齊藤 景子

BACKGROUND AND AIM: Endoscopic retrograde cholangiopancreatography (ERCP) may help detect cholangiocarcinoma in patients with primary sclerosing cholangitis (PSC), but it may be associated with complications. This study was aimed at determining the prognostic impact of ERCP on patients with PSC without cholangitis. METHODS: Patients with PSC without cholangitis were divided into two groups: those who underwent ERCP within three years after diagnosis (ERCP-performed group) and those who did not (non-ERCP group). These groups were compared in terms of clinical outcomes (liver-related death or liver transplantation, endoscopic treatment requirement and repeated cholangitis) and the composite outcome. RESULTS: Of 99 patients with PSC with detailed medical history, 49 were included in the ERCP-performed group and 21 in the non-ERCP group. In Kaplan-Meier analysis, the non-ERCP group was less likely to achieve the three outcomes and the composite outcome, showing statistical significance (endoscopic treatment requirement; p = 0.017 and composite outcome; p = 0.014). A Cox proportional hazards model indicated that ERCP in the asymptomatic state was a significant predictor of endoscopic treatment requirement (hazard ratio [HR]: 4.37, 95% confidence interval [CI]: 1.03-18.59) and the composite outcome (HR: 4.54, 95% CI: 1.07-19.28). CONCLUSION: ERCP in patients with PSC without cholangitis is likely to require further endoscopic treatment and may be associated with poor prognosis.

BACKGROUND AND AIM: The treatment strategy for patients with ulcerative colitis (UC) in clinical remission who have not achieved mucosal healing is unclear. This study aimed to determine the risk factors of relapse in patients in clinical remission with endoscopic activity. METHODS: This retrospective, single-center study included patients with UC who underwent colonoscopy (CS) and were in clinical remission with endoscopic activity. Characteristics were compared between patients who relapsed within 2 years after CS and those who did not. A Cox proportional hazards regression model was used to identify risk factors contributing to clinical relapse. Recent worsening in bowel symptoms was defined as increase in bowel frequency and/or increase in abdominal pain within approximately 1 month based on the descriptions in the medical charts. RESULTS: This study regarded 142 patients in clinical remission with an endoscopic activity of Mayo endoscopic subscore (MES) of ≥1 as eligible, and 33 (23%) patients relapsed during the observation period. Recent worsening of bowel symptoms was a significant risk factor for clinical relapse (hazard ratio [HR]: 3.02, 95% confidence interval [CI]: 1.34-6.84). This was particularly evident in patients with MES of 2 (HR: 5.16, 95% CI: 1.48-18.04), whereas no risk factors were identified in patients with MES of 1. The presence or absence of therapeutic intervention just after CS did not significantly affect clinical relapse. CONCLUSION: Recent worsening in bowel symptoms was a significant risk factor for clinical relapse in patients with UC who were in clinical remission with endoscopic activity.

BACKGROUND: This study aimed to investigate the utility of intensive triamcinolone acetonide (TA) injections after extensive esophageal endoscopic submucosal dissection (ESD). METHODS: This retrospective study included 27 lesions in 27 consecutive patients who underwent ESD (ulcers encompassing ≥3/4 of the esophageal circumference) and received TA injections without oral steroid administration. Groups A and B included patients undergoing ESD with and without complete circumferential resection, respectively. All patients received TA injections (100 mg/session) immediately after ESD. In Group A, weekly based TA injections were performed until near-complete ulcer epithelialization. In Group B, patients did not receive additional injections or received weekly or biweekly TA injections. The primary outcome was stricture rate, and the secondary outcomes were the proportion of patients requiring endoscopic balloon dilation (EBD) and the number of TA injections. RESULTS: Group A included 7 lesions, and Group B included 20 lesions. The median (range) tumor lengths were 40 (30-90) and 45 (30-110) mm in Groups A and B, respectively. In Group A, the median circumferential resection diameter was 40 (20-80) mm. The stricture rate and the proportion of patients requiring EBD were 0 (0%) in Group A and 1 (5.0%) in Group B. The number of TA injection sessions was significantly higher in Group A than in Group B (8 [5-25] vs 1.5 [1-3]; p < 0.001). CONCLUSIONS: Intensive weekly or biweekly based TA injections might aid in preventing post-ESD stricture and the need for EBD in patients undergoing extensive resection involving the entire esophageal circumference.

