INTRODUCTION — Pleural fluid eosinophilia (PFE, also called eosinophilic pleural effusion) is defined as pleural fluid with a nucleated cell count containing more than 10 percent eosinophils [1-4]. It is estimated that approximately 10 percent of exudative pleural effusions are eosinophilic [2,3,5].
The pathogenesis, etiology, and evaluation of eosinophilic pleural effusions will be reviewed here. The general analysis of pleural fluid and evaluation of systemic eosinophilia and pulmonary parenchymal eosinophilia are discussed separately. (See "Diagnostic evaluation of a pleural effusion in adults: Initial testing" and "Approach to the patient with unexplained eosinophilia" and "Overview of pulmonary eosinophilia".)
PATHOGENESIS — The development of pleural fluid eosinophilia (PFE) requires the recruitment of bone marrow-derived eosinophils, since eosinophils are not part of the normal cellular milieu of lung and pleural tissue [6]. The mechanisms that lead to the recruitment of eosinophils to the pleural space have not been clearly elucidated.
The time course of the development of pleural fluid and peripheral blood eosinophilia in response to air or blood in the pleural space suggests that the specific mechanisms that trigger eosinophil accumulation may vary [6]. PFE develops within hours after a spontaneous pneumothorax compared to within 10 to 14 days after the onset of a traumatic or hemorrhagic pleural effusion [6]. In addition, pleural trauma and hemothorax, but not pneumothorax, are often associated with peripheral blood eosinophilia that develops days after PFE.
One hypothesis for the development of PFE is that stimulation of pleural mesothelial cells by nonspecific injury (eg, air, blood, or trauma) leads to the production of cytokines, chemokines, and adhesion molecules that orchestrate eosinophil recruitment [7,8]. In general, the recruitment of eosinophils into tissues is dependent upon mediators, such as eotaxins (eotaxin-1 [CCL11], eotaxin-2 [CCL24], and eotaxin-3 [CCL26]), interleukin (IL)-3, IL-5, granulocyte/macrophage stimulating factor (GM-CSF), and RANTES (Regulated upon Activation, Normal T-cell Expressed and Secreted) [9-11].
These agents cause eosinophil proliferation and maturation in the bone marrow, migration into the circulation, adhesion to the endothelium, and migration across endothelial barriers into tissue. In addition, there must be an interaction between the endothelial cells and circulating eosinophils for migration of eosinophils into the pleural space. Vascular cell adhesion molecule (VCAM-1) and eosinophil very late activation antigen (VLA-4) mediate this important eosinophil-endothelial cell interaction. Soluble VCAM-1 is significantly increased in eosinophilic pleural effusions compared to noneosinophilic effusions and probably contributes to their pathogenesis [12,13].
Eosinophil survival in the pleural space is also thought to be mediated by cytokines, such as IL-3, IL-5, and GM-CSF [13,14]. Eosinophil apoptosis can be inhibited for up to two weeks when eosinophils are incubated with these cytokines; in contrast, eosinophils undergo apoptosis within two days when these cytokines are absent [15].
ETIOLOGY — The spectrum of diseases associated with pleural fluid eosinophilia (PFE) mirrors that of pleural effusions in general and is strongly affected by the diseases prevalent in the population [2,3,16,17]. Pleural irritation or trauma (eg, pneumothorax, hemothorax, thoracic surgery) and malignancy are common causes, but several other processes should be considered in the differential diagnosis (table 1). Even after careful evaluation, the etiology of PFE may remain unknown, especially when compared to non-eosinophilic effusions. As an example, in a meta-analysis of 687 cases of PFE, the most common causes were malignancy (26 percent) followed by idiopathic (25 percent) and parapneumonic (13 percent). The likelihood of an idiopathic effusion was significantly higher in eosinophilic pleural effusion than in non-eosinophilic pleural effusion [17]. (See 'Idiopathic' below.)
Pleural trauma — Pleural irritation or trauma in the form of pneumothorax, hemothorax, thoracotomy, thoracoscopy, or repeated thoracentesis has been associated with pleural fluid eosinophilia in numerous case reports and case series [2,16,18,19]. As examples:
●Air in the pleural space — In many case series, spontaneous pneumothorax is the most common cause of PFE [16,18]. An intense eosinophilic pleuritis follows air entry into the pleural space within hours. One study has reported a time-related rise in percentage of pleural effusion in patients with spontaneous pneumothorax [20].
