• The American Journal of Gastroenterology volume 113, pages S1S5 (2018)
  • doi:10.1038/ajg.2017.451
  • Download Citation
Published online:

P-001 30-Day Readmission After Ileal Pouch Anal Anastomosis Surgery: A Report From ACS-NSQIP Database

Aydinli H. Hande 1, Lynn Patricio1, Aytac Erman2, Grieco Michael1, Remzi Feza3. 1NYU Langone Medical Center, New York, New York, 2Acibadem University School of Medicine, Istanbul, Turkey, 3NYU School of Medicine, New York, New York

BACKGROUND: Ileal pouch anal anastomosis (IPAA) is the preferred surgical option in patients with medically refractory ulcerative colitis (UC) to preserve gastrointestinal continuity. This study aimed to describe 30-day readmission rates, as well as predictive factors for it from a national dataset.

METHODS: Patients who underwent IPAA surgery for UC between 2012 and 2015 were identified from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database using current procedural terminology codes. Demographics, perioperative, and operative variables were collected. Patients were grouped according to the presence of 30-day readmission: (+)/(-).

RESULTS: Query identified 1882 patients, mean age was 40.8± 13.9 years, mean length of stay was 7.2± 5 days and postoperative 30-day morbidity rate was 28% (n=530). Most common complications in the study group were transfusion (7.5%), organ space surgical site infection (SSI) (3.3%), reoperation (3%) and superficial SSI (2%). Twenty-two percent (n=416) were readmitted within 30 days of surgery. Reasons for readmission were: surgical site infection (n=88), dehydration (n=77), small bowel obstruction/ileus (n=38/18) and abdominal pain (n=28). Mean time for readmission was 14± 7 days, 39 patients had second readmission and 4 had a third readmission within 30-days of surgery. Multivariate analysis showed an ASA score of 4 [OR: 14.4 (2.3-89.7), P=0.004] and age<40 [OR: 1.3 (1.08-1.7), P=0.006] were associated with 30-day readmission. Preoperative albumin level of <3.5 was associated with a second readmission [OR:3 (1.1-8.2), P=0.02)].

CONCLUSION(S): IPAA surgery for UC has high morbidity. One fifth of patients were readmitted within 30 days from IPAA surgery for UC and one third of them had a second readmission. This study brings the possibility and consideration for national health care initiative in surgical management of patients with UC, undergoing IPAA surgery.

P-002 United States Claims Database Analysis Comparing Safety, Medical Resource Utilization, and Treatment Costs Associated with Management of Inflammatory Bowel Disease

Long Gráinne1, Tatro Amanda2; Oh Young 2, Reddy Sheila3, Ananthakrishnan Ashwin4. 1Roche Pharmaceuticals Ltd., Welwyn Garden City, United Kingdom, 2Genentech, Inc., South San Francisco, California, 3Partnership for Health Analytic Research, Beverly Hills, California, 4Massachusetts General Hospital, Boston, Massachusetts

BACKGROUND: Pharmacologic options commonly used to control disease and alleviate symptoms in patients with inflammatory bowel disease (IBD) have limited effectiveness. Only a third of patients with IBD treated with commonly used options achieve clinical remission at 1 year, and most experience drug-related adverse events (AEs). Herein we characterized the clinical and economic burden of IBD treatments in terms of AEs of interest, medical resource utilization (MRU), and associated costs.

METHODS: Records in the IMS PharMetrics insurance claims database included in this analysis were for patients aged >18 years who had ≥2 medical claims (≥7 days apart), a diagnosis of ulcerative colitis (ICD-9-CM: 556.x) or Crohn’s disease (Crohn’s) (ICD-9-CM: 555.x), and with >1 qualifying claim in the year preceding treatment within the identification period July 1, 2010, to June 30, 2015. Drug classes of interest were oral corticosteroids (OCS), immunosuppressants (IS), anti–tumor necrosis factor agents (aTNF), or combinations thereof. Patients taking only aminosalicylates (ASA) were used as a reference. We compared AE incidence, MRU, and medical costs in patients with UC or Crohn’s across drug classes. Univariate comparisons included statistical tests of significance (χ2, F test, or Kruskal Wallis). Multivariate analyses included Cox proportional hazards regression, negative binomial regression, logistic regression, or linear regression analysis adjusted by significant sociodemographic, clinical, and disease severity covariates.

RESULTS: This analysis included 30,676 patients (Crohn’s: n=14,528; UC: n=16,148). OCS monotherapy was the strongest predictor of any AE occurring for patients with Crohn’s or UC (Crohn’s: hazard ratio [HR] [95% CI]; HR, 1.62 [1.51-1.73]; UC: HR, 1.57 [1.49-1.66]). A similar pattern was observed for severe infection (Crohn’s: HR, 2.43 [2.07-2.85]; UC: HR, 2.37 [2.07-2.72]) and bone-related conditions (Crohn’s: HR, 1.88 [1.74-2.03]; UC: HR, 1.77 [1.67-1.89]). The strongest predictors for serious hepatic events were IS + OCS (Crohn’s: HR, 2.38 [1.72-3.31]; UC: HR, 2.36 [1.75-3.18]) and OCS (Crohn’s: HR, 2.09 [1.65-2.66]; UC: HR, 1.92 [1.57-2.34]). Compared with patients receiving other therapies, patients with UC or Crohn’s receiving OCS or IS + OCS were more likely to have emergency department visits; IBD-related hospitalization, visits, or procedures; and gastrointestinal surgery. Annualized total medical costs were greatest for aTNF + IS or aTNF therapy in both Crohn’s and UC. However, annualized medical service costs (that exclude IBD drug costs) were highest for patients initiating OCS-containing therapies (Crohn’s: OCS, $27,041 and OCS + IS, $23,332 [P<0.001]; UC: OCS, $19,659 [P<0.001]) followed by other index therapies (Crohn’s: ASA, $10,823, and aTNF + IS, $19,151 [P<0.001]; UC: ASA, $7980, and aTNF + IS, $18,771 [P<0.001]).

CONCLUSION(S): Chronic OCS use was associated with increased risk of severe infection, bone conditions, and serious hepatic events compared with other therapies. Consistent with an increased AE risk, OCS regimens were associated with higher rates of MRU and medical service costs compared with other therapies. Accordingly, treatment decisions should consider alternate options which may provide more favorable long-term benefits.

P-003 Improved Quality of Care for IBD Patients Using HealthPROMISE App: A Randomized, Control Trial

Atreja Ashish 1, Khan Sameer2, Szigethy Eva3, Otobo Emamuzo4, Shroff Hersh4, Chang Helena4, Keefer Laurie5, Rogers Jason2, Deorocki Allyssa4, Ullman Thomas6, Marion James4, Cohen Benjamin4, Maser Elana7, Chapman Mark4, Itzkowitz Steven4, Colombel Jean-Frederic8, Sands Bruce2. 1Mount Sinai School of Medicine, New York, New York, 2Icahn School of Medicine at Mount Sinai, New York, New York, 3Department of Psychiatry, University of Pittsburgh, Pittsburgh, Pennsylvania, 4Icahn School of Medicine at Mount Sinai, New York, New York, 5Mount Sinai Health System, New York City, New York, 6Icahn School of Medicine at Mount Sinai, New York, New York, 7Mount Sinai, New York, New York, 8Dr. Henry D. Janowitz Division of Gastroenterology, Icahn School of Medicine at Mount Sinai, New York, New York

BACKGROUND: Inflammatory bowel disease (“IBD”) is a chronic condition affecting over one million people in the United States (1). The recurrent and potentially debilitating nature of IBD elevates patients’ risk for adverse health outcomes. IBD patients and providers report rushed visits, impersonal communications, and constrained resources as barriers to quality care (2). IBD patients are an ideal population to assess the therapeutic potential of a digital intervention used in conjunction with clinical methods for long-term IBD management. HealthPROMISE is an innovative software platform, developed by Sinai AppLab, comprising a patient mobile application linked to a cloud-based decision support dashboard designed to improve health outcomes and enhance quality of care (“QOC”) by increasing patient engagement, self-management skills, and communication transparency (3,4).