BACKGROUND AND AIM: The measurement of esophageal acid exposure time (AET) using combined multichannel intraluminal impedance-pH (MII-pH) tests is the gold standard for diagnosing gastroesophageal reflux disease (GERD). However, this catheter-based 24-h test can cause considerable patient discomfort. Our aim is to identify factors affecting AET and to develop a scoring model for predicting AET abnormalities before conducting the MII-pH test. METHODS: Of the 366 patients who underwent MII-pH test at two facilities in Japan and Vietnam, 255 patients who also had esophagogastroduodenoscopy and high-resolution manometry were included in this study. Logistic regression analysis was conducted using risk factors for AET > 6% identified from a derivation cohort (n = 109). A scoring system predicting AET > 6% was then constructed and externally validated with a separate cohort (n = 146). RESULTS: Three variables were derived from the prediction model: male gender, Hill grades III-IV, and weak mean distal contractile integrals. Based on these scores, patients were classified into low (0 point), intermediate (1-3 points), and high (4 points) risk groups. The probabilities of having an AET > 6% were 6%, 34%, and 100% for these groups, respectively. A score of < 1 excluded patients with abnormal AET, with a negative predictive value of 93.8% in the derivation cohort and 80.0% in the validation cohort. CONCLUSIONS: We derived and externally validated a prediction model for abnormal AET. This system could assist in guiding the appropriate treatment strategies for GERD.

BACKGROUND AND STUDY AIMS: The infliximab biosimilar CT-P13 was the first biosimilar drug targeting tumor necrosis factor-α. However, its efficacy and safety in real-world clinical situations have remained insufficient. Therefore, we aimed to verify the efficacy and safety of CT-P13 in bio-naïve patients with Crohn's disease. PATIENTS AND METHODS: This retrospective multicenter study compared the remission rate at week 54 between patients with Crohn's disease who were treated with originator infliximab or CT-P13. Endoscopic and laboratory findings were assessed in both groups. A total of 184 (156 originator and 28 CT-P13) patients were analyzed. Of these, 138 originator users and 19 biosimilar users completed 54-week administration. RESULTS: The clinical remission rates in patients taking originator infliximab of CT-P13 at week 54 were 92.5 % and 100 %, respectively. The endoscopic scores of each group significantly decreased from baseline at week 54 in both groups, and the mucosal healing rate at week 54 was 53 % and 64 %, respectively. Laboratory data including C-reactive protein, serum albumin, and hemoglobin significantly improved from baseline to week 14 and 54 in both groups. Adverse events were observed more frequently in the CT-P13 group (25 % vs. 4.5 %, p = 0.0015), but severe adverse events were rare in both groups. CONCLUSION: The efficacy and safety of CT-P13 were comparable with those of originator infliximab in bio-naïve patients with Crohn's disease evaluated by clinical, endoscopic, and laboratory findings. This study establishes the needed groundwork for the development of a strategy for treatment with biologics in patients with Crohn's disease.

BACKGROUND AND AIMS: Delayed bleeding (DB) is a major adverse event associated with colorectal endoscopic submucosal dissection (ESD) that sometimes causes difficulties in making decisions regarding endoscopic hemostasis. This study identified the factors that contribute to follow-up without endoscopic hemostasis when DB is suspected after colorectal ESD. METHODS: In total, 583 patients (603 tumors) who underwent ESD or hybrid ESD for colorectal tumors at Chiba University Hospital between June 2009 and January 2022 were retrospectively registered. Of these, 141 cases (141 tumors) with DB; with hematochezia or hemoglobin decrease ≥2 g/dL after colorectal ESD were analyzed. The DB group was divided into the Hemostasis group (H group; endoscopic hemostasis performed) and no-Hemostasis group (no-H group; no endoscopy performed, or endoscopy performed but no hemostasis performed after hematochezia or hemoglobin decrease). Univariate and multivariate logistic regression analyses were performed to assess the factors contributing to follow-up. RESULTS: Thirty-one patients with 31 tumors were categorized into the H group, while 110 patients with 110 tumors were in the no-H group. Multivariate regression analysis revealed that date from ESD to first hematochezia ≤Day 3 (odds ratio [OR] 4.55, 95% confidence interval [CI] 1.44-14.33; p = 0.010) and bleeding duration ≤1 day (OR 3.35, 95% CI 1.35-8.34; p = 0.009) contributed to follow-up. CONCLUSIONS: In cases of DB after colorectal ESD, a bleeding duration ≤1 day or date from ESD to first hematochezia ≤Day 3 may contribute to follow-up observation without endoscopic hemostasis.

