A Comparison of Three Endoscopic Suturing Devices for Natural Orifice Translumenal Endoscopic Surgery Gastrotomy Closure

Abstract

Background: With the heightened concern for post-surgical leak, the standard practice of repairing hollow viscous injuries with suture should be mimicked in natural orifice translumenal endoscopic surgery (NOTES) closure. This study aimed to compare 3 endoscopic suturing devices in a survival porcine model. Methods: Three groups of 10 pigs each were assigned a NOTES gastrotomy closure by either a tissue plicating device (TPD), tissue anchoring (t-tag) device (TAD), or flexible suturing device (FSD). The TPD requires an individual intubation for each plication applied (2-4 per closure). The TAD and FSD require pre-gastrotomy suture placement to facilitate their tissue closure. A uniform gastrotomy was created, and a brief peritoneoscopy was performed prior to closure. Procedure times were recorded. TAD and FSD times were adjusted for overall comparison secondary to pre- and post-gastrotomy steps required. An upper GI fluoroscopy was administered to assess for gastric leak on post-operative days 0, 2, and 7. Animals were survived to 14 days when a necropsy and gastric burst pressure test was performed. Results: The recorded closure times (min: sec) were 45:00 ± 16: 42 (mean±SD), 25: 27 ± 9: 31, and 35: 35 ± 21: 46 for TPD, TAD, and FSD, respectively. Though not statistically significant, pre-gastrotomy suture placement in the latter groups provided improved gastric insufflation and endoscopic visualization resulting in shorter closure times. Only one fluoroscopic leak was revealed in the TAD group resulting in abdominal sepsis and death on day 7. One other early death also occurred in the TAD group on day 4 from a pulmonary complication unrelated to surgery. Abscesses developed in 10 animals: TPD (n=1), TAD (n=5) and FSD (n=4). Of them, half were large (>2 cm) or diffuse collections: TAD (n=3) and FSD (n=2). Two adjacent organ injuries were also identified: TPD (n=1) and TAD (n=1). Mean bursting pressures were 85 ± 16, 93 ± 33, and 81 ± 22 mmHg for TPD, TAD, and FSD, respectively (p=NS). Conclusion: The 3 endoscopic suturing devices offer feasible closure options by providing a reproducible surgical closure. Though procedure times were statistically similar, placement of a pre-gastrotomy suture provided a technical advantage to the TAD and FSD and trended toward a more efficient closure. Continued experience and engineering improvements of these devices will likely improve results. The prevalence of abscess formation without evidence of leak, however, requires close attention in the ongoing development of closure tools for human NOTES trials.