METHODS: Recruited patients provided informed consent during in-person office visits and were randomized into intervention (HealthPROMISE) or control. Patients completed an intake questionnaire assessing health literacy, disease severity, general health status, and demographic information. The primary endpoint is QOC data based on American Gastroenterological Association’s QOC indicator checklist, which was verified against and conformed to EPIC records. Secondary endpoints include decrease in IBD-related emergency visits and hospitalizations, change in quality of life (“QOL”) score from baseline, and proportion of patients reporting controlled disease status per group. In the app, HealthPROMISE patients update their information and receive a disease summary of QOC metrics and IBD-specific QOL trends. A population health coordinator monitored patient data in real-time on the HealthPROMISE dashboard and communicated with the care team and patients as needed. Ongoing collection of follow-up exit survey data captures overall medication adherence, system usability scale (SUS), SIBDQ, patient activation measures (PAM-13), and general health status (EQ-5D).


RESULTS: 320 patients were enrolled in the study. 162 were randomized to intervention group and 158 to control group (Females 49.1%; White 82.2%; Black 5.3%; Hispanics 9.1%; English as primary language 96.3%; Everyday Computer Usage 93.4%) (Figure 1). IBD-QOL continued to improve among HealthPROMISE patients over a follow-up of 575 days (25.2±11.3 vs. 30.3±11.3 baseline, P<0.001) (Figure 2). Patients reported that uncontrolled anxiety (89.4%) and uncontrolled fatigue (80.9%) were major drivers of poor QOL and disproportionately contributed to disease burden (Figure 3). After an average follow-up of 495±135 days, QOC improved among all patients (78% vs. 59% control), with a more significant increase since baseline observed among HealthPROMISE users (+28 ppt vs. +9 ppt, P<0.01). After a second follow-up of 575±135 days, QOC continued improving (84% vs. 65% control) with a more significant change from baseline observed among HealthPROMISE users (+34 ppt vs. +15 ppt, P<0.01) (Figure 4). After 575 days, Screening Colonoscopy was the most met QOC indicator (92% met) while Smoking Cessation Screening with the least documented QOC indicator (7% met) (Figure 5).

CONCLUSION(S): A significant improvement in QOC was observed among patients using HealthPROMISE. IBD patients engaging with HealthPROMISE reported more equitable participation in their care decision-making process, and showed improved health outcomes compared to patients not using HealthPROMISE. Digital health interventions and IBD remote monitoring can address gaps in QOC, increase patient engagement, and improve health outcomes.

P-004 The Use of Alvimopan as Prophylaxis against Post-Operative Ileus After Bowel Resection in Patients With Inflammatory Bowel Disease

Jang Janice 1, Kwok Benjamin1, Grucela Alexis1, Bernstein Mitchell1, Remzi Feza1, Hudesman David2, Chen Jingjing1, Chang Shannon1. 1NYU School of Medicine, New York, New York, 2New York University School of Medicine, New York, New York

BACKGROUND: Postoperative ileus (POI) is a temporary delay of coordinated intestinal peristalsis following major abdominal surgery, leading to significant symptoms such as nausea, vomiting, abdominal pain, prolonged hospitalization, nosocomial complications, and physical deconditioning. The use of opioids for postoperative pain control further exacerbates the problem. Opioids bind to the mu receptors in the intestinal tract, leading to gut hypomotility. Alvimopan, an oral, peripherally acting mu-opioid receptor antagonist, was FDA approved in 2008 for use before and after bowel resection to help prevent and treat POI. There are no dedicated studies of alvimopan in patients with inflammatory bowel disease (IBD). Therefore, we conducted a study to investigate alvimopan's role in IBD patients who underwent either laparoscopic or open bowel resection.

METHODS: A retrospective chart review was conducted at a 725-bed acute care teaching hospital in New York City between January 2012 and February 2017. Data collected included age, sex, type of IBD, length of stay, post-operative gastrointestinal symptoms (nausea, vomiting, constipation, abdominal distention, first flatus, first bowel movement, PO tolerance), and dose of alvimopan, were collected. The primary outcome was time to GI recovery. Secondary outcomes were: time to first flatus, time to first bowel movement, time to tolerating a liquid diet, time to tolerating solid food, and total length of stay. Descriptive statistics reports were created through a secure web-based application called REDCap (Research Electronic Data Capture), and the data were exported into Stata to run further analyses. Of note, approximately 50% of patients who underwent bowel surgery after March 2015 were placed on a “colon surgery pathway”, which is an order set dedicated to strategies that decrease length of stay and post-operative complications. Key features include early feeding, optimized analgesia regimen to allow patients to ambulate, encouraging use of incentive spirometry, and administration of alvimopan peri-procedurally.

RESULTS: Of 247 patients, 121 received alvimopan (49.0%) and 126 (51.0%) did not. The male to female ratio was 51:49. The mean age of the control group was 44.4 + 16.3 and that of the alvimopan group was 43.2 + 16.4. Patients who received alvimopan had faster GI recovery, with a hazard ratio (HR) of 2.11 (P<0.001), shorter time to first flatus (HR 2.02, P<0.001), shorter time to first bowel movement (HR 1.93, P<0.001), shorter time to tolerating liquid diet (HR 2.48, P<0.001), and shorter time to tolerating solid food (HR 2.00, P<0.001). After controlling for type of bowel resected (large vs. small bowel), laparoscopic vs. open, age, and type of IBD (ulcerative colitis vs. Crohn’s disease) using linear regression, patients who received alvimopan spent 2.59 fewer days in the hospital compared to the control group (P<0.01).

CONCLUSION(S): The results of this study suggest that alvimopan is effective in accelerating the time to GI recovery. Data analysis of all primary and secondary outcomes revealed that alvimopan had a statistically significant benefit during the post-operative period of IBD patients undergoing bowel resection. Length of stay for IBD patients was significantly decreased with peri-operative use of alvimopan.

P-005 Cytomegalovirus Infection in Inflammatory Bowel Diseases in a Large Single Center Cohort

Toruner Murat 1, Manti Bengu2, Keskin Onur2. 1Ankara University School of Medicine, Ankara, Turkey, 2Ankara University School of Medicine, Ankara, Turkey

BACKGROUND: Cytomegalovirus (CMV) viremia and/or infections are frequently seen in patients with inflammatory bowel diseases (IBD) especially in conditions of steroid refractoriness. In this study, we aimed to determine the prevalence of CMV infection/viremia in IBD patients as well as to evaluate underlying risk factors and outcomes of CMV infections in such patients.

METHODS: In this retrospective study, we evaluated 1152 IBD patients admitted to our outpatient gastroenterology clinic upon exacerbating symptoms of IBD. In all patients, CMV viremia was analysed with real-time polymerase chain reaction (PCR) in blood samples. Mucosal samples from these patients were further analyzed with histopathology and CMV immunochemistry (IHC). Patient characteristics including age, gender, disease phenotype, used medications, laboratory test results were abstracted from patient charts. Furthermore, to identify risk factors and clinical outcomes in patients with CMV colitis, 26 patients with CMV colitis were compared with IBD patients with no CMV colitis.

