Abstract:

BACKGROUND: Adhesions are common after conventional surgery; natural orifice transluminal endoscopic surgery (NOTES) avoids peritoneal disruption and may reduce adhesions. OBJECTIVES: To determine whether adhesions (and peritoneal contamination) are less common with NOTES transgastric colon injury and repair (TGCR) than with laparoscopic colon repair (LCR). DESIGN/SETTING: Porcine survival study. INTERVENTIONS: After colon preparation and administration of antibiotics, forty 25-kg male pigs were randomly assigned to either TGCR or LCR. TGCR involved an endoscopic gastrotomy (needle-knife plus balloon dilation), CO(2) pneumoperitoneum, and a 2-cm needle-knife transmural incision of spiral colon. Colotomies were repaired with clips; gastrotomies were closed with clips and a detachable snare. MAIN OUTCOME MEASUREMENTS: Adhesions were assessed at necropsy at 21 days; biopsy specimens were blindly reviewed. A 9-point adhesion score (density/vascularity, width, and extent) was averaged from 3 reviewers. Peritoneal lavage was sent for cell count and culture. RESULTS: Two of 20 TGCR pigs died immediately (unrecognized preoperative autopsy-proven pneumonia). The median procedure times were 70.5 and 19.0 minutes for TGCR and LCR, respectively; weight gains were 7.1 and 8.2 kg, respectively. The median adhesion scores were 4.3 and 3.7, respectively (P = .26); subscores were similar (1.9, 1.5, 1.3 vs 1.7, 1.1, 1.0, respectively (P = .3-.6)). Peritoneal lavage bacterial growth was nonsignificantly lower after TGCR than after LCR (38.9% vs 60.0%, respectively; P = .30); administration of intragastric antibiotics did not decrease contamination. Three TGCR (vs no LCR) pigs had histologic peritonitis. LIMITATIONS: Animal model, colon prepped, injury immediately recognized. CONCLUSION: NOTES colon repair is feasible after transmural injury. Adhesions, histologic peritonitis, and contamination were similar to those with laparoscopy and were not helped by intragastric antibiotics.