Prospective single-arm trial on feasibility and safety of an endoscopic robotic system for colonic endoscopic submucosal dissection

Abstract

Background The development of the EndoMaster “Endoluminal Access Surgical Efficacy” (EASE) system aims to enhance the safety and efficacy of colonic endoscopic submucosal dissection (ESD) through two flexible robotic arms. This is the first clinical trial to evaluate the performance of colorectal ESD using EndoMaster.

Method Patients with early mucosal colorectal neoplasia that was not suitable for en bloc resection with snare-based techniques were recruited. The EndoMaster EASE robotic system consisted of an independently designed flexible robotic platform with two robotic arms. The primary outcome was the complete resection rate using EndoMaster. Secondary outcomes included operating time, hospital stay, procedure-related complications, and oncologic outcomes.

Results 43 patients underwent robotic ESD, with a median robotic dissection time of 49 minutes. The technical success rate was 86.0%, while en bloc resection rate among cases with technical success was 94.6%. The complete resection rate was 83.8% and the median size of specimen was 35 mm (range 15–90 mm). The median hospital stay was 2 (range 1–7) days and there was one delayed bleed after 4 days, which was controlled endoscopically. One patient sustained perforation during the procedure, which was completely closed using clips without sequelae. Two patients were treated by salvage surgery.

Conclusion This first clinical trial confirmed the safety and efficacy of performing colorectal ESD using the EndoMaster EASE robotic system.

Introduction

Colorectal cancer remains the commonest gastrointestinal cancer worldwide [1]. Colorectal laterally spreading tumors (LSTs) serve as an important entity for cancer development through the de novo pathway, while the diagnosis of LSTs is enhanced by advanced endoscopic imaging, as well as improvements in the diagnostic abilities of endoscopists through education and training [2].

According to the Japan Gastroenterological Endoscopy Society (JGES) and European Society of Gastrointestinal Endoscopy (ESGE) guidelines, colonic endoscopic submucosal dissection (ESD) is indicated for: the resection of colorectal neoplasia where en bloc resection is difficult by snare endoscopic mucosal resection (EMR); mucosal tumors with submucosal fibrosis; sporadic tumors with background chronic inflammation; and residual or recurrent local early tumors [3] [4] [5]. Meta-analyses and a recent randomized trial have shown that colorectal ESD achieves significantly higher rates of en bloc resection with lower recurrence rates, when compared with EMR, for the treatment of early colorectal neoplasia [6] [7] [8]. However, ESD is difficult to master, with a long learning curve, especially among inexperienced endoscopists [9]. Hori et al. reported that factors contributing to difficulty performing colorectal ESD included tumors located at the flexure, and the presence of scarring or local recurrence, as well as tumors >50 mm [10]. Saito et al. demonstrated that perforation after colonic ESD was 11.8% among endoscopists whose experience was <50 ESDs, whereas those who had performed >100 ESDs had only a 4.1% perforation rate [11]. Besides experience, lack of effective tissue traction contributed significantly to perforation risk.

Advances in flexible robotics provide a potential solution to bridge this challenge [12]. We previously reported the safe performance of gastric ESD using an endoscopic robotic system [13]. With significant technologic improvements, the feasibility of ESD using the EndoMaster “Endoluminal Access Surgical Efficacy” (EASE) system (EndoMaster Pte, Ltd.) was confirmed in an animal study [14]. In this prospective cohort study, we investigated the clinical feasibility and safety of colonic ESD using EndoMaster.

Method

Study design

This was a prospective phase II clinical trial to assess the feasibility, efficacy, and safety of performing colorectal ESD using the EndoMaster EASE system. The study is reported according to the STROBE guidelines, and was approved by the Joint Chinese University of Hong Kong New Territories East Cluster Clinical Research Ethics Committee.

Participants

Patients aged between 18 and 85 years with an endoscopic diagnosis of early colorectal neoplasia were referred from colorectal surgery and gastroenterology for recruitment; their suitability was assessed by the investigators. The characterization of early colorectal neoplasia was assessed using narrow-band imaging (NBI) according to the NBI International Colorectal Endoscopic (NICE) and Japan NBI Expert Team (JNET) classifications. A single targeted biopsy was taken. Patients were recruited on the basis of the ESGE and JGES guidelines for colorectal ESD, including those with: (i) lesions for which en bloc resection with snare EMR would be difficult to apply; (ii) mucosal tumors with submucosal fibrosis; (iii) sporadic tumors in conditions of chronic inflammation; and (iv) local residual or recurrent early carcinomas after endoscopic resection [5] [6].