RESULTS: Among 1152 patients, 26 patients had CMV colitis confirmed with IHC/Histopathology. Besides, 141 patients had a positive CMV-DNA result (>42 copies/mL). The prevalence of CMV viremia and colitis in IBD patients were 12.2% and 2.25%, respectively. The occurrence of CMV colitis was found to be associated with advanced age (P<0.004), presence of comorbid diseases (P=0.013) and combination treatment of steroids and azathioprine (P=0.049). In multivariate analysis, steroid use was found to increase CMV colitis three-fold. Also, we found that, patients with CMV colitis had longer hospital stays than in patients without CMV colitis (P<0.001). CMV-DNA serum level more than 592.5 copies/mL was found to be predictive for CMV colitis (P=0.014). Colectomy rates did not differ between CMV colitis patients and patients with no CMV colitis. However, mortality rate was observed higher in CMV colitis group compared to the patients with no CMV colitis (11.5% vs 2.3%, P=0.04).

CONCLUSION(S): To our knowledge this study consists of the larger number of CMV colitis patients from a single center. In this retrospective study, CMV viremia and/or colitis was not infrequent in IBD patients. Steroid use, comorbid diseases and advanced age increases the risk of CMV colitis in IBD patients. Mortality rate seems to be higher in patients with CMV colitis.

P-006 Psychiatric Characterization of Patients Enrolled in IBD Subspecialty Home

Szigethy Eva1; Weaver Emily 2, Click Benjamin2, Kogan Jane3, Shrank William3, Mcauliff Katheryn4, Strassburger Meredith2, Regueiro Miguel2. 1Department of Psychiatry, University of Pittsburgh, Pittsburgh, Pennsylvania, 2University of Pittsburgh, Pittsburgh, Pennsylvania, 3UPMC, Pittsburgh, Pennsylvania, 4Boston College, Chestnut Hill, Massachusetts

BACKGROUND: Understanding the psychiatric needs of IBD patients is important for making the integration of behavioral services into medical care cost-effective. One such integrated model of care, the medical home, has increasingly focused on the Four Quadrant model which describes mental and physical health complexity and risk so that integrated treatment can be personalized and scaled accordingly. This study describes the way that psychiatric risk is determined in an IBD subspecialty medical home (SMH) and characterizes the life-time psychiatric diagnoses, current severity of psychopathology, and the baseline physical and mental complexity of IBD patients enrolled for at least one year.

METHODS: Starting in June, 2015, patients were enrolled in an IBD SMH if 1) between ages 18-55; 2) confirmed IBD as primary physical condition; and 3) had UPMC Health Plan insurance. All patients are screened for anxiety (GAD7) and depression (PHQ9) and are offered an intake with a licensed clinical social worker (LCSW) if: 1) self-reported screen scores are above a threshold of >=10; 2) history of psychiatric disorders; 3) on psychotropic medication; or 4) the medical team deems the patient emotionally distressed. The LCSW completes a brief psychiatric diagnostic interview at point of GI care and for a subset with high GAD7 or PHQ9 screen scores, the DSM-V Cross-Cutting Symptom Measure to screen for other mental health domains across psychiatric disorders. The SMH team completes the IBD Complexity Grid to quantify current and past complexity in the biological (IBD), psychological, social, and health systems domains using both patient-reported and objective criteria. Patients who met the 75th percentile cut-off for their IBD Complexity Grid biological and psychological domains were scored as high severity in the quantification of patients into physical and mental quadrants. A psychiatrist is also available within the SMH for more severe or complex psychiatric patients. The patient’s team care-plan is determined by a weekly SMH team conference.

RESULTS: Of 346 patients enrolled in the medical home, mean age was 34.8 (9.7), 42% male, 81% Caucasian, and 61 with Crohn’s disease. 256 patients met with the LCSW and 202 had at least one lifetime psychiatric disorder. The mean PHQ9 and GAD7 screen scores were 7.1 (6.2) and 5.8 (5.7); with 46% and 38% having elevated screen scores respectively. In terms of life-time psychiatric diagnoses, anxiety and depressive disorders followed by somatic symptom disorder, PTSD and ADHD were most common. For the subset with elevated GAD7 or PHQ9, 68% had clinically significant sleep disturbance, 44% irritability, 24% with maladaptive personality traits, 12% with drug or alcohol abuse, 12% with obsessive compulsive and 8% with manic symptoms. In terms of the baseline quadrant characterization, 136 (53%) were low-low and 31(12%) were high-high physical and mental complexity, 46 (18%) were high physical and 44 (17%) were high mental complexity only.

CONCLUSION(S): These findings confirm high rates of psychiatric comorbidity in IBD patients within a SMH at a large tertiary medical center. Characterization of mental health needs within an IBD clinic is feasible. Such quantification allows for data-driven behavioral health resource allocation.

P-007 Exposure-Response to SC Ustekinumab in Moderate –Severe Crohn’s Disease: Results From the IM-UNITI Maintenance Study

Adedokun Omoniyi 1, Xu Zhenhua2, Gasink Christopher1, Jacobstein Douglas2, Szapary Philippe2, Johanns Jewel1, Gao Long-Long1, Davis Hugh3, Hanauer Stephen4, Feagan Brian5, Ghosh Subrata6, Sandborn William7. 1Janssen Research and Development, LLC, Spring House, Pennsylvania, 2Janssen Research & Development, LLC, Spring House, Pennsylvania, 3Janssen Research & Development, LLC, Spring House, Pennsylvania, 4Northwestern University, Chicago, Illinois, 5Robarts Research Institute, University of Western Ontario, London, Canada, 6University of Birmingham, Birmingham, United Kingdom, 7University of California, San Diego, La Jolla, California

BACKGROUND: Subcutaneous (SC) ustekinumab (UST) maintenance was previously shown to maintain clinical response and remission in moderate-severe Crohn’s disease (CD) in the Phase 3 IM-UNITI maintenance trial, but detailed exposure-response (ER) analyses using clinical efficacy, biomarkers of inflammation, and endoscopic substudy data have not yet been presented.

METHODS: Patients achieving clinical response after a single IV dose of UST in the UNITI-1&2 induction studies were randomized 1:1:1 in IM-UNITI to SC placebo (PBO), UST 90 mg q12w, or q8w. Ustekinumab concentration data were categorized into quartiles (Q) and ER relationships were assessed based on wk24 steady-state (ss) trough (or average ss trough) values vs clinical remission, CRP, and endoscopic remission and response (SES-CD<2 or 50% decrease, respectively) in endoscopic substudy patients. The associations between ss trough and efficacy outcomes were assessed using a one-sided Cochrane Armitage trend test. Optimal concentration cut-offs for remission were evaluated using ROC analysis.

RESULTS: Median ss trough concentrations were approximately 3x greater with q8w compared to q12w SC UST (2.11 μg/mL and 0.62 μg/mL, respectively). UST levels were lower in patients with higher baseline CRP (likely a result of greater inflammatory burden), while oral AZA, 6-MP, or MTX use and weight did not result in any notable differences. With both q8w and q12w UST, there were associations with ss trough concentrations (at week 24) and higher proportions of patients achieving clinical remission (ie in q8w Qs2-4 [>0.92 μg/ml] and q12w Qs 3-4 [[>0.64 μg/ml]), lower (ie normalized, <0.3 mg/dL) CRP concentrations at week 24 (ie in combined week 24 ss Qs 3-4 [>1.05 μg/ml], P<0.001]), and endoscopic response as well as endoscopic remission (in combined substudy patient average ss trough Qs 2-4 [>0.5 μg/ml], P=0.006 and 0.054, respectively). By ROC analyses, ss trough concentrations greater than 0.8-1.4μg/mL were associated with higher rates of maintenance of clinical remission.

CONCLUSION(S): In Crohn’s disease patients in the IM-UNITI maintenance study, ustekinumab concentrations above approximately 1μg/mL (across endpoints and analyses, actually ranging from 0.5-1. 4μg/mL) were positively associated with clinical efficacy outcomes and inversely related to CRP concentrations.