BACKGROUND: s: Factors affecting mucosal permeability (MP) in ulcerative colitis (UC) are largely unknown. We aimed to investigate the difference in MP among patients with UC classified according to the colonic locations and to evaluate the correlations between local MP and endoscopic or histological activity of UC. METHODS: The transepithelial electrical resistance (TER), which is inversely proportional to permeability, of tissue samples from the mucosa of the ascending colon, descending colon, and rectum of patients with UC and healthy individuals (HI) was measured by using the Ussing chamber. TERs were compared between patients with UC and HI, and evaluated according to colonic locations and disease activity of UC. RESULTS: Thirty-eight patients with UC and 12 HI were included in this study. Both in HI and patients with UC, MP tends to be higher in the anal side. TER in the ascending colon was significantly lower in patients with UC than in HI (45.3  ±  9.0 Ω × cm2 vs. 53.5  ±  9.7 Ω × cm2, p = 0.01). The increased permeability in UC was observed also in the descending colon, only when the inflammation involved the location. A significant correlation between TER and endoscopic activity was found in the rectum only (r = -0.49, p = 0.002). There were no significant correlations between TERs and UC histology. CONCLUSIONS: The MP in the colon differs according to the colonic location. The ascending colon among patients with UC showed disease-specific changes in MP, whereas the MP is increased in proportion to the endoscopic activity in the rectum.

A 37-year-old woman developed severe colitis with diffuse mucosal erythema and ulcerations throughout the entire colon after the 3rd vaccination of COVID-19. Stool culture was negative, and the pathological findings showed increased lymphoplasmacytic and neutrophilic infiltration in the colonic lamina propria, which were consistent with ulcerative colitis. After the treatment with anti-tumor necrosis factor-α agent, the ulceration markedly improved with development of severe colonic stenosis, which was successfully dilated with endoscopic balloon dilation. In case of COVID-19 vaccination, it should be noted that vaccination could be a trigger for the onset of UC.

This study aimed to investigate the lesion and endoscopist factors associated with unintentional endoscopic piecemeal mucosal resection (uniEPMR) of colorectal lesions ≥ 10 mm. uniEPMR was defined from the medical record as anything other than a preoperatively planned EPMR. Factors leading to uniEPMR were identified by retrospective univariate and multivariate analyses of lesions ≥ 10 mm (adenoma including sessile serrated lesion and carcinoma) that were treated with endoscopic mucosal resection (EMR) at three hospitals. Additionally, a questionnaire survey was conducted to determine the number of cases treated by each endoscopist. A learning curve (LC) was created for each lesion size based on the number of experienced cases and the percentage of uniEPMR. Of 2557 lesions, 327 lesions underwent uniEPMR. The recurrence rate of uniEPMR was 2.8%. Multivariate analysis showed that lesion diameter ≥ 30 mm (odds ratio 11.83, 95% confidence interval 6.80-20.60, p < 0.0001) was the most associated risk factor leading to uniEPMR. In the LC analysis, the proportion of uniEPMR decreased for lesion sizes of 10-19 mm until 160 cases. The proportion of uniEPMR decreased with the number of experienced cases in the 20-29 mm range, while there was no correlation between the number of experienced cases and the proportion of uniEPMR ≥ 30 mm. These results suggest that 160 cases seem to be the minimum number of cases needed to be proficient in en bloc EMR. Additionally, while lesion sizes of 10-29 mm are considered suitable for EMR, lesion sizes ≥ 30 mm are not applicable for en bloc EMR from the perspective of both lesion and endoscopist factors.

