If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The first choice of treatment for rectal neuroendocrine tumors (NETs) ≤10 mm in size is endoscopic resection. However, because rectal NETs usually invade the submucosal layer, achieving R0 resection is difficult. Endoscopic submucosal dissection (ESD) has a high R0 resection rate, and underwater endoscopic mucosal resection (UEMR) was recently introduced to ensure a negative resection margin easily and safely. The aim of this study was to evaluate the efficacy and safety of UEMR versus ESD for rectal NETs ≤10 mm in size.
This retrospective observational study enrolled 115 patients with rectal NETs ≤10 mm in size who underwent ESD or UEMR between January 2015 and July 2019 at the National Cancer Center, Korea. The differences in R0 resection rate, adverse event rate, and procedure time between the ESD and UEMR groups were evaluated.
Of the 115 patients, 36 underwent UEMR and 79 underwent ESD. The R0 resection rate was not different between the UEMR and ESD groups (UEMR vs ESD, 86.1% vs 86.1%, P = .996). The procedure time was significantly shorter with UEMR (UEMR vs ESD, 5.8 ± 2.9 vs 26.6 ±13.4 minutes, P < .001). Two patients (2.5%, 2/79) experienced adverse events in the ESD group and but there were no adverse events in the UEMR group; however, this difference was not statistically significant.
UEMR is a safe and effective technique that should be considered when removing small rectal NETs. Further studies are warranted to define its role compared with ESD.
The use of endoscopic submucosal dissection (ESD) has gradually increased as a result of the high R0 resection rate in the treatment of rectal NETs; however, the technical burden, prolonged procedure time, and related adverse events are considered disadvantages.
During the UEMR procedure, as the air is released and the lumen is filled with water, the mucosa and submucosal tumor separate from the muscularis propria. This creates a pseudopedicle, making it easier to use a snare, as described by Binmoeller et al.
Recently, one multicenter randomized controlled trial reported a significantly increased R0 resection rate with UEMR for 10- to 20-mm sessile colorectal lesions in contrast to that achieved with conventional EMR.
However, few studies and case reports have evaluated the efficacy of UEMR as a treatment modality for rectal NETs ≤10 mm. A previous study reported the R0 resection rate of UEMR for rectal NETs to be 83%,
However, no studies have directly compared the outcomes of UEMR and ESD for rectal NETs. In this study, we aimed to evaluate the efficacy and safety of UEMR compared with that of ESD for rectal NETs.
We retrospectively analyzed patients who underwent UEMR or ESD for rectal NETs at the National Cancer Center in Korea between January 2015 and June 2019. We included NETs ≤10 mm in size as confirmed by pathologic examination. We excluded cases with evidence of lymph node metastasis or distant metastasis on CT or magnetic resonance imaging before the procedure and incomplete pathologic data on margin status. This study was approved by the Institutional Review Board of the National Cancer Center (approval number: NCC2019-0161). Informed consent was obtained from all patients.
Our treatment algorithm is shown in Figure 1. The typical endoscopic feature of a rectal NET is a hard, round, sessile elevation covered with yellow-colored mucosa. If a rectal NET with typical endoscopic features was detected during colonoscopy, the endoscopist evaluated its resectability based on the findings. For lesions ≤10 mm, the first choice of treatment was endoscopic resection. In our center, we do not perform biopsy before endoscopic resection of colorectal lesion owing to the risk of submucosal fibrosis, which renders the lesion endoscopically unresectable. Further, such small lesions may be invisible after biopsy. For lesions >10 mm, we performed a biopsy for pathologic confirmation of the NET. Thereafter, sequential surgical resection, such as transanal excision or low anterior resection, was performed. Patients who underwent a previous biopsy were transferred from outside clinics to our center for endoscopic resection.
The 4 endoscopists (K.S.H., D.S.S., B.C.K., and C.W.H.), who have more than 10 years’ experience in the field of therapeutic colonoscopy performed the endoscopic resection. During the study period (January 2015 to June 2019), 1 endoscopist (K.S.H.) performed UEMR and the others performed ESD for small rectal NETs.
