Feasibility of endoscopic hand suturing on rectal anastomoses in ex vivo porcine models

Prevention of postoperative anastomotic leakage in rectal surgery is still required. This study investigated the feasibility of endoscopic hand suturing (EHS) on rectal anastomosis ex vivo. By using isolated porcine colon, we prepared ten anastomoses 6–10 cm from the virtual anus. Then, we sutured anastomoses intraluminally by EHS, which involved a continuous suturing method in 5 cases and a nodule suturing method with extra corporeal ligation in 5 cases. Completeness of suturing, number of stitches, procedure time and presence of stenosis were investigated. Furthermore, the degree of stenosis was compared between the two suturing methods. In all cases, EHS were successfully completed. The median number of stitches and procedure time was 8 and 5.8 min, respectively. Stenosis was created in all continuous suturing cases whereas none was seen in nodule suturing cases. The shortening rate was significantly greater in the continuous suturing method than in the nodule suturing method. Intraluminal reinforcement of rectal anastomosis by EHS using nodule suturing with extra corporeal ligation is feasible without stenosis, which may be helpful as a countermeasure against possible postoperative anastomotic leakage in rectal surgery.

Endoscopic hand suturing. All procedure was conducted using a multi-bending scope with two working channels (GIF-2TQ260M; Olympus Co. Ltd., Tokyo, Japan). A transparent straight hood (D-201-13, 404; Olympus) was attached to a tip of the endoscope. The suturing procedure described below was conducted by one endoscopist and one assistant who had had clinical experience in EHS.
In five cases, we performed EHS with a continuous suturing method in the same manner as previously introduced 13 . As the first step of EHS, a 3-0 V-loc 180™ absorbable barbed suture (VLOCL0604; Covidien, Mansfield, Massachusetts, USA), which had short barbs to prevent the suture from sliding backward after the tissues being tightened, was delivered through the virtual anus by using the prototype of the flexible needle holder (Olympus) by grasping the thread close to the tail of the needle. The suture thread was once released and the body of the needle was grasped. Then, the needle was passed through the full layer on the proximal part of the anastomosis, followed by passing through the layer on the distal side (Fig. 1c, d). We circumferentially sutured the layers between the anastomosis at intervals of a little less than 1 cm. The remaining portion of the thread was cut with the dedicated scissors forceps (Olympus) and transanally retrieved by grasping the thread part (Fig. 2a).
In other five cases, we performed EHS with a nodule suturing method with extra corporeal ligation. In these cases, the transparent hood attached to the tip of the endoscope had been notched (about 3 mm) at 3 and 9 o' clock in advance (Fig. 1e). The suture thread used was a 3-0 absorbable monofilament suture (Biosyn™; Covidien Japan). The needle was delivered and passed through both sides of anastomosis in the same way as continuous suturing cases. Subsequently, the needle was completely retrieved outside. The endoscopist tied the thread with slipknot and hooked each side of the thread to corresponding notches at 3 and 9 o' clock in the hood. Then, while the assistant was pulling the thread, the endoscopist delivered the knot using the endoscope itself as a knot pusher to the anastomotic site (Fig. 1f). We carried this procedure four times per stitch and cut the remaining thread with scissors forceps. Finally, the needle was retrieved by pulling the thread. The intervals of stitches were set at approximately 1 cm (Fig. 2b, c).
Comparison of degree of stenosis between two suturing methods. After rinsing the isolated colon with tap water, the tract was cut longitudinally and separated every 10 cm. Accordingly, a total of ten colonic www.nature.com/scientificreports/ models in approximately 10 cm squares were prepared and pinned to the urethan boards. An endoscopist manually sutured the mucosa in line eight times. In five models, we sutured with continuous suturing method by using 3-0 V-loc 180™, and other five models were sutured with nodule suturing method by using 3-0 Biosyn™.

Outcome measures.
In the experiment on feasibility, we assessed the completeness of the suturing, the number of stitches, average procedure time of suturing and presence of stenosis after suturing. Changes in procedure time of suturing in the nodule suturing cases were also investigated. The procedure time of suturing was defined as the duration between the first insertion of the needle tip and the final cut of the remaining thread in each case. In the experiment on degree of stenosis, we compared an average of shortening rate of flat colon models between continuous suturing method and nodule suturing method.

