Predictors of severe submucosal fibrosis during endoscopic submucosal dissection in patients with ulcerative colitis: Retrospective cohort study.

OBJECTIVES: Severe submucosal fibrosis is a crucial technical difficulty encountered during endoscopic submucosal dissection (ESD) in patients with ulcerative colitis (UC). We aimed to identify predictors of severe submucosal fibrosis in patients with UC. METHODS: We retrospectively included 55 tumors resected using ESD from 48 consecutive patients with UC. We analyzed the clinicopathological characteristics and treatment outcomes between the F0/1 (none to mild submucosal fibrosis) group (n = 28) and F2 (severe submucosal fibrosis) group (n = 27). RESULTS: No significant difference was found between the F0/1 and F2 groups in en bloc resection rate (100% vs. 96%, P = 0.49), the R0 resection rate (100% vs. 93%, P = 0.24), and the dissection speed (0.18 vs. 0.13 cm2 /min, P = 0.07). Intraoperative perforation was more common in the F2 group (30%) than in the F0/1 group (8%; P = 0.01). Multivariable analysis showed that a longer duration of UC (≥10 years; odds ratio [OR] 6.11; 95% confidence interval [CI] 1.20-31.03; P = 0.03) and scarring of background mucosa of the tumor (OR 39.61; 95% CI 3.91-400.78; P < 0.01) were independent predictors of severe submucosal fibrosis. CONCLUSION: Long UC duration and scarring background mucosa were predictors of severe submucosal fibrosis associated with perforation during ESD.

Impact of endoscopy intervals on metachronous gastric cancer after endoscopic submucosal dissection: Comparison between 1 year and half-a-year.

OBJECTIVES: Japanese guidelines recommend posttreatment endoscopy once or twice a year after endoscopic submucosal dissection (ESD) for early gastric cancer. However, the impact of endoscopy intervals on metachronous gastric cancer (MGC) remains unclear, especially the difference between 1-year and half-a-year intervals. We aimed to investigate this difference. METHODS: This study retrospectively investigated 2429 patients who underwent gastric ESD between May 2001 and June 2019 at our hospital. Patients who developed MGC were classified based on those who underwent the previous endoscopy within at least 7 months (short-interval group) and within 8-13 months (regular-interval group). Propensity score matching (PSM) was used to adjust for possible confounders. The primary outcome was the proportion of MGC beyond curative ESD criteria established in the guidelines. RESULTS: A total of 216 eligible patients developed MGC. The short- and regular-interval groups included 43 and 173 patients, respectively. Overall, no patients in the short-interval group had MGC beyond curative ESD criteria, while 27 patients in the regular-interval group did. The proportion of MGC beyond curative ESD criteria was significantly lower in the short-interval group than in the regular-interval group before (P = 0.003) and after (P = 0.028) PSM. Although not significant, the short-interval group tended to have a higher stomach preservation rate than the regular-interval group (P = 0.093). CONCLUSION: Our study indicated a possible benefit of biannual surveillance endoscopy in the early post-ESD period.

Endoscopic full-thickness resection vs. endoscopic submucosal dissection of residual/recurrent colonic lesions on scars: a retrospective Italian and Japanese comparative study.

BACKGROUND AND AIMS: Endoscopic treatment of recurrent/residual colonic lesions on scars is a challenging procedure. In this setting, endoscopic submucosal dissection (ESD) is considered the first choice, despite a significant rate of complications. Endoscopic full-thickness resection (eFTR) has been shown to be well-tolerated and effective for these lesions. The aim of this study is to conduct a comparison of outcomes for resection of such lesions between ESD and eFTR in an Italian and a Japanese referral center. METHODS: From January 2018 to July 2020, we retrospectively enrolled patients with residual/recurrent colonic lesions, 20 treated by eFTR in Italy and 43 treated by ESD in Japan. The primary outcome was to compare the two techniques in terms of en-bloc and R0-resection rates, whereas complications, time of procedure, and outcomes at 3-month follow-up were evaluated as secondary outcomes. RESULTS: R0 resection rate was not significantly different between the two groups [18/20 (90%) and 41/43 (95%); P= 0.66]. En-bloc resection was 100% in both groups. No significant difference was found in the procedure time (54 min vs. 61 min; P= 0.9). There was a higher perforation rate in the ESD group [11/43 (26%) vs. 0/20 (0%); P= 0.01]. At the 3-month follow-up, two lesions relapsed in the eFTR cohort and none in the ESD cohort (P= 0.1). CONCLUSION: eFTR is a safer, as effective and equally time-consuming technique compared with ESD for the treatment of residual/recurrent colonic lesions on scars and could become an alternative therapeutic option for such lesions.

Cervical esophageal adenocarcinoma of intestinal type in ectopic gastric mucosa.

