- Email: [email protected]
Constrictive pericarditis: Detection of mycobacterium tuberculosis in paraffin-embedded pericardial tissues by polymerase chain reaction
Clinical Biochemistry 40 (2007) 355 – 358
Constrictive pericarditis: Detection of mycobacterium tuberculosis in paraffin-embedded pericardial tissues by polymerase chain reaction Mahmood Zamirian a , Mehdi Mokhtarian a , Mohammad Hosein Motazedian b , Ahmad Monabati c , Gholam Reza Rezaian a,⁎ a
Department of Internal Medicine (Cardiology), Namazee Hospital, Shiraz University of Medical Sciences, Shiraz, Iran b Department of Microbiology, Medical School, Shiraz University of Medical Sciences, Shiraz, Iran c Department of Pathology, Medical School, Shiraz University of Medical Sciences, Shiraz, Iran Received 8 August 2006; received in revised form 3 December 2006; accepted 27 December 2006 Available online 13 January 2007
Abstract Background: Although the utility of polymerase chain reaction (PCR) for diagnosis of acute pleuro-pericardial tuberculosis has been well established, its use for chronic constrictive pericarditis is yet to be reported. Aims: To define the sensitivity and specificity of PCR for diagnosis of tuberculosis (TB) in patients with constrictive pericarditis. Methods: The medical records of 30 consecutive patients with constrictive pericarditis were reviewed. In addition their historical paraffinembedded pericardial tissues were used for new histopathologic examination and PCR amplification for Mycobacterium tuberculosis genome. Results: There were 23 males and 7 females with a mean age of 35 ± 19.5 years. The anticipated causes of constriction included idiopathic (n = 21), tuberculosis (n = 5), cardiac surgery (n = 2) and post traumatic (n = 2). PCR became positive in nine patients. Four out of 5 patients with tuberculous granuloma had a positive test result. In addition all 4 patients with non-tuberculous constrictive pericarditis had a negative test result. Therefore considering the presence or absence of granuloma as a diagnostic criteria, the sensitivity and specificity of PCR were 4/5 (80%) and 20/ 25 (80%), respectively. © 2007 Published by The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. Keywords: Constrictive pericarditis; Mycobacterium tuberculosis; Polymerase chain reaction
Introduction
Material and methods
Tuberculous constrictive pericarditis is often a diagnostic challenge because of its rarity and non-specific clinical manifestations. The need for pericardiectomy and the inefficiency of the conventional diagnostic methods mandate the evaluation of newer tests. Although nucleic acid amplification techniques such as polymerase chain reaction (PCR) have improved the diagnosis of tuberculous pulmonary diseases [1,2], however, its utility for constrictive pericarditis is yet to be examined. Here we report our experience with PCR to diagnose tuberculosis in 30 such patients.
The medical records of 30 consecutive patients with constrictive pericarditis were reviewed. In addition their paraffin embedded pericardial tissues were obtained from the archives of our pathology department for DNA detection of mycobacterium tuberculosis. The tissue blocks were collected between 1970 and 2004. Histopathologic and PCR assessments were independently done by a pathologist and a microbiologist who were blinded to the patient's initial (admission time) clinical and pericardial histopathologic findings. Patients
⁎ Corresponding author. E-mail address: [email protected] (G. Reza Rezaian).
There were 23 (77%) males and seven (23%) females with an age range of 7 to 74 (35 ± 19.5) years. Constrictive pericarditis
0009-9120/$ - see front matter © 2007 Published by The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.clinbiochem.2006.12.005
356
M. Zamirian et al. / Clinical Biochemistry 40 (2007) 355–358
was diagnosed on the basis of clinical manifestations and a combination of echocardiographic and hemodynamic studies and finally confirmed by pericardiectomy and histologic examination. All patients except those with a proven underlying etiology for constriction had received antituberculous therapy for at least 2 weeks prior to pericardiectomy. The mean duration of symptoms before pericardiectomy was 49.6 ± 46.3 months with a minimum of 2 months and maximum of 180 months. The cause of constriction was reported to be idiopathic in 21, tuberculosis in 5, prior cardiac surgery in 2 and penetrating cardiac trauma by shell fragments in the remaining 2 patients. Of the 5 patients with tuberculosis, one had TB adenitis and another had TB osteomyelitis as well.
