Transgastric Peritoneoscopy Versus Laparoscopy for the Detection and Biopsy of Peritoneal Metastases

Abstract

875 Transgastric Peritoneoscopy Versus Laparoscopy for the Detection and Biopsy of Peritoneal Metastases Douglas O. Faigel, Rogier P. Voermans, Mark I. Van Berge Henegouwen, Sean Conlon, James T. Spivey, Renee L. Rowe, Brett C. Sheppard, Paul Fockens The presence of peritoneal metastases is a frequent cause of unresectability for GI malignancies such as pancreatic cancer. Preoperative detection is difficult and laparoscopy (LAP) is frequently required to exclude metastatic disease prior to resection. Transgastric peritoneoscopy (TGP) may be an alternative to LAP, which could be performed at the time of a staging EUS. Aims: Create a model of peritoneal metastases for use in the development of TGP. Employ this model to compare TGP to LAP.Methods: This investigator-initiated protocol was performed in the live anesthetized acute porcine model. 2.5 mm color-coded beads were stapled via LAP to the peritoneum to simulate metastases. Three to 7 beads were placed in each of 12 animals (total 64 beads). Locations and number of beads were randomized and assigned using opaque envelopes. A pre-assessed list of essential locations was used: abdominal peritoneum (14 beads), diaphragm (11), surface of liver (32), and miscellaneous sites (7): duodenal curve, visceral peritoneum, omentum, anterior stomach and pelvis. 3-port LAP was performed by one of 2 surgeons blinded as to the location and number of beads. TGP was then performed with a 2 channel therapeutic upper endoscope using either standard accessories (forceps, cap)(TGP-S) or with a specially designed toolkit (bendable overtube, articulating retractors and graspers)(TGP-T) in randomized order by one of 2 endoscopists also blinded to bead placement. Primary endpoint was visualization and touching the bead with a biopsy forceps within a 30 min time limit. A noninferiority design comparing LAP to TGP was used to calculate the sample size. Fisher’s exact test with Bonferroni’s correction for multiple testing was used to calculate p-values. Results: A total of 64 beads were placed into 12 pigs. LAP found 61 beads (yield Z 95%, 95% CI: 89-100%), TGP-S 39 beads (61%, CI: 49-73%, p ! 0.0002 vs. LAP), TGP-T 40 beads (63%, CI: 51-74%, p ! 0.0002 vs. LAP). TGP-S and TGP-Twere similar in the number, distribution and time to detect beads. Both TGPS and TGP-T were each superior for detecting beads on the abdominal and diaphragmatic peritoneum (n Z 25) than for the other sites (n Z 49): TGP-S 92% vs. 35% (p ! 0.0002), TGP-T 100% vs. 31% (p ! 0.0002). Conclusions: In this first prospective, blinded, comparative trial TGP was inferior to LAP for the detection of simulated metastases. We successfully created a model for peritoneal metastases and established the benchmark for LAP detection. This model will be useful for future device development, which should focus on improved access to the region of the liver and enhanced endoscope optics and performance.