Abstract
Chronic Obstructive Pulmonary Disease (COPD) is the third leading cause of death worldwide. COPD is frequently punctuated by acute exacerbations that are precipitated primarily by infections, which increase both morbidity and mortality and inflates healthcare costs. Despite the significance of exacerbations, little understanding of immune function in COPD exacerbations exists. Natural killer (NK) cells are important effectors of innate and adaptive immune responses to pathogens and NK cell function is altered in smokers and COPD. Using high-dimensional flow cytometry, we phenotyped peripheral blood NK cells from never smokers, smokers, and COPD patients and employed a non-supervised clustering algorithm to define and detect changes in NK cell populations. We identified greater than 1,000 unique NK cell subpopulations across patient groups and describe 13 altered NK populations in patients who experienced prior exacerbations. Based upon cluster sizes and associated fluorescence data, we generated a logistic regression model to predict patients with a history of exacerbations with high sensitivity and specificity. Moreover, highly enriched NK cell subpopulations implicated in the regression model exhibited enhanced effector functions as defined by in vitro cytotoxicity assays. These novel data reflect the effects of smoking and disease on peripheral blood NK cell phenotypes, provide insight into the potential immune pathophysiology of COPD exacerbations, and indicate that NK cell phenotyping may be a useful and biologically relevant marker to predict COPD exacerbations.