BACKGROUND AND AIMS: This randomized controlled trial (RCT) was designed to evaluate the short-term outcomes of underwater endoscopic mucosal resection (UEMR) and endoscopic submucosal dissection (ESD) of 21-30 mm colonic polyps. METHOD: We conducted a single-center RCT. Patients diagnosed with suspected colorectal intramucosal carcinoma (21-30 mm and adaptable for both UEMR and ESD) were randomly assigned to the UEMR and ESD groups at a 1:1 ratio. The primary endpoint was the R0 resection rate. We independently performed one-sample tests against the set threshold for each treatment. The significance level was set at p = 0.224. RESULT: Eleven polyps each in the UEMR and ESD groups, respectively, were analyzed. The R0 resection rate (%) was 36 (95% confidence interval 11-69) and 100 (72-100) for UEMR and ESD, respectively, with a significant difference between the two groups (p = 0.002). The p-value against the set threshold for UEMR was 0.743, whereas that for ESD was < 0.001 (one-sample binomial test). The en bloc resection rates (%) were 82 (48-97) and 100 (72-100) for UEMR and ESD, respectively; however, no significant difference was observed (p = 0.167). The mean treatment time (min) was significantly shorter in the UEMR group (8 ± 6) than in the ESD group (48 ± 29) (p = 0.001). CONCLUSION: ESD could achieve a high R0 resection rate, while the en bloc resection rate was comparable between the two treatment techniques with less burden on patients undergoing UEMR for 21-30-mm colorectal polyps. CLINICAL TRIAL REGISTRATION: The study was registered at the Japan Registry of Clinical Trial as jRCT1030210015 and jRCT1030210177.

Many molecular targeted agents, including biologics, have emerged for inflammatory bowel diseases (IBD), but their high prices have prevented their widespread use. This study aimed to reveal the changes in patient characteristics and the therapeutic strategies of IBD before and after the implementation of biologics in Japan, where the unique health insurance system allows patients with IBD and physicians to select drugs with minimum patient expenses. The analysis was performed using a prospective cohort, including IBD expert and nonexpert hospitals in Japan. In this study, patients were classified into two groups according to the year of diagnosis based on infliximab implementation as the prebiologic and biologic era groups. The characteristics of therapeutic strategies in both groups were evaluated using association analysis. This study analyzed 542 ulcerative colitis (UC) and 186 Crohn's disease (CD). The biologic era included 53.3% of patients with UC and 76.2% with CD, respectively. The age of UC (33.9 years vs. 38.8 years, P < 0.001) or CD diagnosis (24.3 years vs. 31.9 years, P < 0.001) was significantly higher in the biologic era group. The association analysis of patients with multiple drug usage histories revealed that patients in the prebiologic era group selected anti-tumor necrosis factor (TNF)-α agents, whereas those in the biologic era group preferred biologic agents with different mechanisms other than anti-TNF-α. In conclusion, this study demonstrated that both patient characteristics and treatment preferences in IBD have changed before and after biologic implementation.

BACKGROUND AND AIM: Little is known about genetic mutations in the regenerated mucosa (RM) after endoscopic resection (ER) of esophageal carcinoma. Thus, this study investigates the status of genetic variation in RM after ER of esophageal squamous cell carcinoma (ESCC). METHODS: The study cohort included 19 patients with ESCC. We used an esophageal carcinoma panel to identify target sequences for squamous cell carcinoma (SCC), background mucosa (BM), and RM after ER of ESCC. We used OncoKB to check whether each mutation was a putative driver. RESULTS: We identified 77 mutations of 32 genes in SCC, 133 mutations of 34 genes in BM, and 100 mutations of 29 genes in RM. Putative driver mutations were identified in 20 mutations in 14 cases in SCC, 16 mutations in 10 cases in BM, and 7 mutations in 11 cases in RM. The rate of putative driver mutations to total mutations was significantly lower in RM (26% in SCC vs 12% in BM vs 7% in RM, P = 0.009). Additionally, the rate of cases with TP53 putative driver mutations was significantly lower in RM (63% in SCC vs 37% in BM vs 16% in RM, P = 0.011). The percentage of putative driver mutations and the percentage of cases with a putative driver of TP53 were significantly lower in RM. CONCLUSION: Esophageal RM after ER of ESCC could have a lower risk of carcinogenesis.

