Search in the Abstract Database

Abstracts Search 2019

P295 Impact of superimposed cytomegalovirus infection on the outcomes of ulcerative colitis flare-up

H. J. Kim*1, S. J. Oh1, Y-W. Kim2, J. R. Moon1, H-S. Kim3, C. K. Lee1

1Kyung Hee University School of Medicine, Department of Gastroenterology and Hepatology, Seoul, South Korea, 2Kyung Hee University School of Medicine, Department of Pathology, Seoul, South Korea, 3Yonsei University Wonju College of Medicine, Department of Internal Medicine, Wonju, South Korea


The aim of this study was to identify the impact of CMV infection on disease outcome of UC flare-ups and to investigate clinical significance of CMV viral load and antiviral treatment during UC flare-ups.


We retrospectively searched the electronic pathologic database of our tertiary academic hospital. Between January 2007 and July 2017, all colonoscopic biopsies specimens that were assessed for CMV infection were evaluated. CMV colitis was diagnosed as having one or more positive inclusion bodies on histological tests including H&E stain or immunohistochemical stain (IHC) in colonic tissues. CMV viral load was classified as low- or high-grade (5 or more inclusion bodies per section). To classify the CMV viral load, a single, independent gastrointestinal pathologist prospectively reviewed all biopsy specimens. We investigated long-term disease outcomes of UC patients with flare-ups according to their CMV infection status. Poor outcomes were defined as the following: hospitalisation, colectomy, or death. Subgroup analysis was performed according to CMV viral load and antiviral treatment status.


Among 844 cases with final pathologic results for their CMV status, a total of 257 patients with moderate-to-severe UC flare-ups were finally included. Mean age was 43.20 ± 14.68 years and 56.4% were male. Their median follow-up duration was 46 ± 39.01 months. CMV colitis were diagnosed in 36 patients (prevalence of 14%). Compared with patients without CMV colitis, both mean age and mean age at diagnosis were higher in patients with CMV colitis (all p < 0.001). The patients with CMV colitis showed significantly higher disease activity by total Mayo score and Mayo endoscopic sub-score (all p < 0.05), when compared with those without CMV colitis. Additionally, the patients with CMV colitis were more likely to receive systemic steroids, immunosuppressants, and anti-TNF agents (all p < 0.5). Collectively, CMV infection was an independent predictor of poor outcomes (Hazard ratio 2.27, 95% confidence interval 1.12–4.60) and the cumulative probability of poor outcome was significantly higher in the CMV positive group (p <0.001, log-rank test). Twenty-three patients of CMV colitis was graded as low density and 13 patients were high grade. No significant difference was observed in clinical outcome according to CMV density. Despite successful initial treatment with antiviral agents, the rates of CMV recurrence (57.14% vs. 22.73%; p = 0.0361) and hospitalisation (22.73% vs. 64.29%; p = 0.0126) were higher in the treated group.


Superimposed CMV colitis is an independent predictor of poor outcome in moderate to severe UC flare-ups. Antiviral agent does not seem to improve the long-term outcome of UC patients regardless of CMV load.