The resection margin was classified as negative (defined as a margin that unequivocally contained no tumor cells) or positive (defined as a margin with definite tumor cell involvement or which could not be assessed because of piecemeal resection or tangential cutting). R0 resection was deﬁned if both lateral and deep resection margins were negative. R1 resection was defined if any margin was positive regardless of complete endoscopic resection (microscopic positive margin). The submucosal invasion depth was classified as sm1, sm2, and sm3. The cutoff limit between sm1 and sm2 was a submucosal depth of 1000 μm; a submucosal depth of >2000 μm was defined as sm3. The procedure time was measured from the time of identification of the lesion until hemostasis was achieved after resection. Adverse events were early bleeding (within 48 hours), delayed bleeding (after 48 hours), postpolypectomy syndrome, and perforation. Procedural bleeding was not included as an adverse event if immediate endoscopic hemostasis was successfully achieved during the procedure. Because it is known that a previous history of mechanical stimulation, including forceps biopsy or partial endoscopic resection, may affect endoscopic resectability,
subgroup analysis was performed for patients with/without a previous history of mechanical stimulation.
Endoscopic submucosal dissection
The ESD procedure is demonstrated in Video 1 (available online at www.giejournal.org). ESD was performed using a single- or dual- channel colonoscope (CF-H260AL or 2TQ260M and PCF-H290AL, CF-HQ290L; Olympus Optical Co, Tokyo, Japan) and a high-frequency generator with an automatically controlled system (VIO 300D; ERBE, Tuebingen, Germany). A transparent cap was attached to the tip of the scope to provide a constant view and countertraction during the procedure. To lift the tumor or polyp from the muscle layer, 10% glycerin and 5% fructose in 0.9% saline solution (CEROL) mixed with a small amount of indigo carmine and epinephrine or a solution of 0.4% sodium hyaluronate (Endo-Ease; Unimed Pharm Inc, Seoul, Korea) was injected into the submucosal layer around the lesion. The mucosa was then incised with a dual knife (KD-650; Olympus Optical Co), and the submucosal tissue beneath the lesion was dissected gradually from the muscle layer using a dual knife alone or in combination with a hook knife (KD-620LR; Olympus Optical Co). The patient’s position was changed to facilitate visualization of the tissue dissection plane or detachment from the rectal wall. The mucosal defect remained open. The ESD procedures were performed by 4 experienced endoscopists (Fig. 2).
Underwater endoscopic mucosal resection
The UEMR procedure is demonstrated in Video 2. This procedure was performed using a colonoscope (CF-H260AL, CF-HQ290L; Olympus Optical Co). The rectum was completely deflated, and the lumen was filled with distilled water using a water pump (OFP-2; Olympus) or water jet system to make the tumor float. Submucosal injection was not performed. The floating tumor was then snared and resected with a 12.5-mm snare (Short Throw; US Endoscopy, Mentor, Ohio, USA). Electrical cutting and coagulation were performed using an Endocut Q current (effect 2, cut duration 2, cut interval 6) and a soft coagulation current (effect 6, 80 W), respectively, which were generated using a high-frequency generator with an automatically controlled system (VIO 300D; ERBE, Tuebingen, Germany). The mucosal defect remained open. All UEMR procedures were performed by a single endoscopist (Fig. 3).
Bleeding control during the procedure
When bleeding occurred during the endoscopic procedure, the bleeding vessel was grasped with hot biopsy forceps (FD-1U-1, Olympus Ltd, Tokyo, Japan) and the bleeding was coagulated with a soft coagulation current (effect 6, 80 W). We prefer electric coagulation to clipping for bleeding control because (1) in the rectum, it is relatively easy to isolate and grasp the bleeding vessel; (2) the risk of perforation is lower in the rectum than in the colon; (3) during the ESD procedure clipping may disturb the dissection procedure because it can obscure the dissection field; and (4) electric coagulation is more cost-effective and takes less time than clipping.
Postresection management and follow-up
For patients with R0 resection, routine follow-up was recommended. The patients were followed up by colonoscopy, performed 3 to 6 months after resection and annually thereafter, and by annual CT scan of the abdomen and chest. For patients with R1 resection, more intensive colonoscopic examinations were added to the routine follow-up. For patients with vascular invasion, additional surgery was recommended.
