Abstract
Introduction:
Skeletal TB has a paucibacillary nature. It is often found in poorly accessible areas for biopsy purposes. Retrieved samples may have a poor representation of the underlying disease process. Additionally, patients have normally commenced anti-tubercular medication that further decreases the number of bacilli. This has resulted in poor sensitivity and specificity outcomes for the tests that are traditionally done.
The polymerase chain reaction (PCR) has been proven to be a useful test for the demonstration of extrapulmonary TB. It has a high specificity and sensitivity.
Objective:
The study measures the sensitivity and specificity of PCR done on fresh biopsies from patients suspected of a tuberculous spinal infection.
Method and Results:
A retrospective review of results was done of spinal tissue biopsies. A total of 30 consecutive patients were identified. There were 15 males (mean age = 40,3 yrs) and 15 females (mean age = 45,8 yrs). 18 of the patients were HIV positive.
Acid fast bacilli were demonstrated in 4 (13,33%) patients by staining. Culture was positive in 10 (33,33%), in which 2 had commenced antitubercular therapy. Histology reported features in keeping with tuberculosis in 9 (30%) patients. Furthermore, multiple myeloma, adenocarcinoma, osteomyelitis and thyroid carcinoma was reported. PCR was positive in 12 (40%) patients.
Sensitivity of PCR was calculated as 85,7% whilst specificity was 72,7% (the amount correctly classified was 77,8%).
Only 4 patients had positive tests for all 4 methods used for diagnosing tuberculosis.
Conclusion:
Tuberculosis should always be considered in spinal lesions. PCR provides a quick and effective means of identifying it. This allows early commencement of anti-tubercular treatment.
