Endoscopic resection of gastrointestinal tumors: Training levels and professional roles explored

Abstract

In this editorial, we provide commentary on a recently published study by Zhao et al in the World Journal of Gastrointestinal Oncology. The study discusses the clinical characteristics of patients undergoing endoscopic resection for gastric cancers. We feel it is important to engage our endoscopy community in a discussion on the current evidence in the literature on the necessary number of cases for training in endoluminal surgery techniques, particularly endoscopic submucosal dissection. This includes the latest recommendations from the European Society of Gastrointestinal Endoscopy, as well as a summary of key studies on the learning curve for these techniques. Additionally, we explore the impact of an endoscopist’s specialty on endoscopy outcomes, drawing from current evidence in the literature to shape our perspective in this evolving field.

Key Words: Learning curve; Endoluminal surgery; Endoscopic submucosal dissection; Gastroenterologist; Surgeons; Endoscopic mucosal resection

Core Tip: In this editorial, we provide an overview of existing societies’ and associations’ guidelines and data regarding the requisite training standards for conducting endoluminal endoscopic procedures for resection of gastrointestinal tumors. We also emphasize the superior outcomes of endoscopies carried out by gastroenterologists in comparison to those performed by surgeons.

INTRODUCTION

In a recent retrospective study, Zhao et al[1] discussed the characteristics of patients undergoing endoscopic resection for early gastric cancer. In this editorial, we will comment on the existing literature on the training levels and learning curve for endoscopic resection of gastrointestinal tumors. We will share our insights based on our experience, new emerging techniques, and evidence, as well as discuss the debate on whether the endoscopist for the resection of these lesions should be a surgeon or a gastroenterologist.

Endoscopic resection is the primary approach for managing the majority of gastrointestinal polyps. The evolution of this technique has progressed from basic polypectomy to endoscopic mucosal resection (EMR)[2,3], endoscopic submucosal dissection (ESD)[4], and advanced to endoscopic full-thickness resection[5], with some experts emphasizing the growing importance of endoscopic powered resection[6]. A comparative analysis of various endoscopic techniques is provided in Table 1.

Table 1 A comparative analysis of main endoscopic resection techniques.

Technique	Indications	Pros	Cons
Endoscopic mucosal resection	Small, superficial lesions from the mucosal layer (ex. early-stage cancers and adenomas)	Preserves tissue, maintains function, simple technique	Limited size, risk of incomplete resection
Endoscopic submucosal dissection	Larger lesions reaching the submucosal layer	En bloc resection, enables complete resection	More complex, higher complication rate
Endoscopic full-thickness resection	Invasive lesions involving the full-thickness, reaching into the muscularis layer	Minimally invasive with lower recovery time compared to surgery	Complex procedure, significant perforation risk

The choice of resection technique depends on several factors, including the size, location, endoscopic appearance, and histopathological characteristics of the lesion[7]. Additionally, the endoscopist’s familiarity with some techniques may influence the decision and subsequent outcome in certain cases. Similar to other procedures, endoscopic resection entails a learning curve that varies across different resection modalities, with some requiring a substantial number of cases and an extended learning period[8].

Until 2019[9], there were no established guidelines regarding the necessary training and case volume for endoscopists to achieve proficiency in advanced techniques such as EMR and ESD, despite numerous studies exploring the learning curve associated with these procedures.

The incorporation of these techniques into the endoscopy field has been groundbreaking, particularly in reducing the necessity for surgery in nearly 87% of cases involving benign lesions that were initially recommended for surgical intervention, as reported by Grimm[10]. Endoscopic resection has proven superior to surgery due to its lower occurrence of adverse events and mortality despite the higher recurrence rate in surgical patients. Mun and Wagh[11] reported that the overall survival rates are comparable in both groups.

