Feasibility of Zeosuture M for gastric endoscopic submucosal dissection defect closure during anticoagulation: a preclinical porcine study from Japan

Abstract

Background/Aims

Endoscopic submucosal dissection (ESD) in patients undergoing anticoagulant therapy carries a risk of delayed bleeding. Zeosuture M is a novel endoscopic suturing system designed for secure mucosal closure. This study evaluated the feasibility of Zeosuture M and the short-term clinical course after mucosal closure in a warfarinized porcine ESD model.

Methods

Gastric ESD was performed in four pigs under warfarin anticoagulation (three defects per animal; 12 sites). Warfarin administration began 15 days before ESD. All defects were closed using Zeosuture M. Follow-up endoscopy was performed on postoperative days (POD) 1, 2, 4, and 8. Necropsy was conducted on POD 8.

Results

Technical success was 100% (12/12). The mean specimen size was 27.5±3.4 mm, mean number of stitches 3.6±0.7, closure time 28.1±8.0 minutes, and time per stitch 8.0 minutes. No uncontrolled intraprocedural bleeding, delayed hemorrhage, perforation, or peritonitis occurred. Partial loss of apposition was observed on POD 4, and all sites reopened by POD 8 without adverse clinical events.

Conclusions

Zeosuture M provided early stabilization of ESD ulcers during continuous anticoagulation, enabling safe traversal of the high-risk period for delayed bleeding. This represents one of the first preclinical warfarinized porcine evaluations and suggests potential applicability in high-risk patients.

Graphical abstract

INTRODUCTION

Endoscopic submucosal dissection (ESD), which enables en bloc resection of large lesions, has revolutionized the management of early gastric neoplasia.1,2 However, this technique carries procedural risks including perforation and delayed bleeding, which are particularly problematic in patients undergoing anticoagulant or antiplatelet therapy.3-5 With the increasing prevalence of anticoagulant use in the aging population and increased prevalence of cardiovascular disease, strategies to mitigate post-ESD complications in this high-risk group are urgently needed.6

Several endoscopic closure techniques have been developed including clip closure and over-the-scope clips (OTSCs).7 While effective for small lesions, these methods may not always provide durable or stable closure of larger defects or defects in patients receiving anticoagulant therapy.8,9 Endoscopic suturing systems have emerged as promising alternatives that can potentially offer stronger and more reliable defect closure.10-12

Zeosuture M (Zeon Medical) is a double arm-bar automated suturing system designed for use during flexible fiberoptic endoscopy that permits mucosal plication and, depending on the technique used, even full-thickness closure.13-15 However, the outcomes of using this device in patients receiving systemic anticoagulants remain insufficiently characterized. We evaluated the feasibility and short-term safety of defect closure using Zeosuture M in warfarinized pigs, focusing on the early post-ESD period when the risk of bleeding is at its peak.

METHODS

Animal preparation

Four domestic pigs (35–40 kg; IDs Z01–Z04) were administered warfarin orally beginning 15 days prior to the scheduled ESD, at the initial dose of 0.07 mg/kg on day −15, followed by maintenance dosing at 0.04 mg/kg once daily thereafter. Anticoagulation was maintained through postoperative day (POD) 8, and serial international normalized ratio measurements were performed to confirm anticoagulant effects (Table 1). The procedures were performed under general anesthesia.

Table 1.

Baseline and perioperative values of PT-INR in the animals

Pig ID	INR baseline	INR at ESD	INR POD 8	Survival to POD 8
Z01	0.99	3.10	2.53	Yes
Z02	0.95	2.91	1.93	Yes
Z03	0.98	3.06	1.91	Yes
Z04	0.95	2.75	2.00	Yes

INR, international normalized ratio; PT-INR, prothrombin time INR; ESD, endoscopic submucosal dissection; POD, postoperative day; Z01, body posterior 1, antrum anterior 2; Z02, body anterior 1, antrum posterior 2; Z03, body anterior 2, antrum posterior 1; Z04, body anterior 2, antrum posterior 1.

ESD procedure

Standard gastric ESD was performed to create 20 to 30 mm mucosal defects (three per animal; total of 12). The locations were prespecified across the body and antrum to sample differing wall thicknesses (Z01: body posterior 1, antrum anterior 2; Z02: body anterior 1, antrum posterior 2; Z03: body anterior 2, antrum posterior 1; Z04: body anterior 2, antrum posterior 1).

