Advertisement

Clinical features and three-year prognosis of AECOPD patients with different levels of blood eosinophils

  • Wang Ruiying
    Correspondence
    Corresponding author.
    Affiliations
    Department of Pulmonary and Critical Care Medicine, Shanxi Bethune Hospital Shanxi Academy of Medical Sciences, Tongji Shanxi Hospital, Third Hospital of Shanxi Medical University, Taiyuan 030032, China
    Search for articles by this author
  • Zhaoyun
    Affiliations
    Department of Pulmonary and Critical Care Medicine, Shanxi Bethune Hospital Shanxi Academy of Medical Sciences, Tongji Shanxi Hospital, Third Hospital of Shanxi Medical University, Taiyuan 030032, China
    Search for articles by this author
  • Xu Jianying
    Affiliations
    Department of Pulmonary and Critical Care Medicine, Shanxi Bethune Hospital Shanxi Academy of Medical Sciences, Tongji Shanxi Hospital, Third Hospital of Shanxi Medical University, Taiyuan 030032, China
    Search for articles by this author
Open AccessPublished:June 07, 2022DOI:https://doi.org/10.1016/j.hrtlng.2022.05.012

      Abstract

      Background

      Eosinophils are thought to be associated with the frequency and severity of acute exacerbations of chronic obstructive pulmonary disease (AECOPD); however, the role of eosinophilic inflammation in AECOPD is still incompletely understood.

      Objectives

      To investigate the relationship between different levels of blood eosinophils and clinical features, including comorbidities, therapy, and prognosis, and to further explore the optimal eosinophilic cutoff.

      Methods

      We retrospectively collected and analyzed medical data, laboratory findings, chest CT images, treatment, and three-year follow-up data from 984 AECOPD patients with different blood eosinophil (EOS) levels: EOS%<2%, ≥2%; EOS%<3%, ≥3%; eosinophil counts<100 cells/L, ≥100 cells/L.

      Results

      The prevalence of eosinophilia was 36.48% of EOS≥2% (359 cases), 22.87% of EOS≥3% (225 cases), and 48.48% with eosinophil counts≥100 cells/µl (477 cases). EOS was associated with comorbidities, including pulmonary heart disease, arrhythmia (atrial fibrillation), laboratory testing and clinical treatment, including respiratory failure, airway limitation, infectious inflammation, rate of antibiotic use, systemic glucocorticoids, and three mortality rates. The ROC curve showed that the indicators of AUC≥0.5 included chest CT imaging (emphysema 1.8% or ≥100/µl, bronchitis, 1.7% or ≥100/µl), osteoporosis (2.4% or ≥140/µl), mental illness 6.1% (or ≥400/µl), dust exposure (2.2% or ≥240/µl) and ex-smoker (1.3% or ≥100/µl).

      Conclusions

      The relatively higher EOS group (≥2% or ≥100/µl) was associated with fewer complications, mild airflow limitation, a tendency of noninfectious inflammation, and lower 3-year mortality. Eosinophils can not only guide clinical treatment but also be an indicator of predicting clinical outcome and prognosis in AECOPD patients.

      Keywords

      Introduction

      Chronic obstructive pulmonary disease (COPD) is currently one of the top three causes of death worldwide.
      • Halpin D.M.G.
      • Celli B.R.
      • Criner G.J.
      • et al.
      The GOLD Summit on chronic obstructive pulmonary disease in low- and middle-income countries.
      Many people die from its acute exacerbation and its comorbidities.

      Global Initiative For Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease 2021 report. http://www.goldcopd.org

      There is significant clinical heterogeneity in patients with acute exacerbation of COPD (AECOPD),
      • Zhou A.
      • Zhou Z.
      • Zhao Y.
      • et al.
      The recent advances of phenotypes in acute exacerbations of COPD.
      which has attracted the attention of researchers, and some biomarkers are being explored. Blood eosinophilia (EOS) is considered to be a marker of the therapeutic effect of COPD.
      • Prins H.J.
      • Duijkers R.
      • Lutter R.
      Blood eosinophilia as a marker of early and late treatment failure in severe acute exacerbations of COPD.
      However, the whole role of EOS in AECOPD is unclear. Numerous studies have shown the relationship between eosinophilia and clinical outcomes of patients with AECOPD.
      • Zysman M.
      • Deslee G.
      • Caillaud D.
      • et al.
      Relationship between blood eosinophils, clinical characteristics, and mortality in patients with COPD.
      However, the evidence lacks consensus, and the research thresholds are controversial.
      Several studies have shown that high EOS counts are associated with more frequent exacerbations
      • Zysman M.
      • Deslee G.
      • Caillaud D.
      • et al.
      Relationship between blood eosinophils, clinical characteristics, and mortality in patients with COPD.
      ,
      • Vedel-Krogh S.
      • Nielsen S.F.
      • Lange P.
      • et al.
      Blood eosinophils and exacerbations in chronic obstructive pulmonary disease. The copenhagen general population study.
      and predict the response of patients with AECOPD to glucocorticoid therapy.
      • Zysman M.
      • Deslee G.
      • Caillaud D.
      • et al.
      Relationship between blood eosinophils, clinical characteristics, and mortality in patients with COPD.
      ,
      • Xue J.
      • Cui Y.N.
      • Chen P.
      • et al.
      Blood eosinophils: a biomarker of response to glucocorticoids and increased readmissions in severe hospitalized exacerbations of COPD.
      Petite et al.
      • Petite S.E.
      • Murphy J.A.
      Systemic corticosteroid and anti-biotic use in hospitalized patients with chronic obstruc- tive pulmonary disease exacerbation.
      evaluated patients with AECOPD and blood EOS ≥200 cells/μL and/or ≥2% had a lower risk of hospital stay (LOS) and death.
      • Li Q.
      • Larivée P.
      • Courteau J.
      • et al.
      Greater eosinophil counts at first COPD hospitalization are associated with more readmissions and fewer deaths.
      Meanwhile, Dai G et al. indicated that viral-dominant infections may be the main cause of eosinophilic AECOPD.
      • Dai G.
      • Ran Y.
      • Wang J.
      • et al.
      Clinical differences between eosinophilic and noneosinophilic acute exacerbation of chronic obstructive pulmonary disease: a multicenter cross-sectional study.
      The increased EOS count was closely related to pulmonary function, and the cutoff value was 2.963%.
      • Zhou W.
      • Tan J.
      The expression and the clinical significance of eosinophils, PCT and CRP in patients with acute exacerbation of chronic obstructive pulmonary disease complicated with pulmonary infection.
      However, less attention has been given to the association between eosinophils and all comorbidities and prognosis in AECOPD. Yu et al.
      • Yu S.
      • Zhang J.
      • Fang Q.
      • et al.
      Blood eosinophil levels and prognosis of hospitalized patients with acute exacerbation of chronic obstructive pulmonary disease.
      found that respiratory failure was higher when the EOS count was ≤100/µl, but the EOS count was not related to the prognosis at 1-year of AECOPD. Patients with low EOS levels (<2%) develop severe inflammation and increased mortality.
      • Lv M.Y.
      • Qiang L.X.
      • Li Z.H.
      • et al.
      The lower the eosinophils, the stronger the inflammatory response? The relationship of different levels of eosinophils with the degree of inflammation in acute exacerbation chronic obstructive pulmonary disease (AECOPD).
      Therefore, different cutoff values of EOS may be useful biomarkers to predict the clinical features and prognosis of patients with AECOPD.
      Another controversial aspect is the criteria for cutoff of blood EOS counts for diagnostic purposes. Most AECOPD studies used ≥200 cells/μL and/or ≥2% as the threshold to identify hypereosinophilic patients.
      • Wu C.W.
      • Lan C.C.
      • Hsieh P.C.
      • et al.
      Role of peripheral eosinophilia in acute exacerbation of chronic obstructive pulmonary disease.
      ,
      • Hastie A.T.
      • Martinez F.J.
      • Curtis J.L.
      • et al.
      SPIROMICS investigators. Association of sputum and blood eosinophil concentrations with clinical measures of COPD severity: an analysis of the SPIROMICS cohort.
      Other studies on eosinophils in AECOPD followed the criteria of EOS%< 2% and EOS counts < 200 cells/μL, EOS% 2%-2.99% and/or EOS counts of 200-299 cells/μL, EOS%≥3% and/or EOS counts of ≥300 cells/uL.
      • Dai G.
      • Ran Y.
      • Wang J.
      • et al.
      Clinical differences between eosinophilic and noneosinophilic acute exacerbation of chronic obstructive pulmonary disease: a multicenter cross-sectional study.
      Müllerová et al. analyzed clinical characteristics and exacerbations based on EOS counts (<150 and ≥150 cells/µL) and using four categories (<150, ≥150–<300, ≥300–<500, and ≥500 cells/μL).
      • Müllerová H.
      • Hahn B.
      • Simard E.P.
      • et al.
      Exacerbations and health care resource use among patients with COPD in relation to blood eosinophil counts.
      There is no uniform criterion for the cutoff of blood EOS in AECOPD.
      Therefore, based on the different grouping methods of EOS commonly used at present, this study further evaluated the correlation between blood EOS levels and inflammatory indices, pulmonary function, chest CT images, complications, related treatments, and 3-year mortality in AECOPD patients in the real world. The optimal cutoff of EOS was also explored.

