Evaluation of clinico-radiological, bacteriological, serological, molecular and histological diagnosis of osteoarticular tuberculosis

Background: The diagnosis of osteoarticular tuberculosis is clinico-radiological in endemic areas. However every patient does not have the classical picture. Osteoarticular tuberculosis is a paucibacillary disease hence bacteriological diagnosis is possible in 10-30% of the cases. The present study is undertaken to correlate clinico-radiological, bacteriological, serological, molecular and histological diagnosis. Materials and Methods: Fifty clinico-radiologically diagnosed patients of osteoarticular tuberculosis with involvement of dorsal spine (n = 35), knee (n = 8), shoulder (n = 1), elbow (n = 2) and lumbar spine lesion (n = 4), were analyzed. Tissue was obtained after decompression in 35 cases of dorsal spine and fi ne needle aspiration in the remaining 15 cases. Tissue obtained was subjected to AFB staining, AFB culture sensitivity, aerobic/anaerobic culture sensitivity histopathological examination and polymerase chain reaction (PCR) using 16srRNA as primer. Serology was performed by ELISA in 27 cases of dorsal spine at admission and one and three months postoperatively. Results: AFB staining (direct) and AFB culture sensitivity was positive in six (12%) cases. Aerobic/anaerobic culture sensitivity was negative in all cases. Histology was positive for TB in all the cases. The PCR was positive in 49 (98%) cases. All dorsal spine tuberculosis cases showed fall of IgM titer and rise of IgG titer at three months as compared to values at admission. Conclusion: Histopathology and PCR was diagnostic in all cases of osteoarticular tuberculosis. The serology alone is not diagnostic.


INTRODUCTION
O steoarticular tuberculosis involves 2-5% of all tubercular lesions in the body. 1 Out of which 50% affects the spine. 1 The diagnosis of osteoarticular tuberculosis and in particular tuberculosis of the spine is clinico-radiological, particularly in the endemic regions. The typical lesion can be diagnosed clinico-radiologically with support of newer imaging modalities like computed tomography/magnetic resonance imaging (CT/MRI); however, tissue diagnosis is a must when there is a slightest doubt. For accurate diagnosis to be established the tubercular bacteria must be recovered from the lesion. 2 The emerging multidrug resistant strains are posing a threat to cure the tubercular lesion hence the mycobacterium should be isolated and subjected to drug susceptibility test. 2 Osteoarticular tuberculosis, being a paucibacillary disease and some of the patients are already on antitubercular treatment (ATT) at the time of presentation, hence no single modality like AFB culture sensitivity, AFB staining, histopathology are capable of ascertaining the diagnosis for tuberculosis. Attempts have been made in the past to develop serologic methods for detection of antibodies against mycobacterial antigens. [3][4][5] These tests may provide a better indication of activity of disease. Newer methods of diagnosis by molecular methods [polymerase chain reaction (PCR)] have been introduced to partly overcome the problems of traditional methods. The results can be obtained within two to three days, thereby helping in early diagnosis and treatment. [6][7][8][9] To the best of our knowledge no study has correlated conventional diagnostic methods with the newer diagnostic methods in osteoarticular tuberculosis. The present study is an attempt to evaluate all the diagnostic modalities (AFB smear, AFB culture sensitivity, PCR, histology, serology) to ascertain their efficacy in establishing the diagnosis and treatment.

MATERIALS AND METHODS
The study was conducted in a tertiary care hospital. Fifty clinico-radiologically diagnosed cases of osteoarticular tuberculosis with involvement of dorsal spine (n = 35), knee (n = 8), shoulder (n = 1), elbow (n = 2) and lumbar spine lesion (n = 4), were analyzed. Indications for surgery in cases with dorsal spine involvement were deterioration of (n = 14) or static (n = 13) neural deficit and doubtful diagnosis (n = 3). The peripheral limb lesions and lumbar spine cases were aspirated for confirmation of diagnosis. Tissue obtained during surgery and by biopsy was analyzed for (a) direct microscopy and culture of mycobacteria in all cases. Ziehl-Neelsen stain and Lowenstein-Jensen media were used for staining and culture of mycobacteria. 10 The positive cultures were identified with a set of standard proportional tests for species identification and drug susceptibility testing was performed by proportional method for streptomycin, isoniazid, rifampicin, ethambutol and pyrazinamide. (b) Tissue samples were also analyzed for other aerobic and anaerobic bacteria. Gram's stain and culture sensitivity was used for aerobic bacteria in all cases. Anaerobic culture sensitivity was done in Robertson's cooked meat medium. 10 (c) Histopathologically the tissue was stained with hematoxylin and eosin stain (n = 50) and was seen under microscope for acid-fast bacillus (AFB) and epitheloid cell granuloma, Langherhan's cell with or without caseation. 11,12 (d) Polymerase chain reaction for rapid diagnosis of mycobacterium was done using 16srRNA as a primer (n = 50), which is a genus-specific primer, on the granulation tissue/pus obtained. (e) ELISA test was used for serological tests, in all cases. It could be done in only 27 cases of dorsal spine tuberculosis because of paucity of ELISA kits. Serum IgG and IgM levels were measured against 38kd and A-60 tubercular antigen respectively. Blood for serology was taken at the time of admission and then at one and three months postoperatively for serial rise or fall in immunoglobulin titers.

