Proteome profiling reveals novel biomarkers to identify complicated parapneumonic effusions

Abstract

Patients with pneumonia and parapneumonic effusion (PPE) have elevated mortality and a poor prognosis. The aim of this study was to discover novel biomarkers to help distinguish between uncomplicated PPE (UPPE) and complicated PPE (CPPE). Using an iTRAQ-based quantitative proteomics, we identified 766 proteins in pleural effusions from PPE patients. In total, 45 of these proteins were quantified as upregulated proteins in CPPE. Four novel upregulated candidates (BPI, NGAL, AZU1, and calprotectin) were selected and further verified using enzyme-linked immunosorbent assays (ELISAs) on 220 patients with pleural effusions due to different causes. The pleural fluid levels of BPI, NGAL, AZU1, and calprotectin were significantly elevated in patients with CPPE. Among these four biomarkers, BPI had the best diagnostic value for CPPE, with an AUC value of 0.966, a sensitivity of 97%, and a specificity of 91.4%. A logistic regression analysis demonstrated a strong association between BPI levels > 10 ng/ml and CPPE (odds ratio = 341.3). Furthermore, the combination of pleural fluid BPI levels with LDH levels improved the sensitivity and specificity to 100% and 91.4%, respectively. Thus, our findings provided a comprehensive effusion proteome data set for PPE biomarker discovery and revealed novel biomarkers for the diagnosis of CPPE.

Figure 1

Schematic diagram of the experimental design for identifying novel biomarkers in parapneumonic effusions. Pleural effusions were collected from patients with different causes, including transudates, other exudates, malignant, uncomplicated parapneumonic effusion (UPPE) and complicated parapneumonic effusion (CPPE). Four samples from each type of UPPE or CPPE were pooled, processed, and labeled with iTRAQ reagents, following by 2D LC-MS/MS analysis. Peptide and protein were identified and quantified using MASCOT software. Selected proteins were uploaded into the DAVID database and MetaCore software to evaluate possible associations with known pathways and biological processes. The potential newer biomarkers were validated in an extended patients group by ELISA assays. Statistical analysis was performed to determine the efficacy of biomarkers in discriminating between UPPE and CPPE patients.

Figure 2

Box plots of the concentrations of four biomarkers in five types of pleural effusions. The pleural effusion levels of (A) bactericidal permeability-increasing protein (BPI), (B) neutrophil gelatinase-associated lipocalin (NGAL), (C) azurocidin (AZU1), and (D) calprotectin from patients with UPPE, CPPE, transudates, other exudates, and malignant were determined by sandwich ELISA. PPE refer to UPPE and CPPE. non-PPE refer to transudates, other exudates, and malignant. Horizontal lines represent mean values. *P < 0.0001, indicates statistical significance using nonparametric Mann-Whitney U test. PPE: parapneumonic effusion; UPPE: uncomplicated parapneumonic effusion; CPPE: complicated parapneumonic effusion.

Figure 3

The diagnostic power of the pleural fluid BPI levels combined with LDH levels for identifying complicated parapneumonic effusions. A decision tree for CPPE and UPPE discrimination in the first cohort (A) and in the validation cohort (B). In the first cohort, the sensitivity and specificity of the combination of the pleural fluid BPI levels and LDH levels were 100% and 91.4%, respectively. In the second cohort, the sensitivity and specificity of the combination of the pleural fluid BPI levels and LDH levels were 100% and 92.3%, respectively.