      Methods

      Study design and data sources

      This was a retrospective observational study in which we screened the medical records of patients with AECOPD in the Department of Respiratory and Critical Care Medicine of Shanxi Bethune Hospital from September 1, 2013 to July 31, 2017. AECOPD patients were divided into three different groups according to their blood eosinophil level at the time of the index hospitalization. This study was approved by the Ethics Review Committee of Shanxi Bethune Hospital. All participants signed informed consent forms and agreed to the use of the data.

      Study population

      The inclusion criteria were as follows: patients aged > 40 years who had been hospitalized for COPD (main diagnosis ICD-10 J44.100). Also, patients’ lung function tests satisfied the diagnosis of COPD: a postbronchodilator forced expiratory volume in one second (FEV1)/forced vital capacity (FVC) ratio < 0.7, in accordance with 2020 GOLD guidelines.
      • Halpin D.M.G.
      • Celli B.R.
      • Criner G.J.
      • et al.
      The GOLD Summit on chronic obstructive pulmonary disease in low- and middle-income countries.
      The exclusion criteria were defined according to the history of asthma, allergic diseases, hypereosinophilia, blood system-related diseases, and parasitic infections, use of long-term oral corticosteroid, and having received oral or intravenous glucocorticoid administration within 48 h before admission. After three years, we followed up with the patients and recorded the information.

      Blood eosinophil level

      Three different categories were used in this study. According to different eosinophil counts, AECOPD patients were classified into the <100 cells/L or ≥100 cells/L group. Two groups were divided according to the different percentages of eosinophils: EOS%<2%, ≥2%; EOS%<3%, ≥3%.

      Parameters

      Clinical data, laboratory test parameters, and follow-up information were collected from the medical records. The clinical data included demographic characteristics, smoking history, body mass index, comorbidities, use of antibiotic glucocorticoids, and ventilator treatment during hospitalization. Laboratory test parameters included pulmonary function tests, white blood cell counts, neutrophil and lymphocyte counts, C-reactive protein (CRP) levels, D-dimer levels, procalcitonin (PCT), arteriole blood gas analysis, and chest CT imaging. Follow-up information included survival rate, patients' annual admissions to the hospital, their ability to schedule life, and their family's willingness to follow up.

      Statistical analyses

      Data are presented as the mean ± standard deviation and number (percentage). The chi-square test and Fischer's exact test were used to analyze the association of clinical-laboratory-prognosis parameters with AECOPD. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the predictive value of EOS biomarkers in differentiating comorbidities from AECOPD. SPSS 17.0 was used (SPSS Institute, Chicago, IL, USA). Statistical significance was set at a P value less than 0.05.

      Results

      Baseline characteristics of AECOPD patients

      Between September 1, 2013 and July 31, 2017, a total of 1562 AECOPD patients were hospitalized. After screening, data for a total of 984 patients were analyzed (Fig. 1). Table 1 summarizes the main demographic and clinical characteristics of the 984 patients in this study. They were mostly men (77.6%), with a mean age of 70.83 (SD 9.61) years and blood eosinophils of 130 (SD 150) cells/µl. The percentage of male patients in the EOS≥2% group was greater than that in the EOS<2% group (81.74% vs. 76.05%, P < 0.05). There was a significant difference in age between the two groups (EOS<2% 71.29±9.72 vs. EOS≥2% 70.03±9.36, P < 0.001). Eosinophils are relatively lower for those aged more than 80 years (Fig. 2). The proportion of patients in the EOS<100/µl group (507/51.52%) was the highest.
      Fig 1
      Fig. 1A flowchart for screening patients *represents significant differences compared with other groups, red represents significant differences from all other black groups.
      Table 1Baseline characteristics of AECOPD patients.
      VariablesTotalEOS%EOS count
      <2%≥2%<3%≥3%<0.10.1-0.150.15-0.20.2-0.250.25-0.3≥0.3
      Total number of patients984625 (63.52%)359 (36.48%)759 (77.13%)225 (22.87%)507 (51.52%)148 (15.04%)98 (9.96%)65 (6.61%)48 (4.88%)118 (11.99%)
      Gender
      Male764 (78.12%)473 (76.05%)291 (81.74%)592 (78.31%)172 (77.48%)381 (75.45%)123 (83.67%)78 (80.41%)50 (78.12%)42 (87.50%)90 (76.92%)
      Female214 (21.88%)149 (23.95%)65 (18.26%)164 (21.69%)50 (22.52%)124 (24.55%)24 (16.33%)19 (19.59%)14 (21.88%)6 (12.50%)27 (23.08%)
      Age625359507148986548118
      <60 years134 (13.62%)77 (12.32%)57 (15.88%)99 (13.04%)35 (15.56%)58 (11.44%)24 (16.22%)15 (15.31%)9 (13.85%)7 (14.58%)21 (17.80%)
      60-70274 (27.85%)177 (28.32%)97 (27.02%)213 (28.06%)61 (27.11%)151 (29.78%)38 (25.68%)18 (18.37%)22 (33.85%)12 (25.00%)33 (27.97%)
      70-80374 (38.01%)220 (35.20%)154 (42.90%)283 (37.29%)91 (40.44%)181 (35.70%)52 (35.14%)51 (52.04%)24 (36.92%)20 (41.67%)46 (38.98%)
      ≥80202 (20.53%)151 (24.16%)51 (14.21%)164 (21.61%)38 (16.89%)117 (23.08%)34 (22.97%)14 (14.29%)10 (15.38%)9 (18.75%)18 (15.25%)
      Smoking history
      Never227 (23.07%)151 (24.16%)76 (21.17%)169 (22.27%)58 (25.78%)129 (25.44%)24 (16.22%)19 (19.39%)14 (21.54%)7 (14.58%)34 (28.81%)
      Current434 (44.41%)272 (43.52%)162 (45.13%)339 (44.66%)95 (42.22%)209 (41.22%)76 (51.35%)46 (46.94%)32 (49.23%)24 (50.00%)47 (39.83%)
      Ex-smoker323 (32.83%)202 (32.32%)121 (33.70%)251 (33.07%)72 (32.00%)169 (33.33%)48 (32.43%)33 (33.67%)19 (29.23%)17 (35.42%)37 (31.36%)
      Smoking index (pack-year)507148986548118
      <100252 (25.61%)172 (27.52%)80 (22.28%)192 (25.30%)60 (26.67%)144 (28.40%)29 (19.59%)21 (21.43%)16 (24.62%)7 (14.58%)35 (29.66%)
      100-400101 (10.26%)57 (9.12%)44 (12.26%)74 (9.75%)27 (12.00%)42 (8.28%)22 (14.86%)11 (11.22%)7 (10.77%)5 (10.42%)14 (11.86%)
      ≥400631 (64.13%)396 (63.36%)235 (65.46%)493 (64.95%)138 (61.33%)321 (63.31%)97 (65.54%)66 (67.35%)42 (64.62%)36 (75.00%)69 (58.47%)
      Dust exposure
      Yes16 (1.63%)7 (1.12%)9 (2.51%)10 (1.32%)6 (2.67%)7 (1.38%)2 (1.35%)0 (0.00%)1 (1.54%)2 (4.17%)4 (3.39%)
      No968 (98.37%)618 (98.88%)350 (97.49%)749 (98.68%)219 (97.33%)500 (98.62%)146 (98.65%)98 (100.00%)64 (98.46%)46 (95.83%)114 (96.61%)
      Fig 2
      Fig. 2Demographic characteristics based on different EOS groups *represents significant differences compared with other groups, red represents significant differences from all other black groups.