Statistical analysis.
Obtained results are presented as median and range. For comparison of shortening rate between two suturing methods, t-test was used. P values < 0.05 were considered to indicate statistical significances. Statistical analyses were performed using IBM SPSS version 25 (IBM Corp., Armonk, NY, USA).

Results
In the experiment on feasibility of EHS with rectal anastomosis model, the EHS procedure was completed in all cases. The median number of stitches was 8 (range [6][7][8][9], and the median procedure time of suturing was 5.8 min (range 3.6-9.1 min). In the five cases with the continuous suturing method, stenosis was revealed (the first case, mild; the second case, moderate; the third case, severe, the fourth case, moderate; the fifth case, mild). In contrast, no stenoses were seen in the five cases in the nodule suturing method with extra corporeal ligation (Table 1, Fig. 3). The procedure time of suturing in the nodule suturing cases became shorter as the procedure proceeded (Fig. 4).

Discussion
In this ex vivo study, we demonstrated that EHS on rectal anastomosis was technically feasible with the acceptable procedure time. Particularly in the nodule-suturing method, no stenoses were seen, and the suturing speed was accelerated depending on the learning effect.
Regarding the reinforcement of the anastomotic site to reduce the possibility of postoperative leakage, several methods, e.g., trans-anal reinforcing anastomotic sutures using a rigid endoscope 4 and single-stapling with transabdominal reinforcing suture 9,10 , have been reported. However, there remain cases in which anastomosis is difficult to approach either transabdominally or transanally with conventional techniques. Laparoscopic approach will be challenging at the dorsal side of the rectum, particularly in cases with the narrow pelvis. In contrast, difficulty in the transanal suturing with a rigid scope will be increased as the anastomotic site becomes far from the anus. This technique has the flexibility in the suturing area both circumferentially and longitudinally. Therefore, it may be more desirable for reinforcing anastomosis, particularly in the above-mentioned cases.
Another endoscopic technique such as clipping, the most popular mucosal closure method, may work similarly to this technique. However, the previous ex vivo study reported that the strength of the tissue apposition was greater in suturing than in clipping 11,14 . It is expected that EHS provides a comparable strength to surgical suturing because the methodology and materials used in EHS are similar to in surgery.
As an unfavorable event, continuous suturing involved stenosis. In contrast, no stenosis was seen in nodule suturing cases. A possible reason may be that the continuous suturing produces a strong tension in a short axis circumferentially enough to narrow the lumen, whereas the tension in that axis produced by the nodule suturing remains small. In our experiment on the degree of stenosis in each suturing method, the shortening rate was obviously greater in the continuous suturing method than in the nodule suturing method, which implied that the shortening force toward the short axis was stronger in the continuous suturing method than in the nodulesuturing method. Because the stenosis should be avoided to affect the long-term outcome and quality of life 15 , the nodule suturing technique with extra corporeal ligation should be applied. In this study, the procedure time of the nodule suturing became shorter, which suggests that suturing speed will become faster as training and practice are accumulated.
This study has several limitations. First, it was an ex vivo, small-number pilot study using the isolated porcine colon. Consequently, no conclusions can be drawn on whether hemostasis will be adequate in an animate model, and it is uncertain whether the observations on stenosis will also carry over to clinical practice. Second, operators were limited to a single endoscopist and assistant having previous clinical experience of EHS. Third, we did not estimate the sealing strength objectively due to lack of an instrument to measure intraluminal pressure or mechanical strength. Fourth, we have not investigated histological assessments at the sutured anastomoses, which can be done in in vivo models. In vivo experiments would be favorable to directly demonstrate the effectiveness of additional suturing for anastomoses. The results of this ex vivo experiment will be able to accelerate subsequent in vivo experiments as a next step. In conclusion, we demonstrated that EHS of rectal anastomosis was feasible without stenoses by the nodulesuturing method with extra corporeal ligation in ex vivo porcine models. This concept may be one possible option to reduce the risk of anastomotic leakage after rectal surgery.