Endoscope and accessories

We performed closure using a standard single-channel gastroscope equipped with a 3.2-mm working channel (GIF-Q260J; Olympus). An overtube (inner diameter, 20 mm; effective length, 205 mm; TOP Co.) was used during oropharyngeal intubation to facilitate smooth passage and minimize the risk of oropharyngeal injury. Each procedure was conducted by two personnel: one endoscopist and one assistant operating the device.

Device description and closure technique

Zeosuture M is a single-use automated suturing system designed for use with flexible endoscopy that deploys sterile sutures for mucosal plication and is capable of full-thickness closure. In the present study, the target defects were post-ESD ulcer bases; therefore, closure was performed at the mucosal–submucosal level rather than at the full thickness.

1) Closure steps

The two front arms were positioned on both sides of the defect, the rear arm advanced the puncture needle to penetrate the gastric mucosa, and the defect edges were approximated and tensioned with Zeotieupper S (Zeon Medical). After cinching the suture loop in place, the thread was cut between the cinch and the needle using a Hookcutter MI (Zeon Medical). Sequential stitches were placed until an apparent endoscopic closure was achieved (Figs. 1, 2, Supplementary Video 1).

Fig. 1.

Endoscopic suturing devices. (A) Zeosuture M (Zeon Medical). (B) Zeosuture M Mounted on the Distal Tip of the Gastroscope. (C) Zeotieupper S (Zeon Medical). (D) Hookcutter MI (Zeon Medical.).

Fig. 2.

Stepwise mucosal closure using Zeosuture M (Zeon Medical). (A) Post-endoscopic submucosal dissection mucosal defect (ulcer base). (B) The endoscope, equipped with the Zeosuture M at the distal tip, is reinserted; (C) the front arm is positioned at one edge of the defect; (D) the front arm of the opposite side is placed, and the rear arm is used to advance the puncture needle to penetrate the gastric mucosa, passing a suture; (E) the suture traverses both sides of the defect; (F) the suture is tensioned and cinched using the Zeotieupper S through the working channel; (G) after being secured with the suture retainer, the thread is cut with the Hookcutter MI (Zeon Medical); (H) additional stitches are placed in the same manner; (I) complete mucosal closure obtained after three stitches.

Post-procedural assessment

The animals were monitored for 7 days (vital signs, intake, and activity). Endoscopic examinations were performed on POD 1, 2, 4, and 8 to assess suture integrity, mucosal apposition, and the presence of bleeding or perforation. On POD 8, the animals were humanely euthanized and gross inspection of the peritoneal cavity and stomach was performed at necropsy to evaluate the presence of perforation, peritonitis, and/or bleeding.

Statistics

Descriptive site-based analyses (n=12) were conducted using EZR ver. 1.61 (Saitama Medical Center, Jichi Medical University), a graphical user interface for R. Continuous variables are summarized as means±standard deviation with 95% confidence intervals (CIs) (t distribution) and proportions with exact (Clopper-Pearson) 95% CIs. “Time per stitch” was defined as the closure time (minutes) divided by the number of stitches per site.16

Ethics approval and animal welfare

All animal procedures were performed in compliance with the Japanese law and institutional animal welfare guidelines, with the approval of the Aibitec Institutional Animal Care and Use Committee (approval number: IVT21-171).

RESULTS

Technical performance

The technical success rate for mucosal closure was 100% (12/12; 95% CI, 73.5%–100%). The mean specimen size was 27.5±3.4 (95% CI, 25.36–29.64) mm; the mean number of stitches was 3.6±0.7 (95% CI, 3.16–4.01); the mean closure time was 28.1±8.0 (95% CI, 22.99–33.18) minutes. The mean time taken per stitch was 8.0±2.4 (95% CI, 6.44–9.51) min/stitch. No cases of intraprocedural perforation or uncontrolled bleeding were observed (Tables 2, 3).

Table 2.