      EOS was associated with comorbidities (pulmonary heart disease, arrhythmia)

      Table 2 summarizes the 10 comorbidities of the 984 patients, which mainly includes chronic pulmonary heart disease (557/56.61%), coronary heart disease (99/10.06%), all kinds of arrhythmias (121/12.3%), obstructive sleep apnea-hypopnea syndrome (OSHAS), and cancer.
      Table 2Comorbidities of AECOPD patients.
      VariablesTotalEOS%EOS count
      <2%≥2%<3%≥3%<0.10.1-0.150.15-0.20.2-0.250.25-0.3≥0.3
      Total number of patients984625 (63.52%)359 (36.48%)759 (77.13%)225 (22.87%)507 (51.52%)148 (15.04%)98 (9.96%)65 (6.61%)48 (4.88%)118 (11.99%)
      Comorbidities
      Chronic pulmonary heart disease;
      Yes557 (56.61%)375 (60.00%)182 (50.70%)450 (59.29%)107 (47.56%)312 (61.54%)81 (54.73%)54 (55.10%)31 (47.69%)19 (39.58%)60 (50.85%)
      No427 (43.39%)250 (40.00%)177 (49.30%)309 (40.71%)118 (52.44%)195 (38.46%)67 (45.27%)44 (44.90%)34 (52.31%)29 (60.42%)58 (49.15%)
      Coronary heart disease
      Yes99 (10.06%)51 (8.16%)26 (7.24%)83 (10.94%)16 (7.11%)55 (10.85%)13 (8.78%)12 (12.24%)5 (7.69%)5 (10.42%)9 (7.63%)
      No885 (89.94%)574 (91.84%)333 (92.76%)676 (89.06%)209 (92.89%)452 (89.15%)135 (91.22%)86 (87.76%)60 (92.31%)43 (89.58%)109 (92.37%)
      Arrhythmology
      Yes121 (12.30%)90 (14.40%)31 (8.64%)107 (14.10%)14 (6.22%)73 (14.40%)20 (13.51%)12 (12.24%)2 (3.08%)5 (10.42%)9 (7.63%)
      No863 (87.70%)535 (85.60%)328 (91.36%)652 (85.90%)211 (93.78%)434 (85.60%)128 (86.49%)86 (87.76%)63 (96.92%)43 (89.58%)109 (92.37%)
      Atrial fibrillation
      Yes59 (6.00%)48 (7.68%)11 (3.06%)53 (6.98%)6 (2.67%)43 (8.48%)6 (4.05%)3 (3.06%)1 (1.54%)2 (4.17%)4 (3.39%)
      No925 (94.00%)577 (92.32%)348 (96.94%)706 (93.02%)219 (97.33%)464 (91.52%)142 (95.95%)95 (96.94%)64 (98.46%)46 (95.83%)114 (96.61%)
      OSAHS
      Yes37 (3.76%)28 (4.48%)9 (2.51%)33 (4.35%)4 (1.78%)23 (4.54%)4 (2.70%)3 (3.06%)3 (4.62%)1 (2.08%)3 (2.54%)
      No947 (96.24%)597 (95.52%)350 (97.49%)726 (95.65%)221 (98.22%)484 (95.46%)144 (97.30%)95 (96.94%)62 (95.38%)47 (97.92%)115 (97.46%)
      Cancer
      Yes155 (15.75%)100 (16.00%)55 (15.32%)124 (16.34%)31 (13.78%)77 (15.19%)26 (17.57%)18 (18.37%)8 (12.31%)9 (18.75%)17 (14.41%)
      No829 (84.25%)525 (84.00%)304 (84.68%)635 (83.66%)194 (86.22%)430 (84.81%)122 (82.43%)80 (81.63%)57 (87.69%)39 (81.25%)101 (85.59%)
      Hypertension
      Yes298 (30.28%)202 (32.32%)96 (26.74%)235 (30.96%)63 (28.00%)165 (32.54%)36 (24.32%)29 (29.59%)22 (33.85%)11 (22.92%)35 (29.66%)
      No686 (69.72%)423 (67.68%)263 (73.26%)524 (69.04%)162 (72.00%)342 (67.46%)112 (75.68%)69 (70.41%)43 (66.15%)37 (77.08%)83 (70.34%)
      Diabetes
      Yes77 (7.83%)51 (8.16%)26 (7.24%)57 (7.51%)20 (8.89%)39 (7.69%)16 (10.81%)5 (5.10%)1 (1.54%)6 (12.50%)10 (8.47%)
      No907 (92.17%)574 (91.84%)333 (92.76%)702 (92.49%)205 (91.11%)468 (92.31%)132 (89.19%)93 (94.90%)64 (98.46%)42 (87.50%)108 (91.53%)
      Osteoporosis
      Yes25 (2.54%)12 (1.92%)13 (3.62%)15 (1.98%)10 (4.44%)11 (2.17%)2 (1.35%)4 (4.08%)3 (4.62%)0 (0.00%)5 (4.24%)
      No959 (97.46%)613 (98.08%)346 (96.38%)744 (98.02%)215 (95.56%)496 (97.83%)146 (98.65%)94 (95.92%)62 (95.38%)48 (100.00%)113 (95.76%)
      Gastroesophageal reflux
      Yes86 (8.74%)56 (8.96%)30 (8.36%)66 (8.70%)20 (8.89%)50 (9.86%)10 (6.76%)3 (3.06%)4 (6.15%)7 (14.58%)12 (10.17%)
      No898 (91.26%)569 (91.04%)329 (91.64%)693 (91.30%)205 (91.11%)457 (90.14%)138 (93.24%)95 (96.94%)61 (93.85%)41 (85.42%)106 (89.83%)
      Mental illness
      Yes17 (1.73%)10 (1.60%)7 (1.95%)12 (1.58%)5 (2.22%)9 (1.78%)2 (1.35%)0 (0.00%)0 (0.00%)0 (0.00%)6 (5.08%)
      No967 (98.27%)615 (98.40%)352 (98.05%)747 (98.42%)220 (97.78%)498 (98.22%)146 (98.65%)98 (100.00%)65 (100.00%)48 (100.00%)112 (94.92%)
      The EOS>2% or >3% group had less complicated pulmonary heart disease (P < 0.01), arrhythmia (P < 0.01), and atrial fibrillation (P < 0.05). There were no significant differences in the other comorbidities between the two groups. According to different eosinophil count groups, except for chronic pulmonary heart disease (P < 0.05) and atrial fibrillation (P < 0.05), there were no significant differences in other complications among all groups. Within groups, we found that the relatively higher EOS group had a lower incidence of pulmonary heart disease and atrial fibrillation than the EOS <100/µl group (Fig. 3).
      Fig 3
      Fig. 3Comorbidities based on different EOS groups *represents significant differences compared with other groups, red represents significant differences from all other black groups.