P-008 Relationship Between Vedolizumab Concentrations and Deep Remission in Patients With Moderately-to-Severely Active Ulcerative Colitis: A GEMINI 1 Post Hoc Analysis

Sandborn William 1, Colombel Jean-Frederic2, Panaccione Remo3, Dulai Parambir4, Rosario Maria5, Cao Charlie6, Barocas Morris7, Lasch Karen8. 1University of California, San Diego, La Jolla, California, 2Dr. Henry D. Janowitz Division of Gastroenterology, Icahn School of Medicine at Mount Sinai, New York, New York, 3University of Calgary, Division of Gastroenterology, Calgary, Canada, 4University of California San Diego, LaJolla, California, 5Takeda Pharmaceuticals, Cambridge, Massachusetts, 6Takeda Pharmaceuticals USA, Inc., Deerfield, Illinois, 7Takeda Pharmaceuticals U.S.A., Inc., Deerfield, Illinois, 8Takeda Pharmaceuticals USA, Inc., Deerfield, Illinois

BACKGROUND: Deep remission is an important treatment goal for patients with ulcerative colitis (UC).1 Vedolizumab (VDZ), a gut-selective monoclonal antibody to α4β7 integrin, has demonstrated efficacy in achieving clinical remission in patients with UC.2 Here, we report a post hoc analysis of the phase 3 GEMINI 1 trial to determine if VDZ trough serum concentrations correlated with Week 52 deep remission rates in patients with UC.

METHODS: Patients who responded to VDZ induction therapy at Week 6 were re-randomized to placebo (PLA), VDZ every 8 weeks (Q8W), or VDZ every 4 weeks (Q4W) for 46 weeks. Deep remission at Week 52 was evaluated using four different criteria: (1) endoscopic remission and symptomatic improvement (Mayo Clinic endoscopic score [ES]=0, rectal bleeding score [RBS]=0, and decrease in stool frequency score [SFS] or no increase from baseline); (2) endoscopic improvement and symptomatic remission (Mayo Clinic ES ≤1, RBS=0, and SFS=0); (3) total score ≤1 with endoscopic and symptomatic improvement (Mayo Clinic ES ≤1, RBS=0, decrease in SFS or no change from baseline, and total score [ES + RBS + SFS] ≤1); (4) endoscopic and symptomatic improvement (Mayo Clinic ES ≤1, RBS=0, and SFS ≤1). VDZ trough serum concentrations were evaluated at Week 46; VDZ Q8W and Q4W groups were combined for the serum concentration analysis.

RESULTS: At Week 6, 373 patients responded to VDZ and were re-randomized to maintenance PLA (n=126; median age 39.9 years [range 18.0–74.0]; 45.2% female), VDZ Q8W (n=122; median age 39.7 years [range 19.0–78.0]; 43.0% female), or VDZ Q4W (n=125; median age 35.7 years [range 19.0–76.0]; 46.0% female). Significantly more VDZ-treated patients achieved deep remission at Week 52 versus PLA for all definitions of deep remission: (1) PLA 8.7%, VDZ Q8W 27.0% (P=0.0001), VDZ Q4W 28.0% (P<0.0001); (2) PLA 13.5%, VDZ Q8W 32.8% (P=0.0002), VDZ Q4W 31.2% (P=0.0006); (3) PLA 15.1%, VDZ Q8W 36.1% (P=0.0001), VDZ Q4W 36.0% (P=0.0001); (4) PLA 15.9%, VDZ Q8W 43.4% (P<0.0001), VDZ Q4W 43.2% (P<0.0001). There was a trend towards higher Week 52 rates of deep remission with higher Week 46 trough serum concentrations of VDZ for all deep remission definitions. Rates of deep remission for the lowest (<9.26 μg/mL) and highest (>41 μg/mL) quartiles, respectively, were: (1) 30.0% (95% confidence interval [CI 15.8–44.2]) vs 47.5% (95% CI 32.0–63.0); (2) 35.0% (95% CI 20.2–49.8) vs 57.5% (95% CI 42.2–72.8); (3) 40.0% (95% CI 24.8–55.2) vs 67.5% (95% CI 53.0–82.0); (4) 47.5% (95% CI 32.0–63.0) vs 75.0% (95% CI 61.6–88.4).

CONCLUSION(S): VDZ was associated with significantly higher rates of deep remission than PLA at Week 52, regardless of dosing frequency or definition of deep remission. Patients on the highest quartile of trough serum concentration achieved a higher deep remission rate, for all definitions.


1. Zallot C et al. Curr Gastroenterol Rep 2013;15:315.

2. Feagan BG et al. NEJM 2013;369:699.

P-009 Response and Remission After 16 Weeks of Ustekinumab– An All Patients Analysis From the UNITI Crohn’s Studies

Colombel Jean-Frederic 1, Sloan Sheldon2, Gasink Christopher3, Gao Long-Long3, Jacobstein Douglas4, Lee Scott5, Targan Stephan6. 1Dr. Henry D. Janowitz Division of Gastroenterology, Icahn School of Medicine at Mount Sinai, New York, New York, 2Janssen Scientific Affairs, LLC, Horsham, Pennsylvania, 3Janssen Research and Development, LLC, Spring House, Pennsylvania, 4Janssen Research & Development, LLC, Spring House, Pennsylvania, 5University of Washington School of Medicine, Seattle, Washington, 6Cedars Sinai Medical Center, Los Angeles, California

BACKGROUND: Ustekinumab (UST) has been shown to induce & maintain clinical response & remission in moderate-severe Crohn’s disease (CD) in 2 induction [(UNITI-1(anti-TNF failures) & UNITI-2(anti-TNF non-failures) & 1 maintenance (IM-UNITI)] randomized, PBO-controlled Ph3 trials. We evaluated efficacy (response & remission) for all pts who received an IV induction dose of approximately 6mg/kg, including responders (CDAI decrease ≥100) & non-responders, 8wks after first UST maintenance dose of 90mg SC, i.e. 16wks from the IV induction dose.

METHODS: Pts achieving clinical response 8wks after a single IV induction dose were randomized to SC PBO, UST 90mg q12wks or q8wks. UST pts not in clinical response 8wks after IV induction dose were given UST90mg SC & if in clinical response 8wks later were continued on 90mg SC q8w dosing. A total of 458pts were exposed to an IV induction dose of 6mg/kg (UNITI-1, N=249 & in UNITI-2 N=209) with a response rate at wk8 of 37.8% & 57.9% vs. PBO response rate of 20.2% & 32.1% respectively. The remission rate at wk8 in UNITI-1 & UNITI-2 was 20.9% & 40.7 vs. PBO of 7.3% & 19.6% respectively. For this evaluation, response & remission status of the entire population exposed to an IV induction dose of 6mg/kg of UST was evaluated 8wks after the first SC maintenance dose of UST. All pts who received 6mg/kg IV UST induction were included, including responders randomized to SC PBO (who did not receive SC UST at wk8).

RESULTS: Of the 219 pts, not in clinical response to a single UST IV induction dose in UNITI 1&2, 37.6% & 60.5% respectively were in clinical response 8wks after the first maintenance UST dose (90 mg SC). Evaluating all pts exposed to 6mg/kg IV UST induction, response rates 8wks after the first SC injection (16wks after the IV induction dose) for UNITI1&2 were 47.4% & 73.7% respectively. In the sub-population who were anti-TNF naïve upon enrolment into UNITI-2 (n=144), at 8 weeks after the first SC injection, 72.9% were in clinical response and 60.4% were in clinical remission.

CONCLUSION(S): These numbers at wk16 are expected to reflect real world experience in pts who receive the induction dose & one additional maintenance dose 8wks later. The resulting rates of response & remission are higher than previously reported in induction studies across all populations (anti-TNF non-failures & anti-TNF failures). About 73% of anti-TNF non-failures attain clinical response & over half are in remission. The data support the clinical rationale for providing at least one SC maintenance dose of UST irrespective of clinical response 8wks after IV induction to assess UST benefit.

