all 5 criteria were 100%. The sensitivity of Level 1 and Level 2 parenchymal finding was 100%, 93.3%. The sensitivity of Level 1 and Level 2 ductal finding was 93.5%, 95%. The accuracy, sensitivity, and specificity of Level 1 serology were 70.4%, 63%, and 94.1%, respectively. The accuracy, sensitivity, and specificity of Level 2 serology were 22.5%, 27.8%, and 5.9%, respectively. Level 1 other organ involvement (OOI) for Type 1 AIP were 12 patients (21.4%), Level 2 OOI for Type 1 AIP were 13 patients (23.2%), and Level 2 OOI for Type 2 AIP was 1 patient (50%). Level 1 histology for Type 1 AIP was 11 patients, and Level 1 histology for Type 2 AIP was 1 patient. All of patients received steroid therapy (100%) showed resolution or improvement of the pancreatic lesion or OOI clinically and morphologically. Conclusions ICDC is the most sensitive in 5 major criteria, and useful to diagnose and classify AIP with Type 1 and Type 2. However, further studies are necessary to investigate whether it is ideal or not with validation the Level 1 and Level 2. 265b Ammonia(NH3) -Positron Emission Tomography- Computed Tomography (PETCT) in Acute Pancreatitis Abdul Khaliq, Raghav Kashyap, Manish Manrai, Rakesh Kochhar, Anish Bhattacharya, Bhagwant Rai Mittal, Kartar Singh Introduction & aims: Contrast Enhanced Computed Tomography (CECT) is done in acute pancreatitis. Computed tomography severity index (CTSI) correlates with outcome. Renal failure can complicate pancreatitis where CECT or Contrast enhanced magnetic resonance (CEMR) cannot be done. N13-ammonia is well taken in pancreas owing to high perfusion per gram tissue. This inspired us to apply N13-NH3 PET for evaluation of pancreatic perfusion in acute pancreatitis and to compare with CECT and clinical outcome. Methods : Prospective observational study in acute pancreatitis patients. NH3-PET CT followed by CECT was done according to protocol and images analysed by two independent observers. Standard uptake value( SUV) of pancreas( P) and liver(L) were taken and SUV P/L achieved. Clinical parameters, follow-up at 28 days, surgery requirementwere noted. Results : Of 29 patients, CECT was done in 23 patients( 6 patients - deranged RFT precluded CECT) and NH3-PET CT in all cases as well as in 9 controls. Marshall class has good correlation with 28 day outcome( p=0.015). Mean SUV P/L in controls, uninvolved pancreas and necrotic areas were 0.97± 0.14, 0.69±0.27 and 0.15 respectively. CTSI, modified CTSI (MCTSI), PETCTSI and PET-MCTSI all have good correlation amongst themselves ( >0.95) and with 28 day outcome. Additionally, PET CTSI and PET MCTSI correlated with surgical requirement. Inter-observer acceptance of necrosis and grading on NH3 PET CT is good ( κ- 0.93). Conclusion: NH3-PETCT is an alternative cross-sectional imaging for determining necrosis in acute pancreatitis and can be performed even in renal failure. It has good correlation with CECT and with 28-day outcome, requirement of surgery with a good interobserver acceptability. NORMAL PANCREATIC PERFUSION AS SEEN ON PET SCAN CECT & PETCT SHOWING > 50% NECROSIS 266 The Safety Profile of Anti-TNF Therapy in an Older Population With Inflammatory Bowel Disease - a North American Experience Amit P. Desai, Zachary A. Zator, Punyanganie S. de Silva, Deanna D. Nguyen, Vijay Yajnik, Joshua R. Korzenik, Ashwin N. Ananthakrishnan Background: Treatment of inflammatory bowel diseases (IBD) with anti-TNF agents is well established. In increasingly aging populations, awareness of outcomes of older patients treated with biologics is becoming more important. However, such patients are often excluded from clinical trials and few studies to-date have investigated the safety and durability of anti-TNF therapy in this sub-group. The aim of our study was to investigate the safety and outcomes of patients aged 60 years and above initiated on treatment with anti-TNF agents for management of their IBD. Methods: This was a retrospective single-center study of all IBD patients who commenced anti-TNF treatment at age > 60 years. Cases of Crohns disease (CD) and ulcerative colitis (UC) were identified from medical record review. Disease activity, concomitant medications, and Montreal phenotypes were noted at commencement of anti- TNF therapy. Kaplan-Meier survival estimates were used to calculate the probability of remaining on anti-TNF therapy and to identify risk of complications. Results: We identified a total of 52 IBD patients who initiated anti-TNF therapy over the age of 60 years (mean 68, range 60-94 years). 