Categorical variables were analyzed using Pearson's chi-squared test or Fisher's exact test, as appropriate. Continuous variables were analyzed using the Student t test or Wilcoxon rank-sum test after the normality test. P values <.05 were considered statistically significant, and all statistical analyses were performed using SAS version 9.3 (SAS Institute Inc, Cary, NC, USA).
Of the 115 patients, 36 patients underwent UEMR and 79 patients underwent ESD. All resected specimens showed well-differentiated (G1) NETs with Ki-67 expression of <2% according to the pathologic examination. The tumor did not invade the muscularis propria in any of the cases. The mean tumor level (distance from the anal verge) was 6.1 ± 2.3 cm in the UEMR group and 6.4 ± 2.6 cm in the ESD group. The tumor size was similar in both groups with a mean of 5 ± 2 mm. Rescue surgery was performed due to vascular invasion in 3 patients (9.7%) in the UEMR group and 11 patients (13.9%) in the ESD group (Table 1).
R0 resection rates were not significantly different between the UEMR and ESD groups (UEMR vs ESD, 86.1% vs 86.1%, P = .996). The procedure time was significantly shorter in the UEMR group than in the ESD group (UEMR vs ESD, mean 5.8 ± 2.9 minutes vs 26.6 ±13.4 minutes, P < .0001). The incidence of adverse events was 2.5% in the ESD group; no patient in the UEMR group experienced adverse events, but this difference was not statistically significant (Table 2).
Table 2Outcomes of underwater endoscopic mucosal resection and endoscopic submucosal dissection
Outcomes in patients with a previous history of mechanical stimulation
In the subgroup analysis of patients with a previous history of mechanical stimulation including forceps biopsy or partial resection (16 in the UEMR group, 38 in the ESD group), the R0 resection rate was 93.8% in the UEMR group and 86.8% in the ESD group; however, this difference was not statistically significant. Only 2 cases of adverse event were noted in the ESD subgroup with a previous history of mechanical stimulation (Table 3).
Table 3Outcomes of underwater endoscopic mucosal resection and endoscopic submucosal dissection in patients with a previous history of mechanical stimulation
The 16 patients with R1 resection were followed up without surgery (median, 19 months; range, 0-53 months), and no recurrence was noted (Supplementary Table 1, available online at www.giejournal.org). None of the clinicopathologic factors, including tumor size, tumor level, depth of invasion, and previous history of mechanical stimulation, were significantly associated with a positive resection margin (Supplementary Table 2, available online at www.giejournal.org).
Outcomes of cap-assisted endoscopic mucosal resection and precutting endoscopic mucosal resection
We also analyzed data on 77 patients who underwent cap-assisted EMR (EMR-C) or precutting EMR (EMR-P) for rectal NETs at the National Cancer Center in Korea between January 2001 and December 2010. Compared with the outcomes of UEMR, the prevalence of lateral margin positivity was significantly higher in the EMR-C/EMR-P group (UEMR vs EMR-C/EMR-P, 5.6% vs 20.8%, P = .039). R0 resection rates were higher in the UEMR group, but this difference was not statistically significant (UEMR vs EMR-C/EMR-P, 86.1% vs 72.7%, P = .115). The procedure time was significantly shorter in the UEMR group (UEMR vs. EMR-C/EMR-P, mean 5.8 ± 2.9 minutes vs 18.5 ± 11.2 minutes, P < .001) (Table 4).
In both EMR-C and EMR-P, submucosal injection and initial circular incision (precutting) before snaring is applied to achieve a clear margin. In EMR-C, the lesion is lifted by cap-assisted suction before snaring.
In our institution, EMR-C/EMR-P, accounting for a total of 77 cases between 2001 and 2010, were used mainly for rectal NETs ≤10 mm. ESD has been used in most cases since 2011 (total of 185 cases between 2011 and July, 2019).
ESD has the advantage of achieving en-bloc resection of the lesion while securing the resection margin; however, the difficult technique, prolonged procedure time, associated adverse events, and the number of instruments required, including a catheter and solution for submucosal injection, cap, dissection knife, and special electrosurgical unit, discourage its use.
