Robot-assisted endoscopic submucosal dissection versus conventional ESD for colorectal lesions: outcomes of a randomized pilot study in endoscopists without prior ESD experience (with video)

Background and Aims Endoscopic submucosal dissection (ESD) is becoming the preferred method for certain early GI malignancies; however, very few U.S. physicians have adopted this technique. This is in part because of the technically challenging nature of the procedure and the long learning curve. Several endoscopic robots are under development to address these complexities. Methods This is a randomized, controlled, pilot study comparing conventional ESD versus robotic-assisted ESD (RESD) in an ex vivo bovine colon model. Five endoscopists without prior ESD or RESD experience were randomized into 2 groups (group 1, RESD after ESD; group 2, RESD before ESD). A standard template was used to create colonic lesions. The primary outcome was completeness of en bloc resection. Secondary outcomes included differences in procedure time, perforation rate, muscle injury rate, and National Aeronautical and Space Administration Task Load Index (NASA-TLX) to assess physical and mental workload. Results Five endoscopists each performed 4 tissue resections (2 RESD and 2 ESD), for a total of 20 procedures. Complete en bloc resection was achieved in all RESD and in 50% of ESD (P < .0001). The perforation rate was higher in the ESD group (60% vs 30%, P = .18). Total procedure time (34.1 vs 88.6 min, P = .001) and dissection time (27.8 vs 79.4 minutes, P = .002) were lower for RESD. The NASA-TLX also revealed better results for RESD (28.4 vs 47.4, P = .01). Conclusions RESD appears to be more effective in obtaining en bloc resection with shorter procedure times and a lower perforation rate compared with conventional ESD as performed by ESD novices. RESD is also associated with lower physical and mental workloads. Endoscopic submucosal dissection (ESD) is becoming the preferred method for certain early GI malignancies; however, very few U.S. physicians have adopted this technique. This is in part because of the technically challenging nature of the procedure and the long learning curve. Several endoscopic robots are under development to address these complexities. This is a randomized, controlled, pilot study comparing conventional ESD versus robotic-assisted ESD (RESD) in an ex vivo bovine colon model. Five endoscopists without prior ESD or RESD experience were randomized into 2 groups (group 1, RESD after ESD; group 2, RESD before ESD). A standard template was used to create colonic lesions. The primary outcome was completeness of en bloc resection. Secondary outcomes included differences in procedure time, perforation rate, muscle injury rate, and National Aeronautical and Space Administration Task Load Index (NASA-TLX) to assess physical and mental workload. Five endoscopists each performed 4 tissue resections (2 RESD and 2 ESD), for a total of 20 procedures. Complete en bloc resection was achieved in all RESD and in 50% of ESD (P < .0001). The perforation rate was higher in the ESD group (60% vs 30%, P = .18). Total procedure time (34.1 vs 88.6 min, P = .001) and dissection time (27.8 vs 79.4 minutes, P = .002) were lower for RESD. The NASA-TLX also revealed better results for RESD (28.4 vs 47.4, P = .01). RESD appears to be more effective in obtaining en bloc resection with shorter procedure times and a lower perforation rate compared with conventional ESD as performed by ESD novices. RESD is also associated with lower physical and mental workloads.