performed in a thermocycler (Progene, Modle Fprogo5D, Techne, Cambridge UK) employing a protocol with an initial denaturation step for 5 min at 95 °C; 35 cycles each for 0.5 min at 95 °C, 0.5 min at 68 °C, and 1 min at 72 °C; and a final extension segment for 5 min at 72 °C. The products were refrigerated until used. Controls for each PCR run included (i) a positive control, 100 fg of M. tuberculosis DNA; (ii) an additional important negative control sample: DNA sample negative for mycobacterium. PCR products were subjected to electrophoresis in 1.5% agarose gel then visualized by ethidium bromide staining and UV light detection.
New histopathologic studies
Kappa and its P value were calculated using SPSS software (version 11.5) to measure the degree of agreement between the results of PCR and histopathology. A P value <0.05 was considered significant.
New preparations for Ziel-Neelsen (ZN) and Hematoxylin– Eosin (HE) staining were prepared from the paraffin-embedded pericardial specimens and all were reviewed by a pathologist. Ziel-Neelsen staining was done to detect acid fast bacilli.
Statistical methods
Results
Polymerase chain reaction (PCR)
Histopathologic findings
Sample preparation Eight 15 μm sections from each paraffin embedded pericardial tissue were separately placed in two 1.5 mL Eppendorf tubes (4 pieces in each tube) and were sent to the microbiology laboratory. After each sampling the microtome blade and other tools were meticulously washed by hypochlorite solution.
Histopathologic examination showed fibrosis in all patients' (100%) chronic inflammation in 16 (53%), calcification in 9 (30%), and tuberculous granuloma in 5 (17%) patients. No acid fast bacilli, however, could be detected in the tissue samples.
DNA extraction Each sample was deparaffinized by adding 0.5 mL xylene to the Eppendorf tubes and mixing on rotary platform for 30 min at room temperature. The supernatant was then discarded after 10 min of centrifugation at 6000×g. The same process was repeated once more for each sample. To extract the remaining xylene, 1 mL of absolute ethanol was added to each sample and vortexing was done for 5 min. The mixture was then pelleted by centrifugation at 6000×g for 10 min and the ethanol was removed. The xylene extraction procedure was repeated twice for each sample. Two drops of acetone were then added and the tubes were incubated open in a water bath for 15 min at 60 °C. Subsequently the dried samples were resuspended in 200 μL of digestion buffer (50 mmol/lit Tris–HCl [pH8], 1 mmol/lit EDTA, and 0.05% sodium dodecyl sulfate) containing 800 μg/ mL of proteinase K. Mycobacterial DNAs from the samples were amplified using the following primers: IS6110 a (5′-CCTGCGAGCGTAGGCGTCGG-3′) and IS6110 b (5′-CTCGTCCAGCGCCGCTTCGG-3′) [3]. These primers amplify a fragment with a length of 123 bp. PCR amplifications were carried out in 50-μL volumes with 5 μL of prepared sample added. The mixture contained 50 pmol concentrations of each primer; 200 μM each of dNTP; 2 mM MgCl2; 5 μL PCR buffer; and 1 U of TAQ DNA polymerase (TIB Molbiol, Berlin, Germany). PCR runs were
PCR results PCR became positive in 9 patients (Table 1). Four out of 5 patients with tuberculous granulomas had a positive PCR. Therefore the results of PCR and the presence of tuberculous granuloma had a high measure of agreement (Kappa = 0.455 with P = 0.008). In addition, all 4 patients with definite nontuberculous constrictive pericarditis had negative PCR results. The mean duration of performing PCR after pericardiectomy was 10 ± 7.1 years. It was 12.5 ± 7.8 years in those with a positive PCR and 12.5 ± 8.7 years in those with a negative test result (P = NS). The difference was not significant in those with and without granuloma either (13.8 ± 6.1 vs. 13.8 ± 9.2 years P = NS). Table 1 The result of PCR in 30 patients with constrictive pericarditis with and without tuberculous granulomas in tissue biopsy Tuberculous granuloma in biopsy specimen
PCR
+ − Total
PCR = Polymerase Chain Reaction. + = positive. − = negative. Sensitivity: 4/5 = 80%. Specificity: 20/25 = 80%.
