Tuberculous Pleural Effusions

Tuberculous Pleural Effusions

Tuberculous Pleural Effusions· David M. Epstein, M.D., F.C.C.P.; Lewis R. Kline, M.D.; Steven M. Albelda, M.D.; and Wallace T. Miller, M.D. While a n...

802KB Sizes 0 Downloads 150 Views

Tuberculous Pleural Effusions· David M. Epstein, M.D., F.C.C.P.; Lewis R. Kline, M.D.; Steven M. Albelda, M.D.; and Wallace T. Miller, M.D.

While a number of recent reports have documented the changing clinical and radiographic spectrum of parenchymal tuberculosis, relatively little attention has been paid to changes in the patterns of pleural tuberculosis. We therefore reviewed the clinical, laboratory, and radiographic characteristics of 26 adult patients with tuberculous pleural effusions. We found that pleural tuberculosis has become a disease of older adults (median age, 56 years) and that 19 percent (5/26) of the cases were due to postprimary (reactivation) disease. This shift in age led to problems in diagnosis, since many of these older patients had underlying or coexisting disease that could have caused a pleural effusion. Bothspecimens of pleural ftuid and pleural biopsy

were useful in establishing the diagnosis. Examination of sputum was less helpful. All patients who were not anergic had positive cutaneous reactions to first-strength purified protein derivative of tuberculin. Lymphocytosis of the pleural 8uid was not a uniform 6nding; only 62 percent of our patients had greater than 50 percent lymphocytes on their initial examinations of pleural ftuid, and four patients had greater than 90 percent polymorphonuclear cells, Allof the effusions were exudates, and four had glucose levels in the pleural ftuid that were less than 30 mgldl. Pleural tuberculosis is an important diagnostic consideration in adult or elderly patients with exudative pleural effusions.

pleural tuberculosis remains an important treatable cause of exudative pleural effusions. 1.2 Traditionally, pleural tuberculosis has been considered a manifestation of primary tuberculosis seen largely in children and has, in fact, been used by some authors as a criterion for identifying cases of primary tuberculosis. 3The effusions are thought to be small or moderate in size, have high protein levels (above 5 g/dl) and glucose concentrations greater than 60 mglml, and be almost uniformly lymphocytic in nature. 3.4 While a number of recent reports have emphasized a changing clinical and radiographic spectrum of parenchymal tuberculosis in this country,3.5-7 relatively little attention has been paid to changes in the patterns of pleural tuberculosis. We have recently been impressed with the large number of older patients, commonly with evidence of preexisting tuberculosis, who have developed tuberculous effusions. The purpose of this report is to systematically review our recent experience with patients with tuberculous effusions, with special emphasis on the followingquestions: (1) What is the age distribution of pleural tuberculosis? (2) How many effusions could be attributed to primary vs postprimary (reactivation) tuberculosis? (3) How was the diagnosis of pleural tuberculosis successfully made? (4) Were the "traditional" concepts about pleural fluid findings valid in our patients?

MATERIALS AND METHODS

*From the Department of Radiology and the Division of Cardiovascular and Pulmonary Medicine, University of Pennsylvania School of Medicine and Hospital of the University of Pennsylvania, Philadelphia. Manuscript received May 16; revision accepted July 25. Reprint requests: Dr. Epstein, Hospital, University of Pennsylvania, 3400 Spruce Street, Philadelphia 19104

106

We identified 26 patients with a definite diagnosis of pleural tuberculosis over the past seven years (1978 to 1985) from the Hospital of the University of Pennsylvania and the Philadelphia Veterans Administration Hospital. Patients were identified by (1) a review of all cultures of pleural fluid and specimens from biopsy that were positive for Mycobacterium tuberculosis in the last seven years, (2) a computerized search of pathology records for pleural granulomas, and (3) a review of medical records indicating pleural tuberculosis as a diagnosis at discharge. From a review of the charts, each of the patients in this study met at least one of the following criteria for pleural tuberculosis: (1) positive culture for Mycobacterium tubercuiosis from pleural fluid or biopsy specimen; (2)positive smear for acid-fast bacilli from pleural fluid; (3)caseating granuloma in pleural biopsy; (4) culture of sputum positive for Mycobactenum tuberculosis with either an exudative pleural effusion or noncaseating granulomas in a pleural biopsy. We did not accept noncaseating pleural granulomas alone or a positive test with purified protein derivative of tuberculin (PPD) with an exudative pleural effusion as sufficient evidence for pleural tuberculosis, although most of these patients probably had pleural tuberculosis. Patients who met the previous criteria for pleural tuberculosis were then divided into primary and postprimary disease according to the following criteria: The criteria for primary disease were as follows: (1) documentation of a newly positive PPD test; (2) chest roentgenogram within one year showing no evidence of parenchymal tuberculosis: and (3)hilar adenopathy with or without parenchymal disease. The criteria for postprimary (reactivation) disease were (1) a history of a positive PPD test more than one year prior to development of a pleural effusion, and (2) prior treatment or diagnosis of pulmonary tuberculosis. Three patients could not be definitively classified as having primary or postprimary disease and are not included in the subsequent analysis. ~ of the available chest roentgenograms and medical records were reviewed with specific attention to symptoms, underlying disease, PPD status, biochemical and cellular analysis of the pleural fluid, method of diagnosis, additional radiographic findings, and clinical course. Tuberculous PleuralEffusions(Epstein at 81)

