Pleural Effusion in an Asymptomatic Patient* Spectrum and Frequency of Causes and Management Considerations

Transcription

1 special report Pleural Effusion in an Asymptomatic Patient* Spectrum and Frequency of Causes and Management Considerations Nicholas A. Smyrnios, M.D.;t Peter J ]ederlinic, M.D., F.C.C.P.; and RichardS. Irwin, M.D., F.C.C.P. We conducted retrospective chart and literature reviews to analyze the frequency and spectrum of causes of asymptomatic pleural effusion (APE). In our series, 16 percent of patients undergoing thoracentesis for PE were asymptomatic and the spectrum of causes was similar to that for symptomatic patients. Asymptomatic PEs were evenly distributed among transudates, exudates and indeterminate effusions. More symptomatic (S) PE were exudates, although the difference was not statistically significant (p>o.l). In comparison to SPE, APE were more often free Howing and small. In both groups, the four most common diagnoses were malignancy, CHF, parapneumonic and postoperative effusions accounting for greater than 70 percent of each group. Review of the literature demonstrated the following associations with APE: recent childbirth or abdominal surgery, benign asbestos effusion, uremia, malignancy, and tuberculosis. In the uncomplicated postpartum or postoperative setting or in patients with typical findings of left ventricular failure, observation without diagnostic studies is appropriate. In all other situations, APE should be evaluated in traditional fashion. If thoracentesis is nondiagnostic and the effusion is an exudate, closed pleural biopsy and less often, fiberoptic bronchoscopy, should follow. Once malignant or granulomatous pleuritis has been excluded, it may be appropriate to observe for a period of time before proceeding to more invasive procedures. (Chest 1990; 97:192-96) S, APE= symptomatic, asymptomatic pleural effusion P atients with pleural effusion are typically symptomatic. If the disease process is localized to the lungs, the patient will usually complain of cough, dyspnea, or chest pain.' There may also be systemic manifestations such as weight loss, anorexia or fever if the PE is part of a generalized process. The diagnostic work up of symptomatic pleural effusion is well estahlished. 1 2 Although it has been stated that asymptomatic pleural effusion is not uncommon, 1 we are unaware of any articles that deal specifically with the subject. Therefore, we reviewed the medical literature and performed a retrospective chart review in order to determine the frequency and spectrum of causes of APE and how it should be best managed. METHODS Our review mnsisted of two parts. Tht> medical re<- >rds of all patients who underwt>tll thomeentesis at the University of Massachusetts Medieal Center between July 1, 1986 and June 30, 1987 *From the Pulmonary and Ct;tical Care Medicine Division, Department of Medicine, University nf Massaehusells Medical School, \Vorcester. treeipient of a Will Ro~ers Pulmonary Fellowship. Reprint n que.~ts: Dr. ]ederlinic, Pulnwm1ry Division, 55 Lllke AVl'nue North, Worce.Yil'T' were analyzed acmrding In the criteria for retrospedive eharl review established hy Feinstein el al.' ' Patients were divided into three groups based upon the presence, absence or lack of inli>rmalion L'lncerning symptoms of cou~h, dyspnea, and chest pain. Constitutional and other non-chest symptoms were not used in determinin~ asymptomatic status. The SPE ~roup c.'lnsisted of patients reportin~ <- m~, dyspnea and/or chest pain. The APE J.,'Toup consisted of patients who specifimlly denied these symptoms. A third ~roup included patients in whom absence of symptoms was not specifically mentioned in the chart. Only the first thoracentesis performed on any patient durin~ the study period was included. Procedures were reviewed to determine the frequency of transudates and exudates, 5 radio~raphic characteristics and most likely diagnosis of each effusion. Chest radio~raphs were reviewed by one of the investigators. The size of the effusion was estimated on the initial upright inspiratory posteroanterior chest film. A la~e effusion filled greater than 50 percent of the hemithorax, medium filled 25 to 50 percent, and a small effusion less than 25 percent. An effusion was determined to be free-rowing by reviewing lateral decubitus chest films. The