Interferon gamma

Background: Traditional diagnostic tests for pericardial tuberculosis (TB) are insensitive and often require long culture periods, and this has led to more emphasis being placed on biochemical tests such as the pericardial adenosine deaminase (ADA) test. However, controversy exists as to its diagnostic utility. In addition, the use of interferon (IFN)-[gamma], which is a reliable indicator of pleural and peritoneal TB, has not been explored in pericardial effusions. We investigated ADA and IFN-[gamma] levels in pericardial effusions of different etiologies.

Methods and results: A prospective study was carried out from February 1995 to February 1998 at Tygerberg Hospital (South Africa), with pericardial taps being performed under echocardiographic guidance. During this period, 110 consecutive patients presenting with large pericardial effusions were included in the study. Diagnoses were made according to predetermined criteria, and they included TB (n = 64), malignancy (n = 12), nontuberculous infections (n = 5), other effusions (n = 19), and effusions of uncertain origin (n = 10). The median ADA level in the tuberculous group was 71.7 U/L (range, 10.3 to 303.6 U/L), which was significantly higher than that in any other group (p < 0.05). With a cutoff level for ADA activity of 30 U/L, sensitivity was 94%, specificity was 68%, and positive predictive value was 80%. IFN-[gamma] levels were determined in 30 subjects. The median IFN-[gamma] concentration in the tuberculous group was > 1,000 pg/L, which was significantly higher than in any other diagnostic group (p < 0.0005). A cutoff value of 200 pg/L for IFN-[gamma] resulted in a sensitivity and specificity of 100% for the diagnosis of pericardial TB.

Conclusion: Pericardial fluid levels of ADA and IFN-[gamma] are useful in the diagnosis of tuberculous pericarditis.

Key words: adenosine deaminase; interferon-[gamma]; tuberculous pericarditis

Abbreviations: ADA = adenosine deaminase; IFN = interferon; NPV = negative predictive value; PPV = positive predictive value; TB = tuberculosis; ZN = Ziehl-Nielson

**********

Tuberculosis (TB) is a major health problem in South Africa, with an annual incidence rate of 350 per 100,000 population. (1) Approximately 1 to 2% of these cases are complicated by tuberculous pericarditis. (2) In the era before antituberculous therapy, tuberculous pericarditis was rapidly fatal, with an early mortality rate of > 80%. Since the introduction of chemotherapy in 1945, mortality from acute tuberculous pericarditis has decreased significantly (3); the mortality rate in South Africa is 3 to 17%, depending on the prescribed therapeutic regimen. (4,5)

There is considerable urgency in establishing the correct diagnosis so that appropriate treatment can be instituted; however, it is often difficult to establish a definitive bacteriologic diagnosis of tuberculous pericarditis. (3) The probability of obtaining a definitive diagnosis is greatest when pericardial fluid and a pericardial biopsy specimen are examined early in the effusive stage. (6,7) In most patients, this requires many weeks and extensive cultivation by multiple methods,s A normal pericardial biopsy result does not, however, exclude tuberculous pericarditis. In some patients, the examination of the entire pericardium removed at pericardiostomy or autopsy is required to demonstrate clear-cut evidence of TB. (6) Because of the difficulty in isolating the causative organism, pericardial TB often is missed. (9) For this reason, other diagnostic tools, such as pericardial adenosine deaminase (ADA) level, (3,10-12) have been suggested. In addition, interferon (IFN)-[gamma] also has been found to be a reliable marker for the presence of pleural (13-16) and peritoneal TB. (17) Its use in pericardial TB has not been explored.

We determined ADA activity and IFN-[gamma] levels in pericardial fluids from patients with effusions of various origins, thus evaluating the utility of these parameters in the diagnosis of pericardial TB.

MATERIALS AND METHODS

Study Population and Protocol

A prospective study was carried out from February 1995 to February 1998 at Tygerberg Hospital, South Africa. Consecutive patients who presented to the Echocardiography Department with large pericardial effusions were included in the study. All patients gave written informed consent for participation in the study, which was approved by the Ethics Committee of the University of Stellenbosch. Each patient was subjected to a full clinical examination, chest radiograph, ECG, echocardiograph, HIV testing, and sputum Ziehl-Nielson (ZN) stain. A pericardial tap wits performed under echocardiographic guidance through a pigtail catheter, and the fluid was sent for tests of biochemistry, ADA level, microbiology (including a TB culture and ZN stain), and hematology. An aliquot of pericardial fluid was frozen at -70[degrees]C within 30 min for cytokine analyses. The patient was followed up daily for signs of cardiac tamponade and/or recurrence of the effusion.

