Sa1501 The Efficacy of Tacrolimus and the Usefulness of Endoscopy in Predicting Its Efficacy in Patients With Refractory Ulcerative Colitis

screening method. We designed an open-label controlled trial to compare CTC, CS, and BE for diagnosing colorectal cancer or polyps in FOBT-positive patients. Methods: We enrolled 1,500 consecutive patients who had undergone CTC, CS, or BE screening at the 114 hospitals that comprise the Okazaki city medical association. Each institution selected a screening program based on a discussion with the doctor in charge, and diagnoses of colorectal cancer, large (R10 mm) polyps, or smaller polyps were evaluated. CTC was performed using barium-based fecal tagging. Results: The subjects were divided equally into CTC, CS, and BE groups (n Z 500 each, age SD: 68.4 12.8 years, 66.2 12.7 years, and 67.3 13.4 years, respectively; proportion of women: 51%, 42%, and 49%, respectively). The CS group was significantly younger and had more men than the CTC and BE groups. Colorectal invasive cancer was diagnosed in 28 (5.6%) patients for each of the three groups. The detection rate for large polyps was significantly lower in the CTC group than in the CS group (11.0% vs. 21.2%; P! 0.05), and was almost equal to the rate in the BE group (11.0% vs. 16.8%). The detection rate for smaller polyps (5-9 mm) was significantly lower in the CTC group than in the CS group (17.0% vs. 25.0%; P! 0.05), and was almost equal to the rate in the BE group (17.0% vs. 15.6%). The frequency of no polyps was 332 (66.4%), 241 (48.2%), and 305 (61.0%) in the CTC, CS, and BE groups, respectively. The frequency of a diagnosis throughout the colon was significantly lower in the CTC group (382, 76.4%) than the CS group (438, 87.6%) and BE (438, 87.6%) (P! 0.05) groups. Additional examinations were required for the CTC (34, 6.8%), CS (61, 12.2%), and BE (24, 4.8%) groups. The cause of inadequate examination included excessive fluid or stool (33, 97.1%), spasms (8, 23.5%), or both (7, 20.6%) in the CTC group; not acceptable (56, 91.8%) or stenosis due to an advanced tumor (5, 8.2%) in the CS group; and excessive fluid or stool (17, 70.8%), spasms (3, 12.5%), or both (2, 8.3%) in the BE group. Conclusion: CTC provides the same diagnostic sensitivity for colorectal cancer (vs. CS and BE) and for polyps (vs. BE). Therefore, CTC should be used to examine FOBT-positive patients for colorectal cancer and polyps. Sa1501 The Efficacy of Tacrolimus and the Usefulness of Endoscopy in Predicting Its Efficacy in Patients With Refractory Ulcerative Colitis Osamu Watanabe*, Masanao Nakamura, Takeshi Yamamura, Kazuhiro Morise, Masanobu Matsushita, Asuka Nagura, Keiko Maeda, Toru Yoshimura, Arihiro Nakano, Hiroshi Oshima, Junichi Sato, Yasuaki Ueno, Masashi Saito, Rinzaburo Matsuura, Yasuyuki Mizutani, Kazuhiro Furukawa, Kohei Funasaka, Eizaburo Ohno, Ryoji Miyahara, Hiroki Kawashima, Kazuhiro Ishiguro, Yoshiki Hirooka, Takafumi Ando, Hidemi Goto Department of Gastroenterology and Hepatology, Nagoya University Graduate School of Medicine, Nagoya, Japan Background and Aim: Many patients with refractory ulcerative colitis (UC) are likely to require surgery in spite of pharmacological treatment. Tacrolimus, a calcineurin inhibitor, is expected to be an effective alternative drug which allows colectomy to be avoided. We retrospectively investigated patients with refractory UC treated with tacrolimus. Patients and Methods: Forty-seven patients with moderate or severe UC were treated with oral tacrolimus between July 2009 and June 2013 at our hospital. Dosage was adapted to achieve trough levels between 10 and 15 ng/mL for the first two weeks and between 5 and 10 ng/mL after the 3rd week. Clinical disease activity was calculated at baseline and weeks 2, 4 and 12 using the Lichtiger’s clinical activity index (CAI). A CAI score of 4 and below was defined as clinical remission. Sigmoid colonoscopy was performed at baseline and week 2 in 15 patients. Endoscopic activity was assessed by the presence or absence of endoscopic findings (ulcer, bleeding, edema, mucopurulent discharge, vascular pattern) and was also calculated using