© 2003 American Thoracic Society
Yield of Sputum Induction in the Diagnosis of Pleural TuberculosisDivisão de Tisiologia e Pneumologia, Hospital Universitário Clementino Fraga Filho, Universidade Federal do Rio de Janeiro, Brazil; School of Public Health, University of California, Berkeley; and Division of Pulmonary and Critical Care Medicine, University of California, San Francisco, California Correspondence and requests for reprints should be addressed to Marcus B. Conde, M.D., Divisão de Tisiologia e Pneumologia, Hospital Universitário Clementino Fraga Filho, Universidade Federal do Rio de Janeiro, Av. Brigadeiro Trompowsky, s/n°, 3° andar, SME da Pneumologia, Ilha do Fundão, Rio de Janeiro, RJ, Brazil CEP 21941-590. E-mail: marcusconde{at}hucff.ufrj.br
We prospectively evaluated the diagnostic yield of acid-fast bacilli smear and culture for Mycobacterium tuberculosis using sputum induction (SI) in the workup of patients with suspected pleural tuberculosis (TB) who were unable to produce sputum spontaneously. Of the 113 patients studied, a final diagnosis of pleural TB was made in 84 patients (71 HIV seronegative) and a final diagnosis of another disease in 29 patients. Histopathologic examination of the pleural biopsy tissue had the highest diagnostic yield (78%; 66/84). The bacteriologic yield was 62% (52/84) for the pleural tissue, 12% (10/84) for pleural fluid, and 52% (44/84) for sputum cultures obtained by SI. The yield of SI culture for M. tuberculosis was 55% (35/64) in patients with a normal radiograph (except for the pleural effusion) and 45% (9/20) in those with evidence of parenchymal disease suggestive of pulmonary TB (p = 0.6). The yield of sputum cultures obtained by SI is high in patients suspected of having pleural TB even in those cases with no pulmonary parenchymal abnormalities on the chest radiograph.
Key Words: tuberculosis • induced sputum • diagnosis • pleural effusion Tuberculosis (TB) remains one of the most important health problems in the world, with an estimated 8 million new cases annually (1). Pleural TB is a common form of extrapulmonary disease and may occur in the presence or absence of pulmonary parenchymal disease on the chest radiograph (2). In Rio de Janeiro, there were 9,894 new cases of TB reported in 1997, with 1,484 cases of extrapulmonary disease. The pleural effusion associated with TB contains a relatively small number of organisms, making the diagnosis of pleural TB difficult and often requiring invasive procedures like pleural biopsy (2, 3). Maximizing the yield of the diagnostic evaluation is particularly important in areas where the ability to perform pleural biopsy may be limited and/or the rates of drug-resistant disease are high. The presence of Mycobacterium tuberculosis (M.tb) in the respiratory specimen of patients with pleural effusions is diagnostic of TB in someone with a pleural effusion. Sputum specimens are often not evaluated because many of these patients are not able to produce sputum spontaneously. Sputum induction (SI) is a safe procedure with a high diagnostic yield for pulmonary TB in HIV-seronegative and -seropositive patients (4, 5). We prospectively conducted this study to evaluate the diagnostic yield of acid-fast bacilli (AFB) smear and culture for M.tb in induced sputum in the workup of patients with pleural TB in Rio de Janeiro, Brazil.
Patient Selection This study prospectively evaluated patients aged 18 years and older with clinical and radiographic findings consistent with pleural effusion due to TB. All patients were evaluated between January 1, 1998 and April 31, 1999 and were referred from Rio de Janeiro outpatient clinics in the 3–1 Planning Area or from the Ambulatory Service of the Hospital Universitário Clementino Fraga Filho (HUCFF). Patients were not eligible if they had clinical and/or radiographic evidence of renal, cardiac, or liver failure; lung cancer; pregnancy; or had a Karnofsky Performance Status score less than 50%. In addition, patients were excluded if, during the three months before enrollment, they had hemoptysis or had received antituberculosis therapy or anticoagulant therapy for more than 1 week. Informed written consent was obtained from all participants, and the study was approved by the Ethics Committee of the HUCFF/UFRJ.