67% of patients had CD. Infliximab was the anti-TNF agent used in 87% of patients. Mean duration of disease prior to anti-TNF therapy was 18.6 months S-63 AGA Abstracts with mean age at diagnosis of 53 years. A total of 35 patients (68%) discontinued anti-TNF therapy after a mean duration of 28.9 months. The likelihood of remaining on therapy at 12, 24, and 36 months was 0.89, 0.71, and 0.44 respectively. 29 patients (56%) had a Charlson co-morbidity score of 1 or greater but increasing co-morbidity did not influence likelihood of discontinuation (HR 1.01, 95%CI 0.53 - 1.97) (Figure 1). The most common reasons for cessation of treatment were non-response (19%), infection (12%), and drug intolerance (12%). Three patients (6%) required hospitalizations for infectious complications. There was 1 death secondary to Listeria bacteremia in a 94 year old with UC who received infliximab while hospitalized for a severe flare. We then compared the durability of anti- TNF therapy in the older cohort to a second cohort of patients aged 19-45 years at time of initiation of infliximab therapy (mean 31 years). Age > 60 was associated with a three-fold increase in likelihood of cessation of therapy (HR 2.97, 95%CI 1.73-5.11, p<0.001) (Figure 2). Each 10 year increase in age at initiation of anti-TNF was associated with a 20% increase in likelihood of therapy discontinuation (HR 1.20, 95%CI 1.05 - 1.37). Conclusion: The IBD population older than age 60 at the time of initiation of anti-TNF therapy is at higher risk for discontinuation of therapy. They may also be particularly vulnerable to infectious complications requiring hospitalization suggesting need for careful monitoring during therapy. Figure 1: Probability of remaining on anti-TNF therapy among patients 60 or older, by Charlson co-morbidity score Figure 2: Probability of remaining on anti-TNF therapy, by age at initiation 267 Long Term Outcome of a Third Anti-TNF Monoclonal Antibody in IBD After the Failure of Two Other Anti-TNF Agents- Data From a North American Study Punyanganie S. de Silva, Deanna D. Nguyen, Vijay Yajnik, Joshua R. Korzenik, Ashwin N. Ananthakrishnan Background: Treatment of inflammatory bowel disease (IBD) with anti-TNF agents is well established. However, a significant proportion of patients lose response to these drugs. There is limited knowledge of long term outcomes of those who have failed two anti-TNF agents and commenced a third. To-date no North American studies have assessed long term outcomes of a third anti-TNF agent following previous failure or intolerance with two other drugs of the same class. Our aim was to assess long-term efficacy of treatment for inflammatory bowel disease with a third anti-TNF agent after failure or/and intolerance of two other anti- TNFs. Methods: We performed a retrospective study of all IBD patients treated with a third anti-TNF agent after loss of response or intolerance to two prior anti-TNF agents at a single tertiary North American center. Cases of Crohn's disease (CD) and ulcerative colitis (UC) were identified from electronic medical record review. Disease activity, drug therapy and Montreal phenotypes were noted at disease onset and commencement of 3rd anti-TNF agent. Kaplan-Meier estimates were used to calculate the probability of remaining on a 3rd anti- TNF agent and to identify predictors of longer term clinical response. Results: 63 patients (64% women, 57 CD and 6 UC) were included in the analysis. Mean age at initiation of third anti-TNF was 38 years. Mean disease duration was 12 years. Thirty-five (56%) patients discontinued the 3rd anti-TNF after a mean of 13.2 months. Certolizumab pegol was used as third line for 56 (89%) patients. Probability of remaining on the 3rd anti-TNF was 0.69, 0.55, 0.37, and 0.25 at 6, 12, 24, and 36 months respectively. Lack of response (primary or secondary) was the reason for discontinuation of anti-TNFs in 71% on 1st, 40% on 2nd and 70% on 3rd agent. Intolerance caused discontinuation in 10%, 27% and 7% of patients on 1st, 2nd and 3rd anti-TNF agents respectively. Prior primary non-responders to the 1st AGA Abstracts