P-010 Endoscopic and Clinical Efficacy Demonstrated with Oral Ozanimod in Active Crohn’s Disease in Biologic-Naïve and Biologic Experienced Patients

Feagan Brian 1, Sandborn William2, Danese Silvio3, D’Haens Geert4, Levesque Barrett5, Skolnick Brett5, Paul David5, Aranda Richard5, Olson Allan5. 1Robarts Research Institute, University of Western Ontario, London, Canada, 2University of California, San Diego, La Jolla, California, 3Humanitas Clinical and Research Center, Milan, Italy, 4Academic Medical Centre of the University of Amsterdam, Amsterdam, The Netherlands, 5Receptos, a wholly owned subsidiary of Celgene, San Diego, California

BACKGROUND: Ozanimod, an oral, once-daily immunomodulator that selectively targets S1P1R and S1P5R, has demonstrated efficacy in ulcerative colitis (UC) (Sandborn NEJM 2016) is being evaluated in patients with active Crohn’s Disease (CD). This Phase 2 open-label study examined endoscopic and clinical outcomes following treatment with ozanimod 1 mg daily for 12 weeks in biologic-naïve and biologic-experienced CD patients.

METHODS: Patients with active CD defined as Crohn’s Disease Activity Index [CDAI] score 220-450 and simple endoscopic score for CD [SES-CD] ≥6 (isolated ileum disease SES-CD ≥4), were dosed daily with ozanimod 1 mg. Endoscopic assessments were read by an imaging core lab in a blinded manner. Daily electronic diary records were used to collect CD symptoms (including abdominal pain and soft/loose stool frequency). SES-CD was evaluated at baseline and Week 12, and CDAI was assessed at baseline, Weeks 4, 8, 12.

RESULTS: Sixty-nine patients were enrolled and treated. At baseline, mean age was 37.7 years, mean SES-CD was 13.3, and mean CDAI was 320. Mean CD duration was 10.0 years, with 54% of patients having had prior exposure to biologic therapy (i.e., TNF-α, vedolizumab). Patients with available data at baseline and Week 12 were included in the analyses, including SES-CD matched endoscopic segments (n=60 patients; 27, biologic-naïve; 33, biologic-experienced) and CDAI (n=59). Overall, reductions from baseline of ≥25% and ≥50% in SES-CD score were seen in 43.3% and 26.7% of patients, respectively. Reductions from baseline of ≥25% and ≥50% in SES-CD score were seen in 48.1% and 33.3% of biologic-naïve patients, and 39.4% and 21.2% of biologic-experienced patients, respectively Overall, CDAI response (CDAI decrease ≥100) was demonstrated in 66.1% of patients and CDAI remission (CDAI <150) was demonstrated in 45.8% of patients at Week 12. In biologic-naïve patients, CDAI response was demonstrated in 74.1% of patients and CDAI remission was demonstrated in 63.0% of patients. In biologic-experienced patients, CDAI response and remission rates were 59.4% and 31.3%, respectively.

Adverse event and serious adverse event rates were predominantly related to underlying Crohn’s disease with no specific drug attributable serious adverse events demonstrated. The overall safety profile in CD was similar to that observed in UC.

CONCLUSION(S): In this open-label study, ozanimod therapy demonstrated meaningful clinical and endoscopic improvements at Week 12 in both biologic-naïve and biologic-experienced patients with moderate to severe CD. No drug-specific safety signals were identified.

Reference:Sandborn WJ, Feagan BG, Wolf DC, D'Haens G, et al. Ozanimod Induction and Maintenance Treatment for Ulcerative Colitis. N Engl J Med. 2016 May 5;374(18):1754-62.

P-011 Etrolizumab as Induction Therapy in Moderate to Severe Crohn’s Disease: Results From BERGAMOT Cohort 1

Sandborn William 1, Panes Julian2, Jones Jennifer3, Hassanali Azra4, Jacob Rhian5, Sharafali Zaineb4, Oh Young4, Tole Swati4. 1University of California, San Diego, La Jolla, California, 2University of Barcelona, Barcelona, Spain, 3Queen Elizabeth II Health Sciences Center and Dalhousie University, Halifax, Canada, 4Genentech, Inc., South San Francisco, California, 5Roche Products Limited, Welwyn Garden City, United Kingdom

BACKGROUND: Etrolizumab, a humanized anti-β7 monoclonal antibody currently undergoing Phase 3 evaluation in ulcerative colitis (UC) and Crohn’s disease (CD), was demonstrated to be well tolerated and efficacious in patients with moderate to severe UC in the Phase 2 trial EUCALYPTUS. BERGAMOT (NCT02394028), a Phase 3 study, was designed with 3 sequential induction cohorts and a single maintenance cohort to evaluate the safety and efficacy of etrolizumab in patients with moderate to severe CD. This abstract describes the results for the exploratory induction cohort of 300 patients.

METHODS: Eligible patients with moderate to severe CD (confirmed at baseline with centrally read ileocolonoscopy) who were refractory or intolerant to anti–tumor necrosis factor α agents (aTNFs), immunosuppressants, and/or corticosteroids were randomly assigned in a 2:2:1 ratio to receive etrolizumab 105 mg subcutaneously every 4 weeks, etrolizumab 210 mg at weeks 0, 2, 4, 8, and 12, or placebo during a 14-week induction period. End points assessment included CDAI remission (CDAI<150), CDAI-100 response, CDAI-70 response, PRO2 remission (weighted combined score ≤ 11, based on patient report of liquid/very soft stool frequency [SF] and abdominal pain [AP]), symptomatic remission (unweighted SF ≤ 3 and AP ≤ 1), and endoscopic improvement (≥50% reduction from baseline SES-CD using central reading) at week 14.

RESULTS: Three hundred patients (73% aTNF-experienced) with moderate to severe CD (mean CDAI [SD], 315.6 [60.0]; mean SES-CD [SD], 14.1 [7.3]; median fecal calprotectin [range], 918 [30–15451] μg/g; median C-reactive protein [range], 9.75 [0.2-148.0] mg/L) received etrolizumab 105 mg (n=120), etrolizumab 210 mg (n=121), or placebo (n=59). Symptomatic remission (SF ≤ 3 and AP ≤ 1) was observed in a greater proportion of patients treated with etrolizumab 105 mg and 210 mg compared with placebo at week 6 (15.0%, 90% CI [10.4, 21.1] and 25.6% [19.7, 32.6] vs 8.5% [4.2, 16.4], respectively), week 10 (15.8% [11.1, 22.1] and 27.3% [21.2, 34.4] vs 8.5% [4.2, 16.4]), and week 14 (20.8% [15.4, 27.5] and 24.8% [18.9, 31.8] vs 11.9% [6.6, 20.5]). A greater proportion of patients achieved endoscopic improvement with etrolizumab 105 mg (21.0% [15.6, 27.8]) and 210 mg (17.4% [12.4, 23.7]) compared with placebo (3.4% [1.1, 9.7]) at week 14. CDAI remission was achieved at week 14 in 23.3% (17.6, 30.2) of patients receiving etrolizumab 105 mg, 28.9% (22.7, 36.1) receiving 210 mg, and 16.9% (10.4, 26.4) receiving placebo; PRO2 remission was achieved at week 14 in 28.3% (22.1, 35.5), 28.9% (22.7, 36.1), and 20.3% (13.1, 30.2) of patients, respectively. Etrolizumab was well tolerated. The frequency of adverse events was comparable with placebo and no deaths, anaphylaxis, or progressive multifocal leukoencephalopathy occurred.

CONCLUSION(S): In this exploratory induction cohort, treatment with etrolizumab was well tolerated and resulted in clinically meaningful endoscopic improvement, with rapid symptomatic remission as early as week 6 that was sustained through week 14. These early results are indicative of the efficacy of etrolizumab in treating CD. Enrollment into subsequent induction cohorts and into the maintenance phase of BERGAMOT is ongoing.

