Introduction In moderate exacerbations of COPD, patients with the eosinophilic phenotype (>2% of the total leukocyte count) have better outcomes with prednisolone. However, it remains unclear whether patients with severe exacerbations displaying the eosinophilic phenotype have accentuated recovery following corticosteroid therapy compared to non-eosinophilic COPD exacerbations.
Aim Measure the incidence of eosinophilic and non-eosinophilic severe exacerbations of COPD, from a large prospective enhanced recovery multi-centre randomised control trial and investigate severity and recovery between these groups.
Methods COPD patients entering the programme delivered immediately on hospitalisation for an acute exacerbation of chronic respiratory disease to improve long term health outcomes (clinical trial registration ISRCTN05557928) were analysed using admission details, length of stay and proceeding exacerbation history. All patients were dichotomised into eosinophilic (>200 x106 cells/mL and/or >2% of the total leukocyte count) and non-eosinophilic. CRP was measured on admission.
Results There were 243 COPD patients (117 males) identified. The mean (range) age was 71 years (45–93) and the majority of patients (55%) had been hospitalised for an exacerbation of COPD in the previous 12 months. Of all exacerbations, the inpatient mortality rate was 3% (median time to death 12 days, range 9–16) and approximately 90% received both antibiotic and corticosteroid treatment. The incidence of an eosinophilic exacerbation was 25% (median absolute eosinophil count 100 x106cells/ml; range 10 to 1500). In patients with eosinophilic exacerbations compared to non-eosinophilic exacerbations the median (IQR) CRP concentration was significantly lower (12mg/L (5–47) vs. 55mg/L (18–139), p < 0.001); and the presence of an elevated eosinophil count and elevated CRP (>200 x106 eosinophils/mL and CRP>50mg/L) occurred in only 5% of all exacerbations. The length of stay was significantly shorter in patients with eosinophilic exacerbations compared to non-eosinophilic exacerbations (mean (range) 5.0 (1–19) vs. 6.5 (1–33), p = 0.015). The severity of the index exacerbation or the rate of exacerbations or hospitalisations in the following 12 months was not statistically significant between groups.
Conclusions In severe hospitalised exacerbations of COPD, a proportion have an associated eosinophilic phenotype. These exacerbations are usually not associated with an elevated CRP. Eosinophilic exacerbations have better responses to oral corticosteroids with shortened length of stay.