Introduction

Recent advances in high definition endoscopy have increased the diagnostic accuracy for superficial non- ampullary duodenal epithelial tumors (SNADETs). Appropriate endoscopic categorization of these tumors as low grade adenoma (LGA)/high grade adenoma (HGA)/adenocarcinoma (AC) is of extreme importance to plan appropriate management. We report the accuracy and feasibility of a novel ‘Simple Scoring System’ (SSS) for SNADETs to differentiate LGA from HGA/AC.​

Aims & Methods

The aim of this prospective study was to evaluate the accuracy of SSS for diagnosis of SNADETs and to assess the feasibility of its understanding and usage among beginner and expert gastroenterologists. SSS includes four sections which are evaluated and scored: 1. Tumor diameter (10-20mm: +1, >20mm: +2); 2. Color on white light imaging (redness: +1); the presence/absence of irregular surface pattern by magnifying endoscopy using narrow band imaging (ME-NBI) (Present: +1); 4. The presence/absence of irregular vascular pattern on ME- NBI (Present: +1). The score is on a 5-point scale. A score of 3 or more is diagnosed as HGA/AC. Study 1: 319 SNADETs were preoperatively evaluated with SSS from June 2019 to February 2022 and were compared with postoperative pathological results. Diagnostic accuracy according to histological type (gastric, intestinal, or mixed gastrointestinal) was also evaluated. Study 2: Diagnostic accuracy of 20 lesions (10 LGA and 10 HGA/AC lesions) selected randomly were compared before and after a lecture on SSS for 26 expert and beginner gastroenterologists in 7 hospitals. Experts were defined as those with minimum 5 years of endoscopic experience and 3 years of experience with magnifying endoscopy. A total of 4 images of each lesion (white light imaging; indigo carmine; ME-NBI distant view; ME-NBI near view) were presented for evaluation. T test was used to analyze the pre and post lecture outcomes.

Results

Study 1: The overall diagnostic accuracy was 258/319 (80.8%), sensitivity 80.7%, and specificity 81.2%. Diagnostic accuracy for each tissue type was 83.3% for gastric type, 80.3% for intestinal type, and 100% for mixed type. Study 2: The mean diagnostic accuracy before and after the lecture changed from 71% to 79% overall, 70% to 81% for HGA/AC, 72% to 78% for LGA. Diagnostic accuracy of both experts (13) and beginners (13) improved significantly (p value expert: <0.01, beginner: <0.01) after the lecture.

Conclusion

SSS can be understood and used easily. It is an accurate method to diagnose SNADETs endoscopically.

Introduction

In recent years, endoscopic resection of superficial non-ampullary duodenal epithelial tumors (SNADETs), which are considered difficult, has become more widespread due to improvements in techniques and devices. However, large lesions are still highly challenging to treat, and surgical resection may be chosen. Especially endoscopic treatment of protruding lesions that occupy the lumen of the duodenum, is extremely difficult due to the lack of field of view and space for manipulation. On the other hand, large lesions are more likely to be cancerous lesions, and pathological evaluation of R0 resection or not is important. For accurate pathological evaluation, it is necessary to resect the base en bloc. Therefore, we devised two-stage endoscopic resection (TSER)^[1] and evaluate its outcomes.

Aims & Methods

From July 2021 to January 2023, TSER was performed in all 4 cases of type 0-I SNADETs that occupied the lumen of the bulb and could not be observed and manipulated space experienced at our hospital. In the first stage, the bulges were divided by piecemeal polypectomy, and their volume was reduced. It is important that piecemeal polypectomy is performed without damaging the base and surrounding mucosa. Immediately after the piecemeal polypectomy, it was difficult to accurately resect the base en bloc due to bleeding and edema caused by the polypectomy. One month later, the bases became flat lesions with a clear demarcation line. Hence, en bloc resected by endoscopic submucosal dissection (ESD) was performed safely one month later.