P-012 Ozanimod, an Oral S1P Receptor Modulator, Is Effective and Well-Tolerated in the Long-Term Treatment of Moderate to Severe Ulcerative Colitis

Feagan Brian 1, Sandborn William2, D’Haens Geert3, Hanauer Stephen4, Wolf Douglas5, Vermeire Severine6, Paul David7, Petersen AnnKatrin7, Liu Jerry7, Aranda Richard7, Olson Allan7. 1Robarts Research Institute, University of Western Ontario, London, Canada, 2University of California, San Diego, La Jolla, California, 3Academic Medical Centre of the University of Amsterdam, Amsterdam, The Netherlands, 4Northwestern University, Chicago, Illinois, 5Atlanta Gastroenterology Associates, Atlanta, Georgia, 6UZ Leuven Department of Gastroenterology, Leuven, Belgium, 7Receptos, a Wholly Owned Subsidiary of Celgene, San Diego, California

BACKGROUND: Ozanimod, an oral, once-daily immunomodulator that selectively targets S1P1R and S1P5R, has demonstrated clinical efficacy in UC for induction and maintenance therapy in the TOUCHSTONE trial (Sandborn NEJM 2016). The objective of the open-label extension (OLE) of the TOUCHSTONE trial is to evaluate long-term efficacy and safety of daily 1 mg ozanimod in patients with moderate to severe UC who had initially participated in TOUCHSTONE for up to 32 weeks.

METHODS: In TOUCHSTONE,197 patients were randomized (1:1:1) and treated with placebo (n=65), ozanimod 0.5 mg (n=65), or 1 mg (n=67). 170 (86%) entered the OLE. Efficacy data are reported as observed through Week 104 and safety includes all events through the data cut-off of March 2017.

RESULTS: At OLE entry, the partial Mayo Score (pMS) for patients on placebo, ozanimod 0.5 mg, and 1 mg was 4.6, 4.5, and 3.3 respectively, which improved rapidly by OLE Week 4 (-1.7, -1.5 and -0.8), and further improved through OLE Week 8 (–2.3, –1.9 and –1.1). Improvement was sustained up to OLE Week 104 (-2.3, -2.2 and -0.9). Improvement with ozanimod 1 mg was seen in all groups, with the greatest improvement reported in patients who had received placebo or ozanimod 0.5 mg in the TOUCHSTONE trial. All three groups achieved similar long-term improvement in pMS. The rectal bleeding score (RBS) improved rapidly in all groups, particularly in patients who had received placebo prior to OLE: based on observed cases, at OLE entry, 30.9% had no blood in their stools (RBS 0), and 67.3% had RBS ≤1; at OLE Weeks 4 and 8, 66.0% and 76.9% had RBS=0; and 94.3% and 98.1% had RBS ≤1. The effect on rectal bleeding was sustained through OLE Week 104; in all patients combined who remained on ozanimod, 88.2% reported a RBS=0 and 98.7% reported a RBS ≤1, at OLE Week 104. AEs and SAEs were reported in 50.0% and 11.1% patients. The only SAEs in ≥2 patients were anaemia, and ulcerative colitis flare. ALT and AST >3x ULN occurred in 2.9% of patients. All elevations were asymptomatic, <5xULN, transient, and resolving while receiving continued treatment.

CONCLUSION(S): Long-term treatment with 1 mg ozanimod appears to be well tolerated with evidence of rapid onset and durable efficacy.

Reference:Sandborn WJ, Feagan BG, Wolf DC, D'Haens G, et al. Ozanimod Induction and Maintenance Treatment for Ulcerative Colitis. NEJM. 2016;374(18):1754-62.

P-013 Pregnancy Outcomes in Women Exposed to Ustekinumab in the Crohn’s Disease Clinical Development Program

Scherl Ellen 1, Jacobstein Douglas2, Murphy Conor2, Ott Elyssa3, Gasink Christopher4, Baumgart Daniel5, Abraham Bincy6. 1Jill Roberts IBD Center, Weill-Cornell, New York, New York, 2Janssen Research & Development, LLC, Spring House, Pennsylvania, 3Janssen Scientific Affairs, LLC, Spring House, Pennsylvania, 4Janssen Research and Development, LLC, Spring House, Pennsylvania, 5Charite University, Berlin, Germany, 6Houston Methodist, Houston, Texas

BACKGROUND: Ustekinumab (UST) has been approved for moderate to severe Crohn’s Disease (CD) in adult patients (pts). While no adverse developmental outcomes (pre-& postnatal) were observed in animal studies of UST, limited data exist, including previously reported outcomes in psoriasis (PsO) pts, concerning the effects of UST on human pregnancies1. To characterize pregnancy outcomes in women exposed to UST during pregnancy, data from the UST CD clinical development program (CDP) are presented.

METHODS: Pregnancies reported with maternal use of UST (typical terminal half-life of approx 3 weeks) from 5 CD studies were evaluated: 2 Phase 2 (C0379T07: n=131; CERTIFI: n=526) & 3 Phase 3 (UNITI-1: n=769 & UNITI-2: n=640, from which 1,281 continued on maintenance in IM-UNITI).

RESULTS: 877 female pts received ≥1 IV or SC dose of UST, and 26 maternal pregnancies were reported (despite agreeing to adequate birth control). UST treatment was discontinued upon the report of pregnancy in all cases. Mean maternal age was 27.6 years (range 19-43) and mean duration of UST exposure prior to the reported pregnancy was 76±62.1 weeks. Pregnancy outcomes were reported for 24 of 26 pregnancies, including 15 (62.5%) live births (LBs), 4 (16.7%) spontaneous abortions (SAs), and 5 (20.8%) elective abortions. All 4 SAs occurred in the 1st trimester. Mean maternal age was older for pts who had SAs (33.0±2.94 years) vs. LBs (27.6±3.75 years) and median UST treatment duration was longer for pts who had SAs (80 weeks) vs. LBs (56 weeks). Among the LBs, there were no congenital anomalies; 1 infant had a single episode of transient hypoglycaemia treated with oral supplement. No safety signals emerged with neonatal outcomes with gestational age of 38.2±1.3 weeks (n=12), mean 5 min-APGAR of 9.8±0.45 (n=5), and mean birth weight of 6.6±1.6 pounds (n=13).

CONCLUSION(S): While the rate of SA’s was generally comparable to the rate previously reported in PsO data, the small number of pregnancies among women with CD with prenatal exposure to UST precludes definitive interpretation of the data. In this case series, SAs were associated with older maternal age, and longer duration of UST exposure prior to the reported pregnancy was not associated with adverse outcomes. However, the limited available data from the UST CD program requires additional research to determine pregnancy and newborn safety.

1Cather JC, Rahawi KW, Schaufelberger BW, Chan D, Horn EJ, Goyal K. Pregnancy outcomes in women exposed to ustekinumab in the psoriasis clinical development program. Journal of the American Academy of Dermatology. 2014;70(5).

P-014 Ustekinumab IV Induction Results in Crohn’s Disease Symptom Improvement Within the First Week in Anti-TNF Refractory Patients

Sandborn William 1, Yeager Benjamin2, Gasink Christopher3, Hoops Timothy2, Gao Long-Long3, Wang Yanli4, Jacobstein Douglas4, Lee Scott5, Hanauer Stephen6. 1University of California, San Diego, La Jolla, California, 2Janssen Scientific Affairs, LLC, Horsham, Pennsylvania, 3Janssen Research and Development, LLC, Spring House, Pennsylvania, 4Janssen Research & Development, LLC, Spring House, Pennsylvania, 5University of Washington School of Medicine, Seattle, Washington, 6Northwestern University, Chicago, Illinois

BACKGROUND: In both the UNITI-1&2 Crohn’s disease (CD) studies, a single 6mg/kg Ustekinumab (UST) IV infusion showed significantly greater rates of clinical response & remission vs placebo, and significant reductions in CDAI (and >70 pt reduction)1 by the first post-baseline visit at Wk 32. It remains to be determined how soon patients see benefit (ie. before Wk 3).

