range of factors (sex, age, BMI, age at onset of asthma, duration of asthma, presence of atopy, and polyposis) to airflow obstruction (FEV 1 /FVC <70% post-BD), univariate and multivariate regression analyses were performed (Table S2). Regression analysis was performed using SPSS statistical software (IBM Corporate).

| Demographic characteristics
The demographic, functional, clinical, and inflammatory characteristics of the MEGA cohort are summarized in Table 1. Not all data are available for all patients, due to the absence of some data in the CRF, including several comorbidities such as sleep apnea, diabetes, hypertension, or bronchopulmonary mycosis.
A total of 512 patients were included in the study (66% women).
Most patients were Caucasian (92%). Obesity, defined as a BMI > 30, was found in 25% of patients (n = 124); of these, 84% (n = 104) had severe persistent asthma and 16% (n = 20) intermittent or mild persistent asthma (p = 0.0269). Currently, 88% of the patients live in an urban setting, but when these same patients were children this percentage was lower (65.6%), and significantly higher in the most severe asthma patients (p = 0.054). Fifty percent of patients had a first-degree relative with a history of asthma. As for smoking, 53.4% of patients were nonsmokers, 31% ex-smokers, 8.2% smokers, and 7% passive smokers. Asthma severity is not related with smoking habit. Occupational asthma was found in 7.7% of patients, and 13.4% had work-related asthma.
Regarding the severity of asthma, 5% of patients had intermittent asthma (n = 26), 17% mild asthma (n = 90), 33.4% moderate asthma (n = 171), and 39.6% severe asthma (n = 203) ( Figure 1). A total of 22 patients (4.3%) could not be correctly classified due to lack of information in the electronic registry. Bronchiectasis was present in 7% of patients (n = 36), with 67% of them (n = 24) experiencing persistent severe asthma. The age of severe patients is significantly higher than those with less severity.
Over the previous year, 15.5% of patients had developed more than three exacerbations, and only 1.4% did not present any exacerbations in this period. Of all patients who experienced exacerbations, 18.6% required hospital admission during the last year, and 9.8% required at least one admission in the intensive care unit ( Figure 2).

| Inflammatory characteristics
Among the studied patients, 72.3% (n = 370) had at least one positive SPT to the aeroallergen tested. Atopy is significantly lower in moderate and severe asthma compared to mild asthma (p = 0.032 and p = 0.011, respectively). One hundred eighteen patients (23%) had RIAL ET AL.  to Alternaria (Figure 3). Peripheral blood eosinophilia was measured in all patients at the start of the study (Table 1 and Figures 1 and 4). Eighty percent of patients (n = 409) had a peripheral eosinophil count of ≥150 cells/ mm 3 , 53.3% (n = 273) had a count that was equal to or higher than 300 cells/mm 3 , 38.8% (n = 197) had a count of ≥400 cells/mm 3 , and 28.3% (n = 145) had levels above 500 cells/mm 3 (Figure 3) -5 of 10 correlation was found between eosinophil levels above 500 and the presence of atopy (p = 0.0375, relative risk = 0.6884, Katz's approximation). There were no significant differences in eosinophil or FeNO levels according to asthma severity (Figure 4), although a tendency toward increased eosinophil counts was seen in severe asthma compared to intermittent disease (1.5-and 3-fold in blood and sputum, respectively).
Correlations have been made with different levels of eosinophilia in peripheral blood (150, 300, 400, and 500 cells/mm 3 ) and in sputum (>2% and >3%). The results show that there are more patients with above 300 cells/mm 3 eosinophilia in sputum, as shown in Figure 5.
The best correlation is established between more than 300 cells/mm 3 in peripheral blood and more than 2% eosinophils in sputum (Spearman r = 0.5235, p = 0.0002).

| Functional parameters
Lung function test results and different asthma severity levels are shown in Figure 6. Statistically significant differences have been

| Questionnaires for disease control, quality of life, and anxiety/depression
ACT scores of <20 were recorded for 34.5% of patients (n = 134).
The Mini-AQLQ questionnaire showed a mean score of 5.48 ± 1.38.
These results are reflected in Figure 7 and Table 1.
Correlations were made with the Fisher test to determine whether patients with ACT levels <20 had more exacerbations in the previous year, and a statistically significant association was found (p < 0.0001).
Anxiety and depression were assessed using the HADS Questionnaire, revealing statistically significant differences (p < 0.001) between patients with severe asthma and other severity levels.
Patients with severe disease had scores compatible with depression F I G U R E 2 Patients with severe exacerbations according to asthma severity and hospital admissions. ER, emergency room admission; HA, hospital admissions; ICU, intensive care unit admission