Procedural and endoscopic outcomes (n=12 sites)

Site ID	Location	Specimen size (mm)	Stitches (n)	Closure time (min)	Time taken/stitch (min)	POD 1/2	POD 4	POD 8	Adverse event
Z01-A-AW	Antrum, anterior	25	4	42	10.5	Maintained closure	Partial dehisce	Complete dehiscence	None
Z01-A-PW	Antrum, posterior	30	3	23	7.7	Maintained closure	Maintained closure	Complete dehiscence	None
Z01-B-PW	Body, posterior	30	3	26	8.7	Maintained closure	Complete dehiscence	Complete dehiscence	None
Z02-A-PW	Antrum, posterior	30	4	30	7.5	Maintained closure	Complete dehiscence	Complete dehiscence	None
Z02-A-AW	Antrum, anterior	30	4	26	6.5	Maintained closure	Maintained closure	Complete dehiscence	None
Z02-B-AW	Body, anterior	25	3	24	8	Maintained closure	Partial dehiscence	Complete dehiscence	None
Z03-B-PW	Body, posterior	20	3	26	8.7	Maintained closure	Partial dehiscence	Complete dehiscence	None
Z03-B-AW	Body, anterior	25	3	14	4.7	Maintained closure	Maintained closure	Complete dehiscence	None
Z03-A-AW	Antrum, anterior	30	4	23	5.8	Maintained closure	Maintained closure	Complete dehiscence	None
Z04-B-PW	Body, posterior	25	3	42	14	Maintained closure	Complete dehiscence	Complete dehiscence	None
Z04-B-AW	Body, anterior	30	4	27	6.8	Maintained closure	Partial dehiscence	Complete dehiscence	None
Z04-A-PW	Antrum, posterior	30	5	34	6.8	Maintained closure	Maintained closure	Complete dehiscence	None

POD, postoperative day; Z01, body posterior 1, antrum anterior 2; Z02, body anterior 1, antrum posterior 2; Z03, body anterior 2, antrum posterior 1; Z04, body anterior 2, antrum posterior 1; partial dehiscence, any visible gap with exposed submucosa, but with at least one stitch maintaining partial bridging; complete dehiscence, loss of bridging with full re-exposure of the ulcer base (no effective edge-to-edge contact).

Table 3.

Summary of the clinical course and necropsy findings (site-based, n=12)

Variable	Estimate
Technical success rate	12/12 (100.0; 95% CI, 73.5–100.0)
Specimen size (mm)	27.5±3.4 (95% CI, 25.36–29.64)
No. of stitches	3.6±0.7 (95% CI, 3.16–4.01)
Closure time (min)	28.1±8.0 (95% CI, 22.99–33.18)
Intraprocedural uncontrolled bleeding	0/12 (0)
Delayed hemorrhage	0/12 (0)
Perforation	0/12 (0)
Peritonitis	0/12 (0)

Values are presented as number/total number (%) or mean±standard deviation.

CI, confidence interval.

Endoscopic durability

Endoscopy performed on POD 1 and 2 confirmed visible sutures at the closure lines with no active bleeding or perforation. On POD 4, partial loss of mucosal apposition was observed at several sites. On POD 8, all closed defects were reopened. Despite the late loosening of the closure, no cases of delayed hemorrhage or perforation occurred during the observation period (Fig. 3).

Fig. 3.

Follow-up endoscopy after Zeosuture M (Zeon Medical) closure in the porcine models receiving continuous warfarin administration. Z03 antrum anterior: On postoperative day (POD) 4, the closure line remained intact with maintained mucosal apposition and no active bleeding. On POD 8, the defect reopened (complete dehiscence). Z04 body anterior: On POD 4, partial loss of apposition (partial dehiscence) was observed along the closure line, and complete dehiscence was evident on POD 8. No perforations or active hemorrhages were observed in any of the animals.

Clinical course

All animals survived until POD 8 without melena, hematemesis, abdominal distension, or signs of peritonitis. A gross necropsy revealed no peritoneal contamination or active bleeding (Fig. 4).

Fig. 4.

Postoperative day (POD) 8: gross and endoluminal findings. (A) External (serosal) view of the peritoneal cavity showing no turbid ascites, fibrin, or purulent exudate; the serosal surfaces are glistening. (B) External (serosal) view of the stomach around the endoscopic submucosal dissection (ESD) sites showing intact serosa without perforation, hematoma, adhesions, or intraperitoneal hemorrhage. (C) Endoluminal view of the post-ESD ulcer base showing mucosal reopening without active bleeding or visible transmural defect, consistent with the absence of peritonitis on POD 8.