●Blood in the pleural space — In contrast to the rapid development of PFE after spontaneous pneumothorax, PFE is delayed following hemothorax, typically appearing by the tenth day, with eosinophilia appearing in the peripheral blood days later [21,22]. There is significant correlation between the number of red blood cells and the number of eosinophils in the pleural fluid [23]. Bleeding into the pleural space may contribute to the development of PFE in the setting of thoracotomy, chest trauma, benign asbestos pleural effusion, and pulmonary infarction [16].
Nonspecific pleural effusions that develop within 30 days after cardiac surgery are typically hemorrhagic and often eosinophilic. They are more common when the pleura has been excised (eg, for internal mammary grafting). (See "Evaluation and management of pleural effusions following cardiac surgery", section on 'Assess need for diagnostic thoracentesis'.)
●Repeated thoracentesis — Repeated thoracentesis has been suggested as a cause of PFE, but the available data are conflicting [16,21,23,24]. The percentage of eosinophils does not increase in successive thoracenteses [16,23].
Asbestos — Benign asbestos pleural effusions (BAPEs) typically develop in the first two decades following asbestos exposure. Approximately one-third of BAPEs have an increased percentage of pleural eosinophils, ranging up to 50 percent of the total nucleated cell count [25]. BAPEs resolve spontaneously, usually within four to six weeks, although residual pleural thickening may persist indefinitely. (See "Asbestos-related pleuropulmonary disease", section on 'Benign asbestos pleural effusion'.)
Drugs — Approximately 25 drugs have been implicated in the development of PFE (table 2) and other drugs likely have the potential to cause PFE. Symptoms of drug-induced PFE are nonspecific and include pleuritic chest pain, cough, dyspnea, and occasional fever. Concomitant blood eosinophilia has been described with valproic acid, nitrofurantoin, dantrolene, gliclazide, tizanidine, fluoxetine, and warfarin [26-34].
The latency between the development of PFE from first drug exposure is typically several months, with a range from a few days to one to two years (eg, nitrofurantoin) or even 12 years (eg, dantrolene) [6,26]. Herbal preparations (eg, kampo) have also been implicated in PFE [35].
Symptoms usually begin to improve within the first few days after drug discontinuation and the pleural effusion completely resolves over several months [26]. Treatment with oral glucocorticoids has been reported to speed resolution in a few cases [34,36,37].
Infections — Infections caused by bacteria, fungi, mycobacteria, parasites, and viruses can cause PFE [3,16,38-51]. Eosinophils appear late in the course of a parapneumonic effusion, usually three weeks or more, and often after resolution of the pneumonia [38,52]. An eosinophilic parapneumonic effusion is unlikely to progress to empyema [53]. (See "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults".)
The development of PFE in the setting of tuberculosis is now considered rare. However, older reports from populations with an increased prevalence of tuberculosis found that tuberculosis was responsible for up to 40 percent of PFEs [39-41]. However, parasitic pleural effusion has higher degree of pleural fluid eosinophilia [54]. (See "Tuberculous pleural effusion".)
Paragonimiasis is the most common parasite associated with PFE and is endemic to southeast Asia [43,44]. The pleural fluid characteristics of Paragonimus infection are low pH (<7.10) and glucose less than 60 mg/dL (3.3 mmol/L). Cytological examination of the pleural fluid may reveal the parasite. (See "Paragonimiasis".)
Other parasitic infections associated with PFE include sparganosis, toxocariasis, cutaneous myiasis, loiasis, lymphatic filariasis, amebiasis, ascariasis, strongyloidiasis, echinococcosis, Taenia solium, and dracunculiasis [44-50,55]. Parasite-specific IgG antibody detection in serum can be helpful in diagnosing parasitic pleural effusion [55]. These are discussed further in the appropriate topic reviews.
Fungal infections specifically coccidiomycosis can be associated PFE and should be considered as a potential cause of PFE in an endemic area [42].