If no diagnosis could be made within 5 days, a pericardial biopsy was performed under general anesthetic. Biopsy tissue was sent for histology, and bacterial and TB culture. Where applicable, the patient then was started on a regimen of antituberculous therapy with isoniazid (80 mg), pyrazinamide (250 mg), and rifampin (120 mg; 1 tablet per 10 kg body weight) [Rifater; Aventis Phanna; Frankurt am Main, Germany], and the clinical response was evaluated. This treatment was continued for 6 months.

The hospital records of all patients were reviewed, and diagnoses were made according to predetermined criteria. Tuberculous pericarditis was diagnosed if the patient met one or more of the following criteria: (1) identification of the bacillus in pericardial fluid or biopsy specimen by stain and/or by culture, or by the presence of granulomas in pericardial biopsy tissue; (2) positive result of sputum ZN stain and/or culture in the presence of clinical and radiologic evidence of TB and in the absence of any other obvious cause associated with pericardial effusions; and (3) clinical and radiologic evidence of TB in the absence of any other obvious cause with a positive response to antituberculous therapy. Infective effusions included the following: (1) acute febrile illness and responsiveness to antibiotic treatment or identification of the organism in the pericardial fluid; (2) septicemia that is characterized by multisystem involvement in the presence of positive blood cultures; and (3) other obvious infective conditions in the absence of any other cause. Neoplastic effusions were diagnosed by the presence of cytologic and/or histologic evidence of malignancy with the exclusion of any other cause. Other effusions were defined by effusions that were clearly caused by collagen vascular disease, congestive heart failure, renal failure, and various other rare, but well-documented, causes of pericardial effusions. Idiopathic effusions were defined as those that were not due to any demonstrable cause. Full diagnostic workups had been performed on these patients, and all test results were negative. Patients having multiple superimposed diseases or effusions of unknown origin (that is, all possible etiologic causes could not be excluded) were classified as being of indeterminate origin.

Measurement of ADA

ADA activity was determined in all pericardial specimens according to the method described by Giusti. (18) This is a calorimetric method based on the measurement of the formation of ammonia by Berthelot reaction, which is produced when ADA acts on excess adenosine. One unit of ADA is defined as the amount of enzyme required to release 1 [micro]mol ammonia per minute from adenosine under standard assay conditions. The enzyme is stable for at least 24 h at 25[degrees]C, for 7 days at 4[degrees]C, and for 3 months at -20[degrees]C. (19,20)

Measurement of IFN-[gamma]

A human IFN-[gamma] enzyme-linked immunosorbent assay system (Biotrak; Amersham; Buckinghamshire, UK) that employs a quantitative "sandwich" enzyme immunoassay technique was used to determine the pericardial IFN-[gamma] concentration in 30 consecutive pericardial effusions due to TB, malignancy, or infection.

Statistical Analysis

The Wilcoxon two-sample test was employed for the analysis of data. The ADA activity and IFN-[gamma] concentration in tuberculous effusions was compared to the levels of activity in the various other diagnostic groups. The median and range of ADA activity and IFN-[gamma] concentrations in the various diagnostic classes was calculated. In addition, their utility as diagnostic tools for pericardial TB was evaluated at various cutoff levels by calculating the efficiency, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). These values then were compared on the basis of relative operating characteristic curves. (21)

RESULTS

During the 3-year study period, 110 patients presented to the Echocardiography Department with large pericardial effusions. Tuberculous pericarditis accounted for 64 effusions (58%), neoplastic conditions for 12 effusions (11%), infective conditions for 5 effusions (5%), and other effusions for 19 effusions (17%). The group of other effusions included the following etiologic factors: idiopathic (four effusions); post-trauma (four effusions); renal failure (four effusions); congestive heart failure (two effusions); HIV (one effusion); systemic lupus erythematosus (two effusions); systemic sclerosis (one effusion); and rheumatoid arthritis (one effusion). In 10 patients (9%), no diagnosis could be made, and for the purpose of this study these patients were regarded as having conditions that could not be attributed to TB (ie, true-negatives).

In the case of patients having tuberculous pericarditis, three diagnostic subclasses were identified: (1) identification of the bacillus in pericardial fluid or biopsy specimen by stain or by culture, or by the presence of granulomas in biopsy tissue (33 patients, of whom 9 had a positive sputum smear results as well); (2) positive results of sputum ZN stain and/or culture in the presence of clinical and radiologic evidence for TB (6 patients); and (3) clinical and radiologic evidence for TB associated with a response to antituberculous therapy (25 patients).