Patients were excluded when there was: (i) no informed consent available; (ii) carcinoma of colon or rectum with invasion to submucosa or beyond; (iii) evidence of distant metastasis; (iv) another active malignancy present; and if they: (v) were pregnant; (vi) were unfit for general anesthesia; (vii) had a target lesion that could not be reached by the endoscopic platform; (vii) had been recruited into another clinical trial.

Settings

EndoMaster EASE system and clinical team

The EndoMaster EASE System consists of Surgeon Console Unit (SCU), Patient Side Cart (PSC), endoscope, and flexible robotic instruments (Fig. 1s, see online-only Supplementary material). The SCU allows the surgeon to manipulate surgical instruments with controllers. The PSC holds the endoscope with robotic control of surgical instruments. The instruments include right-hand cautery and a left-hand grasper. In addition, the endoscope is equipped with a built-in camera and an additional working channel for ordinary endoscopic accessories.

The clinical team consisted of one endoscopist who performed all robotic ESDs, two assistants who were responsible for fine control of the endoscope and one nurse for periprocedural care. All team members received two sessions of training before commencement of the trial.

Procedure for robotic ESD using the EndoMaster EASE system

All patients underwent robotic ESD under general anesthesia at the endoscopy center of the Prince of Wales Hospital. The targeted lesion was first located and pre-injected using ORISE gel (Boston Scientific Co., Ltd.). The robotic endoscopy platform was then inserted to reach the target. After the scope had been docked into the PSC, the robotic arms were inserted and made ready with a jiggling action. Mucosal incision was first started from the anal side, followed by lifting with the grasper ([Fig. 1]; [Video 1]). The Vio 3 diathermy system (ERBE, Germany) was used with Endocut Q, effect 3, and forced coagulation, effect 3. Further submucosal dissection was performed using the robotic cautery knife. Upon completion, the specimen was retrieved upon withdrawal of the whole system.

Fig. 1 Images during an EndoMaster EASE robotic endoscopic submucosal dissection (ESD) procedure showing: a the robotic arms ready to start the procedure; b the left robotic arm lifting up the mucosa to expose the submucosal plane; c coordinated dissection using electrocautery from the right robotic arm; d further dissection of the submucosa with good exposure provided by traction over the base of the mucosa by the left arm; e the luminal defect after completion of robotic ESD; f the pathologic specimen after robotic ESD.

A rectal endoscopic submucosal dissection (ESD) procedure is performed using the EndoMaster robotic system.Video 1

Outcomes

The primary outcome was the complete (R0) resection rate, defined as one-piece resection with tumor-free lateral and vertical margins. The secondary outcomes included: technical success, and parameters related to the safety and performance of robotic ESD, as well as oncologic outcomes. Technical success was defined as en bloc one-piece excision, irrespective of the histopathology, in the absence of device-related serious adverse events within 30 days.

The clinical variables included post-ESD bleeding and perforation, as well as other adverse events. The variables to assess the performance of ESD included operative time, hospital stay, and device malfunction. The oncologic outcomes included complete resection with clear margins on histologic examination of the resected specimen. Follow-up surveillance colonoscopies were scheduled at 6, 18, and 36 months after ESD to assess for local recurrence and metachronous lesions. Computed tomography (CT) scans of the abdomen and pelvis were performed at 12 months and 36 months after ESD for local recurrence and distant metastasis.

Statistical method of analysis

The data were analyzed according to the principles of per-protocol analysis. Statistical analyses were conducted using the SPSS system. Continuous variables were summarized using descriptive statistics, including number of non-missing observations, medians, quartiles, and ranges (minimum and maximum observed values). Categorical variables were summarized using classical frequency statistics: numbers of non-missing observations and percentages by categories. When applicable, bilateral asymptotic or exact confidence intervals for binomial distributions were calculated at the 95% confidence level.

