Recurrent symptomatic pleural effusions: approaches to diagnosis of a difficult problem.

Pleural effusions, seen in isolation or in association with a number of pulmonary and systemic diseases, are common problems in both the in-patient and out-patient setting. The incidence of pleural effusion in the general population is hard to estimate. Reports suggest an incidence of 0.3% in the general population, with as many as 800,000 cases annually. Pleural effusions frequently complicate malignant disease, heart and liver diseases, infections, and other rare diseases.1 The diagnostic approach to pleural effusion is multifaceted, beginning with a careful clinical assessment and followed by radiologic, laboratory, and cytopathology studies. History provides useful information, such as exposure to asbestos, suggesting mesothelioma or drug-induced pleural disease. Proper studies aid in the diagnosis, recognition of the problem, and disease state. Typical symptoms include dyspnea, cough, and chest pain. Physical signs include asymmetric chest expansion, dullness to percussion, decreased to absent breath sounds, reduced vocal fremitus, and pleura rub. In a cohort of symptomatic patients with respiratory conditions, asymmetric chest expansion and dullness to percussion had adjusted odds ratios of 5.3 and 12.8, respectively, for pleural effusions.2 Except in cases of large pleural effusions, initial imaging confirms and evaluates the size and characteristic of the pleural space. The amount of pleural fluid can be estimated from radiography. Studies have suggested that as little as 75 mL can be detected in the posterior costophrenic sulcus, whereas, a minimum of 175 mL is necessary to obscure the lateral sulcus.3-5 Ultrasound is superior in clarifying the nature of a pleural fluid and the detection of smaller effusions, compared to conventional radiography and physical examination. Studies have found effusions as small as 5–50 mL with careful ultrasound.6 Computed tomography is unequaled in its ability to depict the entire pleural space, simultaneously imaging the pulmonary parenchyma and mediastinum. There are 2 categories of pleural effusions: transudative and exudative. In most circumstances the gross appearance of the pleural fluid can help identify effusion type. Purulent fluid indicates an infection. Milky white fluid could represent a chylous effusion. Bloody fluid or hemothorax results from trauma or instrumentation but may be observed in malignancy. Pleural effusions are always secondary to a disease process. Some effusions can simply be observed; new, unilateral, and enlarging effusions or failure of an effusion to resolve with treatment indicates the need for diagnostic thoracenteses. The investigative work involved in the diagnosis can be mapped out in an algorithm and includes testing.7 The first step consists in establishing type. If the fluid is transudative, the differential diagnosis is narrowed. Typically, exudative effusions require additional testing. Since its original publication, Light’s criteria have been widely used to separate exudates and transudates.8 Perhaps the only limitation of these criteria is the need for concomitant blood draw for measurement of serum lactate dihydrogenate and protein. Other alternative markers have been proposed, although serial different reports have failed to show a significant advantage with other markers.9 Using Light’s criteria, the physician can be confident in correcting almost all exudative effusions. However, a specific problem using Light’s criteria occurs in patients with congestive heart failure treated with diuretics. Another option, particularly helpful when serum levels of lactate dihydrogenate and protein are not available, is to measure cholesterol levels. Meta-analysis suggested that a cholesterol level 45 mg/dL identifies exudates with an accuracy similar to Light’s criteria.10