+
−
4 1 5
5 20 25
M. Zamirian et al. / Clinical Biochemistry 40 (2007) 355–358
Discussion Constrictive pericarditis represents the end stage of any inflammatory process involving the pericardium. In developed countries the etiology is most commonly infectious, post surgical or radiation injury. Acute tuberculous pericarditis although rare, can ultimately result in chronic constriction especially when remained undiagnosed and left untreated. This is particularly true in the developing world. Diagnosis of tuberculous pericarditis is notoriously difficult and because of its high prevalence in our country, empiric therapy is routinely done for many patients with constrictive pericarditis even in the absence of a definite diagnosis of tuberculosis. In fact rapid diagnosis and treatment is crucial to reducing mortality and morbidity from tuberculosis. Adenosine deaminase (ADA) and ã-IFN levels in pericardial fluid have been used for diagnosis of tuberculous pericarditis [4]. The limitation of these tests, however, is that they cannot be utilized in majority of patients with chronic constrictive pericarditis due to lack of any fluid in their pericardial sacs. Polymerase chain reaction (PCR), too has been used for diagnosis of acute tuberculous pericarditis [5]; however, its usefulness for constrictive pericarditis is to be reported. Therefore, we evaluated the usefulness of PCR for diagnosis of tuberculous constrictive pericarditis in paraffin embedded pericardial tissues from 30 patients with constrictive pericarditis diagnosed in our university hospitals during the past 25 years. Paraffin embedded tissues from other sites including lymph node and pleura have been successfully used for PCR [6,7]. AFB staining is insensitive since it requires about 104 organisms per milliliter of specimen to become positive [2], and it lacks specificity as well [2]. In none of our tissues samples even in those with tuberculous granulomas could AFB bacilli be seen. Although the standard detection method is culture, however, it usually takes several weeks for it to become positive, especially if the specimen contains a low number of organisms. Therefore we had to rely on the pathologic findings of granuloma and consider it as the reference test for evaluation of PCR results. The diagnosis of pleural tuberculosis too, which is the second most frequent human tuberculous presentation is usually based on the presence of necrotizing granulomas in pleural biopsy specimens [8]. Sensitivity for PCR was 80% since it was positive in 4 (80%) of the patients who had tuberculous granulomas in their pericardial specimens. In addition it turned positive in another 5 patients with no granulomas and who were previously labeled as having idiopathic chronic constrictive pericarditis. Histopathologic findings, however, could be non-specific even in patients with acute TB pericarditis [5]. The specificity of PCR in our study was 80%; this, however, may be an underestimation because pathology may be inconclusive in up to 30% of patients [5]. So patients with no granulomas should not be considered as truly negative cases especially in areas like ours where tuberculosis is encountered frequently.