      Eos was associated with infectious index and treatment

      Table 3 shows the relevant laboratory results and treatment of the 984 patients. There were 18.09% of patients with white blood cells (WBCs) higher than 10*109/L (mean value 7.71±3.26), 36.28% with increased C-reactive protein (mean value 20.76±41.68), 11.99% with higher procalcitonin (PCT) and 40.2% with type II respiratory failure. Chest CT imaging findings included mainly emphysema (51.22%) and chronic bronchitis (48.78%). There were 10.98% of patients who received systemic glucocorticoids, with 27.03% of patients treated with a ventilator. A total of 78.45% of patients were given antibiotics.
      Table 3Laboratory testing and clinical treatment of AECOPD patients.
      VariablesTotalEOS%EOS count
      <2%≥2%<3%≥3%<0.10.1-0.150.15-0.20.2-0.250.25-0.3≥0.3
      Total number of patients984625 (63.52%)359 (36.48%)759 (77.13%)225 (22.87%)507 (51.52%)148 (15.04%)98 (9.96%)65 (6.61%)48 (4.88%)118 (11.99%)
      Chest computed tomography (CT)
      Emphysema
      Yes504 (51.22%)331 (52.96%)149 (41.50%)377 (49.67%)127 (56.44%)232 (45.76%)93 (62.84%)47 (47.96%)39 (60.00%)26 (54.17%)67 (56.78%)
      No480 (48.78%)294 (47.04%)210 (58.50%)382 (50.33%)98 (43.56%)275 (54.24%)55 (37.16%)51 (52.04%)26 (40.00%)22 (45.83%)51 (43.22%)
      Bronchitis
      Yes430 (43.70%)246 (39.36%)184 (51.25%)322 (42.42%)108 (48.00%)196 (38.66%)77 (52.03%)39 (39.80%)34 (52.31%)25 (52.08%)59 (50.00%)
      No554 (56.30%)379 (60.64%)175 (48.75%)437 (57.58%)117 (52.00%)311 (61.34%)71 (47.97%)59 (60.20%)31 (47.69%)23 (47.92%)59 (50.00%)
      FEV1/pred
      <30831 (84.45%)554 (88.64%)277 (77.16%)658 (86.69%)173 (76.89%)450 (88.76%)129 (87.16%)81 (82.65%)51 (78.46%)34 (70.83%)86 (72.88%)
      30-5065 (6.61%)31 (4.96%)34 (9.47%)42 (5.53%)23 (10.22%)23 (4.54%)6 (4.05%)9 (9.18%)4 (6.15%)9 (18.75%)14 (11.86%)
      50-8068 (6.91%)30 (4.80%)38 (10.58%)46 (6.06%)22 (9.78%)27 (5.33%)9 (6.08%)6 (6.12%)10 (15.38%)3 (6.25%)13 (11.02%)
      ≥8020 (2.03%)10 (1.60%)10 (2.79%)13 (1.71%)7 (3.11%)7 (1.38%)4 (2.70%)2 (2.04%)0 (0.00%)2 (4.17%)5 (4.24%)
      Type II respiratory failure
      Yes396 (40.37%)279 (44.78%)117 (32.68%)322 (42.54%)74 (33.04%)236 (46.64%)58 (39.46%)26 (26.53%)25 (38.46%)12 (25.00%)39 (33.33%)
      No585 (59.63%)344 (55.22%)241 (67.32%)435 (57.46%)150 (66.96%)270 (53.36%)89 (60.54%)72 (73.47%)40 (61.54%)36 (75.00%)78 (66.67%)
      White blood cell count
      <446 (4.67%)21 (3.36%)25 (6.96%)28 (3.69%)18 (8.00%)27 (5.33%)9 (6.08%)6 (6.12%)3 (4.62%)1 (2.08%)0 (0.00%)
      4-10760 (77.24%)451 (72.16%)309 (86.07%)567 (74.70%)193 (85.78%)360 (71.01%)125 (84.46%)81 (82.65%)55 (84.62%)40 (83.33%)99 (83.90%)
      ≥10178 (18.09%)153 (24.48%)25 (6.96%)164 (21.61%)14 (6.22%)120 (23.67%)14 (9.46%)11 (11.22%)7 (10.77%)7 (14.58%)19 (16.10%)
      NEU%
      <4014 (1.42%)6 (0.96%)8 (2.23%)9 (1.19%)5 (2.22%)6 (1.18%)1 (0.68%)1 (1.02%)2 (3.08%)2 (4.17%)2 (1.69%)
      40-75603 (61.28%)291 (46.56%)312 (86.91%)400 (52.70%)203 (90.22%)226 (44.58%)107 (72.30%)76 (77.55%)53 (81.54%)38 (79.17%)103 (87.29%)
      ≥75367 (37.30%)328 (52.48%)39 (10.86%)350 (46.11%)17 (7.56%)275 (54.24%)40 (27.03%)21 (21.43%)10 (15.38%)8 (16.67%)13 (11.02%)
      LYM%
      <20573 (58.23%)435 (69.60%)138 (38.44%)493 (64.95%)80 (35.56%)353 (69.63%)81 (54.73%)42 (42.86%)28 (43.08%)16 (33.33%)53 (44.92%)
      20-50406 (41.26%)187 (29.92%)219 (61.00%)262 (34.52%)144 (64.00%)151 (29.78%)66 (44.59%)56 (57.14%)37 (56.92%)32 (66.67%)64 (54.24%)
      ≥505 (0.51%)3 (0.48%)2 (0.56%)4 (0.53%)1 (0.44%)3 (0.59%)1 (0.68%)0 (0.00%)0 (0.00%)0 (0.00%)1 (0.85%)
      CRP
      <8627 (63.72%)364 (58.24%)263 (73.26%)462 (60.87%)165 (73.33%)283 (55.82%)108 (72.97%)66 (67.35%)48 (73.85%)35 (72.92%)87 (73.73%)
      ≥8357 (36.28%)261 (41.76%)96 (26.74%)297 (39.13%)60 (26.67%)224 (44.18%)40 (27.03%)32 (32.65%)17 (26.15%)13 (27.08%)31 (26.27%)
      D-dimer984
      <500823 (83.64%)494 (79.04%)329 (91.64%)616 (81.16%)207 (92.00%)386 (76.13%)132 (89.19%)93 (94.90%)62 (95.38%)44 (91.67%)106 (89.83%)
      ≥500161 (16.36%)131 (20.96%)30 (8.36%)143 (18.84%)18 (8.00%)121 (23.87%)16 (10.81%)5 (5.10%)3 (4.62%)4 (8.33%)12 (10.17%)
      PCT984
      <0.1866 (88.01%)535 (85.60%)331 (92.20%)659 (86.82%)207 (92.00%)428 (84.42%)136 (91.89%)88 (89.80%)63 (96.92%)46 (95.83%)105 (88.98%)
      0.1-0.586 (8.74%)61 (9.76%)25 (6.96%)71 (9.35%)15 (6.67%)54 (10.65%)9 (6.08%)9 (9.18%)2 (3.08%)2 (4.17%)10 (8.47%)
      ≥0.532 (3.25%)29 (4.64%)3 (0.84%)29 (3.82%)3 (1.33%)25 (4.93%)3 (2.03%)1 (1.02%)0 (0.00%)483 (2.54%)
      antibiotic
      Yes772 (78.46%)522 (83.52%)250 (69.64%)615 (81.03%)157 (69.78%)425 (83.83%)121 (81.76%)68 (69.39%)50 (76.92%)28 (58.33%)80 (67.80%)
      No212 (21.54%)103 (16.48%)109 (30.36%)144 (18.97%)68 (30.22%)82 (16.17%)27 (18.24%)30 (30.61%)15 (23.08%)20 (41.67%)38 (32.20%)
      Inhaled corticosteroids
      Yes 827 (84.04%)526 (84.16%)301 (83.84%)642 (84.58%)185 (82.22%)430 (84.81%)126 (85.14%)83 (84.69%)56 (86.15%)39 (81.25%)93 (78.81%)
      No 157 (15.96%)99 (15.84%)58 (16.16%)117 (15.42%)40 (17.78%)77 (15.19%)22 (14.86%)15 (15.31%)9 (13.85%)9 (18.75%)25 (21.19%)
      Systemic glucocorticoids
      Yes108 (10.98%)84 (13.44%)24 (6.69%)91 (11.99%)17 (7.56%)71 (14.00%)13 (8.78%)5 (5.10%)5 (7.69%)6 (12.50%)8 (6.78%)
      No876 (89.02%)541 (86.56%)335 (93.31%)668 (88.01%)208 (92.44%)436 (86.00%)135 (91.22%)93 (94.90%)60 (92.31%)42 (87.50%)110 (93.22%)
      Ventilator
      Yes266 (27.03%)209 (33.44%)57 (15.88%)230 (30.30%)36 (16.00%)182 (35.90%)29 (19.59%)18 (18.37%)9 (13.85%)10 (20.83%)18 (15.25%)
      No718 (72.97%)416 (66.56%)302 (84.12%)529 (69.70%)189 (84.00%)325 (64.10%)119 (80.41%)80 (81.63%)56 (86.15%)38 (79.17%)100 (84.75%)
      In contrast, the EOS<2% group had a heavier airflow limitation (88.64% vs. 77.16%, P < 0.01), relatively higher D-dimer (20.96% vs. 8.36%, P < 0.01), a significantly higher burden of infectious inflammation (assessed by CT, WBC, CRP, PCT), and respiratory failure (P < 0.05). On CT imaging, either emphysema (58.5% vs. 47.04%, P < 0.001) or bronchitis (51.25% vs. 39.36%, P < 0.001) was more pronounced on chest CT imaging in the >2% group. The use of ventilators and systemic glucocorticoids in the EOS≥2% group was lower than in the EOS<2% group. When the EOS% cutoff was 3%, the statistical results were similar to 2%, except for CT findings and the use of systemic glucocorticoids. According to different eosinophil count groups, FEV1/pred%, WBC, CRP, PCT, D-dimer, type II respiratory failure and chest CT imaging, ventilators, antibiotics, and systemic glucocorticoids were significantly different among the groups (all P < 0.05). Within groups, the EOS<100/µl group had significantly more severe lung function, higher rates of respiratory failure, a higher burden of infectious inflammation, and higher use of ventilators and systemic glucocorticoids (P < 0.05) (Supplementary Fig. 1 and Fig. 4).
      Fig 4
      Fig. 4Clinical treatment based on different EOS groups *represents significant differences compared with other groups, red represents significant differences from all other black groups.