clinical diagnosis made durin~ the patient's hospitalization was used unless there was definite objective evidence to L'lnlradict it. If such information existed, an alternative diagnosis was assi~ned. We attempted to make a dia~osis based upon available information in the remainder. An effusion was l'lnsidered indeterminate when the information required to distinguish it as exudate nr transudate and assi~ a diagnosis was lackiu~. We then reviewed the En~lish literature on pleural effusion durin~ the years 1975 tn 1988 usin~ BRS Collea~ue computerassisted search pro~ram. Index terms were "pleural effusion" and Pleural Effusion in an Asymptomatic Patient (Smymios, Jederlinic, Irwin)

2 Table 1-Compariaon ofltn-of E./}imon Table 3-Diagnoeea of Pleural Ejfuaiona Symptomatic Asymptomatic Symptomatic Asymptomatic Diagnoses Transudate Exudate Indeterminate n=79 18 (23) 44 (56) 17 (21) n=15 "asymptomatic." There were no articles dealing specifically with the topic APE. RESULTS Retrospective Chart Review One hundred fifteen charts were reviewed. Of these, four were excluded from analysis; three represented procedures to drain large pneumothoraces, while one chart lacked essential information. There were 79 patients with SPE and 15 patients with APE. Another 17 patients fell into the third group and were excluded from further analysis. Thus, of the 94 patients with PE that were suitable for review, 79 (84 percent) were SPE and 15 (16 percent) were APE. The percentage of transudates and exudates in each group is shown in Thble 1. The APE group was evenly divided between transudates, exudates, and indeterminate effusions. The SPE group showed a greater percentage of exudates than transudates, although the difference was not statistically significant (p>0.1). The chest radiographic features of the APE group are shown in Table 2. Compared to SPE, APEs were more frequently free-flowing and of smaller size. The diagnoses associated with SPE and APE are shown in Thble 3. In both groups, the three most common diagnoses were malignancy, congestive heart failure, and parapneumonic effusion, accounting for more than 61 percent of the effusions in both groups. If postoperative surgery-related effusions are included, 70 percent of the SPE and 79 percent of the APE were accounted for. Postoperative effusions were more commonly symptomatic, but made up a larger percentage of the APE group. Other causes of APE Table 2-Badiographic IWJtura of Symptomatic and A1ymp1omatic Pleural Ejfuaiona Symptomatic Asymptomatic n=79 n=15 Bilateral 38 (48) 8 (53) Free flowing 30 (38) 11 (73) Loculated 9 (11) 0 (0) Small 15 (19) 10 (67) Medium 12 (15) 4 (27) Large 23 (29) 1 (6) Large= filled >50 percent of hemithorax; medium= filled 25 to 50 percent of hemithorax; small=filled <25 percent of hemithorax. n=79 n=15 Malignancy 25 (32) 4 (26) Congestive heart failure 13 (16) 3 (20) Parapneumonic 11 (14) 3 (20) Postoperative 6 (8) 2 (13) Trauma/hemothorax 8 (10) 0 (O) Empyema 7 (9) 0 (O) Liver disease/ascites 3 (4) 1 (7) Uremia 3 (4) 0 (O) Drug induced lupus 1 (1) 1 (7) Gastric ulcer 0 (O) 1 (7) Infectiousendocwnlltis 1 (1) 0 (0) CAPO related 1 (1) 0 (0) CAPO, continuous ambulatory peritoneal dialysis. included liver disease, gastric ulcer and drug-induced lupus. Review of the Literature Major causes of APE identified by our literature search are shown in Thble 4. While postpartum, postoperative and benign asbestos effusion were most commonly encountered as asymptomatic, it was clear from our review that most causes of SPE can also cause APE in unusual situations. Of the articles used to describe these frequencies, three were prospective6-8 and five retrospective.6 s.-12 None dealt specifically with APE. Frequency DISCUSSION To establish the frequency of APE, we retrospectively analyzed our experience over a one-year period. The frequency with which PE was asymptomatic was 16 percent. Based upon this, we have substantiated the clinical impression of Light that APE are not uncommon. Spectrum of Causes Asymptomatic PE had a similar spectrum of causes compared to SPE. While our study and the literature suggest that any disease that can cause SPE can also Table 4-The Mod Common Cauaa of Alymplomatic Pleural Ejfuaiona Postpartum Postoperative (abdominal) Benign asbestos effusion Uremia Malignancy Tuberculosis Estimated Frequency (%) frequent frequent References 6 8 7, CHEST I 97 I 1 I JANUARY,

