Background and objectives: Differential diagnosis of intestinal tuberculosis(ITB) and Crohn's disease(CD) is difficult, because of clinical and pathologic similarities. Even though many pathologic and molecular studies have been performed, it is still...
Background and objectives: Differential diagnosis of intestinal tuberculosis(ITB) and Crohn's disease(CD) is difficult, because of clinical and pathologic similarities. Even though many pathologic and molecular studies have been performed, it is still difficult to differentiate. PCR for M. tubeculosis (TB-PCR) method has been accepted as an ancillary test for differential diagnosis of ITB and CD. Recently, many kinds of commercial kits have been developed using different targets to detect the M. tuberculosis genome, but their diagnostic accuracy was not sufficiently examined. In this study, we compared the histopathologic features of ITB and CD, and evaluate the diagnostic significance of TB-PCR in formalin fixed and paraffin-embedded specimen of endoscopic biopsy. For this purpose, we used two commercially available TB-PCR kits in Korea and compare the results with in-house PCR method which have been reported in many literatures.
Methods: We selected 97 patients with established diagnosis, 55 cases of ITB and 42 cases of CD, all of them performed colonoscopic biopsy. Microscopic features of ITB and CD were reviewed and eight pathologic parameters were evaluated. For control of TB-PCR test, 9 AFB culture-positive specimens and 10 normal colonic tissues were evaluated as gold standard. PCR assays were done using two commercial kits <A> detected IS6110 and MPB64 and kit <B> detected IS6110 only and manual in-house PCR method in formalin-fixed, paraffin-embedded colonoscopic biopsy specimens.
Results: All evaluated histopathologic criteria such as ulcer with bands of epithelioid histiocytes, size and number of granuloma, giant cell, caseation necrosis, confluent granuloma, discrete granuloma, granuloma with lymphoid cuff were statistically significant for differential diagnosis of ITB and CD. However, 29 cases(52.7%) of ITB showed less than half of these pathologic parameters which resulted in confusion with CD. The sensitivity and specificity of TB-PCR test by kit <A>, kit <B> and in-house PCR method, were 88.9% and 100%, 88.9% and 100%, 66.7% and 100% in gold standard. In endoscopic biopsy specimens of ITB and CD, the sensitivity and specificity of PCR test were 45.5% and 88.1%, 36.4% and 100%, 5.8% and 100%, by kit <A>, kit <B> and in-house PCR method, respectively. Among the 29 cases confused with CD, 10 cases (by kit <A>) and 6 cases (by kit <B>) were TB-PCR positive. A combination of histologic finding and TB-PCR test led to increasing of diagnostic sensitivity from 47.3% to 58.2-65.5%.
Conclusion: Commercially available TB-PCR kits showed higher sensitivity and specificity to detect M. tuberculosis than in-house method. TB-PCR test combined with histopathologic factors appears to be a helpful technique in differential diagnosis of ITB and CD in endoscopic biopsy samples.