Detection of Mycobacterium tuberculosis in formaldehyde solution-fixed, paraffin-embedded tissue by polymerase chain reaction in Pott's disease.

Abstract

Twenty-five formaldehyde solution-fixed, paraffin-embedded tissue blocks from vertebral biopsy specimen materials with presumptive diagnosis of tuberculous spondylitis and nonspecific vertebral osteomyelitis were studied. To evaluate the sensitivity and specificity of polymerase chain reaction in detecting Mycobacterium tuberculosis in formaldehyde solution-fixed, paraffin-embedded tissue samples from histologically proved tuberculous spondylitis. Diagnosis of a mycobacterial infection is a long and tedious process; because of the slow growth rate of mycobacteria on solid media, identification and antibiotic sensitivity testing can take up to 10 weeks, but the sensitivity of culture can be as low as 50%. Direct microscopy is insensitive because clinical samples may contain only few organisms. Recently, polymerase chain reaction has been applied in the rapid amplification and identification of many organisms, including mycobacteria. The DNAs were extracted from 25 paraffin-embedded tissue blocks. An insertion element IS 6110 (Integrated DNA Tec. Inc., Corrallville, IA), a DNA sequence unique to Mycobacterium complex (M. tuberculosis and the subspecies Mycobacterium bovis), was amplified by polymerase chain reaction. Polymerase chain reaction results were compared with those of Mycobacterium culture, acid-fast bacilli staining, and histologic findings. Polymerase chain reaction was positive in 18 cases of 19 tuberculous spondylitis. Three of the polymerase chain reaction test results were positive with concomitant negative culture and positive acid-fast bacilli staining. There were six chronic nonspecific infections, and polymerase chain reaction results were negative in five cases; in the single positive case, DNA amplification results remained positive even after three repeated tests. Polymerase chain reaction has a sensitivity of 94.7%, specificity of 83.3%, positive predictive value of 94.7%, and a negative predictive value of 83.3%. Accuracy was calculated as 92%.