Diagnostic Procedures Pleural fluid, two pleural fragments, and the induced sputum specimens were stained with Ziehl-Neelsen stain, Grocot's methenamine silver stain, and the samples were cultured in Löwenstein Jensen and Sabouraud medium following standard protocols (8). The other three pleural fragments were stained with hematoxylin and eosin, and pleural fluid was stained with Papanicolaou. All specimens that were culture positive for mycobacteria were tested by biochemical methods to distinguish M.tb from other nontuberculous mycobacteria. A diagnosis of pleural TB was made based on growth of M.tb in pleural fluid and/or tissue, in the presence of granulomatous inflammation in histopathologic study of the pleural biopsy tissue, or in resolution of clinical and radiograph abnormalities after 3 months of standard antituberculosis treatment (presumptive diagnosis). A single observer, trained specifically for this, analyzed the prethoracocentesis chest radiograph. Patients were divided into two groups depending on the chest radiograph findings: Group A (presence of parenchymal lesion) and Group B (normal chest radiograph, except for the presence of a pleural effusion). The size of the pleural effusion was estimated from the chest radiograph. An effusion that occupied up to a third of the affected hemithorax was considered small; it was considered medium if it occupied up to half of the hemithorax, and large when more than a half of the hemithorax was affected.
Analysis
Between January 1, 1998 and April 31, 1999, 145 patients were eligible for enrollment with clinical and radiograph findings suspicious of pleural TB. Fourteen patients were excluded because they had no pleural effusion (12 cases) or because the Karnofsky Performance Status score was less than 50% (2 cases). An additional 14 individuals were not included because the respiratory specimens were missing (4 cases), the subjects were unable to produce sputum despite SI (6 cases), or no diagnosis was made (4 cases). Thus, 113 patients were included for analysis. Eighty-four patients (71 HIV seronegative and 13 HIV seropositive) had the final diagnosis of pleural TB, and 29 patients had the final diagnosis of another disease (renal failure in 2 cases, cardiac failure in 2 cases, lymphoma in 2 cases, lung cancer in 10 cases, parapneumonic effusion in 10 cases, systemic lupus erythematosus in 2 cases, and acute pancreatitis in 1 case). Among patients with pleural TB, 67% (56/84) were male. The mean age was 37.2 years (SD 14.6; range = 18–89) and 64% (54/84) were 40 years of age or younger. Ninety-six percent (81/84) of the chest radiographs demonstrated unilateral small or medium-sized effusion. Sixty-four of the patients (54 HIV seronegative and 10 HIV seropositive) with pleural TB had no pulmonary parenchymal opacities identified on the chest radiograph. Of these patients, forty percent (22/54) of HIV-seronegative and 50% (5/10) of HIV-seropositive patients had a history of contact with a pulmonary TB patient within the previous three years. All tuberculous effusions were exudative (84/84), and 84% (71/84) were lymphocytic. The overall yield of the various diagnostic methods is shown in Table 1 . Histopathologic examination of the pleural biopsy tissue had the highest overall diagnostic yield (78%; 66/84). The bacteriologic yield of the pleural tissue was 62% (52/84), whereas that of the pleural fluid was 12% (10/84) and the yield of sputum cultures obtained with SI was 52% (44/84). However, when we included the six patients who could not produce a sputum specimen despite SI, the yield of SI dropped from 52% to 49% (44/90). The only significant difference in the yield of diagnostic methods between HIV-seronegative and HIV-seropositive patients with pleural TB was the yield of AFB smear of pleural biopsy tissue (Table 1).
The yields of SI, pleural fluid, and pleural tissue for an immediate and delayed diagnosis are presented in Table 2 . The immediate diagnosis of pleural TB based exclusively on a sputum specimen obtained by SI was reached in only 2 patients, whereas the delayed diagnosis was obtained in 13 patients. When the results of the AFB smear and culture for M.tb of the specimen obtained by SI and of the pleural fluid were considered together, the yield of both tests combined reached 56% (47/84).
The yield of AFB smear was 11% (7/64) in patients without lesions at chest radiograph and 15% (3/20) in those with parenchymal lesions (p = 0.6), whereas the yield of culture for M.tb was 54% (35/64) in patients without lesions and 45% (9/20) (p = 0.6) in those with parenchymal lesions at chest radiograph. Among patients with no parenchymal opacities seen on the chest radiograph, the pleural effusion was small in 61% (39/64), medium in 34% (22/64), and large in 5% (3/64). Among patients with parenchymal opacities seen on the chest radiograph, the pleural effusion was small in 80% (16/20) and moderate in 20% (4/20). SI was well tolerated by most patients. Adverse effects were reported by 4% (3/84) of patients and included persistent cough (two cases) and headache (one case). All adverse effects resolved spontaneously without intervention. No patient experienced bronchospasm.