Results

From May 2020 to January 2022, a total of 242 colorectal ESDs were performed at the Endoscopy Center of the Prince of Wales Hospital: 199 patients underwent ordinary colorectal ESD; 45 patients with early colorectal neoplasia were initially recruited, but two were excluded after their CT scans, which confirmed tumor extending beyond the submucosa in one patient and liver metastasis in another ([Fig. 2]). The median age of the recruited patients was 66 years (mean age 65.9; range 40–83), and 57.8% were men ([Table 1]). Their lesions were located mostly in the sigmoid (41.9%; 18/43) and rectum (44.2%; 19/43), followed by ascending (1/43), transverse (2/43), and descending colon (2/43). The median tumor size estimated before ESD was 30 mm (range 12–40 mm). Most of the tumors (LSTs) were categorized as NICE II (88.4%), while endoscopic biopsies showed them to be either tubular, tubulovillous, or villous adenomas, with or without dysplasia.

Fig. 2 Study flowchart. EASE, Endoluminal Access Surgical Efficacy; ESD, endoscopic submucosal dissection.

Table 1 Baseline demographics of the 45 patients recruited into the study.

Parameter	n (%), unless otherwise stated [95%CI]
AF, atrial fibrillation; CCF, congestive cardiac failure; COPD, chronic obstructive pulmonary disease; IHD, ischemic heart disease JNET, Japan NBI Expert Team; LST, laterally spreading tumor; NA, not available; NICE, NBI International Colorectal Endoscopic.
Age, median (range), years	66 (40–83) [64–70]
Sex
Male	26 (57.8%) [43.4–72.2]
Female	19 (42.2%) [27.8–56.7]
History of medical co-morbidity	36 (80%) [68.3–91.7]
Hypertension	17 (37.8%) [23.6–51.9]
Diabetes mellitus	10 (22.2%) [10.1–34.4]
COPD/asthma	4 (8.9%) [2.5–21.2]
IHD/ CCF / AF	4 (8.9%) [2.5–21.2]
Other inactive malignancy	9 (20%) [8.3–31.7]
Tumor morphology
LST, granular homogeneous type	18 (40%) [25.7–54.3]
LST, granular mixed nodular type	11 (24.4%) [11.9–37.0]
LST, nongranular type	10 (22.2%) [10.1–34.4]
Wide-based pedunculated polypoid tumor	6 (13.3%) [5.1–26.8]
NICE classification
Type I	6 (13.3%) [5.1–26.8]
Type II	39 (86.7%) [73.2–95.0]
Type III	0
JNET classification
Type 1	3 (6.7) [1.4–18.3]
Type 2A	34 (75.6) [63.0–88.1]
Type 2B	7 (15.6) [6.5–29.5]
Type 3	0
NA	1 (2.2) [0.06–11.8]
Tumor size from colonoscopy report, median (range), mm	30 (12–40) [25–30]

Among the 43 patients who underwent robotic ESD, conversion to conventional ESD occurred in six patients. The reasons for conversion included: severe looping leading to access being impossible (n = 2), significant submucosal fibrosis (n = 2), system malfunction during bleeding (n = 1), and a large polypoid tumor jeopardizing endoscopic visualization (n = 1). The robotic ESD procedures were technically completed in 37 patients (86.0%) ([Table 2]). The median procedure time was 91 minutes (range 30–255), while the median calibration time for the robotic arms before ESD was 4 minutes (range 1–20). The median speed of dissection was 20.6 mm²/min (range 3.3–100). The median length of hospital stay was 2 days (range 1–7). There was one patient who sustained post-ESD bleeding on day 4 and was managed by endoscopic hemostasis, while another patient (2.3%) sustained perforation during ESD with closure by endoclips without sequelae. The median size of specimen was 34.5 mm (range 15 to 90 mm).

Table 2 Details and outcomes of the 43 EndoMaster EASE robotic endoscopic submucosal dissection (ESD) procedures.