Results

4 patients were included. Male: female=3:1, mean age=76.3 (70-82) years. 2 patients were on antithrombotic drugs and were off them during the perioperative period. One patient was on warfarin and was heparinized.
All lesions were at the bulb, anterior/posterior/lesser curvature/greater curvature=1/1/1/1, 0-IIa: 0-IIa+Is=2: 2 at ESD, and the length of the resected specimen was 42.5(25-60)mm. All post-ESD ulcer bottoms were covered with polyglycolic acid sheets and fibrin glue.
Pathological findings; first/second stage = adenoma 1: adenocarcinoma in situ3/ adenocarcinoma in situ 4, second stage lesion length diameter 27.3(9-48)mm, no lymphovascular invasion, and R0 resection.
The histological type was tub1:tub1>pap=1:3, and the mucus type was gastric: mixed (predominantly intestinal) = 3:1. One post procedural bleeding occurred.

Conclusion

TSER can be safely performed for giant protruding tumors at the bulb, with accurate pathologic evaluation and curative results. TSER has the potential to be one of the treatment strategies for ESD of protruding tumors in other organs.

References

[1]Takeuchi N, Ohata K, Chiba H, et al. Two-stage endoscopic resection of a huge duodenal tumor. Dig Endosc. 2023; 35: e39-40

Introduction

Gastric adenocarcinoma of the fundic gland type(GA-FG) originates from the deeper mucosal layer, and predominantly occurs in the upper gastric region. Endoscopic submucosal dissection (ESD) in the upper region often results in a perpendicular approach and can be technically challenging. On the other hand, most GA-FG are relatively small, and snaring resection such as hybrid ESD and endoscopic mucosal resection with a cap-fitted pan endoscope (EMRC) may offer efficacy. However, the safety and feasibility of snaring resection for GA-FG remain unclear.

Aims & Methods

We retrospectively analyzed 77️ cases of GA-FG endoscopically resected between August 2013 and April 2024 at NTT Medical Center Tokyo. Cases were categorized into two groups: (i) ESD group (lesions resected by ESD, n=54) and (ii) Snaring group (lesions resected by hybrid ESD or EMRC, n=23), and treatment outcomes were compared between the two groups.

Results

There was no difference in gender between the two groups. Patients in the snaring group were significantly younger (median age 62.0 vs 56: p<0.05), and the lesions were significantly smaller (median tumor size 5.0 vs 4.0mm: p<0.05). The majority of lesions in both groups were located in the upper gastric region. The two groups had no significant differences in en bloc resection rate, rate of submucosal invasion, lymphovascular invasion, vertical/horizontal margin positivity rate, or R0 resection rate. However, the snaring group exhibited a shorter procedure duration (29.5 vs 12.0 minutes; p<0.05), and reduced hospital stay duration (7.0 vs 5.0 days: p<0.05) compared to the ESD group. Adverse event rates were not significantly different between the two groups. Submucosal tissue was present in 100% of snaring group specimens, with a median resected submucosal layer thickness of 667.0µm (IQR: 527.9-797.2).

Conclusion

Despite comparable rates of en bloc resection, positive vertical margin, and adverse events, the snaring group exhibited significantly shorter procedure time and hospital stay durations. All snaring group specimens contained submucosal tissue, with the resected submucosal layer thickness exceeding 500µm. Endoscopic resection by snaring might be a safe and feasible approach for resection of GA-FG.

References

1. Iwamuro M, Kusumoto C, Nakagawa M, Matsueda K, Kobayashi S, Yoshioka M, et al. Endoscopic features of oxyntic gland adenoma and gastric adenocarcinoma of the fundic gland type differ between patients with and without Helicobacter pylori infection: a retrospective observational study. BMC Gastroenterol. 2022 Jun 12;22(1):294.
2. Ueyama H, Yao T, Nakashima Y, Hirakawa K, Oshiro Y, Hirahashi M, et al. Gastric adenocarcinoma of fundic gland type (chief cell predominant type): proposal for a new entity of gastric adenocarcinoma. Am J Surg Pathol. 2010 May;34(5):609–19.

Introduction

In the endoscopic diagnosis of superficial nonampullary duodenal epithelial tumor (SNADET), it is not easy to differentiate carcinoma from adenoma. On the other hand, SNADETs can be classified into gastric and intestinal phenotypes of mucin phenotype, and it has been reported that carcinomas of the gastric phenotype in particular show a poor prognosis. If the mucin phenotype can be predicted only by endoscopic findings, it may be possible to estimate the biological malignancy of SNADETs. Then, the following studies were conducted, focusing on it as a new endoscopic diagnosis.