CURRENT EVIDENCE ON THE TRAINING OF ENDOSCOPIC RESECTION

European Society of Gastrointestinal Endoscopy guidelines for ESD training

In 2019, the European Society of Gastrointestinal Endoscopy (ESGE) released its latest ESD training recommendation[9]. The guidelines specify that only fully-trained endoscopists with expertise in advanced diagnostic endoscopies should participate in training. Additionally, trainees must demonstrate proficiency in EMR and managing complications like perforations and bleeding before beginning training. Trainees should possess a comprehensive understanding of the ESD concept and procedures before and during the training period. This can be accomplished by participating in live ESD workshops prior to and during training. Additionally, trainees should observe at least 20 cases under the guidance of an ESD expert at a tertiary referral center. They should assist in five cases, preferably within a concentrated timeframe.

It is recommended to begin supervised training on animal or ex vivo models, aiming for at least 20 cases of ESD with successful en bloc resection in at least eight of the last 10 cases, without any instances of perforation. When transitioning to human cases, it is advisable to start with small lesions in the antrum or rectum for optimal results in the first 20 cases, while the first 10 cases should be done under supervision.

The endoscopist must possess comprehensive knowledge of the entire ESD setting, including anatomy, endoscopic procedures, instrumentation, and electrosurgical devices. It is crucial to accurately assess the potential curative nature of ESD for each lesion based on the examination and evaluate the risk-to-benefit ratio compared to alternative curative modalities. Additionally, the ability to differentiate between low, local, and high-risk resections is essential for endoscopists embarking on ESD procedures. It is recommended that endoscopists initially perform cases in established centers equipped to handle complications, with access to ward admission and a proficient surgical team, as well as the presence of an expert gastrointestinal pathologist.

According to the recommendations, maintaining ESD skills requires the completion of 25 cases per year. These cases should demonstrate an en bloc resection rate exceeding 90%, a perforation rate below 3%, and a need for surgical intervention to address less than 1% of complications.

ESGE guidelines for EMR training

ESGE has recently issued its recommendations for EMR training[12]. As outlined in ESD, trainees should demonstrate proficiency in diagnostic procedures, basic polypectomy of pedunculated polyps, the ability to manage complications, and a solid understanding of the concept of EMR and interpretation of lesions. Trainees should also be capable of selecting the appropriate modality, conducting thorough assessments, and effectively ruling out cancer.

It is important to initiate training with in vivo models while being supervised by an experienced trainer. Throughout the training process, the trainee should acquire the ability to discover and manage complications such as bleeding and post-resection defects. Moreover, they should also develop non-technical endoscopic skills, including effectively interacting with the endoscopy team, establishing clear communication with the patient, ensuring that the treatment plan is thoroughly explained, and informing the patient about early detection of any signs of adverse events. The endoscopist is expected to complete 30 cases of EMR before working independently.

The reality check: ESD learning curves in practice

Many studies have analyzed the learning curve for ESD in different settings. These studies have also explored the learning curve for tutored and untutored ESD procedures. The summarized findings of the largest cohorts can be found in Table 2[4,8,13-52].

Table 2 Summary of selected studies on the endoscopic submucosal or mucosal resection learning curve.

