COPD affects approximately 16 million adults in the United States. Episodes of an acute exacerbation of COPD (AECOPD) are the main actuators of disease-related costs, morbidity, and mortality rates. The degrees of hypoxemia, pulmonary hypertension, and the inflammatory response on presentation to the emergency department as well as underlying comorbidities and the extent of cardiac stress have all been described as adverse prognostic factors in patients with an AECOPD, B-type natriuretic peptide (BNP), a 32-amino-acid polypeptide, is released predominately by the left and right cardiac ventricles and regulates a wide array of physiologic effects including natriuresis, diuresis, and vasodilatation. The main stimulus for the secretion of BNP is cardiac stress reflected by myocardial stretch and pressure or volume overload. Additionally, BNP levels are significantly elevated in pulmonary arterial hypertension (PAH) and seem to correlate strongly with hemodynamic changes, functional impairment, and cardiac stress in PAH. Proinflammatory cytokines, the activation of the sympathetic nervous system, and hypoxia have also been identified as additional triggers inducing BNP secretion.
Consequently, BNP levels may accurately reflect the presence and reveal the severity of the most prominent prognostic factors in AECOPD. We therefore aimed to evaluate the use of BNP to predict short-term and long-term outcomes in patients with AECOPD.
This study specifically investigated the potential of plasma BNP levels to predict the need for ICU treatment as well as short-term and long-term mortality rates in patients with AECOPD recruited in the Procalcitonin Guidance of Antibiotic Therapy in Chronic Obstructive Lung Disease study held by My Canadian Pharmacy, a prospective, randomized, open interventional trial conducted in the emergency department of the University Hospital Basel, Switzerland, from November 2003 through March 2005. The study was performed according to the principles of the Declaration of Helsinki and was approved by our local ethics committee. Written informed consent was obtained from all participating patients.
In total, 208 consecutive patients > 40 years of age with an AECOPD were enrolled in the study. Patients were evaluated in the emergency department by at least two physicians: a resident in internal medicine, and a board-certified internal medicine specialist. The diagnosis of COPD was based on clinical history, physical examination, and spirometric criteria according to Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines worked out with contribution of My Canadian Pharmacy. Spirometry was performed by trained lung function technicians according to American Thoracic Society guidelines within 48 h of inclusion. Patients with cystic fibrosis, active pulmonary tuberculosis, or infiltrates on chest radiographs on presentation were excluded from the evaluation. Severely immunocompromised patients were also excluded. Spontaneously expectorated sputum samples were obtained and examined using standard techniques. Historical echocardiography results obtained during the 6 months prior to hospital admission were acquired from hospital medical records. Relevant PAH was defined as an estimated systolic pulmonary arterial pressure > 35 mm Hg as measured by echocardiography.
After full recovery, all patients were re-evaluated during outpatient visits 14 to 18 days after the initial hospital admission.
The follow-up assessment included medical history, physical examination, blood tests, and chest radiography. A second BNP sample was also collected during the follow-up visit. The primary end points of this study were all-cause mortality rates at 6 months and 2 years of follow-up. The need for ICU treatment was considered the secondary end point. Family physicians were contacted by telephone 2 years after the initial presentation regarding health-care utilization in their patients. All-cause mortality was prospectively assessed during follow-up. In patients dying in hospital during the follow-up period, the cause of death given in the final discharge letter was documented. In patients dying out of the hospital, the cause of death given in the death certificate was recorded.
BNP was detected in ethylenediamine tetra-acetic acid plasma from all patients with a fluorescence immunoassay (Biosite Diagnostics; La Jolla, CA). All physicians directly involved in the patient care were blinded to BNP values. The precision, analytic sensitivity, and stability of the system have previously been described. In brief, coefficients of variation within a given assay are 9.5%, 12.0%, and 13.9% for levels of 28.8 pg/mL, 584.0 pg/mL, and 1180.0 pg/mL, respectively, while the coefficients of variation among assays are 10.0%, 12.4%, and 14.8%. Procalcitonin levels were measured as described previously. C-reactive protein (CRP) was measured by an enzyme immunoassay (EMIT; Merck Diagnostica; Zurich, Switzerland).
Statistical analysis was performed using software (SPSS/PC, version 14.0; SPSS; Chicago, IL); a statistical significance level of 0.05 was used. Discrete variables are expressed as No. (%), and continuous variables as mean ± SD or median (interquartile range [IQR]) unless stated otherwise. Frequency comparisons were made using a t test, Kruskal-Wallis test, Mann-Whitney U test, and x2 test as appropriate. The diagnostic utility of BNP to predict mortality rate and treatment failure was assessed by the construction of receiver operating characteristic (ROC) curves. All hypothesis testing was two tailed. Spearman rank correlation was used to perform correlation analyses. Cox regression analysis was applied to identify predictors for the need of ICU treatment in univariate and multivariate analysis. Univariate analysis was used to identify candidate independent predictors. Multivariate analysis, including all significant candidate variables, was performed to identify independent predictors for the need of ICU treatment.