Abstracts of the 12 th Congress of ECCO – European Crohn’s and Colitis Organisation S401 ing data on response of perianal fistulas with UST. However, larger studies are required to confirm these findings. P627 Surrogate markers of mucosal healing in Crohn’s disease patients in clinical remission under biological/immunomodulator treatment S. Siakavellas*, A. Kostas, C. Kosmidis, M. Gizis, G. Papatheodor- idis, G. Bamias National & Kapodistrian University of Athens, Academic Dpt. of Gastroenterology, Athens, Greece Background: Mucosal healing is a desired endpoint in both clini- cal trials and “real-life” practice as it has been associated with bet- ter outcomes in patients with IBD. Lower GI endoscopy is required to determine the presence or absence of mucosal healing. Our aim was to assess specific biomarkers that could accurately predict (ei- ther alone or in combination) the presence of mucosal healing in Crohn’s disease (CD) patients under long-term anti-TNF and/or im- munomodulator treatment. Methods: Eligible patients were those with CD who were on clinical remission for at least 6 months under stable treatment with anti-TNF and/or immunomodulators. Prior to endoscopy all patients were sub- jected to thorough workup every two months with recordings of Harvey-Bradshaw index score and selected laboratory tests that in- cluded fecal calprotectin and serological inflammatory markers. Af- ter the end of this 6 month period, colonoscopy was performed and mucosal healing was determined as present [complete (no inflamma- tory lesions) or partial (minimal inflammatory lesions)] or absent. The predictive value of several clinical and laboratory markers for the presence of mucosal healing was investigated. Results: Twenty-three patients have been recruited so far (Male=9, Age: 40.8±14.3, 19–70, mean ± SD, range, in years). Fourteen pa- tients (60.8%) achieved mucosal healing as evidenced by lower gas- trointestinal endoscopy. Patients in the “no healing” group had sig- nificantly higher fecal calprotectin values when compared to patients with mucosal healing at 2 months prior to endoscopy [“no healing” group 554 μg/gr, 235–1800 (median, interquartile range) vs. mucosal healing group 83, 33–330.5, p=0.012), 4 months prior to endoscopy (“no healing”, 600, 338–600 vs. mucosal healing 134, 22.5–272, p=0.009), as well as at 6 months prior to endoscopy respectively (“no healing”, 265.0, 142–482.5 vs. mucosal healing 64, 13.8–199, p=0.039). No significant differences between the two groups were observed regarding CRP levels. Moreover, higher amylase values were found in the “no healing group” in comparison to the healed mucosa group at 6 months prior to endoscopy (91.1 IU/L ± 24.8 vs. 63.1±27.2, mean ± SD, p=0.02). Finally, smokers had less often mucosal healing (p<0.0001) and higher CRP and fecal calprotectin values as well, than non-smokers. Conclusions: Fecal calprotectin is a better predictor of mucosal heal- ing than CRP in patients with CD in clinical remission. Its use in clinical practice may improve patient management by allowing the identification of patients at higher risk for disease flare, who may require closer follow up and earlier endoscopy. Funding: The present work has been funded by a grant from the Hellenic Society of Gastroenterology to Dr. Bamias. P628 Anti-TNF therapy in refractory pouchitis and Crohn’s disease-like complications of the pouch after ileal pouch-anal anastomosis following colectomy for ulcerative colitis: a systematic review and meta-analysis M. Huguet 1 , B. Pereira 2 , M. Goutte 1,3 , F. Goutorbe 1,4 , C. Allimant 1 , M. Reymond 1 , G. Bommelaer 1,3 , A. Buisson* 1,3 1 University Hospital Estaing, Gastroenterology Department, Clermont-Ferrand, France; 2 University Hospital, Biostatistics Unit, DRCI, Clermont-Ferrand, France; 3 UMR 1071 Inserm/Université d’Auvergne; USC-INRA 2018, Microbes, Intestine, Inflammation and Susceptibility of the host, Clermont-Ferrand, France; 4 Hospital of Bayonne, Gastroenterology Department, Bayonne, France Background: Pouchitis and secondary Crohn’s disease (CD)-like complication of the pouch are the most common complications af- ter ileal pouch-anal anastomosis following colectomy for ulcerative colitis. Data about the effectiveness of anti-TNF agents in these two entities remains sparse. We aimed to perform a systematic review and meta-analysis to eval- uate the efficacy of anti-TNF therapy in differentiating patients with chronic refractory pouchitis and CD-like complications. Methods: Systematic literature search was performed in MEDLINE and from international meetings abstracts. The search process, selec- tion of manuscripts, and data extraction were performed indepen- dently by two physicians according to PRISMA statements. Prevalence and 95% confidence interval (CI) were estimated using random-effects models assuming between and within study variabil- ity. Statistical heterogeneity between results was assessed by exam- ining forest plots, CIand using I 2 and sensitivity analyses were con- ducted. CD-like complications of the pouch were defined as the presence of non-anastomotic fistula and/or non-anastomotic stenosis and/or pre- pouch ileitis. Chronic refractory pouchitis was defined as inflamma- tion limited to the pouch. The short term and the long term responses were evaluated at 8 weeks and 12 months, respectively. Results: We identified 21 articles and three abstracts including 313 patients treated either with infliximab (IFX) (n=194) or adalimumab (ADA) (n=119) for inflammatory complications of the pouch. The rate of complete response (CR) after anti-TNF induction ther- apy for inflammatory complications of the pouch was 0.51 (95% CI [0.39–0.64]; I 2 =0.56). The rate of short-term CR was 0.57 (95% CI [0.38–0.75]; I 2 =0.36) for IFX-treated patients compared to 0.38 (95% CI [0.08–0.72]; I 2 =0.50) for ADA-treated patients (p=0.20). The long-term rate of CR in patients treated with anti-TNF therapy was 0.52 (95% CI [0.39–0.65]; I 2 =0.59), with 0.59 (95% CI [0.45– 0.72]; I 2 =0.30) for IFX-treated patients compared to 0.30 (95% CI [0.15–0.46]; I 2 =0.00) for ADA-treated patients (p=0.19). The rate of CR after anti-TNF induction therapy seemed to be higher for CD-like complications of the pouch 0.64 (95% CI [0.5–0.77]; I 2 =0.18), compared to refractory pouchitis 0.10 (95% CI [0.08– 0.35]; I 2 =0.00) (p=0.06). The rate of long-term CR in patients treated with anti-TNF was 0.57 (95% CI [0.43–0.71]; I 2 =0.32) for CD-like complications of the pouch compared to refractory pouchitis 0.37 (95% CI [0.14–0.62]; I 2 =0.47) (p=0.57). Conclusions: Despite wide heterogeneity of the data, anti-TNF agents have a clear trend to have higher and faster efficacy in CD-like complications of the pouch compared to refractory pouchitis, high- lighting the need to differentiate these two entities in clinical prac- tice. Downloaded from https://academic.oup.com/ecco-jcc/article-abstract/11/suppl_1/S401/2961570 by guest on 27 May 2020