RESULTS
The AFB staining (direct) and AFB culture sensitivity was positive in six (12%) cases. These were of the same specimens. All the cases (100%) had histological features suggestive of tubercular osteomyelitis, confirmed by presence of caseation necrosis, epitheloid cell granuloma and Langerhans giant cells [ Figure 1]. The PCR was positive for mycobacterium tuberculosis complex in 49 (98%) cases [ Figure 2]. All the cases were Gram's stain negative and showed no growth on pus culture sensitivity (Table 1). Mean IgM titer at admission, one month and three months postoperative was 0.9293, 0.6413 and 0.5103. Mean IgG titer at admission, one month and three months postoperative was 0.2974, 0.3027and 0.3341. Serological tests showed fall of IgM and rise of IgG titer at three months as compared to values at admission and at one month post operative ( Table 2).

DISCUSSION
About 30 million people suffer from tuberculosis (TB) throughout the world every year and 1-2% of these patients suffer from osteoarticular tuberculosis. Diagnosis of osteoarticular tuberculosis is difficult since the organism is fastidious and slow-growing. 2 AFB is difficult to isolate in osteoarticular tuberculosis since it is a paucibacillary disease and being a deep-seated lesion it is difficult to procure the tissue. 2 The diagnosis of osteoarticular tuberculosis in endemic areas is clinico-radiological. It is justified to treat the patients clinico-radiologically in classical lesions of the bone. The clinical and radiological response can be observed in 8-12 weeks. However, there are certain cases with doubtful diagnosis, where tissue is required to ascertain diagnosis.
In the bone of the appendicular skeleton, tissue may be procured by fine needle aspiration cytology (FNAC) or core biopsy. Delay of a few days in the treatment of the limb lesion does not give rise to severe consequences, as tuberculosis is a slowly progressive disease. Tuberculosis of the spine is a deep-seated lesion, which if not diagnosed promptly and treated adequately, then consequences would be hazardous, as patient may develop kyphosis and/or neurological complication (paraplegia). Osteoarticular tuberculosis, being a paucibacillary disease some of the patients are already on ATT, hence no single modality like AFB culture, AFB staining, histopathology are capable of ascertaining the diagnosis. 2 The role of serological tests and PCR (molecular methods) are still not well defined in management of osteoarticular tuberculosis.
Sensitivity of AFB staining in various series was reported in the range of 25-75%. 2 Lakhanpal et al., reported 49.53% positivity by AFB culture sensitivity. 2 Other workers have reported in a range of 48.6-80%. 2 In our series AFB staining and AFB culture was positive in six (12%) cases. Out of these six cases four patients were on ATT for more than two months before presentation, one patient was on ATT since last four days and one patient had never taken ATT. In pulmonary TB, since the sputum can be procured easily, direct AFB staining has an important role in its treatment, as the conversion of sputum positive to sputum negative indicates the efficacy of the treatment. However, in osteoarticular tuberculosis, the bacterial diagnosis and absence of mycobacterium tuberculosis on treatment cannot be taken as a criterion of successful treatment. Lakhanpal et al., 2 attributed the lower percentage to the possible effect of preoperative antitubercular therapy. Wilkinson also asserted the same point. Our positivity of AFB staining was very low in spite of 16 (32%) of our cases having never taken ATT before reporting to the hospital. However, status of taking other broad-spectrum antibiotics is not known, which may have an inhibitory effect on mycobacteria. Therefore, the fact that duration of ATT has any influence on the positivity of AFB culture sensitivity could not be substantiated in our study. Factors attributable for less positivity with AFB culture sensitivity are paucibacillary disease (Number of AFB is about 103-104/ml), species present (M. tuberculosis is more likely to be positive than MOTT), patient already on ATT, stain used, observer's experience. The major limitations of AFB culture and sensitivity are that it requires live organisms, has a long incubation period, and low sensitivity in patients already on ATT. Although the culture results are not available for up to four weeks, they prove the diagnosis of TB beyond doubt. 2  A large number of organisms required by all the above methods was the major limitation in detection of M. tuberculosis. A single test, which would amplify the genome, even if a single organism was present, was thought to be ideal for detection of paucibacillary TB cases. The PCR can analyze the expression of genes even from single cells.
The PCR was positive in 49 (98%) cases in our series. One case who was PCR-negative had a histopathology report suggestive of tubercular osteomyelitis. Our results with PCR were comparable with other authors' series. 6 20 Only six (12%) cases of dorsal spine tuberculosis were positive for combined PCR, histopathology, AFB staining and AFB culture sensitivity and serological tests in our series.
We conclude that high sensitivity and specificity of PCR and histopathology can be useful for early diagnosis of paucibacillary osteoarticular tuberculosis. However, once a tissue is procured it should be subjected to AFB staining, AFB culture sensitivity, PCR and histopathology in all the cases.