Malignancy — Prevalence of PFE is low in malignant pleural effusion, reported anywhere from 2.3 to 6.8 percent [1,56]. However, the reported frequency of pleural malignancy in patients with PFE ranges from 6 to 40 percent [2,3,16,57]. Most importantly, malignancy is diagnosed as frequently in eosinophilic effusion as in noneosinophilic effusions [2]. The likelihood of pleural malignancy appears to be inversely related to the percentage of eosinophils in the pleural space [16,17] but there is no absolute eosinophil percentage that would exclude malignancy [3,17] and presence of eosinophils (≥10 percent) cannot be considered an indicator of benignancy [16,23].
Among the malignancies associated with PFE, lung cancer is the most common [3,16]. In a review of 153 pleural fluid samples demonstrating PFE, 47 were associated with malignancy and of these, 23 were lung cancer [16].
Further investigations, such as thoracic computed tomography and pleural biopsy, either closed or thoracoscopic, should be pursued when pleural fluid cytology is negative but the patient has risk factors for underlying malignancy (eg, heavy smoking for many years) [16].
Pulmonary embolism — Pleural effusion accompanies pulmonary embolism in 30 to 50 percent of cases, and 18 percent of these effusions are eosinophilic [4,58,59]. These effusions are typically small (less than one-third of the hemithorax), unilateral exudates. One-half to two-thirds are bloody, due to associated pulmonary infarction. Pleuritic chest pain is typically present in the setting of pulmonary embolism with effusion. (See "Clinical presentation, evaluation, and diagnosis of the nonpregnant adult with suspected acute pulmonary embolism".)
Rare causes — PFE has rarely been reported in association with chronic eosinophilic pneumonia, rheumatoid pleural effusion, and eosinophilic granulomatosis with polyangiitis (EGPA; Churg-Strauss) [14,60]. (See "Chronic eosinophilic pneumonia" and "Overview of pleuropulmonary diseases associated with rheumatoid arthritis" and "Epidemiology, pathogenesis, and pathology of eosinophilic granulomatosis with polyangiitis (Churg-Strauss)".)
Idiopathic — PFE is idiopathic in 8 to 35 percent of patients [2,4,15-18,41,61,62]. The typical patient with idiopathic PFE is a middle-aged man with a small to moderate-sized unilateral pleural effusion [15]. The prognosis is excellent in these patients, even when the effusion persists for months. At least some of these cases may be due to unidentified asbestos exposure, occult pulmonary embolism, viral infection, or a medication.
Because PFE is a diagnosis of exclusion, patients with PFE and no obvious cause should be monitored until the effusion resolves or a known cause becomes apparent.
The approach to a patient with a pleural effusion of undetermined etiology is discussed separately. (See "Diagnostic evaluation of pleural effusion in adults: Additional tests for undetermined etiology".)
EVALUATION — In most cases, the evaluation of pleural fluid eosinophilia (PFE) follows the evaluation protocol for pleural effusions. However, certain clinical features may help to determine the etiology. The routine evaluation of pleural effusions is discussed separately. (See "Diagnostic evaluation of a pleural effusion in adults: Initial testing" and "Diagnostic evaluation of pleural effusion in adults: Additional tests for undetermined etiology".)
History — Historical information that may help to identify the etiology of PFE includes the following:
●A condition in the last 1 to 4 weeks that may have led to entry of air or blood into the pleural space
●List of medications, dietary supplements (eg, L-tryptophan), herbal preparations, and vitamins (see 'Drugs' above)
●History of exposure to asbestos (see 'Asbestos' above)
●Risk factors for lung cancer or history of malignancy (see 'Malignancy' above)
●Signs of infection (eg, fever, sweats, or chills) or travel or residence in endemic areas of fungi (eg, Coccidioides), tuberculosis, or parasites (eg, Paragonimus, Strongyloides, Dracuncula) (see 'Infections' above)
Physical examination — Physical examination findings are generally nonspecific, but evidence suggesting chest trauma or skin lesions suggestive of a fungus should be sought. (See 'Pleural trauma' above.)
Laboratory analysis — Laboratory analysis of peripheral blood is usually nonspecific in patients with PFE, since there is only a weak correlation between the eosinophil percentages in the peripheral blood and pleural fluid [16]. Serologic studies for antibodies to fungi and parasites (eg, Coccidioides, Echinococcus, Schistosoma, Strongyloides, Trichinella spiralis, Wuchereria bancrofti) may be helpful in patients with a history of travel to or residence in endemic areas.