＜文献概要＞クローン病(CD)症例の半数以上が小腸病変を有しており,CD診療では小腸評価が求められる.CDによる慢性炎症に伴い,経過中に小腸狭窄をきたす症例を認めることから,内視鏡による狭窄評価や時に治療を要することがある.小腸内視鏡を用いたバルーン拡張術は手術に比べると侵襲性が低く,腸管切除による短腸症候群を回避する方法として重要な意義をもつ.

BACKGROUND: This study aimed to investigate the validity of pathological diagnosis of early CRC (E-CRC) from the genetic background by comparing data of E-CRC to colorectal adenoma (CRA) and The Cancer Genome Atlas (TCGA) on advanced CRC (AD-CRC). METHODS: TCGA data on AD-CRC were studied in silico, whereas by next-generation sequencer, DNA target sequences were performed for endoscopically obtained CRA and E-CRC samples. Immunohistochemical staining of mismatch repair genes and methylation of MLH1 was also performed. The presence of oncogenic mutation according to OncoKB for the genes of the Wnt, MAPK, and cell-cycle-signaling pathways was compared among CRA, E-CRC, and AD-CRC. RESULTS: The study included 22 CRA and 30 E-CRC lesions from the Chiba University Hospital and 212 AD-CRC lesions from TCGA data. Regarding the number of lesions with driver mutations in the Wnt and cell-cycle-signaling pathways, E-CRC was comparable to AD-CRC, but was significantly greater than CRA. CRA had significantly more lesions with a driver mutation for the Wnt signaling pathway only, versus E-CRC. CONCLUSIONS: In conclusion, the definition of E-CRC according to the Japanese criteria had a different genetic profile from CRA and was more similar to AD-CRC. Based on the main pathway, it seemed reasonable to classify E-CRC as adenocarcinoma. The pathological diagnosis of E-CRC according to Japanese definition seemed to be valid from a genetic point of view.

Video 1Use of a super-soft hood (Space Adjuster; TOP, Tokyo, Japan) for esophageal endoscopic submucosal dissection below an esophageal stricture.

BACKGROUND AND AIMS: Gastric submucosal tumors (SMTs) are treated or monitored according to GI stromal tumor guidelines, but the adequacy of the guidelines has not been thoroughly examined. We investigated the long-term course of gastric SMTs to determine the validity of guideline-based follow-up methods and the factors contributing to their size increase. METHODS: This study included gastric SMTs diagnosed as GI mesenchymal tumors (GIMTs) by using EUS and followed up with EUS. The percentage and speed of GIMT enlargement and factors associated with the enlargement were investigated by using the Cox proportional hazards model. RESULTS: From January 1994 to May 2022, a total of 925 gastric SMTs were evaluated with EGD, and 231 SMTs were diagnosed as GIMTs. Of the 231 GIMTs, 145 were examined by EUS more than twice and were followed up for >6 months. The mean ± standard deviation follow-up period was 5.20 ± 4.04 years (range, 0.5-17.3 years), with 39 (26.9%) of 145 GIMTs increasing in size with a mean doubling time of 3.60 ± 3.37 years. A multivariate analysis of factors influencing tumor growth revealed that irregular extraluminal borders were an increasing factor (hazard ratio, 3.65; 95% confidence interval, 1.26-10.52), initial tumor size ≤9.5 mm (hazard ratio, .23; 95% confidence interval, 0.07-0.77) was a nonincreasing factor, and GIMTs with calcification (n = 13) did not increase in size. CONCLUSIONS: Tumor growth in gastric GIMTs <9.5 mm in diameter and/or with calcification is rare. Follow-up intervals for these lesions could be extended.

