Background: Transgastric peritoneoscopy (TGP) may be a future alternative to the diagnostic laparoscopy (LAP).
Objective: To create a model of peritoneal metastases for development of TGP and to employ this model to compare TGP to LAP.
Methods: Small beads were stapled in porcine peritoneal cavities to simulate metastases. Using a noninferiority design a sample size of 64 beads was determined, which were divided over 12 animals. Randomization was performed for number and location of beads. LAP was performed by one of 2 blinded surgeons. TGP was then performed in the same pig using either standard endoscopic accessories (TGP-s) or a specially designed toolkit (TGP-t) in randomized order by 1 of 2 blinded endoscopists. Primary outcome was number of beads found and touched during peritoneoscopy.
Results: Locations of beads included: abdominal peritoneum (14 beads), diaphragm (11), surface of liver and hepatoduodenal ligament (32), and miscellaneous sites (7). LAP detected 61 beads (yield = 95%), TGP-s 40 beads (63%), and TGP-t 40 beads (63%). TGP-s and TGP-t were both inferior in comparison with LAP (P = 0.8465 and P = 0.7440 respectively). TGP-s and TGP-t were similar in number, distribution and time to detect beads. TGP was superior for detecting beads on the abdominal and diaphragmatic peritoneum than for the liver, namely TGP-s: 23/25 (92%) versus 12/32 (38%) (P < 0.001); TGP-t: 25/25 (100%) versus 11/32 (34%) (P < 0.001).
Conclusion: In this first prospective, blinded, comparative trial TGP was inferior to LAP for the detection of simulated metastases. Future development for NOTES peritoneoscopy should focus on improved access to the region of the liver and enhanced endoscopic optics and performance.