Procedural detail/outcome	n (%), unless otherwise stated [95%CI]
MAC, monitored anesthesia care.
Type of Anesthesia
General	42 (97.7%)
Sedation (MAC)	1 (2.3%)
Duration of robotic arm calibration, median (range), minutes	4.0 (1–20) [3–5]
Duration of resection, median (range), minutes	49.0 (5–202) [39–74]
Duration of whole ESD procedure, median (range), minutes	91 (30–255) [75–107]
Speed of dissection, median (range), mm2/min	20.6 (3.3–100) [16.3–27.3]
Technical success, n (%)	37/43 (86.0%) [72.1–94.7]
Resection rates in technically successful cases
Complete resection	31/37 (83.8%) [68.0–93.8]
En bloc resection	35/37 (94.6%) [81.8–99.3]
Specimen size from pathology report, median (range), mm	34.5 (15–90) [30–40]
Final histopathology of resected specimen
Tubular adenoma with low grade dysplasia	8 (21.6%) [8.4–34.9]
Tubular adenoma with high grade dysplasia	5 (13.5%) [4.5–28.8]
Tubulovillous adenoma with low grade dysplasia	11 (29.7%) [15.0–44.5]
Tubulovillous adenoma with high grade dysplasia	7 (18.9%) [8.0–35.2]
Sessile serrated lesions	4 (10.8%) [3.0–25.4]
Adenocarcinoma	1 (2.7%) [0.07–14.2]
Other	1 (2.7%) [0.07–14.2]
Margin positivity
Deep margin	1 (2.7%) [0.07–14.2]
Circumferential margin	5 (13.5%) [4.5–28.8]
Adverse events
Perioperative mortality	0 (0%)
Post-ESD bleeding	1 (2.3%) [0.06–12.3]
Perforation	1 (2.3%) [0.06–12.3]
Hospital stay, median (range), days	2 (1–7) [2–3]
Follow-up period, median (range), months	24 (3–44) [24–29]
Surveillance colonoscopy performed at 18 months	36
Local recurrence	0 (0%)
Metachronous polyps	27 (75.0%) [60.9–89.1]
Adenomatous polyps	14 (38.9%) [23.0–54.8]

There were 31 patients (83.8%) who had complete resection with tumor-free lateral and deep margins, while the en bloc resection rate was 94.6% among technically successful cases ([Table 2]). Five patients had focal lateral margin involvement by dysplasia and were managed by surveillance colonoscopy ([Table 3]). Three patients received salvage surgery owing to the ESD specimens showing adenocarcinoma with deep margin involvement (n = 1) and adenocarcinoma with close deep margins (n = 1), and dense submucosal fibrosis and failure of robotic ESD (n = 1).

Table 3 Outcomes of the six patients with margin involvement on histopathology after colorectal endoscopic submucosal dissection.

	ESD specimen pathology	Margin involvement		Location	Management	Finding on follow-up colonoscopy1
		Deep	Circumferential
ESD, endoscopic submucosal dissection; HGD, high grade dysplasia; LGD, low grade dysplasia 1 Performed at 18 months.
1	Tubular adenoma with LGD	Clear	Focally involved by LGD	Sigmoid colon	Surveillance colonoscopy	No local recurrence
2	Focal adenocarcinoma in tubulovillous adenoma	Clear	Focally involved by tubulovillous adenoma	Sigmoid colon	Laparoscopic anterior resection	–
3	Tubulovillous adenoma with focal HGD	Clear	Involved with focal HGD	Sigmoid colon	Surveillance colonoscopy	No local recurrence
4	Tubular adenoma with focal HGD	Clear	Focally involved	Rectum	Surveillance colonoscopy	No local recurrence
5	Tubular adenoma with focal HGD	Clear	Focally involved	Rectum	Surveillance colonoscopy	No local recurrence
6	Tubular glands with LGD	Clear	Focally involved	Sigmoid colon	Surveillance colonoscopy	No local recurrence

There were ten incidences of device malfunction (Table 1s). The robotic arms were not recognized or there was control inaccuracy on seven occasions, all of which were resolved immediately by restarting the calibration. There were two occasions of system failure, which resolved immediately upon system restart. The endoscope could not be docked during one procedure, but this was resolved by addition of a fixation tape. None of these device malfunctions resulted in an adverse event for the patient.

The median clinical follow-up was 24 months (range 3–44 months), with 36 patients undergoing surveillance colonoscopy 18 months after their ESD, with no local recurrences detected. There were 27 patients who had metachronous polyps (75.0%), with 38.9% of these being adenomatous polyps.