Aims & Methods

The aim of this study was to clarify the endoscopic findings of SNADET in each mucin phenotypes. Between August 2018 and April 2021, 500 cases of suspected SNADET in our hospital were prospectively entered for analysis. Finally, 401 cases with 401 lesions that were treated endoscopically and pathologically diagnosed as SNADET were included and analyzed. Two pathologists with expertise in gastrointestinal pathology performed histological diagnosis according to the Vienna Classification (VCL) and evaluation of mucin phenotypes based on the results of immunostaining for CD10, MUC2, MUC5AC, and MUC6. Each endoscopic finding was compared between gastric and intestinal phenotype, and multivariate analysis was performed using a logistic regression model.

Results

With regard to pathological evaluation, 289 (72.1%) cases were diagnosed as VCL category 3 and 112 (27.9%) cases were diagnosed as VCL category 4/5. The details of mucin phenotypes in this SNADETs were as follows: 44 pure gastric, 11 gastric predominant, 18 mixed gastrointestinal, 159 intestinal predominant, and 180 pure intestinal phenotypes, respectively, and the percentages of VCL category 4/5 were 63.6%, 54.5%, 38.9%, 30.8% and 16.1%, respectively (P < 0.01, Cochran-Armitage trend test). Compared in the two groups of gastric phenotype (pure gastric and gastric predominant) and intestinal phenotype (pure intestinal and intestinal predominant), there were significant differences in location (from the bulb to the proximal descending part (hereafter, proximal side): 82.2% vs 29.0%, P < 0.01), lesion diameter (20.0 mm vs 15.0 mm, P < 0.05), macroscopic type (Type 0-I: 60.0% vs 14.8%, P < 0.01), redness (present: 68.9% vs 32.8%, P < 0.01), surface structure (closed-loop structure (CLS): 84.4% vs 29.6%, P < 0.01), white opaque substance (WOS) (absent: 51.1% vs 11.0%, P < 0.01). In multivariate analysis, location (proximal side: OR 4.5, 95%CI 1.8-10.9), macroscopic type (Type 0-I: OR 2.5, 95%CI 1.1-5.7), surface structure (CLS: OR 5.4, 95%CI 2.1-13.8), WOS (absent: OR 2.9, 95%CI 1. 3-6.7) were independently and significantly associated with being gastric phenotype of SNADET.

Conclusion

This study showed that SNADETs with gastric and intestinal phenotypes were found to show different endoscopic findings.

Disclosure

There is no conflict of interest in this study.

Introduction

Recent advances in endoscopic imaging have improved the diagnostic accuracy for superficial non- ampullary duodenal epithelial tumors (SNADETs). Endoscopic resection is the treatment of choice for these lesions and endoscopic mucosal resection (EMR) is the commonly used technique. Endoscopic resection with the use of a specific distal attachment (cap-assisted EMR (CEMR)) has been reported as a safe and effective method for SNADETs.¹ However, there is limited data on the technical aspects of the same. Duodenal resections have a high risk of delayed bleeding and perforation probably due to direct exposure of ulcer to biliary and pancreatic juices. Hence, appropriate closure is absolutely essential. We have previously reported safety of closure of duodenal endoscopic resections using over-the-scope and conventional clips.² We now report on the safety and efficacy of using only hemoclips for defect closure of CEMR of SNADETs.

Aims & Methods

This prospective study aimed to evaluate the safety and efficacy of using only hemoclips for closure of SNADETs resected by CEMR. 21 patients from December, 2022 to March, 2023 who underwent the procedure at our center for lesions <10mm in size were included in the study. The pimary outcome was post procedural adverse events that included the rate of delayed bleeding and perforation, and ulcer dehiscence. The secondary outcomes were en bloc and R0 resection rates, complete closure rate, number of clips used, total procedure time, and duration of hospital stay. Descriptive statistics were used to summarize the outcomes.

