Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P226 The first validated post-operative endoscopic Crohn’s disease index: the POCER index— identification of key endoscopic prognostic factors

P. De Cruz*1, 2, M. Kamm1, A. Hamilton2, K. Ritchie3, A. Gorelik2, D. Liew2, I. Lawrance4, J. Andrews5, P. Bampton6, P. Gibson7, M. Sparrow7, R. Leong8, T. Florin9, R. Gearry10, G. Radford-Smith9, F. Macrae2, H. Debinski11, W. Selby12, I. Kronborg1, M. Johnston13, R. Woods14, R. Elliott3, S. Bell3, S. Brown3, W. Connell3, P. Desmond3

1University of Melbourne, Gastroenterology, Melbourne, Australia, 2University of Melbourne, Medicine, Melbourne, Australia, 3St Vincent’s University Hospital, Department of Gastroenterology, Melbourne, Australia, 4University of Western Australia, Gastroenterology, Perth, Australia, 5University of Adelaide, Adelaide, Australia, 6Flinders University, Gastroenterology, Flinders, Australia, 7Monash University, Gastroenterology, Melbourne, Australia, 8University Of Sydney, Concord Repatriation General Hospital, Department of Gastroenterology, Sydney, Australia, 9Queensland Institute of Medical Research, Medicine, Brisbane, Australia, 10University of Otago, Christchurch, Department of IBD, Gastroenterology, Christchurch, New Zealand, 11Monash University, Medicine, Melbourne, Australia, 12University of New South Wales, Gastroenterology, Sydney, Australia, 13St Vincent’s University Hospital, Department of Colorectal Surgery, Melbourne, Australia, 14St Vincent’s University Hospital, Department of Colorectal Surgery, Melbourne, Australia, Melbourne, Australia


The presence and severity of endoscopic recurrence after Crohn’s intestinal resection predicts the subsequent disease course. The Rutgeerts score is specific for postoperative endoscopic recurrence. Although its use has borne the test of time and many clinical studies, a composite score remains un-validated. This study aimed to investigate the predictive value of specific individual early endoscopic findings on subsequent disease course, in the setting of optimised drug therapy, and develop and prospectively validate a postoperative endoscopic index of severity that could be applied clinically and in trials.


In total, 118 patients underwent colonoscopy at 6 and 18 months following intestinal resection. Patients received 3 months metronidazole therapy, and high-risk patients received a thiopurine or adalimumab if thiopurine intolerant. For endoscopic recurrence (Rutgeerts score ≥ i2) at 6 months, patients stepped-up to thiopurine, fortnightly adalimumab with thiopurine, or weekly adalimumab. Central readers calculated Rutgeerts, SES-CD, CDEIS and 4 newly tested parameters of depth (superficial vs deep), number (0, ≤ 2, > 2), size (1–5 mm, ≥ 6 mm), and circumferential extent (< 25%, ≥ 25%) of anastomotic ulcers. A new index was then developed and validated for predicting outcomes at 18 months.


In the study, 58 patients were used for derivation and 60 for validation cohorts. Further, 19 patients in the derivation and 30 in the validation cohorts developed endoscopic recurrence at 18 months. Univariate analysis in the derivation cohort demonstrated a significant (p < 0.001) association between both > 25% circumferential extent and deep lesions and later endoscopic recurrence (Rutgeerts > i2). Regression analysis showed that the new POCER index comprising these 2 endoscopic variables had a significant independent predictive value (OR 1.8; CI 1.2–2.7; p < 0.01) with an area under the receiver operating curve of 0.68 (CI 0.54–0.83) in the derivation and 0.66 (CI 0.52–0.79) in the validation cohorts. The new index score at 6 months post-operatively had 58% sensitivity and 82% specificity for predicting subsequent recurrent disease, and correlated with subsequent CDEIS, SES, and Rutgeerts score at 18 months (p = 0.003, 0.003, and 0.001, respectively).


The new POCER index is the first validated post-operative score, comprising key endoscopic features that portend disease recurrence, and is a good predictor of subsequent endoscopic recurrence in the setting of optimised drug therapy. A high score, comprising the adverse prognostic factors of deep or extensive anastomotic ulceration, may help identify patients who require therapy that is more intensive. The POCER index now needs to be tested in longer-term prospective cohorts.