Discussion

One of the major limitations in performing ESD is the lack of efficient tissue traction for submucosal exposure [8] [9]. The flexible endoscope was designed as a diagnostic tool, while the addition of a therapeutic channel only allows an ESD knife to be passed in coaxial alignment with the imaging system. The difficulty in manipulating ESD devices directly incurs a risk of perforation during submucosal dissection. In a recent prospective study on 54 ESD procedures, an adaptive traction device demonstrated safe and effective tissue traction [15]. Though simple traction devices can achieve good exposure of the dissection plane, it is difficult to reposition them or redirect the traction. We pioneered the development of the “Master And Slave Transluminal Endoscopic Robot” (MASTER) in 2008 [16]. In 2011, MASTER was used to perform gastric ESD in five patients [16]. This MASTER consisted of two robotic arms built on a conventional double-channel gastroscope, which had a major limitation in performance owing to design constraints. The whole system was redesigned to accommodate two new 4-mm robotic arms. These arms are ergonomically placed at the 6- and 9-o’clock positions to balance with best performance.

This is the first clinical trial confirming the feasibility and safety of colonic ESD using the EndoMaster EASE robotic system for colorectal ESD. The median speed of dissection, which was 20.6 mm²/minutes, would be considered as reasonably hastened by the robot, compared with a reported speed of dissection of 37.4 mm²/minutes using assistive traction devices [15]. Hayashi et al. compared 70 patients who underwent ESD for colorectal lesions >50 mm to 402 patients with lesions between 20 and 50 mm [17]. The mean (SD) ESD dissection time by a Japanese expert for >50-mm lesions was 121 (114) minutes, while the ESD dissection time for lesions of 20–50 mm was 54 (50) minutes. In comparison, the use of EndoMaster was able to achieve a relatively shorter time for ESD, especially for larger lesions. In the current study, the complete resection rate achieved for EndoMaster colonic ESD was 83.8%, which is comparable with the results achieved in most studies by conventional colonic ESD.

This study has several limitations. First, all operators were experienced in both ESD and robotic surgery. The results of this trial may not be applicable in other units. Second, the endoscopic platform was able to reach the right colon with careful manipulation in three patients. The routine use of EndoMaster for tumors in the right colon remains to be explored. Moreover, the flexible endoscopic platform should be technologically refined and downsized to allow better manipulability. The optical resolution is limited compared with high resolution endoscopes, therefore we performed diagnostic colonoscopy to locate the lesion, with submucosal injection performed before the robotic ESD procedure. General anesthesia was systematically used in this feasibility trial, but the performance of the robotic system in patients under deep sedation should be evaluated in the future. The recruitment of patients into the current trial could not be consecutive owing to a logistical limitation in the supply of robotic accessories for the trial, which may have led to selection bias.

In summary, this first clinical trial confirmed the safety and efficacy of performing colorectal ESD using EndoMaster. In future, further refinement of flexible robotics should allow better control and more precise manipulation to enhance the performance of ESD. A consecutive randomized prospective study, including cost-effectiveness and carbon footprint analysis, will be needed to accurately identify the place of this promising device in the near future. The next development in endoluminal robotic suturing will allow closure of luminal defects after endoscopic resection to enhance recovery [18].

Conflict of Interest

P.W.Y. Chiu has served as a scientific advisor for EndoMaster Pte Ltd. S.J. Phee and K.Y. Ho are founders of EndoMaster Pte Ltd, Singapore. H.C. Yip, S. Chu, S.M. Chan, L.H.S. Lau, R.S.Y. Tang, and S.S.M. Ng declare that they have no conflict of interest.