None of the patients with pericarditis due to nontuberculous causes had pericardial fluid or biopsy tests that were positive for TB. One patient with a malignant effusion had a false-positive sputum ZN stain result, and two patients who had clinical and radiologic evidence of TB did not respond to antituberculous therapy. The latter patients were classified as having effusions of indeterminate origin.

ADA Activity

No significant difference was found in ADA activity between the TB subclasses, or between HIV-positive and HIV-negative patients with TB. In addition, there was no significant correlation between the number of lymphocytes in the pericardial fluid and the corresponding ADA level.

The median values for ADA activity were determined for each of the diagnostic groups. Patients with tuberculous pleurisy had a median ADA level of 71.7 U/L (range, 10.3 to 303.6 U/L). The corresponding values for patients with infective, malignant, and other effusions were 41.4 U/L (range, 14.6 to 165.8 U/L), 26.8 U/L (range, 1.2 to 260.0 U/L), and 15.0 U/L (range, 0.8 to 103.2 U/L), respectively. The distribution of ADA activity for each of the diagnostic subclasses is shown in Figure 1. ADA activity was significantly higher for patients with tuberculous effusions than for the other diagnostic groups (p < 0.05 for each group).

[FIGURE 1 OMITTED]

Various levels of ADA were tested as the cutoff level for the diagnosis of TB, and 30 U/L was found to yield the best results. This corresponded to a sensitivity, specificity, PPV, and NPV of 94%, 68%, 80%, and 80%, respectively. The diagnostic efficiency of this test at this cutoff level was 83%. There were 15 false-positive cases. Four patients with infective effusions fell into this false-positive group. These patients included two HIV-positive patients with Staphylococcus aureus pneumonia (corresponding ADA levels, 62.4 and 165.8 U/L), one patient with Klebsiella pneumonia (corresponding ADA level, 41.4 U/L), and one patient with subacute bacterial endocarditis (corresponding ADA level, 39.0 U/L). Five patients presenting with neoplastic effusions had ADA levels of [greater than or equal to] 30 U/L, which included three patients having metastatic bronchus carcinoma (corresponding ADA levels, 31.4, 33.2, and 47.4 U/L), one patient with acute lymphoblastic leukemia (corresponding ADA level, 166 U/L), and one patient with mesothelioma (corresponding ADA level, 260.0 U/L). Of the three patients with other types of effusions, the condition of one patient was attributed to HIV (corresponding ADA level, 62.4 U/L), in another it was attributed to rheumatoid arthritis (corresponding ADA level, 65.3 U/L), and in the last one it was attributed to uremia (corresponding ADA level, 104.0 U/L). The remaining three false-positive effusions were classified as being of indeterminate cause. One patient was suspected of having TB (corresponding ADA level, 69.2 U/L) but died before any conclusive diagnosis could be made, while another patient was suspected of having a malignancy (corresponding ADA level, 34.2 U/L). The ADA level in the remaining patient was 51.7 U/L; the cause thereof was unknown.

Using the cutoff level of 30 U/L as being diagnostic of tuberculous pericarditis resulted in four false-negative results. One of these patients was HIV-positive (corresponding ADA level, 16.2 U/L), one patient was culture-positive for TB (corresponding ADA level, 15.8 U/L), and two patients were sputum smear-positive (corresponding ADA values, 10.3 and 18.0 U/L).

IFN-[gamma] Concentration

The distribution of pericardial IFN-[gamma] levels for the various diagnostic groups is shown in Figure 2. The median IFN-[gamma] concentration for the tuberculous group was > 1,000 pg/L and was significantly higher than those in other diagnostic classes (p < 0.0005). There was no statistical difference in the concentration of IFN-[gamma] between the different tuberculous subgroups. In addition, there were no statistical differences in ADA levels between tuberculous effusions in HIV-positive and HIV-negative patients. Using a cutoff level of 200 pg/L as being diagnostic for tuberculous pericarditis resulted in a 100% sensitivity and 100% specificity for TB in this substudy.