Table 3-AnalyN cf PleuralFluid*

Table I-Clinical Data Data

Primary Tuberculosis

Postprimary Tuberculosis

18

5 58 3 5 3 3 3 3

No. of patients Median age, yr Cough Fever Dyspnea Chest pain Weight loss Positive·PPD test Negative PPO test Anergic PPO test unavailable Coexisting disease Alcoholism Malignant disease Other

54

11 11 9 6 5

8 0 4

6 15 5

2

3 1 1 1

4 7

Total 23

14 16 12 9

8 11 0 4 8 18 6

5 8

RESULTS

Primary Tuberculosis

There were 18 patients (15 men and three women) with a median age of 54 years (range, 25 to 82 years) who had a pleural effusion due to primary tuberculosis. As shown in Table 1, the most frequent symptoms were fever and cough, which were observed in 11 patients each. Only one patient was asymptomatic. Fifteen patients had coexisting disease; five patients had alcoholism, three patients had carcinoma of the lung, colon, or bladder; and one patient had a lymphoma. Other identified medical problems included congestive heart failure, rheumatoid arthritis, diabetes mellitus, renal transplantation, asthma, drug abuse, and gastric surgery for ulcer disease. The diagnosis was made equally by findings in the pleural fluid and pleural biopsy (eight each) (Table 2). Twopatients had positive results from both the pleural fluid and biopsy. Invariably, the results from biopsy were available prior to growth of the culture, which takes six weeks. The presence of granulomas alone in the specimen from biopsy was sufficient evidence for treatment until cultures from the biopsy or pleural fluid became available. The diagnosis was missed in

No. of patients Pleural biopsy Granulomas Positive culture Positive smear Pleural 8uid Positive culture Positive smear Positive culture of sputum

Primary Tuberculosis

Postprimary Tuberculosis

Total

18

5

23

10 6 1

3 3 1

13 9 2

7 2

Primary Tuberculosis

8 2 1

Postprimary Thberculosis

4.5 (2.2-6.8) 5.8 (4.1-8.4) Protein, g/dl 62 (0-150) 104 (42-106) Glucose, mgldl 527 (22-5,600) 1,491 (272-2,978) LOR, IV 3,800 (200-198,000) 2,300 (2,300-8,050) WBCs per cu mm Polymorphonuclear 40 (2-97) 31 (5-32) leukocytes, percent 69 (68-84) Lymphocytes, percent 65 (3-97) *Thble data are median values; numbers within parentheses are ranges.

five patients who had a subsequent positive culture from the pleural Huid. Four of these patients never had a pleural biopsy. The other patient had a negative pleural biopsy. The PPD reaction was positive in all patients who were not anergic. Biochemical and cellular analysis of the pleural fluid was available in 16 patients and is summarized in Table 3. Seven (44 percent) of 16 patients had 50 percent or greater polymorphonuclear leukocytes; four of these patients had greater than 90 percent. Three patients had a severely depressed pleural fluid glucose level of less than 30 mgldl. The radiographic findings are presented in Table 4. The one patient with bilateral parenchymal disease had bilateral pleural effusions. This patient was considered to have primary progressive tuberculosis. Only one patient had ipsilateral hilar adenopathy with a pleural effusion. Radiographic follow-up was available in 12 patients. Ten patients showed improvement to complete resolution, with or without treatment within one year: One of these patients had initial improvement with therapy, with a recrudescence of effusion at three months. She then had subsequent resolution of her pleural effusion. One"patient who was not treated showed no change in the pleural effusion at Table 4-Chat Roentgenogram in PlBural 7Uberculoaia Oata No. of patients Pleural effusion Right

Table I-Method ofDitJgnoBia ofPleuralTuberculoBia Data

Data

Left

Bilateral Small Medium Large Parenchymal disease Right

Left

Primary Thberculosis

Postprimary Thberculosis

Total

18

5

23

11 5

4 .1

15 6

3

2 7

2

4 9 4

Bilateral Same side as effusion Opposite side from effusion Adenopathy

1 1 1 1

2

1 2

CHEST I 91 I 1 I JANUARY. 1987

11 4 0 1 3 1 1 1

107

one year; and one patient who also was not treated showed an increase in pleural effusion.