• References

• 1 Dragovich T. Colon Cancer. 2018 Accessed June 27, 2018 at: https://emedicine.medscape.com/article/277496-overview
• 2 Sumimoto K, Tanaka S, Shigita K. et al. Diagnostic performance of Japan NBI Expert Team classification for differentiation among noninvasive, superficially invasive, and deeply invasive colorectal neoplasia. Gastrointest Endosc 2017; 86: 700-709
• 3 Tanaka S, Kashida H, Saito Y. et al. Japan Gastroenterological Endoscopy Society guidelines for colorectal endoscopic submucosal dissection/endoscopic mucosal resection. Dig Endosc 2020; 32: 219-239 DOI: 10.1111/den.13545. (PMID: 31566804)
• 4 Pimentel-Nunes P, Libânio D, Bastiaansen BAJ. et al. Endoscopic submucosal dissection for superficial gastrointestinal lesions: European Society of Gastrointestinal Endoscopy (ESGE) Guideline – Update 2022. Endoscopy 2022; 54: 591-62 DOI: 10.1055/a-1811-7025. (PMID: 35523224)
• 5 Libânio D, Pimentel-Nunes P, Bastiaansen B. et al. Endoscopic submucosal dissection techniques and technology: European Society of Gastrointestinal Endoscopy (ESGE) Technical Review. Endoscopy 2023; 55: 361-389 DOI: 10.1055/a-2031-0874. (PMID: 36882090)
• 6 Arezzo A, Passera R, Marchese N. et al. Systematic review and meta-analysis of endoscopic submucosal dissection vs endoscopic mucosal resection for colorectal lesions. United European Gastroenterol J 2016; 4: 18-29 DOI: 10.1177/2050640615585470. (PMID: 26966519)
• 7 Lim XC, Nistala KRY, Ng CH. et al. Endoscopic submucosal dissection vs endoscopic mucosal resection for colorectal polyps: A meta-analysis and meta-regression with single arm analysis. World J Gastroenterol 2021; 27: 3925-3939
• 8 Jacques J, Schaefer M, Wallenhorst T. et al. Endoscopic en bloc versus piecemeal resection of large nonpedunculated colonic adenomas: a randomized comparative trial. Ann Intern Med 2024; 177: 29-38 DOI: 10.7326/M23-1812. (PMID: 38079634)
• 9 Ebigbo A, Probst A, Römmele C. et al. Step-up training for colorectal and gastric ESD and the challenge of ESD training in the proximal colon: results from a German Center. Endosc Int Open 2018; 6: E524-E530
• 10 Hori K, Uraoka T, Harada K. et al. Predictive factors for technically difficult endoscopic submucosal dissection in the colorectum. Endoscopy 2014; 46: 862-870 DOI: 10.1055/s-0034-1377205. (PMID: 25208032)
• 11 Saito Y, Abe S, Inoue H. et al. How to perform a high-quality endoscopic submucosal dissection. Gastroenterology 2021; 161: 405-410 DOI: 10.1053/j.gastro.2021.05.051. (PMID: 34089735)
• 12 Cui Y, Thompson CC, Chiu PWY. et al. Robotics in therapeutic endoscopy (with video). Gastrointest Endosc 2022; 96: 402-410
• 13 Phee SJ, Reddy N, Chiu PW. et al. Robot-assisted endoscopic submucosal dissection is effective in treating patients with early-stage gastric neoplasia. Clin Gastroenterol Hepatol 2012; 10: 1117-1121
• 14 Chiu PWY, Ho KY, Phee SJ. Colonic endoscopic submucosal dissection using a novel robotic system (with video). Gastrointest Endosc 2021; 93: 1172-1177 DOI: 10.1016/j.gie.2020.09.042. (PMID: 32991869)
• 15 Masgnaux LJ, Grimaldi J, Rivory J. et al. Endoscopic submucosal dissection assisted by adaptive traction: results of the first 54 procedures. Endoscopy 2024; 56: 205-211 DOI: 10.1055/a-2109-4350. (PMID: 37311544)
• 16 Phee SJ, Low SC, Sun ZL. et al. Robotic system for no-scar gastrointestinal surgery. Int J Med Robot 2008; 4: 15-22 DOI: 10.1002/rcs.179. (PMID: 18314917)
• 17 Hayashi Y, Shinozaki S, Sunada K. et al. Efficacy and safety of endoscopic submucosal dissection for superficial colorectal tumors more than 50 mm in diameter. Gastrointest Endosc 2016; 83: 602-607
• 18 Cao L, Li XG, Phan PT. et al. Sewing up the wounds: A robotic Suturing System for Flexible Endoscopy. IEEE Robot Autom Mag 2019; 27: 45-54

Publication History

Received: 25 February 2024

Accepted after revision: 06 September 2024

Accepted Manuscript online:
06 September 2024

Article published online:
06 November 2024