      EOS is associated with mortality according to the follow-up after three years

      Table 4 summarizes the contents of the follow-up data after three years. A total of 441 patients were enrolled in the follow-up study, and 543 patients declined. A total of 187 (42.99%) patients had died. Of the 248 patients with survival information, 17.95% were hospitalized more than twice per year due to AECOPD. There were relatively fewer deaths in the EOS≥2% or 3% group at follow-up (P < 0.01). According to the different eosinophil count groups, compared to the 200≤EOS<250/µl and ≥300/µl groups, the EOS<100/µl group had relatively higher deaths at follow-up (49.57% vs. 29.03%, P < 0.05; 49.57% vs. 22.64%, P < 0.01) (Supplementary Fig. 2)..
      Table 4Three years follow-up of AECOPD patients.
      VariablesTotalEOS%EOS count
      <2%≥2%<3%≥3%<0.10.1-0.150.15-0.20.2-0.250.25-0.3≥0.3
      Total number of patients984625 (63.52%)359 (36.48%)759 (77.13%)225 (22.87%)507 (51.52%)148(15.04%)98(9.96%)65(6.61%)48(4.88%)118(11.99%)
      Follow-up 441

      Gender
      Male342(77.55%)210(75.27%)132(81.48%)263(77.13%)79(79.00%)174(74.68%)42(80.77%)41(80.39%)26(78.79%)20(100.00%)39(75.00%)
      Female99(22.45%)69(24.73%)30(18.52%)78(22.87%)21(21.00%)59(25.32%)10(19.23%)10(19.61%)7(21.21%)0(0.00%)13(25.00%)
      Age
      <6064(14.55%)33(11.87%)31(19.14%)41(12.06%)23(23.00%)28(12.07%)4(7.69%)7(13.73%)6(18.18%)5(25.00%)14(26.92%)
      60-70116(26.36%)76(27.34%)40(24.69%)94(27.65%)22(22.00%)65(28.02%)13(25.00%)10(19.61%)12(36.36%)5(25.00%)11(21.15%)
      70-80168(38.18%)102(36.69%)66(40.74%)134(39.41%)34(34.00%)87(37.50%)21(40.38%)27(52.94%)9(27.27%)8(40.00%)16(30.77%)
      >=8092(20.91%)67(24.10%)25(15.43%)71(20.88%)21(21.00%)52(22.41%)14(26.92%)7(13.73%)6(18.18%)2(10.00%)11(21.15%)
      Death
      No248(57.01%)142(51.26%)106(67.09%)175(52.24%)73(73.00%)117(50.43%)27(54.00%)30(61.22%)22(70.97%)11(55.00%)41(77.36%)
      Yes187(42.99%)135(48.74%)52(32.91%)160(47.76%)27(27.00%)115(49.57%)23(46.00%)19(38.78%)9(29.03%)9(45.00%)12(22.64%)
      Re-admission
      <2192(82.05%)106(77.94%)86(87.76%)131(78.92%)61(89.71%)82(74.55%)21(87.50%)26(86.67%)17(80.95%)9(90.00%)37(94.87%)
      ≥242(17.95%)30(22.06%)12(12.24%)35(21.08%)7(10.29%)28(25.45%)3(12.50%)4(13.33%)4(19.05%)1(10.00%)2(5.13%)
      Daily Living ability (outdoor activity)
      No12(4.84%)10(6.94%)2(1.92%)11(6.21%)1(1.41%)8(6.78%)1(3.70%)1(3.23%)0(0.00%)0(0.00%)2(5.00%)
      Yes236(95.16%)134(93.06%)102(98.08%)166(93.79%)70(98.59%)110(93.22%)26(96.30%)30(96.77%)21(100.00%)11(100.00%)38(95.00%)
      Follow-up intention
      NO431(97.51%)270(96.77%)161(98.77%)331(97.07%)100(99.01%)226(97.00%)51(98.08%)50(98.04%)32(96.97%)20(100.00%)52(98.11%)
      Yes11(2.49%)9(3.23%)2(1.23%)10(2.93%)1(0.99%)7(3.00%)1(1.92%)1(1.96%)1(3.03%)0(0.00%)1(1.89%)

      Exploring the predictive value of eosinophils with AUC>0.5

      There is currently no uniform eosinophil threshold to identify the clinical features of AECOPD, and further analysis was conducted to explore the predictive value of eosinophils. Of all the clinical variables, the ROC curve showed that the indicators of AUC≥0.5 included chest CT imaging findings, osteoporosis, mental illness, dust exposure, and ex-smoker status (Fig. 5). Statistical analyses of all AECOPD patients were conducted again according to the optimal eosinophil threshold determined by the ROC curve. We found that if the EOS% was 1.8% (or ≥100/µl), it was more associated with chest CT with emphysematous manifestations and ≥1.7% (or ≥100/µl) with a bronchitis phenotype (P < 0.01). If the EOS% was ≥1.3% (or ≥100/µl), it was more associated with ex-smoker status. If the EOS% was ≥6.1% (or ≥400/µl), it was more associated with mental disease (P < 0.05). If the EOS% was ≥2.4% (or ≥140/µl), it was more associated with osteoporosis (P < 0.05). If the EOS% was ≥2.2% (or ≥240/µl), it was more associated with dust exposure (P < 0.05). However, the proportions of these three comorbidities measured by different thresholds were low (Table 5).
      Fig 5
      Fig. 5Exploring the predictive value of eosinophils with AUC>0.5.
      Table 5Results of ROC curve.
      VariablesCutoff value(%)AUCSensitivitySpecificityPositive Predictive Valuenegative predictive valuechisqfisher
      Ex-smoker1.30.5050.5460.520.4730.5920.040190.04608
      Dust exposure2.20.5520.5620.6730.0280.9890.060440.08667
      Osteoporosis2.40.6020.520.7090.0450.9830.023890.02423
      Mental illness6.10.5830.2940.9430.0830.9870.0025030.0003983
      bronchitis1.70.5510.4860.6390.5110.6169.03E-050.0001035