357
Negative pathologic results are most common in cases where symptoms have been present for a long period or where treatment has been initiated long before the biopsy or pericardiectomy. The mean duration of symptoms before pericardiectomy was 49.6 ± 46.3 months in our patients and 21 of them had received antituberculous therapy for at least 2 weeks before pericardiectomy. Depending on the gold standard test chosen, PCR has been reported to have sensitivities ranging from 77% to more than 95% and PCR specificities of more than 95% for smear positive specimens [6]. Interestingly, all 4 patients who had post cardiac surgical and traumatic constrictive pericarditis had negative PCR, showing the high specificity of the test. PCR turned to be negative in one of our patients with tuberculous granuloma. False negative results have reported to be associated with the age of paraffinembedded tissues, i.e., the older the sample, the higher is the chance for false negativity [9]. Overall tuberculosis was the most possible underlying etiology of constrictive pericarditis in 33% of our patients by PCR. Our results are similar to that reported in Turkey between the years 1983 and 1993 since tuberculosis was the etiology of pericardial constriction in 38% of their patients [10]. Excluding the patients with tuberculosis (9 with positive PCRs and 1 with falsely negative PCR) and four patients with post cardiac surgery and traumatic constrictive pericarditis, the remaining 16 (53.3%) patients must be considered as having idiopathic constrictive pericarditis which has been reported in 33 to 73% range [11] in other studies. Constrictive pericarditis, even when diagnosed and surgically corrected is associated with a high mortality. In their series of 163 patients with surgically proven constrictive pericarditis, Bertog et al. found 10% perioperative mortalities and another 61 deaths after a median follow-up of 6.9 years among the survivors [12]. Therefore correct diagnosis and appropriate treatment are crucial for the long-term survival of these patients and tuberculous constrictive pericarditis is no exception. Although the conventional technique of direct smear examination with ZN staining is easy to perform and cheap, it has a low sensitivity and specificity [2]. Culture takes a long time specially if the number of organisms is few, and histopathologic findings could be non-specific [3]. New diagnostic tests, therefore, are very much needed and molecular amplification techniques such as PCR could be promising. Thus in areas with a high prevalence of tuberculosis like ours, PCR can be utilized as a useful, rapid and reliable diagnostic test for detection of tuberculous constrictive pericarditis even in patients with long-lasting symptoms and in those with no foot prints of the causative organism in tissue samples. References [1] Kivihya-Ndugga L, Van Cleeff M, Juma E, Kimwomi J, Githui W, Oskam L, et al. Comparison of PCR with the routine procedure for diagnosis of tuberculosis in a population with high prevalences of tuberculosis and human immunodeficiency virus. J Clin Microbiol 2004;42:1012–5.
358
M. Zamirian et al. / Clinical Biochemistry 40 (2007) 355–358
[2] Bogard M, Vincelette J, Antinozzi R, Alonso R, Fenner T, Schirm J, et al. Multicenter study of a commercial, automated polymerase chain reaction system for the rapid detection of mycobacterium tuberculosis in respiratory specimens in routine clinical practice. Eur J Clin Microbiol Infect Dis 2001;20:724–31. [3] Eisenach KD, Donald Cave M. Polymerase chain reaction amplification of a repetitive DNA sequence specific for Mycobacterium tuberculosis. J Infect Dis 1990;161:977–81. [4] Koh KK, Kim EJ, Cho CH, Choi MJ, Cho SK, Kim SS, et al. Adenosine deaminase and carcinoembryonic antigen in pericardial effusion diagnosis, especially in suspected tuberculous pericarditis. Circulation 1994;89: 2728–35. [5] Cegielski JP, Delvin BH, Morris AJ, Kitinya JN, Pulipaka UP, Lema LEK, et al. Comparison of PCR, culture, and histopathology for diagnosis of tuberculous pericarditis. J Clin Microbiol 1997;35:3254–7. [6] Kaul KL. Molecular detection of mycobacterium tuberculosis: impact on patient care. Clin Chem 2001;47:1553–8. [7] Ruiz-Manzano J, Manterola JM, Gamboa F, Calatrava A, Monoso E, Martinez V, et al. Detection of mycobacterium tuberculosis in paraffin-
[8] [9]
[10]
[11]
[12]
embedded pleural biopsy specimens by commercial ribosomal RNA and DNA amplification kits. Chest 2000;118:648–55. Light RW, Ferrer J. Diagnostico de la tuberculosis pleural. Arch Bronchoneumol 1999;35:105–7. Hirunwiwatkul P, Tumwasorn S, Chantranuwat C, Sirichai U. A comparative study of diagnostic tests for tuberculous lymphadenitis: polymerase chain reaction vs histopathology and clinical diagnosis. J Med Assoc Thai 2002;85:320–6. Oh KY, Shimizu M, Edwards WD, Tazelaar HD, Danielson GK. Surgical pathology of parietal pericardium: a study of 344 cases (1993–1999). Cardiovasc Pathol 2001;10:157–68. Ling LH, Oh JK, Schaff HV, Danielson GK, Mahoney DW, Seward JB, et al. Constrictive pericarditis in the modern era: evolving clinical spectrum and impact on outcome after pericardiectomy. Circulation 1999; 100:1380–5. Bertog SC, Thambidorai SK, Parakh K, Schoenhagen P, Ozduran V, Houghtaling PL, et al. Constrictive pericarditis etiology and causespecific survival after pericardiectomy. J Am Coll Cardiol 2004;43: 1445–52.
