Learning curve for endoscopic submucosal dissection of early colorectal neoplasms with a monopolar scissor-type knife: use of the cumulative sum method

Scand J Gastroenterol. 2020 Oct;55(10):1234-1242. doi: 10.1080/00365521.2020.1807597. Epub 2020 Aug 27.

Abstract

Objective: Although scissor-type knives such as the Stag-Beetle (SB) Knife Jr are expected to result in a safe and easy colorectal endoscopic submucosal dissection (CR-ESD), information regarding the learning curve is lacking. Therefore, this study evaluated the learning curve with using SB Knife Jr.

Materials and methods: We retrospectively reviewed 507 CR-ESD procedures performed in 464 patients using SB Knife Jr. The primary endpoint was a learning curve to achieve a satisfactory complete resection rate. The secondary endpoints were learning curves to achieve a satisfactory en bloc resection rate, curative resection rate, and resection speed.

Results: The complete, en bloc, and curative resection rates were 91.9%, 95.9%, and 84.0%, respectively. Moving average analysis showed that 39 cases were required for a complete resection rate of >80%, 41 for an en bloc resection rate of >90%, and 50 for a curative resection rate of >75%. We divided the procedure into three phases using the cumulative sum method: I, II, and III (cases 1-36, 37-119, and 120-507, respectively). Although we found no significant between-phase differences, the complete resection rate showed an increasing trend in Phase III (83.3 vs. 89.2 vs. 93.3%; p = .099). The en bloc resection rate (91.7 vs. 91.6 vs. 97.2%; p = .047) and resection speed (20.5 vs. 7.2 vs. 6.8 min/cm2; p < .001) were greater in Phase III. Despite the larger specimen size (27.3 vs. 38.2 vs. 40.4 mm; p < .001) and more severe fibrosis (p < .001) in Phase III, the procedure time was shorter (73.8 vs. 57.8 vs. 54.2 min; p = .041). The curative resection rate was not significantly different between phases.

Conclusions: SB Knife Jr enables safe and easy CR-ESD during the introductory period compared to the conventional tip-type knife and has an acceptable learning curve. Therefore, using this knife will encourage the widespread adoption of CR-ESD in Asian general hospitals and non-Asian countries.

Keywords: CUSUM; ESD; Learning curve; SB Knife Jr; colorectal ESD; colorectal neoplasm; endoscopic submucosal dissection; moving average analysis; scissor-type knife.

MeSH terms

  • Colorectal Neoplasms* / surgery
  • Endoscopic Mucosal Resection*
  • Humans
  • Learning Curve
  • Retrospective Studies
  • Treatment Outcome