3 cause APE, malignancy, congestive heart failure, pneumonia and abdominal and thoracic surgery accounted for approximately 75 percent of the effusions in both groups. Although postpartum and benign asbestos effusions were not diagnosed in any of our patients, the literature suggests that they commonly present asymptomatically. Therefore, they deserve further discussion. Childbirth is a common, yet infrequently considered cause of APE. Hughson et al6 radiographically examined the prevalence of PE within 24 hours of vaginal delivery. Up to 67 percent of patients had small-to-moderate size effusions which were often bilateral. Almost all of these patients were asymptomatic. They did not have an increased incidence of postpartum complications despite the lack of a specific diagnosis. Although this study requires further confirmation, it seems that asymptomatic effusions in the immediate postpartum period are common and in the absence of other &ndings, these effusions do not require evaluation. Benign asbestos effusion (BAE) is the most common manifestation of asbestos-related pulmonary disease within 20 years of&rst exposure. 7 Epler at al 7 identi&ed PE in 4.8 percent of 1,135 asbestos-exposed workers undergoing a multi-year surveillance, a prevalence &ve times that of the non-exposed control population. Thirty-four of 54 (63 percent) were determined to be BAE, and of these, two thirds were asymptomatic. Characteristics of BAE include: 1) usually unilateral location; 2) small-to-moderate size; 3) negative cultures; 4) negative cytology; and 5) serous or bloody exudative fluid. The prognosis of these patients is excellent despite frequent recurrence of the effusion Management of BAE involves exclusion of other causes of PE and close observation. Why and When ArB Pleural Effusions Asymptomatic? The mechanisms by which symptoms develop in PE are not known. Dyspnea has been attributed to an alteration in chest wall mechanics and to hypoxemia resulting from atelectasis. Lack of improvement in these parameters following removal of600 to 1,800 ml of pleural fluid despite subjective improvement in dyspnea suggests otherwise.13 Coexistent parenchymal disease and splinting from pleuritic pain may also contribute to dyspnea. The mechanism of cough in PE is also unclear. Possible causes include stimulation of pleural or airway cough receptors by the presence of pleural inflammation or atelectasis of the lung causing airway collapse. Also, cough is a common complication of thoracentesis us and may be related to stimulation of pleural or airway cough receptors by removal of a large volume offluid.l6 Chest pain in PE is a manifestation of pleural inflammation. Pain can be either sharp or dull and aching. Its presence implies involvement of the parietal pleura since the visceral pleura is without pain fibers. Chest pain can be localized to the involved area of the pleura or referred. Important sites of referred pain are the abdomen and the ipsilateral shoulder. Abdominal pain occurs because of the presence of intercostal innervation above and below the diaphragm. Shoulder pain indicates involvement of the diaphragmatic pleura in the phrenic nerve distribution. We can only speculate why some PEs are asymptomatic. The absence of a primary inflammatory process involving the pleura may be one reason (postpartum or left ventricular failure effusions). Alternatively, small effusions, even bilateral ones, may not compress airways and lead to symptoms in the sedentary patient. Perhaps symptoms associated with the primary process distant from the chest (eg, colon cancer) absorb most of the patients attention, thus minimizing respiratory symptoms, or cases without pulmonary symptoms may be discovered inadvertently before development of symptoms. Finally, altered mental status associated with underlying disease can result in masking thoracic symptoms. Hepatic encephalopathy as seen in our own series or uremia could lead to this result. Diagnostic Evaluation of Asymptomatic Effusions Our proposed diagnostic work-up of APE is shown in Figure 1. In general, the work-up is similar to that of symptomatic effusion. However, several points deserve emphasis. Since there is no radiographic appearance that is diagnostic, once PE is detected in the asymptomatic patient, careful evaluation of the history and clinical circumstances is required. As outlined in 'Dlbles 3 and 4 multiple causes of PE may be asymptomatic at the time of presentation. Patients with APE following uncomplicated abdominal or thoracic surgery or recent vaginal childbirth can be observed closely. 