Article Text

Download PDFPDF
PTU-016 The Endoscopic Submucosal Dissection Learning Curve: The Experience of a Large Volume Italian Colorectal (CRC) Screening Centre
  1. E Rosa-Rizzotto1,
  2. E Guido1,
  3. D Caroli1,
  4. A Dupuis1,
  5. M Lo Mele2,
  6. M Rugge2,
  7. P Pilati3,
  8. F De Lazzari1
  1. 1Specialized Medicine, St Anthony Hospital
  2. 2Pathology Unit
  3. 3Surgery Unit, Padova, Italy


Introduction Endoscopic submucosal dissection (ESD) is an advanced endoscopic technique. In Eastern countries the learning curve is begun with gastric GI lesions carried out under expert supervision and then goes on to address esophageal and colon lesions. As Early Gastric Cancer (EGC) is a rare disease in Western countries, expert guidance is not commonly available.

Methods All the ESD performed in our Endoscopy Unit in Padua from February 2012 to December 2015 including 12,552 colonoscopies were recruited retrospectively in this study. We considered the learning curve of a single endoscopist who performed 10 ESD on in vivo animal models under expert supervision before starting on human subjects. All the dissections were performed using a Hybridknife needle and ERBEJET2 (ERBE®). ESD was performed if the neoplastic lesion was considered susceptible to ESD regardless to the size. T tests for unpaired data and Pearson’s chi-test were used for statistical analysis.

Results 49 ESD were performed, 28 M(57%), mean age 63 yr. The breadown was: 29 rectum (59%), 12 sigmoid tract (24%), 2 trasverse colon (4%), 4 ascending colon (8%), 2 stomach (4%). The neoplastic lesions were: 36 laterally spreading tumours (73%), 5 polypoid lesions 0 Is (10%), 4 recurrent ton scars (8%), 4 polypoid lesions 0 Isp(10%). Mean polyp area was 17.6 cm2 (range 1–70). Mean intervention time was 98 min (range 20–240). En-bloc dissection was successful in 34/49 (69%) and R0 was reached in 24/33 (72%). The histological features of the polyps were: 10 LGD (20%), 27 HGD (55%), 9 pT1 (18%), 3 pT2 (6%). The procedural complications that took place (14/49 = 28%) included: perforation during the procedure in 10/49 (20%), delayed bleeding in 3/49 (6%), rectal stenosis in 3/49 (6%). No deaths or surgical interventions followed the periprocedural complications. From the 12th procedure onwards the surgical performance became acceptable 22/27 (81%) vs 3/12 (25%) (p < 0.001). From the 30th procedure onwards the surgical performance became good 17/19 (90%, p < 0.05) and the mean execution time was significantly lower 55 vs 122 min (p < 0.0001) with no significant difference in the mean area of the lesions 15.6 vs 18.2 cm2 (p=ns). Only 3 complications occurred after the 30th procedure (p=ns).

Conclusion Our findings demonstrate than an endoscopist can reach a satisfactory level of competence in ESD procedures by beginning training with in vivo animal models (at least 10 procedures) and then should go on to colo-rectal neoplasms (without size limits and no less than 12 procedures). Trainees have probably still not reached a learning curve plateau even after 40 procedures.

Disclosure of Interest None Declared

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.