Chronic obstructive pulmonary disease: exacerbations and mortality : Prognostic value of biomarkers and comorbidities

Sammanfattning: Background: Chronic obstructive pulmonary disease (COPD) is a major cause of morbidity and mortality. COPD is associated with systemic inflammation, and comorbidities are common. A characteristic feature is acute exacerbations (AECOPDs), i.e., episodes of worsening symptoms. AECOPDs are associated with increased mortality.Aim: To find prognostic risk factors for COPD mortality and AECOPDs, focusing on comorbidities and inflammatory biomarkers.Methods: In Paper I, associations between comorbidities, pharmacological treatment, and mortality were analysed in a real-world cohort of almost 18,000 primary care COPD patients. Data from medical records and national registers were analysed in Cox proportional hazards regressions.Papers II–IV were based on the Tools Identifying Exacerbations (TIE) cohort study of 572 COPD patients recruited from primary and secondary care in three Swedish regions. Participants were invited to three yearly visits, including phlebotomy, spirometry, and health questionnaires.In Paper II, the ability of blood neutrophil-to-lymphocyte ratio (NLR) and eosinophils (B-Eos) to predict AECOPDs was analysed with mixed-effects logistic regressions.In Paper III, the ability of C-reactive protein (CRP), fibrinogen, blood leukocytes (B-Leu), and four blood cell indices to predict AECOPDs was analysed with ordinal logistic regressions.In Paper IV, an algorithm for clinical phenotyping previously developed to predict mortality was studied. The algorithm’s ability to predict AECOPDs and mortality was analysed with Cox proportional hazards regressions; additionally, the identified phenotypes were analysed concerning differences in blood-based inflammatory biomarkers.Results: Several comorbidities, including heart diseases, were associated with increased mortality risk. Some pharmacological treatments were associated with increased or decreased mortality risk (Paper I). NLR, B-Eos, CRP, fibrinogen, and B-Leu (Papers II–III) predicted AECOPDs after adjustment for confounders, whereas other blood cell indices were of limited value (Paper III). The clinical phenotyping algorithm predicted AECOPDs and mortality, and the phenotypes had different patterns of inflammatory biomarkers (Paper IV).Conclusions: Comorbidities, particularly heart diseases, are substantial risk factors for mortality in COPD and should be an integral part of management of COPD patients. NLR, B-Eos, CRP, fibrinogen, and B-Leu are independent predictors of AECOPDs and should be further investigated as parts of, e.g., risk prediction tools. A previously developed algorithm for clinical phenotyping predicts mortality and AECOPDs.