Elevated Peripheral Blood Eosinophils (PBE) During Acute Exacerbation

16 Objectives: An elevated peripheral blood eosinophil (PBE) count during acute exacerbation 17 of chronic obstructive pulmonary disease (AECOPD) is a potential predictor of treatment 18 responsiveness and future exacerbation risk. This study aimed to evaluate the prevalence and 19 clinical significance of elevated PBE counts in hospitalized patients with AECOPD in Oman. 20 Methods: This single-center retrospective study included all patients with AECOPD who 21 were admitted to Sultan Qaboos University Hospital between January 2017 and July 2019. 22 The patients were classified as having eosinophilic or noneosinophilic AECOPD based on 23 blood eosinophil counts. An elevated eosinophil count was defined as a blood eosinophil 24 count > 0.3 × 10 9 cells/L on admission. The length of hospital stay, use of oral and inhaled 25 steroids, number of readmissions in a year, and use of mechanical ventilation on admission 26 were compared between the eosinophilic and non-eosinophilic AECOPD groups. Results: Of 27 the 102 patients included in the study, 42.2% had eosinophilic AECOPD. The eosinophilic 28 AECOPD group had a reduced length of hospital stay (P = 0.02) but an increased risk of 29 readmission in a year (P = 0.04). Most patients in both the groups were treated with inhaled 30 and oral steroids. The need for mechanical ventilation did not differ between the groups. 31 Conclusion: Eosinophilia is highly prevalent in patients with AECOPD and is associated 32 with a reduced length of hospital stay but an increased risk of readmission in a year. It can be 33 used as a surrogate marker to predict the health outcomes of patients with AECOPD and 34 select treatment options. 35

The patients were classified as having eosinophilic or noneosinophilic AECOPD based on 23 blood eosinophil counts. An elevated eosinophil count was defined as a blood eosinophil 24 count > 0.3 × 10 9 cells/L on admission. The length of hospital stay, use of oral and inhaled 25 steroids, number of readmissions in a year, and use of mechanical ventilation on admission 26 were compared between the eosinophilic and non-eosinophilic AECOPD groups. Results: Of 27 the 102 patients included in the study, 42.2% had eosinophilic AECOPD. The eosinophilic 28 AECOPD group had a reduced length of hospital stay (P = 0.02) but an increased risk of 29 readmission in a year (P = 0.04). Most patients in both the groups were treated with inhaled 30 and oral steroids. The need for mechanical ventilation did not differ between the groups. 31 Conclusion: Eosinophilia is highly prevalent in patients with AECOPD and is associated 32 Chronic obstructive pulmonary disease (COPD) is characterized by irreversible airflow 51 limitation. Patients with COPD suffer from exacerbation, reduced quality of life, and 52 increased morbidity and mortality. 1-3 53 54 Acute exacerbation of chronic obstructive pulmonary disease (AECOPD) is mainly 55 associated with neutrophilic inflammation; however, predominant eosinophilic airway 56 inflammation has been reported in a subset of patients with COPD. 4-6 Up to 40% of patients 57 with COPD have an eosinophilic phenotype of COPD, defined as peripheral blood eosinophil 58 (PBE) counts ≥ 2%. 4 Several studies have shown that patients with elevated PBE counts are 59 at an increased risk of frequent exacerbations but show a good response to steroid therapy. 1, 4 60 Evidence suggests that circulating eosinophils can be recruited to the lungs and can increase 61 inflammation by the actions of cytokines, immunoregulatory cells, and other 62 proinflammatory mediators. 7 Accordingly, PBE count has been suggested to be useful as a 63 surrogate marker to direct the use of oral steroid therapy in patients with AECOPD and as a 64 predictor of future exacerbation and disease stability. 6 The data assessing the role of PBE on 65 mortality outcome are inconsistent. 8,9 Overall, the role of PBEs in the clinical manifestation 66 of COPD remains highly debatable. 10, 11 No prior study has been conducted in the Middle 67 East to assess the prevalence and clinical significance of eosinophilia during AECOPD. The 68 present study aimed to evaluate the prevalence and clinical significance of elevated PBE 69 counts in hospitalized patients with AECOPD. 70