Constrictive pericarditis: Detection of mycobacterium tuberculosis in paraffin-embedded pericardial tissues by polymerase chain reaction Mahmood Zamirian a , Mehdi Mokhtarian a , Mohammad Hosein Motazedian b , Ahmad Monabati c , Gholam Reza Rezaian a,⁎ a
Department of Internal Medicine (Cardiology), Namazee Hospital, Shiraz University of Medical Sciences, Shiraz, Iran b Department of Microbiology, Medical School, Shiraz University of Medical Sciences, Shiraz, Iran c Department of Pathology, Medical School, Shiraz University of Medical Sciences, Shiraz, Iran Received 8 August 2006; received in revised form 3 December 2006; accepted 27 December 2006 Available online 13 January 2007
Abstract Background: Although the utility of polymerase chain reaction (PCR) for diagnosis of acute pleuro-pericardial tuberculosis has been well established, its use for chronic constrictive pericarditis is yet to be reported. Aims: To define the sensitivity and specificity of PCR for diagnosis of tuberculosis (TB) in patients with constrictive pericarditis. Methods: The medical records of 30 consecutive patients with constrictive pericarditis were reviewed. In addition their historical paraffinembedded pericardial tissues were used for new histopathologic examination and PCR amplification for Mycobacterium tuberculosis genome. Results: There were 23 males and 7 females with a mean age of 35 ± 19.5 years. The anticipated causes of constriction included idiopathic (n = 21), tuberculosis (n = 5), cardiac surgery (n = 2) and post traumatic (n = 2). PCR became positive in nine patients. Four out of 5 patients with tuberculous granuloma had a positive test result. In addition all 4 patients with non-tuberculous constrictive pericarditis had a negative test result. Therefore considering the presence or absence of granuloma as a diagnostic criteria, the sensitivity and specificity of PCR were 4/5 (80%) and 20/ 25 (80%), respectively. © 2007 Published by The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. Keywords: Constrictive pericarditis; Mycobacterium tuberculosis; Polymerase chain reaction
Introduction
Material and methods
Tuberculous constrictive pericarditis is often a diagnostic challenge because of its rarity and non-specific clinical manifestations. The need for pericardiectomy and the inefficiency of the conventional diagnostic methods mandate the evaluation of newer tests. Although nucleic acid amplification techniques such as polymerase chain reaction (PCR) have improved the diagnosis of tuberculous pulmonary diseases [1,2], however, its utility for constrictive pericarditis is yet to be examined. Here we report our experience with PCR to diagnose tuberculosis in 30 such patients.
The medical records of 30 consecutive patients with constrictive pericarditis were reviewed. In addition their paraffin embedded pericardial tissues were obtained from the archives of our pathology department for DNA detection of mycobacterium tuberculosis. The tissue blocks were collected between 1970 and 2004. Histopathologic and PCR assessments were independently done by a pathologist and a microbiologist who were blinded to the patient's initial (admission time) clinical and pericardial histopathologic findings. Patients
⁎ Corresponding author. E-mail address: [email protected] (G. Reza Rezaian).
There were 23 (77%) males and seven (23%) females with an age range of 7 to 74 (35 ± 19.5) years. Constrictive pericarditis
0009-9120/$ - see front matter © 2007 Published by The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.clinbiochem.2006.12.005
356
M. Zamirian et al. / Clinical Biochemistry 40 (2007) 355–358
was diagnosed on the basis of clinical manifestations and a combination of echocardiographic and hemodynamic studies and finally confirmed by pericardiectomy and histologic examination. All patients except those with a proven underlying etiology for constriction had received antituberculous therapy for at least 2 weeks prior to pericardiectomy. The mean duration of symptoms before pericardiectomy was 49.6 ± 46.3 months with a minimum of 2 months and maximum of 180 months. The cause of constriction was reported to be idiopathic in 21, tuberculosis in 5, prior cardiac surgery in 2 and penetrating cardiac trauma by shell fragments in the remaining 2 patients. Of the 5 patients with tuberculosis, one had TB adenitis and another had TB osteomyelitis as well.