F I G U R E 4
Blood eosinophil count and FENO levels (expressed as mean/median and standard error). No statistically significant differences (p < 0.05) were found between blood eosinophil counts and FeNO levels and the severity of asthma F I G U R E 5 Correlation between eosinophilia in peripheral blood and sputum. The best correlation is established between more than 300 cells/mm 3 in peripheral blood and more than 2% of eosinophils in sputum (correlation coefficient = 0.5235, p = 0.0002) F I G U R E 6 Spirometric, pletismography, and bronchial hyperreactivity data. There are statistically significant differences between FEV1%, FVC%, FEV1/FVC, and asthma severity levels *p < 0.05, **p < 0.01, ***p < 0.001. No statistically significant differences were found in FVC%, RV%, TLC%, and PC 20 methacholine and asthma severity. RV, residual volume; TLC, total lung capacity F I G U R E 7 ACT and Mini-AQLQ scores. Patients with severe asthma (191) had worse control with an average ACT score of 18 points (p < 0.001). Statistically significant differences *p < 0.05, **p < 0.01, ***p < 0.001 were found between different levels of asthma severity. ACT, Asthma Control Test; AQLQ, Asthma Quality of Life Questionnaire RIAL ET AL. Adherence to therapy was also measured using the TAI test, revealing somewhat poor or poor adherence in 18% of patients.

| Treatments
During the study period, 386 patients were using an inhaled corticosteroid (ICS)/long-acting beta2-agonist combination, 24% (n = 95) were using a metered-dose inhaler, and more than half of them used a spacer. Seventy-six percent (n = 291) of the subjects used dry A number of findings make this cohort of asthmatic patients one of great interest. Significant associations were found between airflow obstruction and various demographic and clinical features, such as age, asthma duration, and the presence of CRSwNP (Table S2). These data contrast with those of other cohorts, in which a strong association was found with female sex, BMI, and presence of atopy. 24 In our cohort, these variables seem to be associated with airflow obstruction, but do not have predictive ability, since the adjusted R 2 is only 3.4%. Random forest modeling was used to complete the regression analysis, though this did not improve predictability. It will be interesting to know whether these associations change over the follow-up period and whether the predictive capacity of these associations can be assessed.
Air trapping is a characteristic of the severely asthmatic population, and the RV rises with increasing severity. 25 Furthermore, air trapping has been more frequently associated with neutrophilic phenotypes and poor response to ICS with the presence of persistent bronchial obstruction. In our population, however, there was no statistical difference across patients in terms of RV or TLC according to severity classification.
A novel aspect of this study is the evaluation of rhinosinusitis by means of the SNOT-22 questionnaire. As expected, we found that higher SNOT scores (more severe rhinosinusitis) were correlated with more severe asthma. It is also important to note that NSAID intolerance was strongly associated with asthma severity in this cohort. When interpreting these data, it is important to note the good results obtained with regard to the inhalation technique (adequate in 97%) and adherence (adequate in 82%) in comparison to other cohorts. 8,[19][20][21][22] However, these results may be due to the fact that the patients were recruited in specialized centers and may not reflect real life in primary care, where treatment is often inadequate. 26 We found elevated peripheral blood eosinophil counts in 53.3% of patients, as indicated by mean values above 300 cells/mm 3 , and 28.3% had levels above 500 cells/mm 3 . An assessment of eosinophilic inflammation based on differential cell count in induced sputum samples revealed that 73% of the patients presented more than 2% of eosinophils, indicating levels above other cohorts described in the literature. 19,22,27,28 We investigated whether an eosinophil count of more than 400 cells/mm 3 is a risk factor for having worse asthma control (ACT < 20), though no statistically significant differences were detected. Furthermore, no differences were found between patients with eosinophil counts above or below 400 cells/mm 3 and early-onset disease debuting in childhood or adolescence. Furthermore, we have found a positive correlation between the levels of eosinophilia in blood and in sputum when the cut-off point of eosinophils in sputum was set at more than 2% and in peripheral blood higher than 300 cells/mm 3 , a better correlation index was obtained than establishing the sputum cut-off point at more than 3% (correlation index 0.5235 vs. 0.4819). The latter may be due to the high proportion of patients with CRSwNP (29%) and atopy included in the sample (71%). Blood eosinophil counts above 500 cells/mm 3 correlated with the presence of atopy. It is important to note that in this cohort, eosinophil levels are not associated with asthma control or age at disease onset.
In asthma, sensitization to fungal, perennial, or seasonal allergens increases the risk of uncontrolled symptoms, exacerbations, and poor disease outcomes. 10 In severe asthma, 20% to 29% of patients typically show sensitization to ≥1 fungal allergen, with Aspergillus being one of the most common. 29 These patients have worse lung function, increased risk of oral corticosteroid use and hospitalization, and a higher degree of airflow obstruction than patients not sensitized to fungal allergens. 30 Nevertheless, our patients presented a similar rate of sensitizations to Aspergillus and Alternaria at different severities.
A similar finding was observed with other allergens.
The study has a number of limitations, such as the lack of a control group and the fact that patients were recruited in specialized centers with a higher number of severe asthma patients than mild cases. In addition, the diagnosis of CRSwNP was based on data obtained from the clinical history.
To conclude, among the patients with asthma included in this cohort, eosinophilic asthma was the predominant phenotype, and many were atopic. An increase in disease severity was associated with a higher number of comorbidities, more exacerbations, worse disease control, a greater tendency to experience anxiety and depression, and worse lung function. The characteristics of the patients included are consistent with those previously reported in other cohorts. This cohort is useful for characterizing different asthma phenotypes and in identifying associated biomarkers as well as the stability of these biomarkers over time.
Continued follow-up of these patients will shed light on the longterm factors that may influence disease severity and will likely provide insight into the treatments that can influence the progression of the disease, identifying the possible causes of the exacerbations and how such exacerbations affect the clinical course of the disease.