Endoscopic Resection for Colorectal Tumors

Colorectal cancer (CRC) ranks third among the most prevalent cancers and is second leading cause of cancer-related deaths worldwide [1]. However, its mortality and morbidity can be reduced by colonoscopy and the subsequent removal of precursor lesions and early cancers [2]. Although endoscopic resection (ER) has been in existence for several decades since snare polypectomy was first established, ER techniques continue to evolve and improve. Similarly, innovative closure techniques and devices have been developed to address the resection of large and complex lesions using a more aggressive technique. In this review, we summarize the latest issues in ER techniques for colorectal tumors and recent advances in mucosal defect closure techniques. We also discuss the current status of ER of T1-CRC, which is increasing in number.

Before performing ER, a thorough evaluation of all colorectal lesions must be conducted in order to select the appropriate resection technique. Choosing the optimal resection technique depends on the tumor morphology, size, and predicted histology. In addition to white-light endoscopy, magnifying endoscopy using advanced endoscopic imaging, such as narrow-band imaging and chromoendoscopy, should be considered for histological prediction. The Japanese Narrow-band imaging Expert Team (JNET) classification [3] is commonly used to predict the histology and determine the presence of features associated with submucosal invasion. JNET type 2B often indicates intramucosal carcinoma (high-grade dysplasia) or shallow submucosal invasive cancer and such lesions should undergo en bloc resection. However, JNET type 2B sometimes includes deep submucosal invasive cancer along with type 3. Hence, both JNET types 2B and type 3 diagnosed with low confidence require further investigation of the Kudo pit pattern classification [4] using magnifying dye-based chromoendoscopy to determine if surgery is needed. The treatment strategy using the JNET and pit pattern classification is shown in Figure 1.

Cold polypectomy has gained increasing popularity in recent years. This procedure avoids thermal damage and can reduce the time and risk of adverse events (AEs), such as delayed bleeding [5‒7]. Cold snare polypectomy (CSP) is now the standard of care for the resection of nonpedunculated lesions <10 mm in size [5‒7]. Alternatively, cold forceps polypectomy, which uses jumbo forceps to bite the lesion, can be used for lesions ≤3 mm. It is well known that the complete resection rate for lesions <10 mm using CSP is non-inferior to that using hot snare polypectomy (HSP) [5‒7]. However, the resection depth in CSP is reportedly shallower than that in HSP or endoscopic mucosal resection (EMR). It is considered that the majority of the resected specimen does not contain submucosal tissue. Further, only part of the muscularis mucosae can be resected, indicating incomplete mucosal layer resection with CSP [7]. Therefore, CSP is inadequate for lesions suspected to be carcinomas, and a careful endoscopic diagnosis is recommended before CSP. Meanwhile, according to the latest European guidelines, CSP, including piecemeal CSP, is now recommended for sessile serrated lesions without dysplasia of any size [6]. This is because multiple studies have demonstrated similar efficacy for piecemeal CSP and HSP or EMR, with a better safety profile.

In the past, hot polypectomy was routinely performed along with EMR. Owing to the emergence of cold polypectomy, the use of hot forceps polypectomy has fallen out of favor, and CSP has increasingly replaced the use of HSP for non-pedunculated lesions <10 mm. However, HSP remains the standard method for removing pedunculated lesions [5‒7]. For lesions with large heads or stalks, postprocedural bleeding remains a risk, and prophylactic clipping or endoloop placement before resection may be considered. In cases where malignancy is suspected, endoscopists should be aware of the possibility of stalk invasion and resection should be performed as close as possible to the base of the stalk.

EMR has been a well-established technique for at least 2 decades. This method involves the submucosal injection of fluid beneath the lesion and resection with electrocautery. Globally, EMR is the most commonly used technique for lesions ≥10 mm [5, 6]. However, the en bloc resection rate of EMR for lesions ≥20 mm is reported to be 16%–48%, and en bloc EMR is generally limited to lesions <20 mm [6]. Several modified EMR techniques, such as tip-in EMR and pre-cutting EMR, have been developed to maximize the chances of en bloc resection. Indeed, their efficacy over conventional EMR in terms of en bloc resection and recurrence rates has been reported [8]. In tip-in EMR or anchored snare-tip EMR, a small mucosal incision is created on the oral side of the lesion after submucosal injection and the tip of the snare is subsequently anchored into the mucosal incision site, allowing the entire lesion to be grasped, while the tip of the snare remains anchored. Pre-cutting EMR is a technique in which snaring is performed after a circumferential mucosal incision using the tip of a snare or an endoscopic submucosal dissection (ESD) knife.

