EPEs account for 5% to 16% of the exudative pleural effusions, and IEPE is an important cause of EPEs which can almost always be treated medically [4, 5]. However, limited well-documented cases have been described [7–11]. In this study, we tried to analyse clinical characteristics of IEPE and to make diagnostic procedure clear.
Archontogeorgis K et al.  firstly investigated the diagnostic approach in 10 patients with IEPE, but the clinical characteristics of IEPE was not assessed. We tried to describe the clinical features of IEPE in spite of 11 cases prospectively collected. Shortness of breath is one of major symptoms; Moreover, fever, productive cough, fatigue, lymphadenopathy, splenomegaly and ascites often exist [4, 5]. Most patients had bilateral pleural effusion [7–9, 11], but some had just unilateral effusion . We found all the 11 patients developed respiratory symptoms which were similar to IEPE symptoms. Among these cases, 7 had bilateral effusion and 4 had lung involvement. Previous reports [7–11] showed eosinophils seemed always significantly elevated, reaching up to 3.5×109/L. But the number of eosinophils were normal or slightly elevated. Due to its nonspecific characteristics and laboratory tests, some patients were initially misdiagnosed.
Current investigations on pleural effusions emphasise the use of a diagnostic algorithm or recommend the use of a stepwise approach [14–17]. Thoracocentesis was performed to ascertain the nature of pleural effusion and to differentiate it from other conditions. Consistent with the results of a previous study , pleural effusions in the 11 patients were exudative according to Light’s criteria. Pleural CEA, ADA and LDH were nonspecific in these 11 cases. Prospective studies with more patients need be conducted to evaluate the diagnostic value of such effusion parameters.
A meta-analysis concluded that the most common cause of EPEs is malignancy (26%) . Therefore, malignancy must be excluded for IEPE diagnosis. CEA, as a tumour marker, plays a role in MPE differentiation. Pleural CEA is often positive in suspected patients with malignancy [18, 19]. In this study, CEA in the serum and pleural effusion were at normal level (<5 mg/mL). Pleura biopsies seem to be mandatory when malignancy is excluded. Archontogeorgis K et al. emphasised that pleuroscopy is mandatory in diagnosing IEPE. In this study, pleural samples were collected by using combined ultrasound-guided cutting needle biopsy and standard pleural biopsy, without thoracoscopic assessment. Enough and pleura biopsies were obtained, and the sensitivity and accuracy reached up to 88.6% and 93.8%, respectively , which were close to thoracoscopy examination . Eosinophilic infiltration was found in 6 cases. However, there were no evidence of tuberculosis or malignancy in these patients.
Except for malignancy, the causes of EPE vary and complicated, including parapneumonic effusions, pleural air/blood, tuberculosis, transudate, and collagen vascular disease [4, 5]. Therefore, a thorough work-up is required to rule out known and obvious causes of EPE. We did not identify any related medications, autoimmune disease and chest trauma in these cases. In our previous study , we confirmed that in patients with unexplained pleural effusion, parasite-specific IgG antibody detection had to be done when pleural fluid testing showed EPE. Physicians should consider a diagnosis of PPI should be considered when parasite-specific IgG antibody is positive. Based on this, we excluded PPI diagnosis.
Hypereosinophilic syndrome (HES) was redefined in 2010 as more than 1500/mm3 eosinophils without a discernible secondary cause (eg, HIV infection, parasite or worm infection, allergic diseases, drug allergies, and nonhematologic malignancies) . Idiopathic HES (IHES) sometimes presented with EPE , but the causality of IHES and EPE was not reached. Although the absolute eosinophil count was 2.16×109/L in case 2, IHES diagnosis was excluded after comprehensive haematological determinations. Echocardiography displayed pericardial effusion in a case, while lung involvement was shown in 4 cases. So, we think PET/CT or transbronchial lung biopsy to verify the diagnosis of chronic eosinophilic pneumonia in lung involvement needs confirmation. When EPE has with no apparent etiology, the diagnosis of IEPE should be considered.
This study had limitations. Because this is a single-center experience with a small number of patients was quite small, the characteristics IEPE were not well defined. A large, multicentre, prospective study is needed for validation of the findings.