surgery (2 perforations, 2 vascular compromise), 35 presented with acute incarceration and were treated with endoscopic or radiographic guided decompression prior to surgery. Patients presenting acutely were older (73±15y vs. 69±11y, p=0.048) and had a decreased BMI (25±11kg/m 2 vs. 29±7kg/m 2 , p=0.001). Heartburn, Regurgitation, early satiety and dysphagia were more common in ER (p≤0.005 each). Chest pain was more frequent with AR (69% vs. 44%, p=0.004). There was no difference in preoperative Charlson Comobidity Index (CCI), age adjusted CCI, and ASA Score. PEH type III and type IV were distributed similarily between groups (type III: 77% vs. 81%, p=0.726; type IV: 23% vs. 13%, p=0.140, for AR and ER, respectively). Intrathoracic stomach >75% and mesoaxial rotation were more common in AR (69% vs. 48%, p=0.02; 13% vs. 3%, p=0.003). Operative time was increased with similar blood loss (173±72min vs. 152±39min, p=0.002). Hill repair and Gastrostomy were performed more often in AR (p=0.015 and p=0.001). There was no difference in postoperative complications, Clavien-Dindo Severity Scores, or 30day-readmission rate. AR was associated with an increased length of hospital stay (6±3d vs. 4±2d, p<0.001). One patient in the ER group died, overall mortality 0.2%. Postoperative regurgitation and heartburn were less common after AR vs. ER (0% vs. 11% and 3% vs. 16%, p=0.04 each), while chest pain and dysphagia were similar (13% vs. 10% and 9% vs. 17%, p>0.3 each). Barium swallow studies were available 5-7 months after surgery in a majority of patients. Reflux and recurrent hernia were similar for AR and ER (12% vs. 22% and 16% vs. 15%, p>0.05 each). Conclusion: Acute presentations associated with PEH can be managed successfully with comparable outcomes to elective operations in high volume centers. Many patients can and should be treated with guided decompression for acute incarceration and obstruction and subsequently undergo semielective repair. Patients with large PEH, recurrent chest pain, and especially those with mesoaxial rotation should routinely be considered for elective repair. 1012 ENDOSCOPIC SUBMUCOSAL DISSECTION FOR ESOPHAGEAL ADENOCARCINOMA AND A NORTH AMERICAN PERSPECTIVE Philippe Bouchard, Jonathan Cools-Lartigue, Jonathan Spicer, Carmen L. Mueller, Lorenzo E. Ferri Introduction: Given the low rate of lymph node metastasis, organ-sparing endoscopic resec- tion of early esophageal cancer is justified. Endoscopic Submucosal Dissection (ESD) has been shown in Asian studies with esophageal squamous cell carcinoma to be superior to Endoscopic Mucosal Resection (EMR) in terms of the ability to resect larger lesions with negative margins. However, there is very limited data on the application of ESD in Western countries and for esophageal adenocarcinoma. We sought to review our experience employing ESD in patients with early esophageal cancer from a North American surgical perspective. Methods: A prospectively maintained database (2005-16) on all patients with esophageal cancer at a North American surgical referral center was queried for those undergoing ESD for adenocarcinoma or HGD. Patient demographics, pre-resection tumour characteristics, endoscopic resection technical variables, pathologic results, and short and long term outcomes were recorded. Data presented as Median (IQR). Results: Of 650 patients in the database 26 underwent 27 ESDs for adenocarcinoma /HGD between 2012-2016. The median age was 66 (16.5), most were male (78%:21/27). The majority (67%) had pre-treatment EUS and clinical staging was nodular HGD (11%)/cT1a (44%)/cT1b (44%). Procedure time was 86 (50) minutes and 25/27 (93%) were performed en-bloc. Although bleeding was common during the procedure, there were no post-ESD bleeding events requiring re-intervention. Perforation occurred in 2/27 (7%) one of which required operative repair. LOS was 1 (1) day. Final pathology revealed nodular HGD (4/27), pT1a(7/27), pT1b(15/27), pT2(1/27). Complete RO resection was achieved in 18/27(67%). Of the 9 R1 cases, all were deep margin positive and occurred solely in pT1b/pT2 lesions. Salvage laparoscopic esophagectomy was performed in 6 patients for positive deep margins and 1 for multifocal invasive cancer. At follow up of 13.5 (2-30) months recurrent cancer occurred in only 1 patient (1/21 of unresected pts), who was managed successfully with a repeat ESD. Conclusions: In this largest single institution North American series to date we have found that although technically challenging, ESD represents a safe and effective treatment of early esophageal adenocarci- noma, even in many T1b lesions. ESD should be more widely adopted as an important tool in the therapeutic armamentarium for the treatment of early esophageal adenocarcinoma in Western countries. 