71
Methods 72 This retrospective cohort study was conducted at Sultan Qaboos University Hospital (SQUH), 73 a 500-bed multi-specialty tertiary hospital that provides health care for Muscat and Al-74 Batinah governorates' residents. It is also considered a major referral center for many 75 specialties that provide high-quality care for patients referred from the entire country of 76 Oman. 12 77 78 All patients with AECOPD admitted to SQUH between January 2017, and July 2019 were 79 included in the study. We have used the Global Initiative for Chronic Obstructive Lung 80 Disease (GOLD) criteria to ascertain COPD diagnosis for all included patients. In addition, 81 we have included the most recent lung function test before the index hospitalization when 82 there is more than one lung function test. Data were collected from electronic patient records 83 using a standardized electronic data collection sheet. The patients were classified as having 84 eosinophilic or non-eosinophilic AECOPD based on their blood eosinophil counts. An 85 elevated eosinophil count was defined as a blood eosinophil count greater than 0.3 × 10 9 86 cells/L at the time of admission. In addition, the length of hospital stay, use of oral and 87 inhaled steroids, number of readmissions in a year, and use of mechanical ventilation on 88 admission were compared between the eosinophilic and non-eosinophilic AECOPD groups. 89 90 Categorical variables were reported as numbers and percentages, while continuous variables 91 were expressed as means ± standard deviations (SDs) for normally distributed data and as 92 medians and interquartile ranges (IQRs) for non-normally distributed data. Continuous 93 variables between the groups were compared using Student's t-test for normally distributed 94 data and Wilcoxon's rank-sum test for non-normally distributed data. Fisher's exact test was 95 used to assess the association between categorical variables (given the small sample size during the study period. Twenty-three patients were found to be asthmatic, while three 105 patients were lost to follow-up. Thus, 102 patients were included in the study. The mean age 106 of the patients was 72.9 ± 10.9 years, and 79.4% of the patients were male. Approximately 107 93.1% of the patients had a history of smoking (current or ex-smoker). Both the groups had a 108 severely reduced forced expiratory volume in one second (FEV1) (34.8 ± 17.8 %). In total, 109 42.2% of the patients had eosinophilic AECOPD. Patients with non-eosinophilic AECOPD 110 stayed in the hospital for a longer duration than those with eosinophilic AECOPD (p = 0.02). 111 Patients with eosinophilic AECOPD had a significantly higher number of readmissions in a 112 year than those with non-eosinophilic AECOPD (P = 0.04). Most patients in both groups 113 received systemic steroids (92.2%) and were on inhaled steroids (78.4%) before admission. 114 There was no significant difference in the need for mechanical ventilation between the groups 115 (P = 0.32). Moreover, the eosinophil count before discharge did not differ significantly 116 between the groups (P = 0.33). 117 118

Discussion 119
The present study is the first to assess the prevalence and clinical significance of eosinophilia 120 in hospitalized COPD patients in the Middle East, where most patients are of Arabic 121 ethnicity. We found that patients with eosinophilic AECOPD have a reduced length of 122 hospital stay but are at an increased risk of readmission in a year. 123

124
The prevalence of eosinophilic AECOPD ranged from 10% to 37% in previous studies. 13-15 125 These differences in the prevalence of eosinophilia during AECOPD could be explained by 126 the difference in patients' ethnicity, use of corticosteroids before admission, and difference in 127 cut-off values used to define eosinophilia. 1, 16-18 The most commonly used cut-off value to 128 define eosinophilic COPD is 2%, which corresponds to 150 cells/μL. However, the absolute 129 eosinophil count might be more accurate because the white cell count can differ significantly 130 for various reasons. 1 In the present study, we used a cut-off value greater than 300 cells/μL to 131 define eosinophilic AECOPD, which has been validated in previous studies. 1, 7, 16, 19, 20 The 132 prevalence of eosinophilic AECOPD in our cohort was 42.2%, which is higher than most of 133 the previously reported values. This higher prevalence of eosinophilic AECOPD could be 134 related to the high prevalence of smoking in our cohort. 135

136
In patients with COPD and under certain circumstances, the PBEs are recruited to the lungs 137 prompting cascades of inflammatory responses, including secretion of chemokines, 138 cytokines, and cytotoxic granular products. 21 Most patients in both the groups were treated 139 with inhaled and oral steroids; however, patients with elevated PBE counts showed a better 140 response, as evidenced by a reduced length of hospital stay (p = 0.02), which could be 141 explained by the anti-inflammatory role of corticosteroids on patients with eosinophilic 142 AECOPD. Our finding of reduced length of hospital stay of eosinophilic AECOPD is similar 143 to that of previous studies. 16, 22 144 145 About 40% of patients with eosinophilic AECOPD required mechanical ventilation compared 146 to 50.9% of patients with non-eosinophilic AECOPD, which may be explained by the poor 147 response of non-eosinophilic AECOPD to corticosteroids. Also, previous studies 148 demonstrated that non-eosinophilic AECOPD is strongly associated with infections and 149 worse outcomes, which may be explained by the higher need for mechanical ventilation. 21 150 151 There were no significant differences in age, sex, FEV1, and smoking status between the 152 groups ( Table 1). In addition, before discharge from the hospital, the eosinophil count did not 153 differ significantly between eosinophilic and non-es eosinophilic AECOPD, which could be 154 explained by a high percentage of patients who were treated with steroids in both groups. 155 156 This finding may provide insight into using oral and inhaled steroids in patients with 157 AECOPD based on the eosinophil count to avoid risks associated with the indiscriminate use 158 of steroids in such patients. 23 There was no difference in the need for mechanical ventilation 159 between the groups (p = 0.32). In the present study, the mortality outcome was not assessed 160 because of the small sample size; however, previous studies have suggested that eosinophilic 161 AECOPD is associated with a lower inpatient mortality rate, but the data are conflicting. There are several limitations to the present study. First, it was a single-center retrospective 172 study. Second, the inpatient mortality rate in the groups could not be assessed because of the 173 small sample size. Third, the study included hospitalized patients with AECOPD; however, it 174 did not include patients with mild and moderate exacerbations who were managed in the 175 outpatient setting. Moreover, potential cofounders, including heart failure, ischemic heart 176 disease, and hypertension, were not considered.  0.02 4 (6-3) 5 (4-7) 4 (3-7) Length of hospital stay (days), median (IQR) 0.33 0.2 (0.0-0.5) 0.1 (0-0.3) 0.1 (0.0-0.3) Eosinophil count on discharge (× 10 9 /L) median (IQR) 0.04 1 (0-3) 0 (0-1) 0 (0-2) Readmission in a year (n),median (IQR) *forced expiratory volume in one second. 279 280