performed in a thermocycler (Progene, Modle Fprogo5D, Techne, Cambridge UK) employing a protocol with an initial denaturation step for 5 min at 95 °C; 35 cycles each for 0.5 min at 95 °C, 0.5 min at 68 °C, and 1 min at 72 °C; and a final extension segment for 5 min at 72 °C. The products were refrigerated until used. Controls for each PCR run included (i) a positive control, 100 fg of M. tuberculosis DNA; (ii) an additional important negative control sample: DNA sample negative for mycobacterium. PCR products were subjected to electrophoresis in 1.5% agarose gel then visualized by ethidium bromide staining and UV light detection.
New histopathologic studies
Kappa and its P value were calculated using SPSS software (version 11.5) to measure the degree of agreement between the results of PCR and histopathology. A P value <0.05 was considered significant.
New preparations for Ziel-Neelsen (ZN) and Hematoxylin– Eosin (HE) staining were prepared from the paraffin-embedded pericardial specimens and all were reviewed by a pathologist. Ziel-Neelsen staining was done to detect acid fast bacilli.
Statistical methods
Results
Polymerase chain reaction (PCR)
Histopathologic findings
Sample preparation Eight 15 μm sections from each paraffin embedded pericardial tissue were separately placed in two 1.5 mL Eppendorf tubes (4 pieces in each tube) and were sent to the microbiology laboratory. After each sampling the microtome blade and other tools were meticulously washed by hypochlorite solution.
Histopathologic examination showed fibrosis in all patients' (100%) chronic inflammation in 16 (53%), calcification in 9 (30%), and tuberculous granuloma in 5 (17%) patients. No acid fast bacilli, however, could be detected in the tissue samples.
DNA extraction Each sample was deparaffinized by adding 0.5 mL xylene to the Eppendorf tubes and mixing on rotary platform for 30 min at room temperature. The supernatant was then discarded after 10 min of centrifugation at 6000×g. The same process was repeated once more for each sample. To extract the remaining xylene, 1 mL of absolute ethanol was added to each sample and vortexing was done for 5 min. The mixture was then pelleted by centrifugation at 6000×g for 10 min and the ethanol was removed. The xylene extraction procedure was repeated twice for each sample. Two drops of acetone were then added and the tubes were incubated open in a water bath for 15 min at 60 °C. Subsequently the dried samples were resuspended in 200 μL of digestion buffer (50 mmol/lit Tris–HCl [pH8], 1 mmol/lit EDTA, and 0.05% sodium dodecyl sulfate) containing 800 μg/ mL of proteinase K. Mycobacterial DNAs from the samples were amplified using the following primers: IS6110 a (5′-CCTGCGAGCGTAGGCGTCGG-3′) and IS6110 b (5′-CTCGTCCAGCGCCGCTTCGG-3′) [3]. These primers amplify a fragment with a length of 123 bp. PCR amplifications were carried out in 50-μL volumes with 5 μL of prepared sample added. The mixture contained 50 pmol concentrations of each primer; 200 μM each of dNTP; 2 mM MgCl2; 5 μL PCR buffer; and 1 U of TAQ DNA polymerase (TIB Molbiol, Berlin, Germany). PCR runs were
PCR results PCR became positive in 9 patients (Table 1). Four out of 5 patients with tuberculous granulomas had a positive PCR. Therefore the results of PCR and the presence of tuberculous granuloma had a high measure of agreement (Kappa = 0.455 with P = 0.008). In addition, all 4 patients with definite nontuberculous constrictive pericarditis had negative PCR results. The mean duration of performing PCR after pericardiectomy was 10 ± 7.1 years. It was 12.5 ± 7.8 years in those with a positive PCR and 12.5 ± 8.7 years in those with a negative test result (P = NS). The difference was not significant in those with and without granuloma either (13.8 ± 6.1 vs. 13.8 ± 9.2 years P = NS). Table 1 The result of PCR in 30 patients with constrictive pericarditis with and without tuberculous granulomas in tissue biopsy Tuberculous granuloma in biopsy specimen
PCR
+ − Total
PCR = Polymerase Chain Reaction. + = positive. − = negative. Sensitivity: 4/5 = 80%. Specificity: 20/25 = 80%.
