Histopathology and TB-PCR kit analysis in differentiating the diagnosis of intestinal tuberculosis and Crohn's disease

Abstract

Aim: To compare the histopathologic features of intestinal tuberculosis (ITB) and Crohn's disease (CD) and to identify whether polymerase chain reaction for Mycobacterium tuberculosis (TB-PCR) would be helpful for differential diagnosis between ITB and CD.

Methods: We selected 97 patients with established diagnoses (55 cases of ITB and 42 cases of CD) who underwent colonoscopic biopsies. Microscopic features of ITB and CD were reviewed, and eight pathologic parameters were evaluated. Nine cases of acid fast bacilli culture-positive specimens and 10 normal colonic tissue specimens were evaluated as the positive and negative control of the TB-PCR test, respectively. PCR assays were done using two commercial kits: kit <A> detected IS6110 and MPB64, and kit <B> detected IS6110 only; a manual in-house PCR method was also performed on formalin-fixed, paraffin-embedded colonoscopic biopsy specimens.

Results: Statistically significant differences were noted between ITB and CD with regard histopathologic criteria: size of granulomas (P = 0.000), giant cells (P = 0.015), caseation necrosis (P = 0.003), confluent granulomas (P = 0.001), discrete granulomas (P = 0.000), and granulomas with lymphoid cuffs (P = 0.037). However, 29 cases (52.7%) of ITB showed less than five kinds of pathologic parameters, resulting in confusion with CD. The sensitivities and specificities of the TB-PCR test by kit <A>, kit <B>, and the in-house PCR method were 88.9% and 100%, 88.9% and 100%, and 66.7% and 100% in positive and negative controls, respectively. The PCR test done on endoscopic biopsy specimens of ITB and CD were significantly different with kit <A> (P = 0.000) and kit <B> (P = 0.000). The sensitivities and specificities of TB-PCR were 45.5% and 88.1%, 36.4% and 100%, and 5.8% and 100%, for kit <A> and kit <B> and in-house PCR method on endoscopic biopsy specimens. Among the 29 cases of histopathologically confusing CD, 10 cases assayed using kit <A> and 6 cases assayed using kit <B> were TB-PCR positive. A combination of histologic findings and TB-PCR testing led to an increase of diagnostic sensitivity and the increase (from 47.3% to 58.2) was statistically significant with kit <B> (P = 0.000).

Conclusion: The TB-PCR test combined with histopathologic factors appears to be a helpful technique in formulating the differential diagnosis of ITB and CD in endoscopic biopsy samples.

Figure 1

Histopathological parameters of intestinal tuberculosis (ITB) and Crohn’s disease (CD). A: Confluent granuloma in ITB; B: Confluent granuloma with caseation necrosis and Langhans giant cells in ITB; C: Granuloma with lymphoid cuff in ITB; D: Vague granuloma in CD; E: Band of epithelioid histiocytes in ulcer base in ITB; F: Small granuloma in CD.

Figure 2

Polymerase chain reaction (PCR) product by kit and kit . A: Number 3 and 7 showing positive result by kit ; B: Number 2, 4, 6, 7 showing positive result by kit .