This study demonstrates the relatively high yield of cultures obtained by SI for M.tb in patients with tuberculous pleural effusions, even in those patients with no visible pulmonary parenchymal abnormalities on the chest radiograph. This study confirms that the best method to diagnose pleural TB is the histologic evaluation of the pleural tissue (78%) followed by culture of the pleural biopsy (62%). However, unlike previous studies, our data have demonstrated that SI culture has the next highest yield: the yield of induced sputum culture was 6 to 13 times greater (55%) than that described previously among patients with a normal chest radiograph (4% to 9%), and similar to those described among patients with both pulmonary and pleural TB (2, 10, 11). Although the presence of M.tb in the respiratory specimen of patients with pleural effusions is strongly suggestive of the diagnosis of pleural TB, the diagnostic value of cultures for M.tb in the sputum of patients with pleural TB is generally considered low (9). Berger and Mejia (2) reported that 30% (12/40) of patients with pleural TB grew M.tb from sputum culture or gastric contents. Ten of these patients had evidence of pulmonary parenchymal involvement on the chest radiograph. Only 9% (2/22) of the patients in whom no evidence of parenchymal lesions could be identified on the chest radiograph had a positive culture sputum or gastric aspirate. Levine and colleagues (13) reported that 33% (7/21) of patients with tuberculous pleural effusions had positive sputum, gastric, or bronchial lavage specimen. In another study, Arrington and colleagues (14) reported that none of 25 patients with tuberculous pleuritis had a positive sputum or gastric aspirate culture. The fact that most patients with pleural TB do not produce sputum spontaneously may be the reason for the reportedly low yield of sputum culture in this setting. We believe that the higher yield reported in our study is due to the fact we prospectively used SI to collect specimens in all patients. Previous studies have described the sensitivity of cultures obtained by SI to be more than 60% for the diagnosis of pulmonary TB in patients who were unable to produce sputum spontaneously (4, 5). The presence of M.tb in pulmonary specimens in the setting of an otherwise normal chest radiograph (55%, 35/64) is likely explained by the presence of subpleural foci of disease. These lesions are not normally visualized on conventional radiograph, but have been well documented in histopathologic studies and through computed tomography (3, 12). The yield of pleural fluid cultures for M.tb was 11% (9/84) in this study, whereas other investigators have reported values of 10 to 35% (2, 10, 11). The yield of SI and pleural fluid bacteriologic tests combined (56%; 47/84) was similar to that observed with the pleural tissue bacteriologic analysis alone (64%; 54/84) (McNemar = 0.2). Another important point demonstrated by the finding of a relatively high frequency of sputum AFB smear and mycobacterial culture positivity is the potential risk of M.tb transmission to contacts. Unlike other forms of extrapulmonary tuberculosis, pleural TB can result in transmission to others. Additional studies are needed to evaluate the degree of infectiousness of these patients. It is important to point out that only one investigator read all of the chest radiographs. Therefore, we may have missed some parenchymal opacity that would have been detected by another independent chest radiograph reader. However, we believe that this is unlikely because the investigator was trained specifically for the study and had both PA and lateral decubitus views available for review. We conclude that the yield of sputum cultures obtained by SI is high in patients suspected of having pleural TB even when there is no radiograph evidence of pulmonary parenchymal disease. Although SI should not substitute for pleural biopsy, it offers another approach to the diagnosis of suspected pleural TB, particularly in areas where pleural biopsy and/or histopathologic facilities are not readily available. Given the implications for the diagnostic work-up of suspected pleural tuberculosis, additional studies are needed to confirm the results of this study.
This work was supported in part by a grant from the Fogarty Institute International AIDS Training Grant (TW00043) and from Fundacao Carlos Chagas Filho de Amparo a Pesquisa do Estado do Rio de Janeiro/FAPERJ (E-26/170.206/2001). Received in original form November 13, 2001; accepted in final form November 8, 2002
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
