Hydrogen leak test is minimally invasive and highly specific for assessment of the integrity of the luminal closure after natural orifice transluminal endoscopic surgery procedures (with video).

Background: Leak-resistant closure of transluminal access is a major challenge facing natural orifice transluminal endoscopic surgery (NOTES).

Objective: To evaluate a hydrogen (H(2))-based leak test for assessment of transluminal-access closure integrity after NOTES procedures.

Setting: Nine acute porcine experiments.

Design And Intervention: After gastric-wall puncture and balloon dilation, peritoneoscopy was performed, followed by transmural closure of the gastric opening. The animals were randomly assigned to complete or incomplete closure groups. The H(2) leak test was performed by using 1000 mL of 4% H(2) gas mixture and the Hydrogen Leak Detector H2000+. The animals were then euthanized for a methylene blue (MB) test of gastric closure integrity.

Main Outcome Measurement: Intraperitoneal H(2) concentration after gastric insufflation with H(2).

Results: The H(2) leak test was quick and easy. Intraperitoneal H(2) concentrations in parts per million in both groups were similar at baseline (mean +/- SD, 0.18 +/- 0.29 parts per million [ppm] vs 0.22 +/- 0.35 ppm, P = .97) and after balloon dilation (414.8 +/- 198.5 ppm vs 601.3 +/- 116.1 ppm, P > .99). Postclosure intraperitoneal H(2) concentrations dropped to 0.01 +/- 0.77 ppm in the complete-closure group, similar (P = .81) to matched-pairs preopening levels and significantly lower than in the incomplete-closure group (162.0 +/- 83.0 ppm, P < .02). On necropsy, the MB test was negative in all 5 animals of the complete-closure group and positive in all 4 animals of the incomplete-closure group. A cutoff of 25 ppm in intraperitoneal H(2) concentration after closure gave 100% sensitivity, specificity, and positive and negative predictive values for MB leakage.

Limitations: Nonsurvival animal experiments.

Conclusions: The H(2) leak test is highly accurate for detection of leakage after NOTES procedures and could become a substitute for currently used MB leak tests.