In patients with other features to suggest rheumatoid arthritis or eosinophilic granulomatosis with polyangiitis (EGPA; Churg-Strauss), measurement of cyclic citrullinated peptide or antineutrophil cytoplasmic antibody (ANCA) titers may be helpful. (See "Overview of pleuropulmonary diseases associated with rheumatoid arthritis" and "Epidemiology, pathogenesis, and pathology of eosinophilic granulomatosis with polyangiitis (Churg-Strauss)".)
Pleural fluid analysis — PFE is defined as pleural fluid with a nucleated cell count containing more than 10 percent eosinophils [1-4]. The specific percentage of eosinophils in pleural fluid does NOT help differentiate between the different causes of PFE. The likelihood of pleural malignancy appears to be inversely related to the percentage of eosinophils in the pleural space [16] but there is no absolute eosinophil percentage that would exclude malignancy [3].
Determination of whether the fluid is an exudate or transudate is helpful. A list of transudative and exudative causes of PFE is provided in the table (table 1). The differentiation of transudates and exudates is discussed separately. (See "Diagnostic evaluation of a pleural effusion in adults: Initial testing".)
Pleural fluid cultures for bacteria, fungi, and mycobacteria should be performed. Cytological analysis should include examination for malignant cells and parasites.
Undetermined etiology — When the etiology of a pleural effusion with eosinophilia remains undetermined after the above studies, further evaluation will typically include computed tomographic scanning and pleural biopsy. These tests are discussed separately. (See "Diagnostic evaluation of pleural effusion in adults: Additional tests for undetermined etiology" and "Medical thoracoscopy (pleuroscopy): Diagnostic and therapeutic applications".)
After a comprehensive evaluation, a minority of patients have no apparent cause and are considered to have idiopathic PFE. (See 'Idiopathic' above.)
DIAGNOSIS — Diagnosis of pleural fluid eosinophilia (PFE) is based on pleural fluid cell count >10 percent eosinophils. The underlying cause of PFE is often identified through history and routine studies described above, but may not identified in a small proportion of patients.
TREATMENT AND MONITORING — The treatment of PFE varies with the cause and is discussed on the appropriate topic reviews. Because PFE is a diagnosis of exclusion, patients with PFE and no obvious cause should be monitored until the effusion resolves or a known cause becomes apparent.
SUMMARY AND RECOMMENDATIONS
●Pleural fluid eosinophilia (PFE) is defined as pleural fluid with a nucleated cell count containing greater than 10 percent eosinophils. (See 'Introduction' above.)
●Lung and pleural tissue do NOT normally harbor eosinophils. As a result, the development of PFE requires the recruitment of bone marrow derived eosinophils. Nonspecific injury (eg, air, blood, or trauma) may stimulate pleural mesothelial cells and lead to production of the cytokines, chemokines, and adhesion molecules that orchestrate eosinophil recruitment. (See 'Pathogenesis' above.)
●The most common causes of PFE are the entry of air or blood into the pleural space malignancy, and infection. Less common causes include drugs, asbestos, pulmonary thromboembolism, chronic eosinophilic pneumonia, eosinophilic granulomatosis with polyangiitis (EGPA; Churg-Strauss), and rheumatoid pleuritis. (See 'Etiology' above.)
●Approximately 25 drugs have been implicated in the development of PFE (table 2). Concomitant peripheral blood eosinophilia is sometimes present. (See 'Drugs' above.)
●PFE is idiopathic in 8 to 35 percent of patients. Because PFE is a diagnosis of exclusion, patients with PFE and no obvious cause should be monitored until the effusion resolves or a known cause becomes apparent. (See 'Idiopathic' above.)
●Bacterial (including parapneumonic effusions), fungal, mycobacterial, parasitic, and viral infections can cause PFE. (See 'Infections' above.)
●The frequency of pleural malignancy in PFE is variable and ranges between 6 to 40 percent, depending on the prevalence of malignancy in the population studied. Among malignancies, lung cancer is the most common. (See 'Malignancy' above.)