Ref.	Modality	Number of cases	Tutored	Number of operators	Adverse events	En bloc rates	R0 rates	Number of cases when competency achieved	Notes
Hadjinicolaou et al[13]	ESD	60	No	1	Total: 16.7%, including perforation 13.3% and bleeding 3.3%	93%	80.7%	40-50
Khalaf et al[14]	ESD	503	No	1	Total: 3.2%, including two intraprocedural and 16 post-procedures	81.9%	71.1%	Between 250 and 350 procedures
Aliaga Ramos et al[15]	ESD	98	No	1	Total: 4%, including three cases of perforations (3%) and one case of bleeding (1%)	95.9%	80.6%	NA
Zorron Cheng Tao et al[16]	ESD	514	Yes	26	Perforation (experts = 3.7%; trainees = 6.9%), delayed bleeding (experts = 2.9%; trainees = 4.4%)	Experts = 95.6%; trainees = 94.7%	Experts = 85.7%; trainees = 83.6%	20-50	Colorectal ESD
Ramos-Zabala et al[17]	ESD	80	No	1	Perforations 7.5% and bleeding 3.7%	98.7%	75%	80	Colorectal ESD
Boda et al[18]	ESD	420	Yes	31	Total: 3%, including intraoperative perforation 2% and delayed bleeding 1%	94%	85%	20	Colorectal ESD
Spychalski et al[19]	ESD	298	Yes	1	Total: 16.78%, perforation 8.4%, bleeding 3.36%	NA	84.9%	36	Colorectal ESD
Zhang et al[8]	ESD	460	Yes	1	Total: 3.5%, including perforation 0.5%, delayed bleeding 2.1%, and others 0.7%	90%	80%	250	The learning thresholds in the study refer to proficiency rather than competence (which requires less stringent benchmarks such as en bloc resection > 80% rather than > 90%)
Rönnow et al[20]	ESD	301	Yes	1	Complications occurred in 24 patients (8%) divided into 12 immediate perforations, five delayed perforations, one immediate bleeding and six delayed bleedings. Six patients (2%), all with proximal lesions, had emergency surgery	80%	69%	120	Colorectal ESD was described in this study. Proficiency was used instead of competency
Yamamoto et al[21]	ESD	71	Yes	2	Total: 7.0% including two perforations, two post-operative bleeding, and one delayed perforation	80.3%	70.4%	20 for the first endoscopist and 51 for the second endoscopist	Colorectal ESD
Spychalski et al[22]	ESD	228	Yes	1	Total: 16.67%, perforation 7.9%, bleeding 4.4%	79.39%	86%	76	Colorectal ESD
Chong et al[23]	ESD	71	Yes	1	Perforation 15.5%	81.2%	58%	35	Colorectal ESD
Ohata et al[24]	ESD	80	Yes	2	Perforation 3%, delayed bleeding 3%	98%	100%	20	Colorectal ESD
Jacques et al[25]	ESD	62	Yes	2	Perforation 6.4%, delayed bleeding 9.6%	100%	85.5%	NA	Colorectal ESD
Yoshida et al[26]	ESD	334	Yes	7	Perforation 1%	100%	95%	30
Aslan et al[27]	ESD	95	No	1	Total: 3% including Perforation in one patient (1%), and bleeding in two patients (2%) after the procedure	93%	92%	NA
Jeon et al[28]	ESD	93	No	1	Overt perforation (2.15%) and micro perforation (4.3%)	89.25%	83.87%	50	Colorectal ESD
Pioche et al[29]	ESD	64	Yes	2	Perforation 6.2%	89.1%	92.2%	16	Colorectal ESD, animal study
Shiga et al[30]	ESD	120	Yes	2	Perforation 6.7%, postoperative bleeding 1.7%	94.2%	80%	40	Colorectal ESD
Spychalski and Dziki[31]	ESD	70	Yes	1	Perforation 5.7%, bleeding 5.7%	66%	96%	50	Colorectal ESD