METHODS: Patient (pt) CDAI diary daily data (day -7 to +14) from the UNITI-1 study of pts who had previously failed TNF antagonists were compiled and analyzed post-hoc for the 3 pt-reported CDAI components (stool frequency[SF], abdominal pain[AP], & general well-being[WB]). Mean change in these daily scores with IV UST 6 mg/kg and 130mg were compared vs. placebo (PBO), as were 2-item SF+AP PRO2 weekly, over the 7d prior, weighted either 1:1 or as a CDAI subscore (assessed by mean change, and as % of pts with >50pt improvement). Ranked transformation was used to compare groups for all analyses.

RESULTS: IV UST induced significant improvement in all 3 components within the first 2 wks, with AP first significantly better than PBO on d2 for both UST doses, and consistently significantly better than PBO from d6 through d14 for 6mg/kg and from d8 through d14 for 130mg. Mean improvement in SF was first significantly better than PBO on d7 for UST both doses, while this occurred on d8 for WB. Week 1 and 2 SF+AP with CDAI weighting was significantly improved for both UST doses vs PBO at d7 and d14, & SF+AP added with equal 1:1 weight was significantly improved for both UST doses at d14. 29.3% of 6mg/kg & 31.4% of 130 mg groups attained >50pt improvement in CDAI solely based on SF&AP components over the second week vs 18.8% in the PBO group (P<0.05 and P<0.01, respectively).

CONCLUSION(S): Even in the refractory CD population of previous anti-TNF failures in UNITI-1, symptom relief based on individual pt-reported CDAI components began to show significant improvement as early as 1 day post UST infusion, and was seen consistently among all 3 components by d8 with both IV UST doses, confirmed by consistent PRO2 benefit in the second week. These findings support previously reported significant early efficacy seen at the post-baseline (Wk3) visit in the UNITI induction studies.2

1Best WR, et al. Gastroenterology 1976;70:439-44.

2Feagan BG, et al. N Engl J Med 2016;375:1946-60.

P-015 Maintenance of Remission with Tofacitinib in Patients With Ulcerative Colitis: Subpopulation Analysis from an Open-Label, Long-Term Extension Study

Colombel Jean-Frederic 1, Reinisch Walter2, Osterman Mark T3, Thorpe Andrew J4, Nduaka Chudy I4, Zhang Haiying4, Lawendy Nervin4, Friedman Gary S4, Su Chinyu5. 1Dr. Henry D. Janowitz Division of Gastroenterology, Icahn School of Medicine at Mount Sinai, New York, New York, 2Medical University of Vienna, Dept Internal Medicine III, Vienna, Austria, 3Penn Presbyterian Medical Center, Division of Gastroenterology, Philadelphia, Pennsylvania, 4Pfizer Inc, Collegeville, Pennsylvania, 5Pfizer Inc, Collegeville, Pennsylvania

BACKGROUND: Tofacitinib is an oral, small molecule Janus kinase inhibitor that is being investigated for ulcerative colitis (UC). The efficacy and safety of tofacitinib was demonstrated in three Phase 3, randomized, placebo-controlled studies (OCTAVE Induction 1, NCT01465763; OCTAVE Induction 2, NCT01458951; OCTAVE Sustain, NCT01458574) in patients with moderate to severe UC.1 An ongoing Phase 3, multicenter, open-label, long-term extension study (OLE; NCT01470612) included subjects rolling over from OCTAVE induction 1 or 2, or OCTAVE Sustain. We present data from the cohort of patients who were previously in remission at the end of OCTAVE Sustain (Week 52).

METHODS: Patients in remission (Mayo score ≤2 with no individual subscore >1, and rectal bleeding subscore of 0) at Week 52 of OCTAVE Sustain were to receive tofacitinib 5 mg twice daily (BID) in the OLE study. Remission and mucosal healing (Mayo endoscopic subscore of 0 or 1) at Months 2 and 12, based on local-read assessment, are presented. Safety data up to 3 years of treatment (as of July 8, 2016) are reported for all patients who received tofacitinib 5 mg BID in the OLE study.

RESULTS: Out of 914 patients enrolled in the OLE study, 144 (mean age: 45 years old; gender: 47.2% female) were in remission, based on central endoscopic reading, at Week 52 of OCTAVE Sustain. Of these, 58 and 69 had received tofacitinib 5 and 10 mg BID, respectively, in OCTAVE Sustain, and 15 (10.4%) discontinued in the OLE study (reasons included: study-drug-related adverse event [AE], n=4; insufficient clinical response, n=6). Remission rates based on observed data were 141/144 (97.9%), 113/135 (83.7%), and 54/64 (84.4%) at baseline, Month 2, and Month 12 of the OLE study, respectively. Remission proportions were similar whether patients had received tofacitinib 5 or 10 mg BID in OCTAVE Sustain (baseline: 96.6% vs 98.6%; Month 2: 88.2% vs 82.1%; Month 12: 85.7% vs 82.8%, respectively). Mucosal healing was achieved by 141/144 (97.9%), 125/138 (90.6%), and 59/65 (90.8%) patients at baseline, Month 2, and Month 12, respectively. Among the 156 patients receiving tofacitinib 5 mg BID in the OLE study, 101 (64.7%) reported treatment-emergent AEs. Serious and severe AEs occurred in 7.1% and 4.5% of patients, respectively. The most frequent treatment-emergent AEs by system organ classes were ‘infections/infestations’ and ‘gastrointestinal disorders’, and by preferred term were ‘nasopharyngitis’ and ‘worsening of UC’. Four (2.6%) patients had serious infections AEs and seven (4.5%) had herpes zoster AEs (all mild to moderate in severity). No malignancies (excluding non-melanoma skin cancer) were reported with tofacitinib 5 mg BID.

CONCLUSION(S): The majority of patients with moderate to severe UC who achieved remission at Week 52 of OCTAVE Sustain maintained remission and mucosal healing with tofacitinib 5 mg BID, over time, up to Month 12 in the OLE study. Compared with OCTAVE Sustain, no new safety concerns associated with long-term exposure to tofacitinib emerged in this subpopulation.

1. Sandborn WJ et al. N Engl J Med 2017;376:1723–36.

P-016 Young Investigator Complications After Colonoscopy in Patients With Inflammatory Bowel Disease

Wang Peiqi 1, Ngamruengphong Saowanee2, Selaru Florin2, Parian Alyssa3, Melia Joanna4, Lazarev Mark4, Hutfless Susan5. 1The Johns Hopkins University, Baltimore, Maryland, 2Johns Hopkins University, Baltimore, Maryland, 3Johns Hopkins University, Baltimore, Maryland, 4Johns Hopkins University School of Medicine, Baltimore, Maryland, 5Division of Gastroenterology, Department of Medicine, Johns Hopkins University Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland

BACKGROUND: Colonoscopy is important for disease activity monitoring and colorectal cancer surveillance in patients with inflammatory bowel disease (IBD). The IBD community has called for greater utilization of colonoscopy to achieve mucosal healing as the ideal clinical outcome. However, colonoscopy-related complication profile has not been comprehensively depicted for the IBD population. Most studies evaluating colonoscopy safety, including the Centers for Medicare and Medicaid Services (CMS)-approved quality measure for colonoscopy, exclude procedures performed on IBD patients.

