DOP091. Intestinal Epstein–Barr virus is associated with mucosal lymphoproliferation and subsequent intestinal surgery in inflammatory bowel disease patients
L.H. Nissen1, D.J. de Jong1, I.D. Nagtegaal2, W. Kievit1, M.G. Lynch2, J.H.J. van Krieken2, F. Hoentjen1, 1Radboud UMC, Gastroenterology & Hepatology, Nijmegen, Netherlands, 2Radboud University Nijmegen Medical Centre, Pathology, Nijmegen, Netherlands
Thiopurine therapy increases the risk of (Epstein–Barr virus associated) lymphomas for Inflammatory Bowel Disease (IBD) patients up to four times. Epstein–Barr virus (EBV) can cause a wide spectrum of lymphoproliferative reactions, ranging from morphologically benign with normal B lymphocytes (BL) and lymphoplasmacytic infiltrate in the lamina propria (LI) to aggressive lymphomas with atypical BL and LI.
EBV can be detected in colonic mucosa in up to 60% of the IBD patients, but there is no consensus on when to perform EBV testing on intestinal mucosa. We hypothesized that EBV testing can be guided by histological features including morphology of BL and LI.
The aim of this study was to determine the value of the histology of the inflammation in predicting EBV presence in intestinal mucosa and to correlate EBV positivity with clinical endpoints such as intestinal surgery and development of lymphoma.
All IBD patients who underwent EBV testing by EBV-encoded RNA - in situ hybridization (EBER) in intestinal biopsies between January 2005 and October 2013 in our centre were identified. All biopsies were revised by a blinded, expert gastro-intestinal pathologist and scored on three histological features: number of EBV positive cells per high power field (HPF); normal or atypical LI and normal or atypical BL. Demographic and clinical data were collected from patient charts. Adverse events that were registered included intestinal surgery and lymphoma. We used the Chi square test or Fisher's exact test to identify an association with EBV positivity.
58 IBD patients were included, 28 were EBV positive and 30 were EBV negative. Ulcerative colitis was more frequent in the EBV positive group (82.1% versus 56.7%; p = 0.052).
EBV positive patients had significantly more frequent atypical LI (57.1% versus 3.3%; p les; 0.001). The specificity for predicting EBV presence of the atypical LI is high (96.7%), just as its positive predictive value (94.1%). The sensitivity and specificity of atypical BL for predicting EBV are 42.9% and 83.3%. At time of biopsy, EBV positive patients used more often combinations of two or more anti-inflammatory drugs (5-aminosalicytes excluded; 50% versus 16.7%; p = 0.007).
Eighteen EBV positive patients (64.29%) had 20 pre-defined complications (18 colectomies, 2 lymphomas). Within the group of EBV positive patients, those who developed complications had a significantly higher EBV load (50% versus 10%; p = 0.048), expressed as the frequency of 10 or more EBV positive cells per HPF.
In the present study, atypical LI was associated with mucosal EBV in IBD patients. A high EBV load correlated with adverse events.