Postprimary Tuberculosis There were five patients (19 percent) with documented pleural effusion due to postprimary tuberculosis (four men and one woman). The median age was 58 years (range, 51 to 63 years). All five patients were symptomatic (Table 1). Three of the five patients had coexisting disease, including gastric surgery for peptic ulcer disease, chronic lymphocytic leukemia, and chronic alcoholism. Two of the five patients had had prior treatment with artificial pneumothorax for tuberculosis, and the other three had a positive PPD test for greater than one year prior to the development of pleural effusion. One of these patients was also treated with isoniazid. The method of diagnosis is summarized in Table 2. The diagnosis was missed in one patient who had a positive culture of pleural fluid and negative biopsy. Biochemical analysis of the pleural fluid was available in four patients and is presented in Table 3. Cell counts were available in three of the four patients, and all cell counts revealed a strong lymphocytic predominance. None of the effusions had a glucose level of less than 30 mg/dl. The radiographic findings are listed in Table 4. Interestingly, one patient with unilateral parenchymal disease had his pleural effusion on the opposite side. Three of the five patients showed improvement to complete resolution of the pleural effusion within one yean One patient had no change at four months and was lost to follow-up. The last patient showed residual loculated pleural effusion at two years after diagnosis. This patient was the only one who did not receive antituberculosis chemotherapy. DISCUSSION

When a tuberculous pleural effusion occurs in the absence of radiologically apparent tuberculosis, it is usually considered to be the sequel to a primary infection three to six months previously; however; tuberculous pleural effusion may occur at any stage of active infection and may be seen with both primary and postprimary disease. Tuberculous pleural effusion is thought to result from rupture of a subpleural caseous focus in the lung into the pleural space.Y The fluid is generally a serous exudate but may be serosanguineous and usually contains few tubercle bacilli. It accumulates most probably as a hypersensitivity reaction to tuberculoproteins. Rarely, tuberculous pleural effusion may be the result of hematogenous dissemination or contamination by adjacent infected lymph nodes. Although tuberculous pleurisy initially is a self-limited disease, 65 percent of the untreated patients develop active pulmonary or extrapulmonary disease within 108

five years of its occurrence. 2.8 'Iraditionally, primary tuberculosis has been considered a disease of childhood. Therefore, an isolated pleural effusion in an adult, and particularly in an elderly patient, is often thought to be due to a disease process other than tuberculosis; however, as the overall incidence of tuberculosis has been declining, increasing numbers of individuals are entering adulthood without prior tuberculous exposure.v" The increased prevalence of pulmonary tuberculosis in young adults is thought to be related to healthy susceptible individuals living or working in crowded environments or under poor socioeconomic conditions. Primary tuberculosis in the elderly may be related to these same poor living conditions, with increased susceptibility caused by chronic debilitating disease and immunosuppression." Several authors have noted an increased incidence of primary tuberculosis in older patients and have attributed some of the "unusual radiographic manifestations" to actually represent the more usual aspects of primary disease seen in the unsuspected adult population." Indeed, our observations support this concept; the median age of our patients with primary tuberculous pleurisy was 54, with six (33 percent) of 18 patients greater than 60 years of age. The shift in age in the incidence of pleural tuberculosis can lead to problems in diagnosis. Many older patients have underlying or coexisting disease which may cause a pleural effusion. Congestive heart failure, cancer, pneumonia, and pulmonary embolism are common diseases in the elderly that can cause pleural effusions. Interestingly, five of our patients older than 55 years with primary tuberculosis had such coexisting diseases that actually obscured the correct diagnosis. These patients included two with hepatic disease and ascites, two with congestive heart failure (one of whom also had rheumatoid arthritis), and one with histiocytic lymphoma. None had their diagnosis recognized and were lost to follow-up without receiving appropriate antituberculosis chemotherapy. The correct diagnosis was subsequently established from a positive culture of pleural fluid six weeks later. In our series an almost equal number of patients had the diagnosis of pleural tuberculosis established from smear and culture of the pleural fluid as from pleural biopsy including microscopy and culture. Three patients had a positive pleural biopsy and pleural fluid. In most series the diagnostic yield from culture of pleural fluid is less than 30 percent, although Sibley" reported positive cultures in 70 percent of his eases.v' We had positive cultures of pleural fluid in 40 percent. Pleural biopsy was performed in 16 patients, 14 of whom had a positive diagnosis established with either granulomas or culture (88 percent). These results are comparable to other series in the literature. 2.4 Although we did not accept the isolated finding of noncaseating granulomas Tuberculous Pleural Effusions (Epstein fit aJ)