Moreover, because most rectal NETs are small lesions located in the rectum, ESD for these tumors is relatively easy to perform and has a low adverse event rate in contrast to that with large lesions located in the colon. In this study, the mean procedure time for the 79 ESD procedures performed between 2015 and 2019 was 29.4 ± 14.9 minutes, and only 2 patients experienced adverse events, including 1 case of minor perforation and 1 delayed bleeding.
UEMR was recently introduced as a method to secure resection margins safely in sessile colorectal lesions less than 2 cm in size.
Another advantage of UEMR is that it does not require special equipment other than a snare. UEMR has been performed in our institution selectively for rectal NETs ≤10 mm in size since 2015, accounting for a total of 36 cases between 2015 and July, 2019. According to the present study, the R0 resection rates of UEMR and ESD did not differ statistically; however, the procedure time was significantly shorter in the UEMR group. Moreover, no adverse events were noted in the UEMR group. Taken together, our results suggest that the UEMR procedure is an effective and safe method for curative resection of rectal NETs.
In our previous studies, we reported that mechanical stimulation, such as biopsy, can induce submucosal fibrosis in endoscopically resectable colorectal lesions, which in turn might make endoscopic resection difficult.
In our study, although it was not statistically significant, UEMR showed a numerically higher R0 resection rate than ESD in the subgroup with rectal NET and a previous history of mechanical stimulation such as biopsy or partial resection. In addition, 2 cases of adverse events were noted only in the ESD group with a history of previous mechanical stimulation. Although this result was not statistically significant, it suggests that endoscopic resection for rectal NETs in patients with a history of previous mechanical stimulation is more likely to lead to adverse events; therefore, UEMR may be more beneficial in such cases. However, further studies with more cases are needed to validate this finding. Despite these advantages, there are patient-dependent limitations when performing the UEMR technique. If bowel preparation is insufficient, the lesion may not be identified well enough, and it may sometimes be necessary to change the position of the patient owing to the dependency of water on gravity.
EMR-L, although never performed in our center, is known to be an effective method for securing a deep resection margin of rectal NETs. According to previous reports, the R0 resection rate of EMR-L is higher than that of conventional EMR and is comparable with ESD for small rectal NETs.
Endoscopic full-thickness resection (EFTR) using special devices, such as a full-thickness resection device, has been introduced to replace conventional EMR in challenging situations, such as recurrent adenoma, nonlifting adenoma, and subepithelial lesions.
We have no experience in the EFTR technique, and no reports on implementation of the EFTR technique for rectal NETs are available. Nevertheless, the EFTR technique may be useful for rectal NET lesions in selected cases, and further research on this aspect is warranted.
Our study has some limitations that should be taken into account when interpreting the findings. First, the retrospective study design has a possibility of selection bias. However, we tried to overcome this limitation by including consecutive patients during the study period. Second, the number of UEMR cases was relatively small, and all procedures were performed at a single institute. Lastly, the follow-up duration was relatively short to obtain an accurate assessment of the recurrence rate.
Despite these limitations, our study is the first to directly compare the safety and efficacy of UEMR and ESD in the treatment of rectal NETs. Based on our findings, UEMR and ESD have a similar R0 resection rate and low adverse event rates; however, UEMR is associated with a significantly shorter procedure time than ESD. In conclusion, UEMR is a safe and effective technique that should be considered when removing small rectal NETs. Further studies are warranted to define its role compared with ESD.
This work was supported by a grant from the National Cancer Center, Korea (NCC-1810060 and 1810192-1).
Supplementary Table 1Details of patients with R1 resection
Mixed solution was injected into the submucosal layer around the lesion. The mucosa was then incised with a dual knife, and the submucosal tissue beneath the lesion was dissected gradually from the muscle layer using a dual knife alone or in combination with a hook knife.
The rectum was completely deflated, and the lumen was filled with distilled water to make the tumor float. The floating tumor was then snared and resected using electrical cutting and coagulation. When bleeding occurred from the resected surface, the bleeding vessel was grabbed and coagulated with hot biopsy forceps.
Endoscopic submucosal resection with a ligation device is an effective and safe treatment for carcinoid tumors in the lower rectum.
We read with great interest the article by Park et al.1 Those authors showed comparable R0 resection rates of underwater EMR (UEMR) and endoscopic submucosal dissection (ESD) for <10 mm rectal neuroendocrine tumors (NETs) but a shorter procedure time with UEMR.