METHODS: Using 2014 all-payer administrative claims data of six geographically and racially diverse states (California, Florida, New York, Georgia, Nebraska, and Vermont), we identified colonoscopy procedures performed in ambulatory surgery centers for IBD and non-IBD patients. IBD patients were those who had Crohn’s disease (CD) or ulcerative colitis (UC) coded on the colonoscopy visit. We tracked patients’ unplanned visits (including emergency department visits and unplanned hospital admissions according to the CMS definition) within 7 and 30 days after colonoscopy, via linkage of encrypted patient identifier and utilization date. We examined the diagnoses present on these unplanned visits and compared the rates of unplanned visits due to any cause, perforation, sepsis, any infection, and Clostridium difficile infection (CDI) between IBD and non-IBD populations. To establish a temporal relationship for infection-related complications, we excluded procedures where an infection was coded on the procedure visit. We evaluated the risk factors for post-colonoscopic unplanned visits among IBD patients using multivariable logistic regression models.

RESULTS: We identified 36,175 colonoscopy procedures performed on IBD patients (15,273 CD and 20,902 UC) and 1,340,501 procedures on non-IBD patients. Compared with non-IBD patients, the 7-day all-cause unplanned visit rate was 1.6 and 1.3 times higher for CD and UC patients (both P<0.0001; rate per 1,000 procedures: non-IBD 14.6, CD 22.7, UC 19.2). CD patients were more likely than non-IBD to require unplanned care for perforation within 7 days (relative risk(RR)=3.64, P<0.0001), sepsis (RR=2.59, P=0.004), and any infection (RR=2.32, P<0.0001). Compared with non-IBD patients, UC patients had similar rates of perforation and sepsis but higher rates of any infection (RR=1.87, P<0.0001). We identified a 7-fold and 13-fold increase in CDI rate after colonoscopy for CD and UC patients compared with non-IBD. Hospitalization or endoscopic procedure within 30 days prior to the colonoscopy were the strongest risk factors for an unplanned visit within 7 days of the colonoscopy (hospitalization odds ratio(OR)=4.55, 95% confidence interval(CI) 3.38-6.11; endoscopic procedure OR=2.38, 95% CI 1.50-3.79). Younger age (10-39 years) or age above 80 years (reference group: 50-59 years), having an EGD concomitantly with the colonoscopy and having one or more co-morbidities were also associated with an increased odds of all-cause unplanned visits. Therapeutic procedure was not significantly associated with higher unplanned visit rates compared with diagnostic procedure.

CONCLUSION(S): Patients with IBD have significantly higher unplanned visit rates after colonoscopy than the non-IBD population. CDI was 7 and 13 times more likely for CD and UC patients even after excluding those with preexisting infections. Providers should consider IBD patients’ recent history of hospitalization and endoscopic procedures before performing a procedure. Substituting colonoscopy with non-invasive testing modalities could potentially benefit high-risk patients.

P-017 Advanced Age is Associated with Decreased Long Term Functional Outcomes Following Ileal Pouch Anal Anastomosis

McKenna Nicholas, Mathis Kellie, Pemberton John, Lightner Amy. Mayo Clinic Rochester, Rochester, Minnesota

BACKGROUND: Ileal pouch-anal anastomosis (IPAA) is the preferred surgical treatment for patients with ulcerative colitis (UC). There is limited literature regarding long term functional and quality of life outcomes following pouch construction at an advanced age. The purpose of this study was to determine if long-term functional outcomes and quality of life are altered in the elderly population.

METHODS: A retrospective review was performed of all patients undergoing IPAA for UC at our institution between 2002 and 2013. Patients were stratified into two groups based on age at time of IPAA (elderly age>50 years and non-elderly age ≤50 years) and analysis was performed on short-term post-operative complications and long-term functional outcome and quality of life.

RESULTS: 911 patients who underwent IPAA (542 male, 369 female) were included. 178 patients (19.5%) were defined as elderly (>50 years), and 733 (80.5%) as non-elderly (<50 years). Older patients had a higher ASA score (p<0.0001), increased rates of obesity (p<0.0001) and dysplasia or cancer at the time of colectomy (p=0.008), and were less often taking steroids at the time of their operation (p=0.002). There were no differences in 30-day hospital readmission, superficial surgical site infection, partial small bowel obstruction/ileus, complete small bowel obstruction, intensive care unit (ICU) admission, or venous thromboembolism (VTE). Pelvic sepsis was more frequent in the younger patients (p=0.048) and urinary retention was higher in elderly patients (p=0.03). Over a median follow-up of 5.1 years and 5.0 years in elderly and non-elderly patients, respectively, elderly patients reported worse functional outcomes with increased frequency of daytime incontinence (43% versus 27%, p=0.002) and pad usage (47% versus 22%, p<0.001). However, quality of life was similar between the two groups with no difference in family relationships, recreation, and work around the house (all p>0.05) but increased restrictions with respect to social activity (p=0.01) and sexual activity (p=0.01).

CONCLUSION(S): Performing an IPAA on patients over the age of 50 is not associated with increased 30-day postoperative complications, but it is associated with worse long-term outcomes including increased daytime incontinence, pad use, and social life restriction compared to a younger cohort. While safe to perform IPAA in the elderly, it is important to set appropriate expectations regarding long-term outcomes.

P-018 Transabdominal Pouch Salvage for Failed Laparoscopic Ileal Pouch-Anal Anastomosis

Aydinli H. Hande 1, Aytac Erman2, Ashburn Jean3, Kessler Hermann4, Remzi Feza5. 1NYU Langone Medical Center, New York, New York, 2Acibadem University School of Medicine, Istanbul, Turkey, 3Cleveland Clinic, Cleveland, Ohio, 4Cleveland Clinic, Cleveland, Ohio, 5NYU School of Medicine, New York, New York

BACKGROUND: While laparoscopy is accepted as the preferable technique for most of the abdominal operations especially for patients with virgin abdomen, use of laparoscopy for restorative rectal operations has been under debate in terms their general applicability and outcomes. This paper specifically focuses on the characteristics, management strategy and outcomes of transabdominal redo ileal pouch-anal anastomosis (IPAA) surgery in patients who underwent laparoscopic IPAA creation.

METHODS: Between 4/2007 and 7/2016, data regarding patients undergoing transabdominal redo surgery for failed laparoscopic IPAA were reviewed. Patient demographics, primary diagnosis, technical details of the primary and redo IPAA, indications for redo surgery, perioperative and postoperative outcomes were evaluated.

RESULTS: There were 76 [n=26 (34%) males] patients with a median age of 31 years (13-76 years) and a median body mass index of 23 kg/m2 (15-32 kg/m2) at the time of revision surgery. Diagnoses at the time of redo surgery were ulcerative colitis (n= 67, 88%), Crohn’s disease (n=4, 5.2%), familial adenomatous polyposis (n=4, 5.2%), and colonic inertia (n=1, 1.6%). Median time to redo surgery was 2 years (0.2-12) after laparoscopic IPAA creation. 73 (71%) patients required a diverting loop ileostomy prior to or during redo IPAA surgery. A new pouch was created in 57% (n=43) of patients, reuse of the previous pouch was done in 32 patients (42%) and repair of the pouch was done in 1 patient. Indications for redo IPAA surgery were leak and fistula (n=40, 52.6%), obstruction (n=21, 27.6%) and pelvic perianal abscess (n=17, 22.3%). 19 patients (25%) diagnosed with long remnant rectal cuff (>2cm from the dentate line) and 6 patients (7.8%) with mesenteric twist at the time of redo pouch surgery. Mean operative time was 270 minutes (SD: ± 141 minutes) and length of stay was 7.7 days (SD: ±3.9 days) after redo IPAA surgery. 30-day morbidity was 45% (n=34). At a median follow-up of 3 (0.1-8) years after redo surgery, a total of 25 patients (32.8%) were diagnosed with redo pouch failure.

CONCLUSION(S): The outcomes after redo IPAA surgery are promising in patients with failed laparoscopic IPAA. While pelvic sepsis is the common cause of failure, long remnant rectal cuff and mesenteric twist seems technically preventable problems causing failure after laparoscopic creation of the index IPAA.