+
−
4 1 5
5 20 25
M. Zamirian et al. / Clinical Biochemistry 40 (2007) 355–358
Discussion Constrictive pericarditis represents the end stage of any inflammatory process involving the pericardium. In developed countries the etiology is most commonly infectious, post surgical or radiation injury. Acute tuberculous pericarditis although rare, can ultimately result in chronic constriction especially when remained undiagnosed and left untreated. This is particularly true in the developing world. Diagnosis of tuberculous pericarditis is notoriously difficult and because of its high prevalence in our country, empiric therapy is routinely done for many patients with constrictive pericarditis even in the absence of a definite diagnosis of tuberculosis. In fact rapid diagnosis and treatment is crucial to reducing mortality and morbidity from tuberculosis. Adenosine deaminase (ADA) and ã-IFN levels in pericardial fluid have been used for diagnosis of tuberculous pericarditis [4]. The limitation of these tests, however, is that they cannot be utilized in majority of patients with chronic constrictive pericarditis due to lack of any fluid in their pericardial sacs. Polymerase chain reaction (PCR), too has been used for diagnosis of acute tuberculous pericarditis [5]; however, its usefulness for constrictive pericarditis is to be reported. Therefore, we evaluated the usefulness of PCR for diagnosis of tuberculous constrictive pericarditis in paraffin embedded pericardial tissues from 30 patients with constrictive pericarditis diagnosed in our university hospitals during the past 25 years. Paraffin embedded tissues from other sites including lymph node and pleura have been successfully used for PCR [6,7]. AFB staining is insensitive since it requires about 104 organisms per milliliter of specimen to become positive [2], and it lacks specificity as well [2]. In none of our tissues samples even in those with tuberculous granulomas could AFB bacilli be seen. Although the standard detection method is culture, however, it usually takes several weeks for it to become positive, especially if the specimen contains a low number of organisms. Therefore we had to rely on the pathologic findings of granuloma and consider it as the reference test for evaluation of PCR results. The diagnosis of pleural tuberculosis too, which is the second most frequent human tuberculous presentation is usually based on the presence of necrotizing granulomas in pleural biopsy specimens [8]. Sensitivity for PCR was 80% since it was positive in 4 (80%) of the patients who had tuberculous granulomas in their pericardial specimens. In addition it turned positive in another 5 patients with no granulomas and who were previously labeled as having idiopathic chronic constrictive pericarditis. Histopathologic findings, however, could be non-specific even in patients with acute TB pericarditis [5]. The specificity of PCR in our study was 80%; this, however, may be an underestimation because pathology may be inconclusive in up to 30% of patients [5]. So patients with no granulomas should not be considered as truly negative cases especially in areas like ours where tuberculosis is encountered frequently.
357
Negative pathologic results are most common in cases where symptoms have been present for a long period or where treatment has been initiated long before the biopsy or pericardiectomy. The mean duration of symptoms before pericardiectomy was 49.6 ± 46.3 months in our patients and 21 of them had received antituberculous therapy for at least 2 weeks before pericardiectomy. Depending on the gold standard test chosen, PCR has been reported to have sensitivities ranging from 77% to more than 95% and PCR specificities of more than 95% for smear positive specimens [6]. Interestingly, all 4 patients who had post cardiac surgical and traumatic constrictive pericarditis had negative PCR, showing the high specificity of the test. PCR turned to be negative in one of our patients with tuberculous granuloma. False negative results have reported to be associated with the age of paraffinembedded tissues, i.e., the older the sample, the higher is the chance for false negativity [9]. Overall tuberculosis was the most possible underlying etiology of constrictive pericarditis in 33% of our patients by