Agapov and Dvoinikova[32]	ESD	44	No	1	Perforation 11.4%	84.1%	84.1%	NA	Colorectal ESD
Berr et al[33]	ESD	50	No	1	Complications were 2 bleedings (4%) and 7 perforations (14%), 5 clipped and 2 (4%) operated	76%	63%	22
Białek et al[34]	ESD	58	No	1	Total: 5.7%, including one case (1.9%) with moderate grade bleeding (delayed bleeding)	86.5%	81.1%	25	Colorectal ESD
Hong et al[35]	ESD	112	No	1	The bleeding rates of each quarter were 71% (2/28), 4.0% (1/25), 10.7% (3/28), and 3.3% (1/30), respectively. Only one perforation was reported in the last quarter	En bloc resection rates were 964% (27/28), 100% (25/25), 96.4% (27/28), and 96.7% (29/30) in each quarter	Complete resection rates were 96.4%, 100%, 100%, and 100%	60	Gastric ESD
Rahmi et al[36]	ESD	45	Yes	91	Perforation 18%, delayed bleeding 13%	64%	53%	40	Colorectal ESD
Shiga et al[37]	ESD	80	Yes	4	Perforation 7.5%, postoperative bleeding 3.8%	93.8%	75 %	40	Colorectal ESD
Hsu et al[38]	ESD	50	Yes	1	Perforation 6%	86%	86%	NA	Colorectal ESD
Iacopini et al[39]	ESD	60	Yes	1	Perforation 5%	68%	78%	20	Colorectal ESD
Inada et al[40]	ESD	518	No	6	Perforation 3.8%, postoperative hemorrhage 23%	91.5%	NA	NA	Colorectal ESD
Niimi et al[41]	ESD	115	Yes	2	Total: 7.7% including bleeding 1.1.%, and perforation 6.6%	Endoscopist A: 92.4%; endoscopist B: 95.7%	Endoscopist A: 73.9%; endoscopist B: 65.2%	23	Colorectal ESD
Ohata et al[42]	ESD	182	Yes	4	Perforation rate for the trainees ranged 2%-6.3%	The en bloc resection rate was 100% for all trainees	The R0 resection rates for trainees A, B, C and D were 94.0%, 93.8%, 94.0% and 98.0%, respectively	NA	Colorectal ESD
Probst et al[43]	ESD	82	No	2	Perforation (1.3%) and bleeding (7.9%)	81.6%	69.7%	25
Yamamoto et al[44]	ESD	1430	Yes	13	Total: 4% including bleeding 3.2% and perforation 0.8%	NA	95.5%	20
Sakamoto et al[45]	ESD	101	Yes	2	Perforation 2%	100%	29.4%	30	Colorectal ESD
Hotta et al[46]	ESD	115	Yes	1	The perforation rates for the first, second and third periods were 12.5% (5/40), 5% (2/40) and 5% (2/40), respectively	93%	85%	80	Colonic ESD
Saito et al[47]	ESD	1090	No	101	Total: 6.5% including perforation 5.2% and postoperative bleeding 1.5%	88%	NA	NA	Colorectal ESD
Saito et al[48]	ESD	198	No	NA	Perforation 5%, and postoperative bleeding 2 %	84%	83%	NA	Colorectal ESD
Fujishiro et al[4]	ESD	186	No	2	Perforation 6% and bleeding 1%	91.5%	71%	NA	Colorectal ESD
Tamegai et al[49]	ESD	70	No	NA	Perforation 1.4%	98.6%	95.6%	NA	Colorectal ESD
Tanaka et al[50]	ESD	70	No	2	Perforation 10%	NA	80%	NA	Colorectal ESD
Hurlstone et al[51]	ESD	42	No	1	Perforation 1.4%	78.6%	74%	NA	Colorectal ESD
Choi et al[52]	EMR	80	No	1	Three perforations were reported for the first 20 cases (15%), whereas only one was reported for the next 60 (1.7%). The frequency of major bleeding was 3.75% (3/80) and was not related to quartile	En bloc resection rates were 55%, 45%, 85%, and 85%	Complete resection rates were 65%, 60%, 90%, and 85%	40