DISCUSSION

In the warfarinized pigs, use of Zeosuture M enabled immediate technical defect closure and an uneventful early post-procedural course from POD 1 through 8, the period during which the risk of post-ESD bleeding is at its peak.3,5,6,8 The POD-4 partial loosening and POD-8 complete loosening of the sutures align with prior observations in porcine models that purely mucosal stitches may not persist beyond a week in the stomach, where submucosal remodeling and wall thickness predispose to late relaxation; thus, the clinical goal of mucosal closure lies in bridging the early high-risk window rather than ensuring long-term apposition.12 Our data support this interpretation; securing early mucosal approximation resulted in a zero incidence of bleeding/peritonitis and a stable clinical course in warfarinized porcine models.

Implications in anticoagulated patients

Patients receiving anticoagulation therapy constitute a cohort at a substantially elevated risk of delayed bleeding after gastric ESD, with warfarin carrying the highest weight in the BEST-J risk model.17 Moreover, clinically relevant bleeding under continued low-dose warfarin therapy was observed on POD 2–4,18 highlighting the first postoperative week as a vulnerable window. Within this context, our warfarinized porcine model showed that immediate mucosal approximation with Zeosuture M was technically successful and the clinical course remained uneventful through POD 1–8, despite later loosening of mucosal stitches. These findings support a bridging rationale, whereby secure early closure may stabilize the ulcer base and mitigate bleeding risk during the high-risk period in anticoagulated settings. We acknowledge that human outcome data are limited and controlled clinical studies are needed to confirm whether early endoscopic closure reduces delayed bleeding in anticoagulated patients.

Positioning among the available closure options

Clip-based methods and the use of OTSCs can reduce the risk of selected defects; however, their usefulness may be limited by the defect size or unfavorable tension vectors, particularly in the stomach.7,9 Accumulating clinical data support robust apposition achieved with endoscopic hand suturing after ESD, including in patients undergoing antithrombotic therapy10,11; non-suturing mechanical approaches such as the reopenable-clip over-the-line method, endoscopic ligation with O-ring closure, and endoloop-and-clip closure are feasible for larger defects and high-risk settings.19-21 These trends are consistent with contemporary complication reviews and technology updates that summarize the indications and device performances across platforms.22-24 Zeosuture M represents an operator-controlled suturing option; although the device offers the capability for full-thickness closure, our protocol intentionally targeted mucosal/submucosal approximation in ESD ulcers.13-15

Strengths and limitations

1) Strengths

A high-risk setting was modeled with the use of systemically warfarinized pigs; a standardized Zeosuture M technique was applied to 12 gastric ESD defects (body/antrum). Protocolized endoscopy on POD 1/2/4/8 with a predefined grading of closure (intact/partial/complete) characterized the time course under continuous anticoagulant therapy, and POD-8 necropsy confirmed the absence of peritoneal contamination or bleeding.

2) Limitations

Operator experience and learning curve: The operators had experienced only a few pilot cases before this porcine experiment; therefore, a formal learning curve could not be determined. Nevertheless, complete defect closure was technically feasible from the outset, and after a few practical cases, the procedure could be performed reproducibly.

(1) Study design

This was a small, single-arm preclinical study without a comparator, the follow-up duration was short (until POD 8), and we did not use histopathological or long-term healing endpoints. The analyses were site based, with multiple sites assessed per animal. Therefore, controlled comparative studies are required to assess the safety, efficacy, durability, and cost.

Conclusion

Zeosuture M enabled reliable early stabilization of gastric ESD defects in porcine models receiving continuous warfarin administration. Although mucosal apposition was loosened by POD 8, the absence of any events of delayed hemorrhage or peritoneal complications supports the device’s role in bridging the early high-risk period in anticoagulated patients and warrants further clinical evaluation in randomized controlled studies.

Supplementary Material

Supplementary Video 1.

Defect closure using the Zeosuture M.

Supplementary materials related to this article can be found online at https://doi.org/ce.2025.329.