AIM: To evaluate the feasibility, reproducibility and efficacy of a new tissue anchoring device in a porcine survival model.

METHODS: Gastrotomies were performed using a needle-knife and balloon dilator in 10 female Yorkshire pigs weighing 30-35 kg. Gastric closure was attempted using a new tissue anchoring device. The tightness of the closure was confirmed by means of air insufflation and the ability to maintain gastric distension with stability in peritoneal pressure measured with a Veress needle. All animals were monitored daily for signs of peritonitis and sepsis over 14 d. During necropsy, the peritoneal cavity and the gastric access site were examined.

RESULTS: Transgastric access, closure and 14 d survival was achieved in all pigs. The mean closure time was 18.1 ± 19.2 min and a mean of 2.1 ± 1 devices were used. Supplementary clips were necessary in 2 cases. The closure time was progressively reduced (24.8 ± 13.9 min in the first 5 pigs vs 11.4 ± 5.9 min in the last 5, P = NS). At necropsy, the gastric access site was correctly closed in all cases with all brace-bars present. One device was misplaced in the mesocolon. Minimal adhesions were observed in 3 pigs and signs of mild peritonitis and adhesions in one.

CONCLUSIONS: The use of this new tissue anchoring device in porcine stomachs is feasible, reproducible and effective and requires a short learning curve.

Natural orifice transluminal endoscopic surgery (NOTES) has changed the approach to the peritoneum in the last few years[1-5]. This novel technique permits access to the peritoneal organs through the mouth, rectosigmoid or vagina with diagnostic and therapeutic purposes. Numerous hybrid NOTES procedures (combining NOTES with laparoscopy) have been described in the last five years[6-9], but it was not until 2007 that the first pure NOTES procedures in humans were reported[10-13]. Although the transluminal approach holds great potential, secure access site closure remains a critical issue[14]. In recent cases and series, endoscopic closure is substituted by use of rigid instruments, using the transvaginal access in almost all cases. However, this approach excludes the male population.

Considering the safety of laparoscopy, studies are mandatory to evaluate secure and reproducible closure methods in NOTES procedures[15]. Several closure techniques have been tested[16], including clips[1,17-19], septal occluders[20], T-tags[21-23], more complex suturing devices[24-26], and linear endoscopic staplers[27]. T-tags have been tested recently to treat gastrogastric fistulas in humans[28]. However, most of these devices are time consuming and often difficult to implement endoscopically and the current data do not allow definitive conclusions regarding the different options[16,29].

The aim of this study was to assess the feasibility, reproducibility and efficacy of a new tissue anchoring device as a gastric suture system in a porcine survival model.

All transgastric accesses were achieved with no difficulties and a mean time of 5.7 ± 3.6 min (range 2-14). No major complications occurred. Minor complications included an accidental injury to the anterior abdominal wall and 4 minor bleedings during the use of the needle-knife. A 30 min peritoneoscopy was possible in all animals.

The brace-bar was used in all cases and closure was easily achieved in 18.1 ± 19.2 min. This time was reduced to 11.4 ± 5.9 min when considering only the last 5 cases and was less than 9 min in the last 3 cases, whereas it was 24.8 ± 13.9 min in the first 5 (P = 0.1). Details of all procedures are described in Table 1. A total of 21 sets of sutures (mean 2.1 ± 1, range 1-4) were used to achieve closure but only 15 (mean 1.5 ± 0.5, range 1-2) could be completely tightened. Therefore, in 5 cases, the incision was closed with only 1 suture (Figure 3), but in two of them 1 and 3 clips respectively, were added.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 The incision looks completely sealed after the insertion of one brace-bar and the stomach is able to maintain air distension.

In total, 6 sutures (29%) were ineffectively positioned. In 5 attempts one of the tags did not stay attached to the mucosa either immediately after the tag release or when tightening the suture. The remaining failure was caused by a thread rupture after steady tightening.

Immediately after gastrotomy closure, all brace-bars seemed well positioned and gastric distension with air was possible in all cases without changes in intraperitoneal pressure, suggesting the closure was correct (mean peritoneal pressure before and after the closure: 14.3 ± 3.3 mmHg, range 3-15).

The mean procedure time, including gastric access creation, peritoneoscopy and gastrotomy suture, was 63.7 ± 18.2 min.