¹Endoscopic submucosal dissection data over the numbered centers but trainee numbers not defined.

ESD: Endoscopic submucosal dissection; EMR: Endoscopic mucosal resection; NA: Not appliable.

PROFESSIONAL ROLES OF ENDOSCOPISTS

The quality standards set by the British Society of Gastroenterology do not specifically address the specialization of the endoscopist[53]. However, there has been an ongoing debate regarding which specialty yields the best results in endoscopies.

Various metrics were used to assess the outcomes of different specialties in diagnostic endoscopies, particularly colonoscopies, such as cecal and terminal ileal intubation rates and polyp detection rates.

In a meta-analysis conducted by Mazurek et al[54], 36 studies involving a total of 3500832 colonoscopy results were examined. The authors found that endoscopy quality was significantly influenced by the endoscopist’s specialty. Gastroenterologists had a higher adenoma detection rate and cecal intubation rate compared to surgeons. These findings contrast with the recommendations of the Society of American Gastrointestinal and Endoscopic Surgeons, which suggested that specialty does not impact the outcomes of gastrointestinal endoscopic procedures[55]. The findings were derived from studies indicating that improved outcomes in adenoma detection rate are linked to accreditation and a high volume of annual endoscopies rather than the specific specialty[56].

Despite all of the previous studies, there is a lack of studies comparing the outcomes of different specialties performing third-space endoscopy procedures, particularly ESD.

In a review article by Wong and Sommovilla[57], the authors emphasized the significance of surgeons acquiring skills in endoluminal surgery. Many surgeons are hesitant to learn ESD due to the steep learning curve, the high cost of specialized tools, and the limited opportunities for fellowships in these techniques, leading to intense competition. Nonetheless, the authors recommended that all surgeons learn endoluminal surgery techniques to decrease the incidence of unnecessary colectomies. The ultimate goal is to collaborate with gastroenterologists to improve patient care rather than to engage in competition.

CONCLUSION

Endoscopic resection is now widely recognized as the primary approach for managing gastrointestinal polyps and, more recently, early cancers. The recent study by Zhao et al[1] examined patient characteristics and analyzed the adverse events following endoscopic resection for early gastric cancers. However, it is important to also consider the qualifications of the endoscopist in such procedures. In this editorial, we summarized the current evidence regarding the necessary training for performing ESD for gastrointestinal polyps. To the best of our knowledge, the training levels required to manage early cancer or complex polyps have not been clearly outlined in guidelines or large cohorts. Future longitudinal studies are needed to highlight training levels required to manage early cancer or complex polyps. We believe it is important to address the current evidence based on our experience. While we support the ESGE guidelines regarding the potential usefulness of ex-vivo models, they are not a substitute for live animal training, which is essential before performing procedures on humans. As for the recommendation of a minimum of 20 cases, we find it challenging to set a specific number because the experience varies among endoscopists. Drawing from our experience, we argue that incorporating other third space techniques before or during training significantly enhances the learning curve, providing valuable experience and confidence for the endoscopist and a better understanding of the layers of the gastrointestinal tract and tunneling techniques. We propose a higher minimum number of annual cases than suggested by the ESGE and suggest at least 60 cases annually to maintain proficiency. However, we acknowledge that achieving this number may be unfeasible in some centers with low case volumes. In instances where there is a high likelihood or suspicion of early-stage cancer, it is imperative that only proficient endoscopists undertake the resection of these lesions. This also applies to circumferential lesions and complex polyps. The existing evidence indicates that the outcomes achieved by gastroenterologists and surgeons in terms of endoscopy specialization are largely comparable, with a slight advantage for gastroenterologists. However, we emphasize the importance of collaboration and teamwork. Surgeons bring valuable technical skills and expertise in managing potential adverse events, while gastroenterologists demonstrate better outcomes in preventing colorectal cancer through higher polyp detection rates and access to additional training opportunities for endoluminal surgery. This could help prevent unnecessary colectomy procedures, providing better care for patients with adenomatous lesions or early cancers. Although we regard this editorial as one of the few contributions to the endoscopic community that discusses that sensitive issue, it is essential to acknowledge certain limitations. This manuscript may not be enough to form an opinion since all authors are gastroenterologists, and the existing literature is deficient in comparative studies regarding the outcomes of endoscopic resection performed by gastroenterologists vs those performed by surgeons.