[FIGURE 2 OMITTED]

DISCUSSION

ADA (enzyme classification code, 3.5.4.4.), a polymorphic enzyme that is involved in purine metabolism, catalyzes the deamination of adenosine to inosine and ammonium. (22) Although it is found in most tissues, ADA activity is greatest in the lymphoid tissues, (22) its activity being 10 to 20 times more active in T lymphocytes than in B lymphocytes. (23) The presence of ADA in pleural fluid and other fluids reflects the cellular immune response in the fluid compartment, especially the activation of T lymphocytes. (24)

In this study, ADA levels were highest among patients in the TB group. Although infective conditions also may be associated with high ADA levels, a relative cell count can be used to distinguish between these two entities. (25) Tuberculous effusions are characterized by a relative lymphocytosis, and infective effusions are characterized by a neutrophil predominance. (26) High lymphocyte counts also can be found in effusions secondary to connective tissue disorders and malignancies, particularly those secondary to hematologic malignancies. (26)

Very few studies regarding the use of ADA in tuberculous pericarditis have been conducted. (3,10-12) The results of these studies are summarized in Table 1. The present study supports the use of ADA as a diagnostic tool for tuberculous pericarditis. Based on relative operating characteristic curves, (21) the best results are yielded at a cutoff level of 30 U/L, which corresponds to a sensitivity, specificity, PPV, NPV, and diagnostic efficiency of 94%, 68%, 80%, 89%, and 83%, respectively.

In this substudy, significantly elevated levels of IFN-[gamma] were demonstrated in patients with tuberculous pericarditis compared to those with both infective and malignant pericardial effusions (p < 0.0005). Elevated IFN-[gamma] concentrations in patients with tuberculous pleural and peritoneal effusions have been reported in a number of studies. (13-17) This has lead to the proposal that IFN-[gamma] be used as a diagnostic tool for tuberculous exudates. (14,16) Using a cutoff level of 200 pg/L as being diagnostic for tuberculous pericarditis resulted in a 100% sensitivity and 100% specificity for TB in this substudy.

Tuberculous pericarditis is traditionally diagnosed by the identification of Mycobacterium tuberculosis in pericardial fluid or biopsy specimens. Tuberculous pericarditis was diagnosed in only 33 of 64 patients (52%) in this manner. A further 15 patients (23%) were characterized by a positive result of sputum ZN staining in the presence of clinical and radiologic evidence (9 of these patients also had a TB-positive pericardial fluid or biopsy specimen). The diagnosis of TB in the remaining 25 patients (39%) was based on clinical and radiologic evidence for TB that was associated with a response to empirical antituberculous therapy. The results of these findings are summarized in Table 2.

In addition to the poor diagnostic efficiency of these traditional methods for diagnosing TB, long culture periods often result in clinical and therapeutic decisions being made before these laboratory results become available. (27) The use of pericardial ADA levels and pericardial IFN-[gamma] concentrations thus provides a rapid and accurate means of detecting tuberculous pericarditis, especially in high-prevalence areas, thereby expediting the initial decision-making process. IFN-[gamma] should be used to aid in the rapid diagnosis of tuberculous pericarditis. When IFN-[gamma] is not routinely available as a diagnostic test, we advocate a strategy of using ADA as a screening test and IFN-[gamma] as a confirmatory test, if diagnostic uncertainty for tuberculous pericarditis remains.

ACKNOWLEDGMENTS: We are indebted to the Provincial Administration, Western Cape, for the use of facilities and Annemarie Jacobs for assistance with manuscript preparation.

REFERENCES

(1) Colvin M, Abdool SSK. HIV infection among patients with tuberculosis in KwaZulu-Natal, South Africa [letter]. Int J Tuberc Lung Dis 1998; 2:172

(2) Larrieu AJ, Tyers GFO, Williams EH, et al. Recent experience with tuberculous pericarditis. Ann Thorac Surg 1980; 29:464-468

(3) Koh KK, Kim EJ, Cho CH, et al. Adenosine deaminase and carcinoembryonic antigen in pericardial effusion diagnosis, especially in suspected tuberculous pericarditis. Circulation 1994; 89:2728-2735

(4) Desai HN. Tuberculous pericarditis: a review of 100 cases. S Afr Med J 1979; 55:877-880

(5) Strang JIG, Kakaza HHS, Gibson DG, et al. Controlled clinical trial of complete open surgical drainage and of prednisolone in treatment of tuberculous pericardial effusion in Transkei. Lancet 1988; 2:759-764

(6) Barr JF. The use of pericardial biopsy in establishing etiologic diagnosis in acute pericarditis. Arch Intern Med 1955; 96: 693-696

(7) Sagrista-Sauleda J, Permanyer-Miralda G, Soler-Soler J. Tuberculous pericarditis: ten-year experience with a prospective protocol for diagnosis and treatment. J Am Coll Cardiol 1988; 11:724-728