      Discussion

      The study fully evaluated the role of eosinophils in AECOPD. We found that the EOS group (≥2% or ≥3% or ≥100/µl.) was associated with a tendency toward noninfectious inflammation, mild airflow limitation, fewer comorbidities, and lower 3-year mortality.
      Our findings are consistent with previous studies on infection and antibiotics. The EOS, CRP, and PCT have high clinical guiding value in judging the type of infection and predicting the treatment effect for AECOPD.
      • Xue J.
      • Cui Y.N.
      • Chen P.
      • et al.
      Blood eosinophils: a biomarker of response to glucocorticoids and increased readmissions in severe hospitalized exacerbations of COPD.
      ,
      • Bafadhel M.
      • Davies L.
      • Calverley P.M.
      • et al.
      Blood eosinophil guided prednisolone therapy for exacerbations of COPD: a further analysis.
      • Tanrıverdi H.
      • Örnek T.
      • Erboy F.
      • et al.
      Comparison of diagnostic values of procalcitonin, C-reactive protein and blood neutrophil/lymphocyte ratio levels in predicting bacterial infection in hospitalized patients with acute exacerbations of COPD.
      • Gao S.
      • Duan Y.
      • Chen J.
      • et al.
      Evaluation of blood markers at admission for predicting community acquired pneumonia in chronic obstructive pulmonary disease.
      Many studies used the EOS grouping method (EOS%< 2% and/or EOS counts < 200 cells/μ; EOS% 2%-2.99% and/or EOS counts 200-299 cells/μL; EOS% ≥ 3% and/or EOS counts ≥ 300 cells/μL) and believed that the hypereosinophilic group (≥2% or ≥200 cells/μL) had a tendency toward noninfectious inflammation.
      • Dai G.
      • Ran Y.
      • Wang J.
      • et al.
      Clinical differences between eosinophilic and noneosinophilic acute exacerbation of chronic obstructive pulmonary disease: a multicenter cross-sectional study.
      ,
      • Wu C.W.
      • Lan C.C.
      • Hsieh P.C.
      • et al.
      Role of peripheral eosinophilia in acute exacerbation of chronic obstructive pulmonary disease.
      ,
      • MacDonald M.I.
      • Osadnik C.R.
      • Bulfin L.
      • et al.
      Low and high blood eosinophil counts as biomarkers in hospitalized acute exacerbations of COPD.
      ,
      • Choi J.
      • Oh J.Y.
      • Lee Y.S.
      • et al.
      The association between blood eosinophil percent and bacterial infection in acute exacerbation of chronic obstructive pulmonary disease.
      However, the mechanism of EOS and infection in AECOPD is unclear. Eosinophils may phagocytize bacteria through phagocytosis similar to neutrophils,
      • Weller P.F.
      • Goetzel E.J.
      The human eosinophil: roles in hostdefense and tissue injury.
      ,
      • Ravin K.A.
      • Loy M.
      The eosinophil in infection.
      and eosinophil granule protein has antibacterial properties.
      • Bystrom J.
      • Amin K.
      • Bishop-Bailey D.
      Analysing the eosinophil cationic protein–a clue to the function of the eosinophil granulocyte.
      • Persson T.
      • Andersson P.
      • Bodelsson M.
      • et al.
      Bactericidal activity of human eosinophilic granulocytes against Escherichia coli.
      • Yousefi S.
      • Gold J.A.
      • Andina N.
      Catapult-like release of mitochondrial DNA by eosinophils contributes to antibacterial defense.
      Eosinophils in infection may be related to the initiation of the innate immune response to pathogens by releasing extracellular DNA traps
      • Mukherjee M.
      • Lacy P.
      • Ueki S.
      Eosinophil extracellular traps and inflammatory pathologies-untangling the web!.
      and expressing specific pattern recognition receptors.
      • LeMessurier K.S.
      • Samarasinghe AE.
      Eosinophils: nemeses of pulmonary pathogens?.
      ,
      • Kvarnhammar A.M.
      • Cardell LO.
      Pattern-recognition receptors in human eosinophils.
      Bacterial infection may influence corticosteroid responsiveness by changing the inflammatory response of neutrophils and eosinophils.
      • Kolsum U.
      • Donaldson G.C.
      • Singh R.
      • et al.
      Blood and sputum eosinophils in COPD; relationship with bacterial load.
      ,
      • Bass D.A.
      Behavior of eosinophil leukocytes in acute inflammation. II. Eosinophil dynamics during acute inflammation.
      Some studies have shown that blood eosinophils can guide systemic corticosteroid exposure in AECOPD,
      • Xue J.
      • Cui Y.N.
      • Chen P.
      • et al.
      Blood eosinophils: a biomarker of response to glucocorticoids and increased readmissions in severe hospitalized exacerbations of COPD.
      ,
      • Walters J.A.
      • Tan D.J.
      • White C.J.
      • et al.
      Systemic corticosteroids for acute exacerbations of chronic obstructive pulmonary disease.
      ,
      • Kichloo A.
      • Aljadah M.
      • Vipparla N.
      • et al.
      Optimal glucocorticoid dose and the effects on mortality, length of stay, and readmission rates in patients diagnosed with acute exacerbation of chronic obstructive pulmonary disease (AECOPD).
      ,
      • Sivapalan P.
      • Lapperre T.S.
      • Janner J.
      • et al.
      Eosinophil-guided corticosteroid therapy in patients admitted to hospital with COPD exacerbation (CORTICO-COP): a multicentre, randomised, controlled, open-label, non-inferiority trial.
      and systemic corticosteroids can improve clinical outcomes in AECOPD associated with serum eosinophils ≥2%
      • Bafadhel M.
      • McKenna S.
      • Terry S.
      • et al.
      Blood eosinophils to direct corticosteroid treatment of exacerbations of chronic obstructive pulmonary disease: a randomized placebo-controlled trial.
      or ≥0.34*109 cells/L.
      • Vedel-Krogh S.
      • Nielsen S.F.
      • Lange P.
      • et al.
      Blood Eosinophils and exacerbations in chronic obstructive pulmonary disease. The Copenhagen general population study.
      However, other studies found that systemic glucocorticoids did not improve ICU mortality and shorten LOS,
      • Abroug F.
      • Ouanes-Besbes L.
      • Fkih-Hassen M.
      • et al.
      Prednisone in COPD exacerbation requiring ventilatory support: an open-label randomised evaluation.
      and the failure rate was almost 14.5–39%.
      • Crisafulli E.
      • Torres A.
      • Huerta A.
      • et al.
      Predicting in-hospital treatment failure (≤ 7 days) in patients with COPD exacerbation using antibiotics and systemic steroids.
      • Daniels J.M.
      • Snijders D.
      • de Graaff C.S.
      • et al.
      Antibiotics in addition to systemic corticosteroids for acute exacerbations of chronic obstructive pulmonary disease.
      • Soler N.
      • Esperatti M.
      • Ewig S.
      • et al.
      Sputum purulence- guided antibiotic use in hospitalised patients with exacerbations of COPD.
      Systemic glucocorticoids also increase the risk of side effects. EOS≥2% in severe AECOPD patients requires a lower systemic corticosteroid dose.
      • Serafino-Agrusa L.
      • Scichilone N.
      • Spatafora M.
      • et al.
      Blood eosinophils and treatment response in hospitalized exacerbations of chronic obstructive pulmonary disease: a case-control study.
      Inhaled corticosteroids alone can replace systemic corticosteroids in AECOPD treatment.
      • Sun X.
      • He Z.
      • Zhang J.
      • et al.
      Compare the efficacy of inhaled budesonide and systemic methylprednisolone on systemic inflammation of AECOPD.
      In our study, 84.04% of patients received aerosolized glucocorticoids, and 10.98% of patients received systemic glucocorticoids. The EOS group (≥ 2% or 3%) used lower doses of steroids. It could be that the low eosinophil group had more symptoms and a worse response to systemic corticosteroids. There is still a gap between the implementation of international guidelines and our treatment in the management of AECOPD. The mechanism between eosinophils and glucocorticoids is relatively clear.
      • Bloom J.W.
      • Chacko J.
      • Lohman I.C.
      • et al.
      Differential control of eosinophil survival by glucocorticoids.
      ,
      • Wechsler M.E.
      • Munitz A.
      • Ackerman S.J.
      • et al.
      Eosinophils in health and disease: a state-of-the-art review.
      There are few studies on AECOPD that address comorbidities or that follow-up with patients. We found that hypereosinophilia (≥ 2% or ≥3% or ≥100/µl) was associated with less cor pulmonale, respiratory failure, and atrial fibrillation, which was similar to previous studies, with EOS >2% or 340/ul.
      • Zhou W.
      • Tan J.
      The expression and the clinical significance of eosinophils, PCT and CRP in patients with acute exacerbation of chronic obstructive pulmonary disease complicated with pulmonary infection.
      ,
      • Wu H.X.
      • Zhuo K.Q.
      • Cheng D.Y.
      Peripheral blood eosinophil as a biomarker in outcomes of acute exacerbation of chronic obstructive pulmonary disease.
      ,
      • You Y.
      • Shi G.C.
      Blood eosinophils and clinical outcome of acute exacerbations of chronic obstructive pulmonary disease: a systematic review and meta-analysis.
      However, there is no relevant research to analyze the types of arrhythmias that were observed. Meanwhile, the prognostic indicators of AECOPD patients are being explored. Eosinopenia (<50/µL) was a powerful predictor of 12-month
      • MacDonald M.I.
      • Osadnik C.R.
      • Bulfin L.
      • et al.
      Low and high blood eosinophil counts as biomarkers in hospitalized acute exacerbations of COPD.
      or 18-month mortality.
      • Mao Y.
      • Qian Y.
      • Sun X.
      • et al.
      Eosinopenia predicting long-term mortality in hospitalized acute exacerbation of COPD patients with community-acquired pneumonia-a retrospective analysis.
      Patients with eosinophils larger than 150/µL had lower all-cause mortality.
      • Zhang Y.
      • Liang L.R.
      • Zhang S.
      Blood Eosinophilia and its stability in hospitalized COPD exacerbations are associated with lower risk of all-cause mortality.
      We followed up with patients after 3 years and found that low eosinophil levels (<100/µL or <2%) were associated with higher mortality. Eosinophils can be used as a prognostic indicator of mortality in AECOPD. In addition, this study found that chest CTs showed more obvious emphysema or bronchitis with EOS ≥100/µl. Researchers
      • Hastie A.T.
      • Martinez F.J.
      • Curtis J.L.
      • et al.
      SPIROMICS investigators. Association of sputum and blood eosinophil concentrations with clinical measures of COPD severity: an analysis of the SPIROMICS cohort.
      found that QCT measures for emphysema were higher in the sputum eosinophil group (≥1.25%). Smoking COPD subjects are more likely to develop emphysema and have fewer eosinophils.
      • Salvi S.S.
      • Brashier B.B.
      • Londhe J.
      • et al.
      Phenotypic comparison between smoking and non-smoking chronic obstructive pulmonary disease.
      Emphysema does not seem to be associated with eosinophils,
      • Makita H.
      • Nasuhara Y.
      • Nagai K.
      • et al.
      Hokkaido COPD Cohort Study Group. Characterisation of phenotypes based on severity of emphysema in chronic obstructive pulmonary disease.
      which was inconsistent with our result. Our study did not further distinguish the severity of chest CT results. Therefore, the relationship between EOS and chest CT phenotypes remains unclear, and further subgroup analyses are needed.
      Hypereosinophils seem to play a protective role in AECOPD infection, comorbidity and three-year prognosis. However, there are some limitations to the study. First, it was a retrospective design with a single-center study. There were missing data on some clinical features and the design lacked more performance metrics. Second, the sample was relatively small, and the proportion of patients who participated in follow-up was low. The predictive value of EOS by the ROC curve was unstable. Further validation in large samples is needed. Third, we did not further evaluate the severity of chest CT subtypes. Fourth, we did not collect EOS in AECOPD patients' sputum, so it is difficult to directly compare the consistency of the relationship between sputum specimens in the relevant literature and our study. We will continue to collect relevant clinical data on patients with AECOPD, further explore the relationship between EOS and AECOPD through multicenter research, and conduct subgroup analyses of CT phenotypes. In the future, we will elaborate on the mechanism of EOS in AECOPD in combination with basic research.