Recently, underwater EMR (U-EMR) has emerged as an alternative to conventional EMR. U-EMR, first described in 2012, eliminates submucosal injection by using the buoyancy effect of water immersion in the mucosa and submucosa, which leads to separation from the muscularis propria. In addition to the buoyancy effect, water immersion minimizes luminal distension, flexural angulation, and loop formation, thereby enabling improved endoscopic maneuverability. U-EMR has been shown to be superior to conventional EMR for en bloc and R0 resection of lesions with various sizes, without increasing the risk of AEs [5, 6]. Meanwhile, the submucosal excision depth of U-EMR has been discussed compared to conventional EMR. In a post hoc study of a multicenter randomized controlled trial, U-EMR enabled obtaining a decent thickness of submucosal tissue [9]. U-EMR is likely to be sufficient for the resection of intramucosal neoplasia. Still, there is room for debate regarding whether U-EMR allows for the resection of the submucosal tissue with an adequate margin in T1-CRC lesions. Thereby, a modified U-EMR technique combining submucosal injection (underwater injection EMR) has been described [10]. This technique aims to achieve deeper submucosal excision by taking advantage of both submucosal injection and the underwater environment for lesions that are worrisome for T1-CRC. The detailed outcomes of this modified technique remain unknown and therefore further studies are required. A good indication of U-EMR is reported to be residual/recurrent lesions after ER [5, 6]. Submucosal injection during conventional EMR for these lesions usually makes it difficult to capture the lesion with a snare because only the surrounding mucosa elevates around the non-lifting lesion. Nonetheless, ESD would likely be preferred to securely obtain an en bloc resection for larger or cancerous lesions.

ESD, a technique initially developed for early gastric cancer in Japan, allows en bloc resection regardless of tumor size and enables accurate pathological evaluation (Fig. 2). In Japan, ESD has been introduced more than 20 years ago and is now considered a mature and safe resection technique that can be performed throughout the gastrointestinal tract, including the colon, with the aid of advanced endoscopic skills, tools, and accessories, such as traction devices. In 2022, a large-scale, multicenter, prospective cohort study by the Colorectal ESD Activation Team of Japan (CREATE-J) demonstrated excellent short- and long-term outcomes of colorectal ESD for lesions ≥20 mm [11, 12]. ESD attained en bloc resection rate of 97% and R0 resection rate of 90.4%, with only 0.5% of patients requiring surgery for AEs. In addition, the local recurrence rate was 0.5% (median follow-up period: 46.0 months). In the West, piecemeal EMR is still commonly used for large lesions; however, ESD is increasingly being performed. While studies before 2017 have shown lower en bloc and R0 resection rates of ESD in Western countries, a recent meta-analysis reported an improvement in outcomes during the last 5 years [13].

Several studies have demonstrated that ESD has higher en bloc resection rates and lower local recurrence rates for lesions ≥20 mm, at the cost of longer procedure times and higher rates of AEs (although most can be managed conservatively) [5, 6, 14]. A recent large French randomized controlled trial [15] also showed a significantly lower 6-month recurrence rate for laterally spreading tumors (LSTs) ≥25 mm with ESD compared to that of piecemeal EMR (0.6% vs. 5.1%). ESD exhibited a higher rate of AEs (35.6% vs. 24.5%), but patients requiring surgery for complications were few and similar to those of piecemeal EMR (1.1% vs. 0%). From an economic standpoint, an Australian group suggested that a selective ESD for high-risk lesions rather than universal ESD was the most cost-effective strategy in their analysis of wide-field EMR versus ESD for all LSTs >20 mm [16]. However, recent studies from Japan and France have shown the potentially favorable cost-effectiveness of ESD for all LSTs >20 mm when compared to piecemeal EMR [17, 18]. According to these studies, superior cost-effectiveness might be observed for ESD over time, driven by a lower recurrence rate and reduced need for surveillance colonoscopy.

