Search in the Abstract Database

Abstracts Search 2017

* = Presenting author

P293 Histological risk factors to predict clinical relapse in ulcerative colitis in mucosal healing

Ozaki R.*1, Kobayashi T.1, Saito E.1, Toyonaga T.1, Okabayashi S.1, Umeda S.1, Nakano M.1, Matsuoka K.1, Morinaga S.1, Hisamatsu T.2, Hibi T.1

1Kitasato University Kitasato Institute Hospital, Minato, Japan 2Kyorin University School of Medicine, Mitaka, Japan

Background

Currently, the target of treatment for ulcerative colitis (UC) have been changing from clinical remission to include mucosal healing, and finally histological remission. We evaluated the correlation between future clinical relapse and histological findings in patients with UC who achieved mucosal healing.

Methods

Patients with UC who underwent colonoscopy with biopsies and achieved mucosal healing were enrolled into a retrospective cohort. Data were collected from April 2013 to October 2016. Mucosal healing was defined as Mayo endoscopic subscores 0 or 1. Histological findings based on Japanese criteria for the diagnosis of UC, were evaluated as the presence or absence of inflammatory cell infiltration, erosion, crypt abscess, goblet cell depletion and crypt distortion. Assessment of histologic inflammation was done by two experienced pathologists. Clinical relapse was defined by the addition of medical treatment for the clinical symptoms. Multivariate analysis by Cox proportional hazards regression tests were used to generate a prediction model of clinical relapse.

Results

A total of 189 UC patients were enrolled. Clinical relapse occurred in 64 patients (33.9%) during the observation period. There was no difference in the time to relapse between Mayo endoscopic subscores 0 and 1 (p=0.3613). The rate of Clinical relapse was significantly higher in patients with erosion and/or crypt distortion compared to patients without them (p=0.0244; and p=0.0178, respectively). No significant difference was observed in patients who have inflammatory cell infiltration, crypt abscess, and/or goblet cells depletion.

Multivariate analysis indicated that the presence of erosion (hazard ratio, 1.910; 95% confidence interval, 1.105–3.302, p=0.02052) and crypt distortion (hazard ratio, 2.611; 95% confidence interval, 1.125–6.059, p=0.02547) significantly affect the time to relapse, respectively.

Conclusion

Presence of erosion and crypt distortion are considered to be histological risk factors for future clinical relapse in UC patients in remission who achieved mucosal healing, while Mayo endoscopic subscores did not show significant predictive values for relpase.