To gain a better understanding of the effects of biologics, we evaluated clinical outcomes in patients with moderate to severe exacerbations of ulcerative colitis (UC). This retrospective, multicenter study retrieved the entire clinical courses of UC patients who began treatments between 2004 and 2018. All exacerbations and clinical parameters, including treatment details for exacerbations and both remission and re-exacerbation dates, were identified during the observation period. Two different endpoints, the cumulative incidence rates of surgical resection and re-exacerbation, were evaluated separately in moderate to severe exacerbation events. Among 1401 patients, 1626 exacerbation events were determined according to a partial Mayo score (remission: < 2, mild: 2-4, moderate: 5-7, and severe: > 7). During the observation period, as administration rates of biologics increased, both surgical resection and hospitalization rates decreased, for 959 moderate to severe exacerbation events. We confirmed that biologics significantly reduced the cumulative re-exacerbation rate in moderate to severe exacerbation events during the study period compared with suboptimal therapies (a 0.507-fold decreased risk according to COX regression analysis, P < 0.001). However, they had not enough impact in reducing the cumulative incidence rate of surgical resection in moderate to severe exacerbation events that were corticosteroid-refractory or dependent (a 0.878-fold decreased risk according to COX regression analysis, P = 0.606). Biologics may improve remission duration, but these agents had no significant impact in reducing the risk of surgical resection in moderate to severe active UC.

This pilot study aimed to investigate the utility of texture and color enhancement imaging (TXI) with magnified endoscopy (ME) for the preoperative diagnosis of superficial nonampullary duodenal epithelial tumors (SNADETs). We prospectively evaluated 12 SNADETs. The visibility for ME-TXI, ME with indigo carmine (ICME)-white-light imaging (WLI), ICME-TXI compared to ME-NBI (narrow-band imaging) was scored (+ 2 to - 2 ME-NBI was set as score 0) by 3 experts. Scores + 2 and + 1 were defined as improved visibility. The intra-observer and interobserver agreement for improved visibility of surface structure (SS) was evaluated. Sensitivity, specificity, and positive predictive value (PPV) for Vienna Classification (VCL) C4/5 associated with the preoperative diagnosis of ICME-TXI were analyzed. The SS visibility score of ICME-TXI was significantly higher than that of ME-NBI, ME-TXI, and ICME-WLI (P < 0.001 respectively). The kappa coefficients of reliability for intra-observer and interobserver agreement for the SS visibility improvement with ICME-TXI were 0.96, 1.00, 1.00 and 0.70, 0.96, 0.96 respectively. All endoscopists preferred ICME-TXI for visualizing SS mostly for all lesions. The sensitivity, specificity, and PPV (%) of ICME-TXI for VCL C4/5 were 80, 66.7, and 63.2, respectively. ICME-TXI facilitates the visibility of the SS of SNADETs and may contribute to their preoperative diagnosis.

BACKGROUND AND AIM: Anti-tumor necrosis factor (TNF)α antibody (ATA) and biologics/molecular targeted agents with other mechanisms (non-ATA) are currently available for refractory ulcerative colitis (UC). However, the knowledge about optimal drug selection after the initial treatment with ATA failure is lacking. This study assessed whether the response to the initial ATA could be a basis for selecting subsequent agents in UC patients. METHODS: Ulcerative colitis patients treated with ATA or non-ATA as the subsequent biologic after the failure of initial ATA were retrospectively analyzed. The efficacy at 14 weeks was examined according to the response to initial ATA. RESULTS: Of 163 patients treated with the first ATA, the efficacy of subsequent ATA and non-ATA was evaluated in 63 and 36, respectively. Remission and response to subsequent-line therapy, regardless of ATA or non-ATA, were lower in patients with primary nonresponse (PNR) to initial ATA than in patients with efficacy to initial ATA (33.3% vs 69.2%, P < 0.01). In patients with PNR to initial ATA, the remission rate with subsequent ATA was significantly lower than with subsequent non-ATA (4.3% vs 26.3%, P = 0.04). In patients who showed efficacy to initial ATA, the remission rate with subsequent ATA was also lower than that with subsequent non-ATA (30.6% vs 56.3%, P = 0.08). PNR with initial ATA was the predictor of PNR to subsequent ATA (odds ratio: 5.62, 95% confidence interval: 1.50-21.7). CONCLUSION: Non-ATA may be suitable in UC patients as the subsequent biologics regardless of the outcome of the first ATA.