6 8 Patients with &ndings of left ventricular failure (increased cardiac shadow, infiltrates in butterfly pattern, or vascular redistribution, S 3 gallop) can also be observed during diuresis. All others should have a diagnostic thoracentesis if the fluid is accessible since clinically useful information will be obtained in over 90 percent of cases. IS When the cause of APE is unclear following thoracentesis, the next diagnostic step is that of closed pleural biopsy. In experienced hands, the diagnostic yield is increased in cases of suspected tuberculous pleurisy or malignant pleuritis by combining multiple closed pleural biopsies with repeat thoracentesis for culture or cytology. The most common complications are pneumothorax and pain

4 Asymptomatic Pleural Effusion /~ Uncomplicated CHF Postoperative Postpartum + Observe All Others Lateral DJubitus I Film Others LOCitted Uhrasound """ Thoracentesis Diagnostic lor - Exudate Transudate Malignancy + lnlection Pancreatic Process or Esophageal Rupture """ Small / Evaluate for CHF Cirrhosis Nephrotic Syndrome Pulmonary Embolus Non-diagnostic Repeat Thoracentesis - Malignancy + and Closed Pleural Biopsy Granuloma Evaluate and Treat lor + Other Systemic Disease - Non-specific Pleuritis - Observe Collagen Vascular Disease or Normal Pulmonary Embolus I Subdiaphragmatic Disease t Unsuspected Chest Disease Repeat Thoracentesis - Malignancy and Closed Pleural Biopsy Granuloma Observe -- Asbestos - Non-specific Pleuritis Exposure / ~ ~ ~ Uremia Stable Unstable ~ Dialysis Leslie and Kinasewitz 18 have defined a group of patients with non-diagnostic pleural biopsies who do not require further evaluation. They classified patients on the basis of pleural biopsy specimens into three groups: granulomatous pleuritis, malignant pleuritis, and nonspecific pleuritis. Repeat pleural biopsy had a ~ ~ Observe Open Pleural FIGURE 1. Diagnostic algorithm for evaluation Biopsy of asymptomatic pleural effusion. diagnostic yield of 55 percent. Five criteria were predictive of an eventual diagnosis of granulomatous or malignant pleuritis: weight loss greater than 4.5 kg; fever greater than 38 C; positive skin test for tuberculosis; effusion occupying greater than one-half the hemithorax; and pleural fluid lymphocytosis of greater FIGURE 2. Posteroanterior chest roentgenogram. (A, left). Left sided pleural effusion (see text). Computed tomography of the upper abdomen. (B, right). Large subcapsular splenic hematoma. The effusion resolved completely following splenectomy. CHEST I 97 I 1 I JANUARY,

5 than 95 percent. Ninety-seven percent of cases of malignancy or tuberculosis had at least one of these criteria. When all of these were absent, there was a 94 percent chance that the pleural disease was nonspecific pleuritis. This was supported by an average of 33 months offollow-up. If pleural biopsy is nondiagnostic and none of the above criteria is present, patients can be followed clinically unless there is a parenchymal abnormality on chest radiograph. Computed tomography (CT) of the chest and upper abdomen should be part of the evaluation of APE when thoracentesis and initial closed pleural biopsy have been nondiagnostic, when uremic effusions do not clear with dialysis, and before accepting a diagnosis of BAE. Detection of loculation, radiographically inapparent parenchymal lesions, or subdiaphragmatic pathology by CT can help focus the subsequent evaluation. We have recently seen a young man with an asymptomatic left PE and positive result of PPD test. After a nondiagnostic thoracentesis and pleural biopsy, CT scan of the chest and upper abdomen demonstrated a previously unsuspected subcapsular splenic hematoma, the result of trauma several years earlier (Fig 2). On the basis of this study, the effusion was considered to be sympathetic and resolved completely following splenectomy. Fiberoptic bronchoscopy has been useful in evaluating PE in the presence of hemoptysis or associated roentgenographic abnormality. 