Eosinophilic pleural effusion: incidence, etiology and prognostic significance.

INTRODUCTION Eosinophilic pleural effusion (EPE) has been associated with less risk for malignancy with a potential causal relationship with the presence of air and/or blood in the pleural space. However, these theories have fallen by the wayside in the light of recent publications. OBJECTIVES To determine the incidence and etiology of EPE and to observe whether the eosinophils in the pleural liquid (PL) increase in successive thoracocenteses. PATIENTS AND METHODS We analyzed 730 PL samples from 605 patients hospitalized between January 2004 and December 2010. RESULTS We identified 55 samples with EPE from 50 patients (8.3%). The most frequent etiologies of EPE were: unknown (36%) and neoplasm (30%). There were no significant differences in the incidence of neoplasms between the non-eosinophilic pleural effusions (non-EPE) (25.9%) and the EPE (30%) (p=0.533). One hundred patients (16.5%) underwent a second thoracocentesis. Out of the 9 who had EPE in the first, 6 maintained EPE in the second. Out of the 91 with non-EPE in the first thoracocentesis, 8 (8.8%) had EPE in the repeat thoracocentesis. The percentage of eosinophils did not increase in the successive thoracocenteses (p=0.427). In the EPE, a significant correlation was found between the number of hematites and eosinophils in the PL (r=0.563; p=0.000). CONCLUSIONS An EPE cannot be considered an indicator of benignancy, therefore it should be studied as any other pleural effusion. The number of eosinophils does not seem to increase with the of repetition of thoracocentesis and, lastly, the presence of blood in the PL could explain the existence of EPE.