P460 The POCER index: application of a novel endoscopic score in a real-life cohort of patients with Crohn's disease after surgery

Abstract

Abstract Background The Rutgeerts score (Rs) is the most widely used score that offers essential prognostic information in the field of postoperative management of patients (pts) with Crohn's disease (CD). A recent study has introduced the POCER postoperative index (Pi), which includes the circumferential extent and depth of anastomotic ulcers, which can be an additional tool to better define the prognosis of this category of pts Methods Our study aims to evaluate the Pi in a real-life cohort, its correlation with Rs, and its prognostic value. We conducted a single center retrospective study, evaluating different outcomes of pts with CD who underwent ileocolonic resection. All pts underwent the first endoscopic evaluation within 6-12 months to assess postoperative recurrence (POR). We applied both the Rs and the Pi. POR was defined as Rs≥2 or Pi≥2. A clinical and objective assessment was performed at 18-24 months. Clinical Recurrence (CR), evaluated in all pts, was defined with a Harvey-Bradshaw Index (HBI)≥5. Objective Recurrence (OR), evaluated in a subgroup, was defined by the presence of endoscopic activity and/or radiological activity (assessed by ultrasound (US) when a wall thickness≥4 mm was present) Results We included 103 pts who underwent ileocolonic resection, with a median age at surgery of 38 years old. The main indication for surgery was a stenotic disease. 63% of pts had ≥1 risk factor for POR. All pts underwent prophylactic biologic therapy. At 6 month, POR was observed in 50 (48.3%) and 17 (16.5%) pts, according to Rs and Pi, respectively. Mean Rs and Pi were 1.39 (± 1.31) and 0.70 (± 1.08); a moderate correlation (r=0.523) between the two was observed. We assessed the ability of Rs and Pi at 6-12 months to predict CR and OR at 18-24 months. A Rs≥2 predicted CR with an AUROC of 0.8345 (sensitivity 93.3% and specificity 59.3%). A Pi≥2 predicted CR with an AUROC of 0.693 (sensitivity 33.33% and specificity 84.88%). When evaluating the performance of the mean of the two scores, we observed that a mean score≥1.75 could predict CR with an AUROC of 0.807 (sensitivity 80.00% and specificity 68.6%). A mean score≥1.75 could also predict OR with an AUROC of 0.662 (sensitivity 45.45% and specificity 74.29%). We also evaluated the subgroup of pts with Rs 1-2 at 6 months and the significance of the addition of the Pi: although no statistical significance was observed, a Pi<1 had a specificity of 100% in predicting the absence of CR Conclusion Rs and Pi can be used to assess POR and predict CR and OR. The combination of the two might outperform each of them with a good compromise between sensitivity and specificity. The addition of the Pi in pts with Rs 1-2, might be helpful in clarifying the prognostic significance of this result