Although colorectal ESD requires a demanding technique with a long learning curve compared to esophageal and gastric ESD, a wide variety of dissection techniques and devices, such as lifting agents, and electrosurgical devices have been developed to facilitate the procedure. Improvements in dissection techniques mainly encompass traction-assisted method, pocket creation method (PCM), and underwater ESD/water pressure method (WPM). The application of traction devices augments the stability of the mucosal flap and facilitates submucosal dissection by increasing the separation of the planes exerted by traction. This traction may be applied with external traction or internal devices and commonly used are clips with thread, S-O clips, and loop-attached rubber band traction. Traction-assisted ESD is considered more efficient in procedure time and safer in perforation rates than conventional ESD, with comparable en bloc resection rates [19]. PCM is performed by entering and dissecting the submucosa through a minimal mucosal incision and creating a pocket beneath the lesion before opening the margins. The advantage of the PCM is its improved scope stability and angle of access to the submucosa, with less impact on respiratory movement. Compared with conventional ESD, PCM exhibits significantly higher R0 and en bloc resection rates, along with shorter procedure times and lower rates of AEs [19]. In underwater ESD/WPM, the underwater environment or the use of water pressure via the water jet creates traction and facilitates easier submucosal dissection. It is expected that underwater ESD/WPM to be particularly useful in challenging cases, such as lesions with submucosal fibrosis or poor endoscope maneuverability [19].

The necessity for prophylactic closure for mucosal defects after colorectal ESD remains controversial. However, there are suggestions that the closure after ESD defects is associated with a lower incidence of delayed bleeding or AEs [19, 20]. Recent evidence also shows that prophylactic closure prevents delayed bleeding after EMR in proximal lesions >20 mm [5, 6].

Closure techniques and devices for mucosal defects after ER mainly include through-the-scope clips (TTSCs), over-the-scope clips (OTSCs), and endoscopic suturing. TTSCs are commonly used and several methods using TTSCs for closing large mucosal defects exist. However, many of them result in the closure of only the mucosal layer, creating a dead space beneath the clips (submucosal dead space). Several techniques have recently been developed to overcome this limitation. One is the reopenable-clip over-the-line method, which uses a nylon line and a hole in the tooth of reopenable clip (SureClip, Micro-Tech, Nanjing, China) [21]. In this method, the clip with line is repeatedly placed at the edge of the mucosal defect and nearby muscle layer and gathered by pulling the free end of the line. Instead of using reopenable clips, reloadable clips (EZ Clip, Olympus Medical Systems, Tokyo, Japan) used in the same manner have also been described, which are easier, cheaper, and ecologically friendly (Easy and Eco clip over-the-line method) [22]. Another technique is the origami method, which is a double-layered suturing technique that involves closing not only the mucosal layer but also the muscle layer at the center of the mucosal defect [23]. A novel TTS dual-action tissue clip (Micro-Tech Endoscopy, Ann Arbor, MI, USA) with twin arms was recently developed [24]. This device has a center post-flanked by two arms, which can operate independently, with an overall maximum opening width of 15 mm, and facilitate tissue approximation. The OTSCs (Ovesco Endoscopy, Tubingen, Germany; Padlock Clip, STERIS, Ireland) consists of biocompatible nitinol clips with a bear-claw design, which is mounted on a distal attachment on the tip of an endoscope [20]. This device allows for full-thickness, robust closure; however, accurate placement is required, and misplacement can limit subsequent rescue. Endoscopic suturing includes the OverStitch system (Apollo Endosurgery, Austin, TX, USA) [24] and the endoscopic hand suturing system [20]. The OverStitch consists of a curved needle driver attached to the tip of the endoscope, a catheter-based suture anchor, and an actuating handle attached near the endoscopic controls. The endoscopic hand suturing uses a flexible through-the-scope needle holder (Olympus Medical Systems) and a V-Loc 180 absorbable barbed suture for its suturing. These suturing devices provide robust closure but are technically demanding and require additional training and expertise. In addition, their use in the proximal colon or narrow lumen is limited. A common drawback of OTSCs and suturing devices is that they both require endoscope withdrawal and reinsertion. To overcome this limitation, X-Tack (Apollo Endosurgery), a TTS suture-based device system, was recently developed [20, 24]. This device closes defects by repeatedly placing a thread with an anchor into the mucosa to fix it in place and implanting the next anchor into the mucosa at the edge of the defect. Although its feasibility and safety have been demonstrated, the robustness of the closure requires further investigation. In summary, each closure technique has its own benefits and limitations, and its cost-effectiveness must also be considered.