(8) Cegielski JP, Devlin BH, Morris AJ, et al. Comparison of PCR, culture and histopathology for the diagnosis of tuberculous pericarditis. J Clin Microbiol 1997; 35:3254-3257

(9) Fowler NO. Tuberculous pericarditis. JAMA 1991; 266:99-103

(10) Martinez-Vazquez JM, Ribera E, Ocana I, et al. Adenosine deaminase activity in tuberculous pericarditis. Thorax 1986; 41:888-889

(11) Telenti M, Fdez J, Susano R, et al. Tuberculous pericarditis: diagnostic value of adenosine deaminase. Presse Med 1991; 20:637-640

(12) Komsuoglu B, Goldeh O, Kulan K, et al. The diagnostic and prognostic value of adenosine deaminase in tuberculous pericarditis. Eur Heart J 1995; 16:1126-1130

(13) Soderblom T, Nyberg P, Teppo AM, et al. Pleural fluid interferon-[gamma] in tuberculous and rheumatoid pleurisy. Eur Respir J 1996; 9:1652-1655

(14) Valdes L, San Jose E, Alvarez D, et al. Diagnosis of tuberculous pleurisy using the biologic parameters adenosine deaminase, lysozyme and interferon gamma. Chest 1993; 103:458-465

(15) Villena V, Lopez-Encuentra A, Echave-Sustaeta J, et al. Interferon-[gamma] in 388 immunocompromised and immunocompetent patients for diagnosing pleural tuberculosis. Eur Respir J 1996; 9:2635-2639

(16) Ogawa K, Koga H, Hirakata Y, et al. Differential diagnosis of tuberculous pleurisy by measurement of cytokine concentrations in pleural effusion. Tuber Lung Dis 1997; 78:29-34

(17) Ribera E, Martinez-Vazquez JM, Ocana I, et al. Diagnostic value of ascites gamma interferon levels in tuberculous peritonitis: comparison with adenosine deaminase activity. Tubercle 1991; 72:193-197

(18) Giusti G. Adenosine deaminase. In: Bergmeyer HU, ed. Methods of enzymatic analysis. New York, NY: Academic Press, 1974; 1092-1096

(19) Ellis G, Goldberg DM. A reduced nicotinamide adenine dinucleotide-linked kinetic assay for adenosine deaminase activity. J Lab Clin Med 1970; 76:507-517

(20) Heinz F. UV-method. In: Bergmeyer HU, ed. Methods of enzymatic analysis. 3rd ed. Weinheim, Germany: Verlag Chemie, 1984; 315-323

(21) Beck JR, Schultz EK. The use of relative operating characteristic (ROC) curves in test performance evaluation. Arch Pathol Lab Med 1986; 110:13-20

(22) Van der Weyden MB, Kelley WN. Human adenosine deaminase distribution and properties. J Biol Chem 1976; 251:5448-5456

(23) Barton RW, Goldshneider I. Nucleotide metabolizing enzymes and lymphocytic differentiation. Mol Cell Biochem 1979; 28:135-147

(24) Ocana I, Martinez-Vazquez JM, Segura RM, et al. Adenosine deaminase in pleural fluids: test for diagnosis of tuberculous pleural effusion. Chest 1983; 84:51-53

(25) Burgess LJ, Maritz FJ, Le Roux I, et al. Use of adenosine deaminase as a diagnostic tool for tuberculous pleurisy. Thorax 1995; 50:672-674

(26) Meyers DG, Meyers RE, Prendergast TW. The usefulness of diagnostic tests on pericardial fluid. Chest 1997; 111:1213-1221

(27) De Wit D, Maartens G, Steyn L. A comparative study of the polymerase chain reaction and conventional procedures for the diagnosis of tuberculous pleural effusion. Tubercle 1992; 73:262-267

* From the Departments of Chemical Pathology (Dr. Taljaard) and Cardiology (Drs. Burgess, Carstens, and Doubell, and Mr. Reuter), Tygerberg Hospital and University of Stellenbosch, Tygerberg, South Africa.

Manuscript received April 17, 2001; revision accepted February 6, 2002.

Correspondence to: Lesley J. Burgess, MMed, PhD, Tygerberg Hospital and University of Stellenbosch, PO Box 19174, Tygerberg 7505, South Africa; e-mail: lburgess@iafrica.com.

COPYRIGHT 2002 American College of Chest Physicians
COPYRIGHT 2003 Gale Group