as diagnostic of pleural tuberculosis, we did consider this a positive result when there was other corroborating evidence of tuberculosis including a positive culture, positive sputum, or positive pleuralfluid. In the six patients where the diagnosis of tuberculosis was missed, all had positive cultures of pleural fluid. This emphasizes the importance of careful follow-up of results of cultures, even with a negative biopsy. Cultures of sputum were useful in only 20 percent of the patients with coexisting parenchymal tuberculosis. The analysis of pleural fluid from our patients revealed a number of interesting findings. While previous investigators have found that the great majority of their patients had highly lymphocytic effusions (Berger and Mejia' report only 88 percent of effusions had more than 50 percent lymphocytes), only 62 percent (12/19) of our patients had more than 50 percent lymphocytes on their initial examination of pleural fluid. This was particularly impressive in those patients with primary tuberculosis, where seven (43 percent) of 16 had a polymorphonuclear leukocytic predominance. Although it is well known that serial specimens of pleural fluid may reveal a progression from predominantly polymorphonuclear leukocytes, we were impressed by the number of patients with granulocytic predominance on the initial examination. Unfortunately, we have no information on serial changes. As might be expected, all pleural effusions were exudative either by protein or lactic dehydrogenase (LDH) criteria, but only 10 (50 percent) of 20 had greater than 5 g of protein. Furthermore, four patients had a pleural fluid glucose level of less than 30 mg/dl. This very low glucose level was not seen in any of the patients of Berger and Mejia." In their experience, such a low glucose level was only seen in patients with malignant neoplasms, rheumatoid arthritis, and empyema. Two of our four patients may have had such coexisting diagnoses with their pleural tuberculosis. One of these patients had rheumatoid arthritis and the other a bronchopleural fistula with presumed empyema that was culture negative except for Mycobacterium tuberculosis. The pleural fluid glucose level was greater than 50 mg/dl in 13 (65 percent) of 20 of our patients, which is considered typical of tuberculous pleural effusions. Mesothelial cells are said to be

absent or scarce in tuberculous effusion. 2 We could not document this observation. In summary, pleural tuberculosis is an important diagnosis to be considered in adult or elderly patients presenting with an exudative pleural effusion. Despite their age, most of these patients will have primary tuberculosis, and pleural effusions without hilar adenopathy may be the sole radiographic finding in adults with primary tuberculosis; however; it should be remembered that tuberculous effusion can occur in association with reactivation tuberculosis. Many of these patients are debilitated or immunosuppressed. Coexisting underlying disease which can commonly produce pleural effusions may hamper establishment of the correct diagnosis. Furthermore, pleural fluid . lymphocytosis is not a reliable indicator of pleural tuberculosis. We suggest that a PPD test with anergy panel, pleural fluid culture, and pleural biopsy be performed in any adult with an undiagnosed pleural effusion to rule out the possibility of pleural tuberculosis. REFERENCES 1 Hirsch A, Ruffie ~ Nebut M, Bignon J, Chretien J. Pleural effusion: laboratory tests in 300 cases. Thorax 1979; 34:106-12 2 Light RW Tuberculous pleural effusions. In: Light RW, ed. Pleural diseases. Philadelphia: Lea and Febiger, 1983:119-25 3 Khan M, Kovnat D, Bachus B, Whitcomb M, BrodyJ, Snider G. Clinical and roentgenographic spectrum of pulmonary tuberculosis in the adult. JAMA 1977; 62:31-8 4 Berger HW, Mejia E. Tuberculous pleurisy. Chest 1973; 63: 88-92 5 Miller WI: MacGregor RR. Thberculosis: frequency of unusual radiographic findings. Am J Roentgenoll978; 130:867-75 6 Choyke PL, Sostman HD, Curtis AM, Ravin CE, Chen rn: Godwin JD, et ale Adult-onset pulmonary tuberculosis. Radiology 1983; 148:357-62

7 Hadlock F~ Park SK, Awe RJ, Rivera M. Unusual radiographic findings in adult pulmonary tuberculosis. AmJ RoentgenoI1980; 134:1015-18 8 Sibley JC. A study of 200 cases of tuberculous pleurisy with effusion. Am Rev Tuberc 1950; 62:314-23 9 Woodring JH, MacVandiviere H, Fried AM, Dillon ML, Williams TD, Melvin IG. Update: the radiographic features of pulmonary tuberculosis. Am J Roentgenoll986; 146:497-506 10 Stead WW, Kerby GR, Schlueter D~ Jordahl CW The clinical spectrum of primary tuberculosis in adults: confusion with reinfection in the pathogenesis of chronic tuberculosis. Ann Intern Med 1968; 68:731-45

CHEST I 91 I 1 I JANUARY. 1987

109