Search in the Abstract Database

Abstracts Search 2014

* = Presenting author

P099. What is the real value of acid fast staining, PCR, and culture of endoscopically obtained tissue samples in diagnosis of gastrointestinal tuberculosis in patients with ileocecal inflammation? Comparative data from an Eastern European registry

Y. Erzin1, I. Hatemi1, M. Kuskucu2, G. Aygun2, K. Midilli2, S.N. Esatoglu3, S. Sadri3, A.F. Celik1, 1Istanbul University Cerrahpasa Medical Faculty, Gastroenterology, Istanbul, Turkey, 2Istanbul University Cerrahpasa Medical Faculty, Microbiology, Istanbul, Turkey, 3Istanbul University Cerrahpasa Medical Faculty, Internal Medicine, Istanbul, Turkey


Most of the available data on gastrointestinal tuberculosis (GITB) comes from either Asian or African countries where the high incidence of TB and extensive use of test therapeutic anti-tuberculosis (anti-TB) treatment (Tx) exist. In addition, PCR results mainly come from fixed tissue samples and the presence of concomitant extra-intestinal TB (ExITB) commonly is not documented.


To investigate the diagnostic values of acid fast stain (AFS), M.tuberculosis (MT) cultures (MICIT and LJ) and MT-PCR and stress on the factors influencing the results of this diagnostic tools. This is a retrospective assessment, recorded in a prospective manner.


Between 2003–13, 20 GI-TB, and 109 consecutive Crohn's disease (CD, n = 90) and GI Behçet's disease (GIBD, n = 19) patients (pts) with ileocecal involvement were selected from our database. Colonoscopic punch biopsies (nearly all from the ileocecal area, first 2–4 in saline, an additional 2–4 for pathology in formaldehyde) were freshly stained, cultured and PCR for MT was performed. All pts were enrolled as a new pt at their first admission and were HIV(−). Pts who had objective findings of GITB [tissue culture positive, and/or TB-PCR and/or AFS(+)] were given anti-TB Tx. Only one pt being (−) for all tests but clinically suggestive for TB got Tx too. Pts who were in clinical and complete endoscopic remission after Tx (9–12 mo.) were confirmed to have GITB and accepted as the gold standard for the calculations. The sensitivities (SENSs) and specificities (SPECs) of all tests were calculated. Sampling/processing: in freshly taken samples, before starting even a single dose of anti-tuberculosis Tx, immediate processes, tests from homogenized samples, both MICIT and LJ 7/20 (35%) pts also had additional pulmonary TB, 4/20 (20%) had miliary TB. False positivity for TB-PCR was in 4/109 pts with non TB ileocecal inflammation (all in house nested TB-PCR). Lower part of table indicates the effect of ExITB on the test results. However, even only in intestinal TB pts, SENSs of culture and PCR were, 80% and 65% respectively without dropping in their SPECs.

(n = 20)
(n = 20)
(n = 18) a
Sensitivity (95% CI)35 (15–59)90 (68–98)83 (58–96)
Specificity (95% CI)100 (96–100)100 (96–100)96 (90–99)
False negative13/202/202/13 in-house
1/5 RT-PCR
a Two samples dropped; inhibitor and insufficient tissue.
Extraintestinal (n = 11)Intestinal (n = 9)p
EZN6 (55%)1 (11%)0.07
PCR11 (100%)4 (57%)0.043
Culture11 (100%)7 (78%)0.19


In contrast to previous non European reports culture of colonoscopic biopsy specimens is an highly SENS/SPEC to differentiate GITB from other ileo-cecal inflammations. Different methodological details, in sampling and processing, and the presence of exITB may have some influence on the results. Higher theoretical contamination chance with in house nested PCR may increase false (+) rate.