All the pigs completed the 14 d follow-up period. They had a weight gain of 3.3 ± 3.2 kg. At necropsy, the gastric access site was completely closed in all cases and all brace-bars were present (Table 2). The tags were usually attached at the gastric serosa (n = 15) or inside the gastric wall (n = 14). In the first case, 1 tag was misplaced inside the mesocolon. In this case, an omental patch had been added to the suture pulling the omentum inside the stomach through the incision. Minimal adhesions were observed in 3 pigs and signs of mild peritonitis and adhesions in one.

NOTES holds great appeal as a less invasive alternative to laparoscopic surgery. As NOTES heads toward human trials, it is essential that the creation and closure of transluminal incisions be performed in a safe, rapid, and reproducible manner[14-16,28].

In this study, we assessed the feasibility of a new generation tissue anchoring device with relatively good results. Moreover, it turned out to be easy and intuitive to use and the time for placement was short and progressively reduced. It was not necessary to use complementary clips when we gained experience with the system. One of the advantages of this device (and the main difference with the former prototype used by Sumiyama et al[22]) is that it can be used with a single channel endoscope. The same needle catheter is used for releasing the tags and tightening them later without need for a different forceps grasper, and this makes the procedure shorter. Furthermore, the depth of the needle insertion is limited to 20 mm and this might decrease the risk of complications.

However, we still found some problems with the device: the needle had to be loaded extracorporally after each set of tissue anchors was applied and this prevented sequential stitching. Moreover, we observed a dysfunction of the needle after several attempts which could explain the high rate of sutures being ineffectively positioned because the tags could not be released deeply enough. We think that pre-charged and non-reusable devices might improve the procedure time and security. On the other hand, we did not drop any tags in the peritoneal cavity and, since each pair of tags is attached to a thread, we think that the possibility of dropping one in the peritoneum is extremely low.

The possibility of an inadvertent injury of organs and structures outside of the gut wall has been described as a possible limitation of T-tag based systems[30,31]. Sumiyama et al[22] produced 12 gastric perforations in 6 pigs that were closed with 48 tissue anchor sets and three of the 24 used in the anterior gastric wall (12.5%) penetrated surrounding organs (2 penetrated the liver and 1 the anterior abdominal wall). However, as mentioned above, with the new brace-bar prototype the depth of the needle insertion is limited to 20 mm and this fact was crucial in the low incidence of surrounding structure injuries in our series (1 tag out of 21 sets, 4.8%). This was the first case and we tried to perform an omental patch pulling the omentum through the incision. During this maneuver, the mesocolon was probably unsafely moved towards the gastric wall causing this complication. In the remaining cases in which the omental patch was not attempted no lesions occurred at the adjacent structures. The importance of the depth of the needle to avoid complications has been demonstrated very recently by Park et al[32] These authors performed needle punctures of 1-1.5 mm using a different anchor-based endoscopic system (the TAS o tissue apposition system) and they did not have any adjacent organ penetration with a 100% of closure effectiveness.

Although the ex-vivo study of Voermans et al[33] suggested that t-tag based methods do not permit the serosa to serosa approach and leaking pressure is lower than with other devices, the surviving pigs showed a good postoperative course. This fact could be explained because physiological intraluminal pressures are much lower than pressures obtained in acute bursting tests and, therefore, might not be a necessary objective test for viscerotomy closure[34]. From a clinical standpoint, the critical test for a gastric closure is animal survival without clinical signs of leakage or complications.

Previous studies have demonstrated that peritoneal contamination occurs when using transgastric access. A conservative interpretation of these findings is that the current “aseptic” technique may require further refinement, as suggested by Rolanda et al[18] and Ryou et al[35]. In fact, we had some difficulties in completely cleaning some of the stomachs and it was very common to notice residual liquid near the incision. Only one pig did not have a satisfactory recovery and the necropsy showed the presence of fibrin exudates in the upper abdomen. Although the incision site was seen completely sealed at necropsy, we cannot totally exclude the possibility that an initial suture failure occurred but it could be also related with a potential contamination of the abdominal cavity by the stomach content. Because in this case we used only one brace-bar, we now consider it prudent to use two devices to ensure a safer closure.

The present study has some limitations: first, the number of cases is low and we did not include a control group. Second, the 14 d survival period might be short to evaluate late complications. Finally, the use of complementary clips in two cases might modify the results of the study.

In conclusion, the use of a brace-bar in a gastric porcine model is easy, fast, and reproducible after a short learning curve and permits the use of a single channel endoscope. We believe this tissue anchoring system holds tremendous potential as a suturing method for both iatrogenic and intentional perforations of the gastric wall. Unfortunately, it is still far from a safe application in humans. Further studies and more technological improvements are still mandatory before expanding its use to humans.