BACKGROUND AND AIMS: The appropriate selection of endoscopic resection (ER) for relatively small superficial nonampullary duodenal adenomas (SNADAs) considering recurrence is not completely clarified. Therefore, this study investigated ER utility [endoscopic mucosal resection (EMR), underwater EMR (UEMR), and cap-assisted EMR (EMRC)] for SNADAs from the viewpoint of recurrence as well as short-term outcomes. METHODS: We retrospectively analyzed patients with sporadic SNADAs who underwent EMR, UEMR, and EMRC at Chiba University Hospital between May 2004 and March 2020 and were observed for ≥12 months after ER (EMR: 34 patients, 36 lesions; UEMR: 54 patients, 55 lesions; and EMRC: 45 patients, 48 lesions). Outcomes were evaluated using weighted logistic regression analysis. The logistic regression analysis was weighted using propensity scores. RESULTS: EMRC showed a significantly higher en-bloc and R0 resection rate than EMR. EMRC showed a significantly higher en-bloc and R0 resection rate than EMR. All techniques were equally safe. Only one case each of intraoperative perforation and postoperative perforation (in two different patients) occurred, which were associated with EMRC. UEMR resulted in higher R0 resection and lower postbleeding rates than EMR. Moreover, patients who underwent UEMR showed no perforation. Median observation period per lesion (months) after ER was 84 (16-199) for patients who underwent EMR, 25 (12-60) for patients who underwent UEMR, and 63 (12-180) for patients who underwent EMRC. No significant difference was observed between EMR vs. UEMR and between EMR vs. EMRC in terms of recurrence [odds ratio (95% confidence interval): 0.20 (0.01-2.86), p = 0.24 and 0.78 (0.09-6.84), p = 0.82, respectively]. CONCLUSIONS: The recurrence risk was not different for EMR, UEMR, and EMRC. Therefore, UEMR, a simple and safe procedure, could be the first choice for relatively small SNADAs. With larger prospective studies, UEMR data may turn out to be more robust, corroborating it as the endoscopic modality of choice for certain SNADAs.

BACKGROUND AND AIMS: The application of underwater EMR (UEMR) for nonpolypoid superficial nonampullary duodenal epithelial tumors (SNADETs) has not been comprehensively assessed. Therefore, the current study aimed to validate the efficacy of UEMR versus conventional EMR and cap-assisted EMR (EMRC) for SNADETs measuring ≤20 mm. METHODS: We retrospectively analyzed patients with sporadic nonpolypoid SNADETs measuring ≤20 mm undergoing EMR, EMRC, or UEMR at Chiba University Hospital between May 2004 and October 2020 (EMR, 21 patients and 23 SNADETs; UEMR, 60 patients and 61 SNADETs; EMRC, 45 patients and 48 SNADETs). A weighted logistic regression analysis was performed to analyze outcomes. Univariate and multivariate logistic regression models were used to identify the predictors of RX/1 and piecemeal resection. The recurrence rate of lesions observed ≥12 months after resection was assessed. RESULTS: Both UEMR and EMRC had a significantly higher R0 resection rate than EMR. UEMR had significantly lower multiple resection and postbleeding rates than EMR. Only 1 patient (2.1%) who underwent EMRC experienced intraoperative and postoperative perforation. EMR was involved in RX/1 and piecemeal resection. The recurrence rates of EMR, UEMR, and EMRC were 4.3%, 2.0%, and 6.3%, respectively. CONCLUSIONS: UEMR had significantly higher R0 resection and lower postbleeding rates than EMR. Moreover, it was safer than EMRC and was associated with a lower incidence of recurrences. The significant results of the retrospective analysis suggest a randomized controlled study with adequate numbers needs to be conducted to confirm the superior efficacy of UEMR before it is recommended for primary treatment option for SNADETs measuring ≤20 mm.