Rachmilewitz’s endoscopic index (EI) at baseline and week 2. Results: After four weeks of tacrolimus therapy, 28 patients (60%) showed a complete response to this therapy, 6 (13%) had mild to moderate disease activity, and 12 (27%) showed no response. One patient discontinued treatment due to light-headedness. Of the 20 of 47 patients with severe disease, 13 (65%) obtained complete remission. Fifteen of the 20 patients with severe disease were fasted for the first two weeks, of whom 12 (80%) entered complete remission, whereas 5 severe patients with oral intake were unresponsive to therapy without one patient. After 12 weeks of tacrolimus therapy, 25 of 28 patients who responded at the fourth week remained in remission. Corticosteroids (CS) were then tapered and discontinued. The mean dosage of CS was 18.5 mg/day before tacrolimus therapy and 2.1 mg/day after 12 weeks (p! 0.001). Sigmoid colonoscopy was performed in 15 patients at baseline and week 2. Eleven of 15 patients were in clinical remission at week 4. Mean Endoscopic Index score was 8.7 1.7 at baseline and 4.6 2.1 at week 2. .EI at baseline was not correlated with clinical remission at week 4, but endoscopic findings at week 2 were predictive. Patients with partial existence of normal vascular pattern at week 2 significantly achieved clinical remission at week 4. Conclusion: Oral tacrolimus was effective in AB240 GASTROINTESTINAL ENDOSCOPY Volume 81, No. 5S : 2015 inducing remission in patients with steroid-refractory UC. Oral intake of food might be detrimental to the efficacy of tacrolimus in patients undergoing remission induction with this agent. Endoscopic findings (partial existence of normal vascular pattern) at week 2 may be predictive of remission at week 4. Sa1502 INR Value Not Platelet Count Predicts Outcomes in Emergent Endoscopic Variceal Ligation Cyril Varghese*, Larissa L. Fujii-Lau, Patrick S. Kamath Mayo Clinic, Rochester, MN; Gastroenterology, Washington University in St. Louis, St Louis, MO Purpose: The risk of re-bleeding and mortality associated with platelet counts and INR following both emergent and elective Endocsopic Variceal Ligation (EVL) were evaluated. Methods: Three hundred forty patients (mean age 56; 68% male) were entered into a prospectively collected database. EVL was carried out electively in 242 patients and emergently in 98 patients. Re-bleeding and mortality was assessed in elective versus emergent EVL for different platelet and INR values. Results: In the elective group, mean age was 56; 68% male with etiology of cirrhosis being 23% NASH; 15% viral; 15% cholestatic; 10% alcoholic; 9 % both viral and alcoholic; and rest due to other causes. In the emergent group, mean age was 55, 68% male with etiology of cirrhosis 31% alcoholic; 16% both viral and alcoholic; 14% viral; 9% NASH; 4% cholestatic and rest due to other causes. The median INR in the elective group was 1.2 (IQR 1.1-1.3) and, in the emergent group, 1.4 (IQR 1.2-1.6); platelets were 89,000 (IQR 62,500-129,500) in the elective group and 85,500 (IQR 60,000135,500) in the emergent group. In the elective group, upper gastrointestinal (GI) bleeding occurred in 11 patients within 6 weeks. No patient died. In the emergent group, GI re-bleeding occurred in the first 5 days in 9 patients and between 5 days and 6 weeks in 12 patients. Twenty one patients died in the emergent group within 6 weeks of procedure. Eleven (52.3%) died of an upper GI re-bleeding event. An INR of R1.4 but not the platelet count was associated with re-bleeding and mortality in emergent EVL patients (Table 1 and 2). Conclusion: There are currently no clear guidelines determining safe limits for platelet counts and INR for endoscopic variceal ligation (EVL), either for emergent or elective procedures. Through this study we conclude that for emergent EVL, re-bleeding and mortality are associated with INR R 1.4 but not platelet counts. However, for elective EVL, upper GI bleeding is related neither to platelet count nor INR values. Table 1. Platelet and INR Association with Re-bleeding in Elective and Emergency EVL Elective EVL Emergent EVL N* Bleed No Bleed pvalue N* Bleed