In patients with T1-CRC, lymph node metastasis (LNM) occurs in approximately 10% of cases (25), and ER of T1-CRC is considered non-curative depending on the following four pathological features of the resected specimen: depth of submucosal invasion, lymphovascular invasion, differentiation, and tumor budding [5, 6, 25]. When deemed non-curative, additional surgery with LNM dissection must be considered. However, LNM occurs in only 11%–15% of the patients who undergo additional surgery [25]. Consequently, surgery may be overtreatment for a subset of patients with T1-CRC, and optimal management must strike the balance between oncological safety and minimizing treatment-induced morbidity and mortality. In rectal lesions, surgery may reduce the quality of life because of the need for stoma creation or impaired anal function. In a phase II trial, adjuvant concurrent capecitabine-based chemoradiotherapy in patients with high-risk T1 lower rectal cancer after local excision showed a 5-year disease-free survival of 94.2% (53/57 patients) [26]. Given these promising results, the Japan Clinical Oncology Group is currently investigating the efficacy and safety of adjuvant chemoradiotherapy for patients with high-risk rectal T1-CRC after local resection (including ER and surgical local resection) in a prospective, multicenter, single-arm confirmatory trial (RESCUE study). This less invasive strategy including ER can be the standard of care for rectal T1-CRC depending on the results of this trial.

The number of ERs for T1-CRC, including those intended for total excisional biopsy, has been increasing owing to the aging society and improvements in ER skills. In fact, the oncological outcomes of additional surgery for T1-CRC are not impaired by prior ER [25]. Thus, the initial ER of suspected T1-CRC can be a viable option when the specimen is resected with R0, and the pathological findings can be accurately assessed. Recently, a new nomogram was developed to accurately predict the risk of LNM in T1-CRC after ER, based on a nationwide multi-institutional study in Japan [27]. This nomogram can be helpful in personalized approaches and can further expand the potential of ER as a total excisional biopsy for T1-CRC. Although EMR and ESD can be used for resection, the vertical margin in deep submucosal invasive CRC (T1b-CRC) may be problematic depending on the case. The following ER techniques have emerged as options for obtaining abundant submucosal tissue and securing a free vertical margin.

Per-anal endoscopic myectomy or endoscopic intermuscular dissection is a recently developed technique that involves dissection of the outer longitudinal and inner circular muscle layers. This technique was first established for rectal lesions with severe fibrosis and muscle retraction signs [28] and is now proposed for the resection of T1b rectal cancers. In a recent prospective cohort study conducted in the Netherlands [29], 67 patients with suspected T1b rectal cancer were treated (median lesion size: 25 mm). The technical success rate (defined as en bloc resection with successful intermuscular dissection) was 96% and R0 resection rate was 81%. Endoscopic intermuscular dissection also enabled a high R0 resection rate, even for pathological T1b cancers (36/40 cases, 90%). Only minor AEs were seen in 12% of the cases in this study although further studies are required to confirm the safety and feasibility of this procedure.

In Western countries, endoscopic full-thickness resection is indicated for lesions that are not amenable to conventional resection techniques, such as those with scarring from previous incomplete resection, submucosal tumors, and lesions at difficult anatomical locations (appendiceal orifice and diverticulum) [30]. This technique has attracted attention as a potentially valid diagnostic and therapeutic option also for T1-CRC. A full-thickness specimen is likely to be optimal for the evaluation of submucosal infiltration depth and vertical margin. The most commonly used device is a full-thickness resection device (Ovesco Endoscopy). In a Dutch multicenter, retrospective study including 330 endoscopic full-thickness resections for T1-CRC, technical success (defined as macroscopic complete en bloc resection) was 87.0%, and R0 resection rate was 85.6% [30]. Accurate pathological risk assessment was possible in almost all cases (99.3%). The R0 rates also did not differ significantly between primary treatment (82.0%) and secondary treatment after previous incomplete ER (88.0%). However, the full-thickness resection device device has limitations in terms of the size and is generally accepted for smaller lesions of 15–20 mm.

Over the past decades, ER techniques have been refined and optimized due to improvements in endoscopic techniques and devices. Further innovations, such as developments in robot-assisted endoscopy, may enable easier and safer procedures, with the potential to extend the therapeutic options.

The authors declare no conflict of interest associated with this manuscript.

This work was supported in part by the National Cancer Center Research and Development Fund (2023-A-15).

Y.H. conceptualized the article and drafted the article. N.T. and Y.S. contributed to the article with their expertise and reflective improvements. All authors approved the final version for submission.