      Conclusions

      Our study compared common eosinophil classification methods for the first time and discussed the optimal threshold of AECOPD using an ROC curve. The group with relatively high EOS levels (≥2% or ≥100/µl) was associated with fewer complications, mild airflow restriction, a tendency toward noninfectious inflammation, and less 3-year mortality. In particular, the critical value of EOS of 100/µl seems to be more clinically relevant to AECOPD. Eosinophils can not only guide clinical treatment but also be used as an index to predict the clinical outcome and prognosis of AECOPD patients

      Ethics approval and consent to participate

      All participants agreed to participate in the experiment, and this study was approved and authorized by the Shanxi Bethune Hospital Affiliated to Shanxi Medical University. Ethical review approval number: SBQLL-2020-006.

      Declaration of Competing Interest

      The authors declare that there are no conflicts of commercial interest related to this paper.

      References

        • Halpin D.M.G.
        • Celli B.R.
        • Criner G.J.
        • et al.
        The GOLD Summit on chronic obstructive pulmonary disease in low- and middle-income countries.
        Int J Tuberc Lung Dis. 2019; 23: 1131-1141
      1. Global Initiative For Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease 2021 report. http://www.goldcopd.org

        • Zhou A.
        • Zhou Z.
        • Zhao Y.
        • et al.
        The recent advances of phenotypes in acute exacerbations of COPD.
        Int J Chron Obstruct Pulmon Dis. 2017; 12: 1009-1018
        • Prins H.J.
        • Duijkers R.
        • Lutter R.
        Blood eosinophilia as a marker of early and late treatment failure in severe acute exacerbations of COPD.
        Respir Med. 2017; 131: 118-124
        • Zysman M.
        • Deslee G.
        • Caillaud D.
        • et al.
        Relationship between blood eosinophils, clinical characteristics, and mortality in patients with COPD.
        Int J Chron Obstruct Pulmon Dis. 2017; 12: 1819-1824
        • Vedel-Krogh S.
        • Nielsen S.F.
        • Lange P.
        • et al.
        Blood eosinophils and exacerbations in chronic obstructive pulmonary disease. The copenhagen general population study.
        Am J Respir Crit Care Med. 2016; 193: 965-974
        • Xue J.
        • Cui Y.N.
        • Chen P.
        • et al.
        Blood eosinophils: a biomarker of response to glucocorticoids and increased readmissions in severe hospitalized exacerbations of COPD.
        Zhonghua Jie He He Hu Xi Za Zhi. 2019; 42 (Chinese): 426-431
        • Petite S.E.
        • Murphy J.A.
        Systemic corticosteroid and anti-biotic use in hospitalized patients with chronic obstruc- tive pulmonary disease exacerbation.
        Ann Pharmacother. 2019; 53: 144-150
        • Li Q.
        • Larivée P.
        • Courteau J.
        • et al.
        Greater eosinophil counts at first COPD hospitalization are associated with more readmissions and fewer deaths.
        Int J Chron Obstruct Pulmon Dis. 2019; 14: 331-341
        • Dai G.
        • Ran Y.
        • Wang J.
        • et al.
        Clinical differences between eosinophilic and noneosinophilic acute exacerbation of chronic obstructive pulmonary disease: a multicenter cross-sectional study.
        Mediators Inflamm. 2020; 20201059079
        • Zhou W.
        • Tan J.
        The expression and the clinical significance of eosinophils, PCT and CRP in patients with acute exacerbation of chronic obstructive pulmonary disease complicated with pulmonary infection.
        Am J Transl Res. 2021; 13: 3451-3458
        • Yu S.
        • Zhang J.
        • Fang Q.
        • et al.
        Blood eosinophil levels and prognosis of hospitalized patients with acute exacerbation of chronic obstructive pulmonary disease.
        Am J Med Sci. 2021; 362: 56-62
        • Lv M.Y.
        • Qiang L.X.
        • Li Z.H.
        • et al.
        The lower the eosinophils, the stronger the inflammatory response? The relationship of different levels of eosinophils with the degree of inflammation in acute exacerbation chronic obstructive pulmonary disease (AECOPD).
        J Thorac Dis. 2021; 13: 232-243
        • Wu C.W.
        • Lan C.C.
        • Hsieh P.C.
        • et al.
        Role of peripheral eosinophilia in acute exacerbation of chronic obstructive pulmonary disease.
        World J Clin Cases. 2020; 8: 2727-2737
        • Hastie A.T.
        • Martinez F.J.
        • Curtis J.L.
        • et al.
        SPIROMICS investigators. Association of sputum and blood eosinophil concentrations with clinical measures of COPD severity: an analysis of the SPIROMICS cohort.
        Lancet Respir Med. 2017; 5: 956-967
        • Müllerová H.
        • Hahn B.
        • Simard E.P.
        • et al.
        Exacerbations and health care resource use among patients with COPD in relation to blood eosinophil counts.
        Int J Chron Obstruct Pulmon Dis. 2019; 14: 683-692
        • Bafadhel M.
        • Davies L.
        • Calverley P.M.
        • et al.
        Blood eosinophil guided prednisolone therapy for exacerbations of COPD: a further analysis.
        Eur Respir J. 2014; 44: 789-791
        • Tanrıverdi H.
        • Örnek T.
        • Erboy F.
        • et al.
        Comparison of diagnostic values of procalcitonin, C-reactive protein and blood neutrophil/lymphocyte ratio levels in predicting bacterial infection in hospitalized patients with acute exacerbations of COPD.
        Wien Klin Wochenschr. 2015; 127: 756-763
        • Gao S.
        • Duan Y.
        • Chen J.
        • et al.
        Evaluation of blood markers at admission for predicting community acquired pneumonia in chronic obstructive pulmonary disease.
        COPD. 2021; 18: 557-566
        • MacDonald M.I.
        • Osadnik C.R.
        • Bulfin L.
        • et al.
        Low and high blood eosinophil counts as biomarkers in hospitalized acute exacerbations of COPD.
        Chest. 2019; 156: 92-100
        • Choi J.
        • Oh J.Y.
        • Lee Y.S.
        • et al.
        The association between blood eosinophil percent and bacterial infection in acute exacerbation of chronic obstructive pulmonary disease.
        Int J Chron Obstruct Pulmon Dis. 2019; 14: 953-959
        • Weller P.F.
        • Goetzel E.J.
        The human eosinophil: roles in hostdefense and tissue injury.