1013 INCIDENCE, SURVIVAL, AND PREDICTORS OF LYMPH NODE METASTASIS IN EARLY STAGE GASTRIC SIGNET RING CELL CARCINOMAS IN THE UNITED STATES Sridevi K. Pokala, Zhengjia Chen, Parit Mekaroonkamol, Anthony Gamboa, Steven Keilin, Qiang Cai, Field F. Willingham Introduction: The use of endoscopic resection in the management of early stage gastric signet ring cell carcinoma has been proposed in Eastern countries. The role for endoscopic resection for gastric signet ring cell carcinoma in the US has not been defined. Methods: Cases of gastric signet ring cell carcinoma were extracted from the national SEER database for the years 2004-2013. Cases with unknown tumor characteristics, unknown patient characteristics, metastatic disease, neoadjuvant radiation, and lack of surgical resection or microscopic evaluation of lymph nodes were excluded. Early stage disease was defined as Tis, T1a, T1b, and T1 NOS. Relative survival was calculated using the Kaplan-Meier method. Univariate and multivariate analysis were performed to identify predictors of lymph node involvement. Results: 10,624 cases of gastric signet cell carcinoma were abstracted from the SEER database from 2004-2013. After exclusions, 1,332 cases were included for incidence analysis. The incidence of gastric signet ring cell carcinoma from 2004-2013 was .198 per 100,000 person- years for all stages and .094 per 100,000 person-years for early stages. 1,099 cases were included in the survival analysis. The 5-year survival was 88.6% for early stage tumors and 27.6% for advanced stage tumors. 511 cases were analyzed for predictors of nodal metastasis. On multivariate analysis, T1b tumors had a higher rate of lymph node spread than T1a (OR 3.7, 95% CI 2.1-6.5 p < .001) and tumors ≥4 cm had a higher rate than tumors <1 cm (OR 8.1 95% CI 2.7-24.3, p<.001). Grade was not a significant predictor of nodal metastasis (p=.372). Of T1a tumors, 3.7% at 0-1cm, 6.8% at 1-2cm, 9.3% at 2-3cm, 15.0% S-1231 SSAT Abstracts at 3-4cm and 28% at ≥ 4cm had nodal metastasis (p<.05). Of T1b tumors, 11.1% at 0- 1cm, 28.7% at 1-2cm, 28.6% at 2-3cm, 39.0% at 3-4cm and 45.7% at ≥ 4cm had nodal metastasis (p=.055). Conclusion: Signet ring cell carcinoma is an aggressive cancer, however, the 5-year survival for early stage tumors in the US population was greater than 85% and T1a tumors < 3 cm in size had less than a 10% rate of lymph node metastasis. For highly selected patients, it may be possible to consider organ-sparing endoscopic resection approaches in US cohorts. Rate of nodal metastasis in gastric signet ring cell carcinoma by stage and tumor size Multivariate association between variables and nodal metastasis in gastric signet ring cell carci- noma 1014 INCREASED METABOLIC BENEFIT FOR ELDERLY PATIENTS UNDERGOING ROUX-EN-Y GASTRIC BYPASS VS SLEEVE GASTRECTOMY FOR MORBID OBESITY Katherine D. Gray, Maureen D. Moore, Gregory Dakin, Rasa Zarnegar, Alfons Pomp, Cheguevera Afaneh Background Elderly patients with morbid obesity and metabolic syndrome are at high-risk for end-organ complications. It is not well studied whether bariatric surgery can improve the metabolic profile of these patients. We sought to assess outcomes at intermediate follow- up in a cohort of elderly patients undergoing laparoscopic sleeve gastrectomy (LSG) or laparoscopic Roux-en-Y gastric bypass (LRYGB). Methods A retrospective review was con- ducted of all patients ≥ age 60 undergoing LSG or LRYGB at our institution between 2007- 2014. Patients with prior bariatric operations were excluded. Demographics, body mass index (BMI), American Society of Anesthesiology (ASA) score, metabolic comorbidities, and operative details were collected. Post-operative complications were scored as major if ≥ Clavien-Dindo class 3. Comorbidity resolution was measured by number of oral medications for diabetes mellitus (DM), hypertension (HTN), and hyperlipidemia (HL). Percent total weight loss (%TWL) was also measured. Results A total of 134 patients who underwent LSG (n=65) or LRYGB (n=69) were identified. There were no statistical differences between groups with respect to age (mean 64, range 60-75), BMI (mean 44.0 +/- 6.1), or ASA score SSAT Abstracts