PCR. Our results are similar to that reported in Turkey between the years 1983 and 1993 since tuberculosis was the etiology of pericardial constriction in 38% of their patients [10]. Excluding the patients with tuberculosis (9 with positive PCRs and 1 with falsely negative PCR) and four patients with post cardiac surgery and traumatic constrictive pericarditis, the remaining 16 (53.3%) patients must be considered as having idiopathic constrictive pericarditis which has been reported in 33 to 73% range [11] in other studies. Constrictive pericarditis, even when diagnosed and surgically corrected is associated with a high mortality. In their series of 163 patients with surgically proven constrictive pericarditis, Bertog et al. found 10% perioperative mortalities and another 61 deaths after a median follow-up of 6.9 years among the survivors [12]. Therefore correct diagnosis and appropriate treatment are crucial for the long-term survival of these patients and tuberculous constrictive pericarditis is no exception. Although the conventional technique of direct smear examination with ZN staining is easy to perform and cheap, it has a low sensitivity and specificity [2]. Culture takes a long time specially if the number of organisms is few, and histopathologic findings could be non-specific [3]. New diagnostic tests, therefore, are very much needed and molecular amplification techniques such as PCR could be promising. Thus in areas with a high prevalence of tuberculosis like ours, PCR can be utilized as a useful, rapid and reliable diagnostic test for detection of tuberculous constrictive pericarditis even in patients with long-lasting symptoms and in those with no foot prints of the causative organism in tissue samples. References [1] Kivihya-Ndugga L, Van Cleeff M, Juma E, Kimwomi J, Githui W, Oskam L, et al. Comparison of PCR with the routine procedure for diagnosis of tuberculosis in a population with high prevalences of tuberculosis and human immunodeficiency virus. J Clin Microbiol 2004;42:1012–5.
358
M. Zamirian et al. / Clinical Biochemistry 40 (2007) 355–358
[2] Bogard M, Vincelette J, Antinozzi R, Alonso R, Fenner T, Schirm J, et al. Multicenter study of a commercial, automated polymerase chain reaction system for the rapid detection of mycobacterium tuberculosis in respiratory specimens in routine clinical practice. Eur J Clin Microbiol Infect Dis 2001;20:724–31. [3] Eisenach KD, Donald Cave M. Polymerase chain reaction amplification of a repetitive DNA sequence specific for Mycobacterium tuberculosis. J Infect Dis 1990;161:977–81. [4] Koh KK, Kim EJ, Cho CH, Choi MJ, Cho SK, Kim SS, et al. Adenosine deaminase and carcinoembryonic antigen in pericardial effusion diagnosis, especially in suspected tuberculous pericarditis. Circulation 1994;89: 2728–35. [5] Cegielski JP, Delvin BH, Morris AJ, Kitinya JN, Pulipaka UP, Lema LEK, et al. Comparison of PCR, culture, and histopathology for diagnosis of tuberculous pericarditis. J Clin Microbiol 1997;35:3254–7. [6] Kaul KL. Molecular detection of mycobacterium tuberculosis: impact on patient care. Clin Chem 2001;47:1553–8. [7] Ruiz-Manzano J, Manterola JM, Gamboa F, Calatrava A, Monoso E, Martinez V, et al. Detection of mycobacterium tuberculosis in paraffin-
[8] [9]
[10]
[11]
[12]
embedded pleural biopsy specimens by commercial ribosomal RNA and DNA amplification kits. Chest 2000;118:648–55. Light RW, Ferrer J. Diagnostico de la tuberculosis pleural. Arch Bronchoneumol 1999;35:105–7. Hirunwiwatkul P, Tumwasorn S, Chantranuwat C, Sirichai U. A comparative study of diagnostic tests for tuberculous lymphadenitis: polymerase chain reaction vs histopathology and clinical diagnosis. J Med Assoc Thai 2002;85:320–6. Oh KY, Shimizu M, Edwards WD, Tazelaar HD, Danielson GK. Surgical pathology of parietal pericardium: a study of 344 cases (1993–1999). Cardiovasc Pathol 2001;10:157–68. Ling LH, Oh JK, Schaff HV, Danielson GK, Mahoney DW, Seward JB, et al. Constrictive pericarditis in the modern era: evolving clinical spectrum and impact on outcome after pericardiectomy. Circulation 1999; 100:1380–5. Bertog SC, Thambidorai SK, Parakh K, Schoenhagen P, Ozduran V, Houghtaling PL, et al. Constrictive pericarditis etiology and causespecific survival after pericardiectomy. J Am Coll Cardiol 2004;43: 1445–52.