        Am J Pathol. 1980; 100: 791-820
        • Ravin K.A.
        • Loy M.
        The eosinophil in infection.
        Clin Rev Allergy Immunol. 2016; 50 (Apr): 214-227
        • Mukherjee M.
        • Lacy P.
        • Ueki S.
        Eosinophil extracellular traps and inflammatory pathologies-untangling the web!.
        Front Immunol. 2018; 9: 2763
        • LeMessurier K.S.
        • Samarasinghe AE.
        Eosinophils: nemeses of pulmonary pathogens?.
        Curr Allergy Asthma Rep. 2019; 19: 36
        • Kvarnhammar A.M.
        • Cardell LO.
        Pattern-recognition receptors in human eosinophils.
        Immunology. 2012; 136: 11-20
        • Bystrom J.
        • Amin K.
        • Bishop-Bailey D.
        Analysing the eosinophil cationic protein–a clue to the function of the eosinophil granulocyte.
        Respir Res. 2011; 12: 10
        • Persson T.
        • Andersson P.
        • Bodelsson M.
        • et al.
        Bactericidal activity of human eosinophilic granulocytes against Escherichia coli.
        Infect Immun. 2001; 69: 3591-3596
        • Yousefi S.
        • Gold J.A.
        • Andina N.
        Catapult-like release of mitochondrial DNA by eosinophils contributes to antibacterial defense.
        Nat Med. 2008; 14: 949-953
        • Kolsum U.
        • Donaldson G.C.
        • Singh R.
        • et al.
        Blood and sputum eosinophils in COPD; relationship with bacterial load.
        Respir Res. 2017; 18 (May 8): 88
        • Bass D.A.
        Behavior of eosinophil leukocytes in acute inflammation. II. Eosinophil dynamics during acute inflammation.
        J Clin Invest. 1975; 56: 870-879
        • Walters J.A.
        • Tan D.J.
        • White C.J.
        • et al.
        Systemic corticosteroids for acute exacerbations of chronic obstructive pulmonary disease.
        Cochrane Database Syst Rev. 2014; 9Cd001288
        • Kichloo A.
        • Aljadah M.
        • Vipparla N.
        • et al.
        Optimal glucocorticoid dose and the effects on mortality, length of stay, and readmission rates in patients diagnosed with acute exacerbation of chronic obstructive pulmonary disease (AECOPD).
        J Investig Med. 2019; 67: 1161-1164
        • Bafadhel M.
        • McKenna S.
        • Terry S.
        • et al.
        Blood eosinophils to direct corticosteroid treatment of exacerbations of chronic obstructive pulmonary disease: a randomized placebo-controlled trial.
        Am J Respir Crit Care Med. 2012; 186: 48-55
        • Vedel-Krogh S.
        • Nielsen S.F.
        • Lange P.
        • et al.
        Blood Eosinophils and exacerbations in chronic obstructive pulmonary disease. The Copenhagen general population study.
        Am J Respir Crit Care Med. 2016; 193: 965-974
        • Sivapalan P.
        • Lapperre T.S.
        • Janner J.
        • et al.
        Eosinophil-guided corticosteroid therapy in patients admitted to hospital with COPD exacerbation (CORTICO-COP): a multicentre, randomised, controlled, open-label, non-inferiority trial.
        Lancet Respir Med. 2019; 7 (Aug): 699-709
        • Crisafulli E.
        • Torres A.
        • Huerta A.
        • et al.
        Predicting in-hospital treatment failure (≤ 7 days) in patients with COPD exacerbation using antibiotics and systemic steroids.
        COPD. 2016; 13: 82-92
        • Daniels J.M.
        • Snijders D.
        • de Graaff C.S.
        • et al.
        Antibiotics in addition to systemic corticosteroids for acute exacerbations of chronic obstructive pulmonary disease.
        Am J Respir Crit Care Med. 2010; 181: 150-157
        • Soler N.
        • Esperatti M.
        • Ewig S.
        • et al.
        Sputum purulence- guided antibiotic use in hospitalised patients with exacerbations of COPD.
        Eur Respir J. 2012; 40: 1344-1353
        • Abroug F.
        • Ouanes-Besbes L.
        • Fkih-Hassen M.
        • et al.
        Prednisone in COPD exacerbation requiring ventilatory support: an open-label randomised evaluation.
        Eur Respir J. 2014; 43: 717-724
        • Serafino-Agrusa L.
        • Scichilone N.
        • Spatafora M.
        • et al.
        Blood eosinophils and treatment response in hospitalized exacerbations of chronic obstructive pulmonary disease: a case-control study.
        Pulm Pharmacol Ther. 2016; 37: 89-94
        • Sun X.
        • He Z.
        • Zhang J.
        • et al.
        Compare the efficacy of inhaled budesonide and systemic methylprednisolone on systemic inflammation of AECOPD.
        Pulm Pharmacol Ther. 2015; 31: 111-116
        • Bloom J.W.
        • Chacko J.
        • Lohman I.C.
        • et al.
        Differential control of eosinophil survival by glucocorticoids.
        Apoptosis. 2004; 9 (Jan): 97-104
        • Wechsler M.E.
        • Munitz A.
        • Ackerman S.J.
        • et al.
        Eosinophils in health and disease: a state-of-the-art review.
        Mayo Clin Proc. 2021; 96: 2694-2707
        • Wu H.X.
        • Zhuo K.Q.
        • Cheng D.Y.
        Peripheral blood eosinophil as a biomarker in outcomes of acute exacerbation of chronic obstructive pulmonary disease.
        Int J Chron Obstruct Pulmon Dis. 2019; 14: 3003-3015
        • You Y.
        • Shi G.C.
        Blood eosinophils and clinical outcome of acute exacerbations of chronic obstructive pulmonary disease: a systematic review and meta-analysis.
        Respiration. 2021; 100: 228-237
        • Mao Y.
        • Qian Y.
        • Sun X.
        • et al.
        Eosinopenia predicting long-term mortality in hospitalized acute exacerbation of COPD patients with community-acquired pneumonia-a retrospective analysis.
        Int J Chron Obstruct Pulmon Dis. 2021; 16: 3551-3559
        • Zhang Y.
        • Liang L.R.
        • Zhang S.
        Blood Eosinophilia and its stability in hospitalized COPD exacerbations are associated with lower risk of all-cause mortality.
        Int J Chron Obstruct Pulmon Dis. 2020; 15 (May 19): 1123-1134
        • Salvi S.S.
        • Brashier B.B.
        • Londhe J.
        • et al.
        Phenotypic comparison between smoking and non-smoking chronic obstructive pulmonary disease.
        Respir Res. 2020; 21: 50
        • Makita H.
        • Nasuhara Y.
        • Nagai K.
        • et al.
        Hokkaido COPD Cohort Study Group. Characterisation of phenotypes based on severity of emphysema in chronic obstructive pulmonary disease.
        Thorax. 2007; 62: 932-937