19 However, the diagnostic yield of bronchoscopy in the absence of these associated findings is low. Rigid or fiberoptic thoracoscopic examination of the pleura allows direct visualization for biopsy and its sensitivity in the diagnosis of tuberculosis and malignancy approaches 100 percent in some series. However, it does require hospitalization overnight and its advantage over conventional means of diagnosis is unclear. 110 We ~mmend thoracotomy in only a small minority of patients undiagnosed after closed pleural biopsy since the procedure does not guarantee the diagnosis A definitive diagnosis is made in only 52 percent of cases.21 Furthermore, Ryan et al 23 found that in approximately 61 percent of patients with nonspecific pleuritis at thoracotomy, the course was entirely benign without recurrence of effusion or development of a specific etiology. The same finding has been noted by others. 12 Those who remain asymptomatic and whose effusions are stable can be observed. Patients who develop symptoms or whose effusions reaccumulate or increase in size are referred for open biopsy. In conclusion, we found the frequency of APE to be 16 percent in our series of patients undergoing thoracentesis. When compared to symptomatic patients, the spectrum of causes was similar, with malignancy, CHF, pneumonia and recent surgery the major 118 contributors. Review of the literature revealed asbestos exposure, recent childbirth, tuberculosis and uremia as other conditions commonly associated with APE. When the clinical setting indicates need for diagnostic evaluation, the approach is similar to that for SPE. ACKNOWLEDGMENT: The authors thank Linda Cormier for her secretarial assistance. REFERENCES 1 Light Rw. Pleural diseases. Philadelphia: Lea and Febiger, Sahn SA. The pleura. Am Rev Respir Dis 1988; 138: Feinstein AR, Pritchett JA, Schimpf CR. The epidemiology of cancer therapy: III. The management of imperfect data. Arch Intern Med 1969; 123: Feinstein AR, Pritchett JA, Schimpf CR. The epidemiology of cancer therapy: IV: The extraction of data from medical records. Arch Intern Med 1969; 123: Light Rw, Macgregor I, Luchsinger PC, Ball WC. Pleural effusions: The diagnostic separation of transudates and exudates. Ann Intern Med 1972; 77: Hughson WG, Friedman PJ, Feigin DS, Resnik R, Moser KM. Postpartum pleural eijusion: A common radiologic &nding. Ann Intern Med 1982; 97: Epler GR, McCloud TC, Gaensler EA. Prevalence and incidence of benign asbestos pleural eijusion in a working population. JAMA 1982; 247: Light Rw, George RB. Incidence and signi&cance of pleural effusion after abdominal surgery. Chest 1976; 69: Chernow B, Sahn SA. Carcinomatous involvement of the pleura. Am J Med 1977; 63: Epstein DM, Kline LR, Albelda SM, Miller WT. Tuberculous pleural eijusions. Chest 1987; 91: Berger HW, Rammohan G, Neff MS, Buhain WJ. Uremic pleural effusion. Ann Intern Med 1975; 82: Gaensler EA, Kaplan AI. Asbestos pleural eijusion. Ann Intern Med 1971; 74: Brown NE, Zamel N, Aberman A. Changes in pulmonary mechanics and gas exchange following thoracentesis. Chest 1978; 74: Seneff MG, Corwin RW; Gold LH, Irwin RS. Complications associated with thoracocentesis. Chest 1986; 89: Collins TR, Sahn SA. Thoracocentesis: Clinical value, complications, technical problems and patient experience. Chest 1987; 91: Corwin Rw, Irwin RS. Thoracentesis. In: Rippe JM, Irwin RS, Alpert JS, Dalen JE, eels. Intensive care medicine. Boston: Little, Brown and Co. 1985; Cope C, Bernhardt H. Hook-needle biopsy of the pleura, pericardium, peritoneum and synovium. Am J Med 1963; 35: Leslie WK, Kinasewitz GT. Clinical characteristics of the patient with nonspeci&c pleuritis. Chest 1988; 94: Feinsilver SH, Barrows AA, Braman SS. Fiberoptic bronchoscopy and pleural eijusion of unknown origin. Chest 1986; 90: Faurschou P, Madsen F, Viskum K. Thoracoscopy: Influence of the procedure on some respiratory and cardiac values. Thorax 1983; 38: Douglass BE, Carr DT, Bernatz PE. Diagnostic thoracotomy in the study of idiopathic pleural elfusion. Am Rev Tuberc 1956; 74: Schiess JM, Hamson HN, Wier JA. The role of thoracotomy in the differential diagnosis of pleural eijusion 23 Ryan CJ, Rodgers RT, Unni KK, Hepper NG. The outcome of patients with pleural eijusion of indeterminate cause at thoracotomy. Mayo Clin Proc 1981; 56: Pleural Eftullion In an Asymplomallc Patient (SmymJos, Jedeillnlc, Irwin)