Results

16 males and 5 females with median age of 59(38-75) years underwent CEMR for SNADETs and closure with hemoclips only. Paris classification morphology of 67.7% lesions was 0-IIa and 33.3% lesions was 0-IIc. All lesions of median size 4(2-6) mm were successfully resected en bloc. The median specimen size was 14(12-25) mm and the median number of hemoclips used for complete closure were 7(5-11). The mean resection and closure times were 5.7+1.8 and 9.3+3.9 minutes respectively. The mean total procedure time was 14.9+4.9 minutes. Complete closure rate was 100% with no dehiscence seen on check endoscopy (done one day post index procedure). Pathology revealed 42.9% of the lesions having low grade dysplasia and 57.1% having high grade dysplasia with R0 resection rate being 100%. There was no delayed bleeding or perforation. The mean duration of post procedure hospitalization was 2.2+0.9 days.

Outcomes of closure of CEMR for SNADETs with hemoclips only

Conclusion

Our single center study is the first to demonstrate an excellent safety profile for the use of only hemoclips in closing post CEMR defects of SNADETs. It is easy to perform, requires no additional accessories, and reduces procedure time and cost compared to other available methods of closure. Furthermore, in lesions under 10mm, CEMR was effective at en-bloc resection with favorable procedure time and duration of post procedural hospital stay. Larger scale studies may be warranted to confirm this method of resection for SNADETs and safety of closure technique.

References

1. Jamil LH, Kashani A, Peter N, et al. Safety and efficacy of cap-assisted EMR for sporadic nonampullary duodenal adenomas. Gastrointest Endosc 2017; 86:666-72.
2. Ohata K, Sakai E, Suzuki Y et al. Risk factors of delayed bleeding after endoscopic resection of superficial non-ampullary duodenal epithelial tumors and prevention by over-the-scope and conventional clipping. Dig Endosc 2021; 33: 390–8.

Introduction

Endoscopic resection (ER) of duodenal tumors carries a high risk of intraoperative or delayed perforation because of the thin duodenal wall and exposure to bile and pancreatic juice. Delayed perforation can particularly cause serious condition due to the leakage of the digestive fluids. The detailed management and clinical course following duodenal perforation remains unclear. This study aimed to clarify the optimal management of duodenal perforation following ER.

Aims & Methods

This single-center retrospective study included cases of perforation among patients who underwent ER for duodenal tumors at our institution between April 2013 and October 2024. The study endpoint was the clinical course after perforation, including the hospitalization period after ER for duodenal tumors. Perforation was defined as the presence of extraluminal gas on computed tomography (CT) scan. Delayed perforation was defined as a perforation that was not endoscopically recognized during the procedure but was diagnosed postoperatively. In contrast, intraoperative perforation was defined as a case in which a perforation was endoscopically identified during the endoscopic procedure. Long hospitalization was defined as a hospital stay of 20 days or longer.

Results

Of the 1,475 cases with duodenal tumors underwent ER during the study period, intraoperative perforation occurred in 8 cases (0.5%) and delayed perforation in 10 cases (0.7%). The median age of patients with perforation was 62 (range: 47–80) years, and 50% (9 patients) were male. The 18 cases included 6 cases of endoscopic papillectomy (EP), 7 cases of endoscopic mucosal resection (EMR), 2 cases of endoscopic submucosal dissection, and 3 cases of endoscopic submucosal resection with a ligation device. Among the cases of intraoperative perforations, long hospitalization, 61 days of conservative treatment, was required in one case who developed a large perforation during scope insertion of single‑balloon enteroscopy for duodenal adenoma resection. The remaining 7 cases achieved successful endoscopic closure intraoperatively and hospitalization periods were within 17 days. Of the 10 delayed perforation cases, 2 occurred after EP and 8 after EMR for non-ampullary tumors. The median time from the end of endoscopic resection to the diagnosis of perforation was 24 hours (4–46 hours). The median hospitalization period of the cases with delayed perforations was 32 days (9–81 days), and significantly longer than that of intraoperative perforation (13 days) (p<0.05). Long hospitalization was required in 8 cases (80%) of delayed perforation. None of the delayed perforation cases required surgical intervention, and endoscopic nasobiliary and nasopancreatic duct drainage (ENBPD) was performed in 7 cases (70%). Ascites was observed on CT scan at the time of diagnosis of delayed perforation in 7 of the 8 patients who required long hospitalization. One case without ascites at the diagnosis of delayed perforation still required long hospitalization, possibly due to the 10-hour delay from the diagnosis to ENBPD. On the other hand, two cases who did not need long hospitalization showed no ascites at the diagnosis of delayed perforation.

