Abstract
BACKGROUND: Differentiating pleural effusion remains a challenge in clinical practice. Serum and effusion chemistries help pulmonologists assess the risk of underlying causes and select further diagnostic procedures. Amino acids (AAs) in body fluids are promising tools for estimating the probability of underlying causes. This proof-of-concept study aimed to investigate the accessibility of serum and pleural fluid AAs in distinguishing the four major etiologies of pleural effusions: malignancy, heart failure, tuberculosis, and pneumonia. METHODS: This study prospectively enrolled 153 patients with pleural effusion and unknown causes on admission. The concentrations of 11 AAs in both pleural fluid and serum were measured using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Logistic regression was used to investigate the potential of the AAs panel. The receiver operating characteristic (ROC) curve was used to assess the performance of the AAs panel. Decision curve analysis (DCA) was used to evaluate the net benefit of the panel for patients with pleural effusion. RESULTS: Our study included pleural effusions caused by heart failure (n = 23, 15%), malignancy (n = 66, 43%), pneumonia (n = 32, 21%), tuberculosis pleurisy (n = 20, 13%), and other causes (n = 8, 5%). Significant differences were observed among various AAs, including alanine and phenylalanine. The areas under the curves (AUCs) for the 11-AA panel in serum to identify four different causes were all ≥0.75. Among them, the AUC for heart failure-related pleural effusion was 0.90 (95% CI: 0.83-0.97). Furthermore, the decision curves for the AAs panels were consistently above the reference lines. CONCLUSIONS: This proof-of-concept study finds that AAs detection in pleural effusion and serum is possible and feasible, but currently has a minor to moderate added diagnostic role.