PTH-018 Learning curve and outcomes of endoscopic submucosal dissection in a uk centre

IntroductionIt can be difficult for Endoscopists in the UK to gain experience in endoscopic submucosal dissection (ESD) partly due to the lack of suitable easier gastric cases, time constraints, and fear of complications. Studies have suggested that 20 colorectal cases in a Western setting may be sufficient to achieve satisfactory outcomes 1, but the applicability of this to a training model in the UK is unknown. The aim of this study is to report on initial experience and a training model of ESD in a single UK centre.MethodA total of 34 cases were referred for ESD between May 2013 and January 2017 to the University Hospital Southampton. The authors underwent a structured training schedule; training and practice on ex-vivo pig models followed by training on live pig models with expert tuition (at ASGE institute, Chicago), progressing to human cases in the colon and stomach, utilising knife assisted resection techniques (KAR) when required. The authors employed a local buddy system during cases to share learning, and one author (PB) attended a further upskilling course on live pig models (at ESGE Workshop, Salzburg) after 20 cases. Outcomes were prospectively recorded and the first 20 cases were compared with the subsequent 14 cases.ResultsIndications were suspected early rectal (14/36), sigmoid (1/36) or gastric cancer (4/36), gastric NETs (2/36), scarred colonic adenomas (7 rectal, 1 transverse colon), or adenomas abutting dentate line (5/36). Mean diameter of lesion was 3.5 cm (1–7 cm). Median total procedure time fell between study periods (110 v 90mins, p=NS). R0 resection rates significantly improved after 20 cases from 45% to 85.7% (p=0.016) and use of KAR techniques significantly reduced (65% v 21.4%, p=0.012). R0 resection rates were significantly reduced when KAR was used (43% v 78%, p=0.042). When procedures using KAR were excluded, time/cm2 fell from 43.8 to 30 mins/cm2, p=NS). Perforation was seen in one patient in the first 20 cases (sealed with clips, surgery not required). No significant post procedural bleeding was seen. 30/34 patients have completed 3 month follow up; recurrence was 7.1% after R1% and 0% after R0 en-bloc resection. 1 patient underwent surgery for sm3 disease.ConclusionLearning ESD can be achieved safely in the UK using primarily rectal cases. Use of KAR techniques can help operators gain experience of knives but are associated with increased risk of R1 resection. Using our model of training after 20 cases, outcomes includi