Conclusion

Compared to intraoperative perforation, delayed perforation was associated with significantly prolonged hospitalization, particularly in cases with ascites. Early detection of ascites and prompt initiation of ENBPD may help mitigate the clinical burden. Establishing an emergent backup system for ENBPD is essential for the safe management of delayed perforation after duodenal ER.

Introduction

The ulcers after duodenal endoscopic resection (ER) are exposed to bile and pancreatic juice, therefore delayed perforation and bleeding are common. As a preventive measure, the suture of the ulcer after ER has been reported to be effective. Our data are shown as follows. From January 2007 to January 2018, ER was performed for 219 duodenal tumors. We compared 211 lesions in the suture group with 8 lesions in the non-suture group. Delayed perforation was 0% in the suture group and 13% (1/8) in the non-suture group, showing a significant difference (p=0.0365). Delayed bleeding was 1% (3/211) in the suture group and 25% (2/8) in the non-suture group, showing a significant difference (p=0.01). Therefore, the suture of the ulcer after ER is essential. If the ulcer is small after ER, it can be sutured endoscopically, but if it exceeds 3 cm, it is difficult to suture with the endoscope alone.

Aims & Methods

The aim of this study is to investigate the outcome of duodenal laparoscopic and endoscopic co-operative surgery (D-LECS) for superficial non-ampullary duodenal epithelial tumors (SNADETs).
First, tissue around duodenum is dissected laparoscopically. Second, endoscopic submucosal dissection (ESD) is performed using saline injection, Hook knife, and VIO 300D or VIO3. Sites with poor maneuverability endoscopically were resected with laparoscopic assistance. Third, ulcer after ER is sutured laparoscopically. If inadequate, endo clip is added.
Thirty-five patients with SNADETs who underwent D-LECS from January 2014 to March 2023 were investigated outcome of procedure. The patient’s characteristics were as follows. male/female: 23/12, median age 65 (37-83) years old, location of the tumor (bulbs/2^nd/3rd portion): 3/27/5, major macroscopic type (0-I/0-IIa/0-IIc): 1/29/5. pancreatic side/ non-pancreatic side of duodenum: 26/9.
SM1 was defined as submucosal invasion of less than 500 μm, and SM2 was defined as deeper. R0 was defined as lateral and vertical margin negative, and lymph-vascular involvement negative. Median follow-up period was 13 (0-42) months.

Results

1. Median procedure time was 351 (195-673) minutes. In it, median endoscopic procedure time and laparoscopic procedure time were 205(132-429) and 131(67-420) minutes, respectively. Median diameter of resected specimen and tumor were 40 (26-70) and 27 (14-53) mm, respectively. Intramucosal tumor (Vienna classification category 3 to 5.1) /T1bSM2 (category 5.2): 31/4. The median length of hospital stay was 9 (8-15) days.
2. The en bloc resection and R0 resection rate were 94% (33/35), 83% (29/35), respectively.
3. Complications: intraoperative perforation was 8% (3/35). The intraoperative perforation site was sutured laparoscopically after ESD was completed. Intraoperative bleeding requiring blood transfusion, delayed perforation, and delayed bleeding were 0%.
4. The rate of complete closure by laparoscopy was 0% (0/26) for pancreatic side, and all of them were sutured with endo-clips and polyglycolic acid sheets. And that for non-pancreatic side was 100% (9/9).
5. Prognosis
Two cases of T1bSM2 were underwent pancreaticoduodenectomy, and there were no residual tumor and no lymph node metastasis. There has been no recurrence during twenty-eight and forty-two months after DLECS,
T1aM and T1bSM1 were followed up, and no local or metastatic recurrence was observed.

Conclusion

DLECS enabled complete closure of ulcers after ER regardless of size or location. Therefore, it is a safe and effective treatment for large SNADETs.