A 45-year-old man underwent esophagogastroduodenoscopy because of symptoms of laryngopharyngeal discomfort. We found a protruded reddish lesion adjacent to the ectopic gastric mucosa (EGM) in the cervical esophagus, and a biopsy revealed that it was a tubular adenocarcinoma. We diagnosed the patient with intramucosal cancer and performed endoscopic submucosal dissection. Esophageal endoscopic submucosal dissection was performed under general anesthesia using a conventional procedure. The resected tumor measured 23 × 14 mm and was adjacent to the EGM. Histologically, the tumor cells showed moderately well-differentiated adenocarcinoma confined to the muscularis mucosa with no lymphovascular infiltration. Immunohistochemically, the tumor cells were positive for intestinal markers, namely MUC2 and CD10, and negative for gastric markers, namely MUC5AC and MUC6. The patient had no post-endoscopy submucosal dissection stenosis and remained disease-free without local recurrence. EGM of the cervical esophagus develops from the columnar epithelium during embryonic development. There are few reports on endoscopic submucosal dissection for mucosal cancer. Of these, immunostaining was performed in three cases. All were positive for MUC5AC and MUC6 and negative for MUC2 and CD10. Usually, EGM shows gastric type epithelium, but occasional cases with intestinal metaplasia, which show positivity for MUC2 and CD10, have been reported. Therefore, we consider this to be an extremely rare case of esophageal adenocarcinoma arising from intestinal metaplasia within the EGM.

Endoscopic resection is feasible for high-grade dysplasia in patients with ulcerative colitis.

BACKGROUND: Endoscopic resection (ER) is feasible for treating well-circumscribed dysplasia in patients with ulcerative colitis (UC). However, long-term prognosis of ER for high-grade dysplasia (HGD) in patients with UC remains unclear. We aimed to evaluate the long-term prognoses of ER for HGD compared with low-grade dysplasia (LGD) and verify the feasibility of ER and follow-up with surveillance colonoscopy for HGD. METHODS: An observational, single-center retrospective study included 38 and 22 patients with LGD and HGD who were followed-up with surveillance colonoscopy after ER. We evaluated the cumulative incidence rate of metachronous HGD or colorectal cancer (CRC) and identified the characteristics of metachronous dysplasia. RESULTS: The median follow-up period was 56 months, and surveillance colonoscopies were performed 3.6 times (mean). The 5-year cumulative incidence rate of HGD/CRC was relatively high in HGD (24.6%) than in LGD (13.7%), but the difference was not significant (p = .16). In HGD cases, six metachronous dysplasia lesions (two LGD and four HGD) were detected 11.6-40.5 months after ER. However, these patients did not progress to CRC. All metachronous lesions were well-circumscribed and with no invisible dysplasia surrounding them; they were 'endoscopically resectable' lesions. Two of the four metachronous HGD lesions were treated endoscopically and two, by colectomy. No synchronous HGD or CRC was detected in the colectomy specimens. CONCLUSIONS: Our results suggest that ER and follow-up with surveillance colonoscopy is feasible in patients with HGD when histological complete resection is achieved.

Learning curve analysis for duodenal endoscopic submucosal dissection: A single-operator experience.

BACKGROUND AND AIMS: Superficial duodenal epithelial tumors are emerging targets for endoscopic submucosal dissection (ESD). However, it is unknown how competence is achieved in duodenal ESD. This study aimed to elucidate the learning curve for duodenal ESD. METHODS: This retrospective observational study included 100 consecutive patients who underwent duodenal ESD by a single endoscopist between March 2014 and September 2021. The primary outcome was to define the learning curve for duodenal ESD by an endoscopist with sufficient non-duodenal ESD experience. Cumulative sum (CUSUM) curve analysis was used to assess the learning curve in terms of procedural speed. Comparative analyses of phases identified using the CUSUM method were performed. RESULTS: In total, 98 patients were included in the analysis. Evaluation of the cumulative sum curve revealed four distinct phases in the graph: phase I, cases 1-25 (learning phase); phase II, cases 26-47 (proficiency phase); phase III, cases 48-72 (mastery phase); and phase IV, cases 73-98 (after introduction of general anesthesia). The median procedural speed was significantly faster in phase II than in phase I (11.1 mm2 /min vs 7.0 mm2 /min, P = .002). Clinically significant intraoperative perforation tended to decrease through phase II to phase IV (22.7%, 12.0%, and 3.8% in phases II, III, and IV, respectively). Delayed perforation occurred only in phases I and II. CONCLUSIONS: Duodenal ESD requires 25 cases to gain proficiency and 50 to achieve mastery even for an endoscopist with extensive non-duodenal ESD experience.

Magnifying endoscopy is useful for tumor border diagnosis in ulcerative colitis patients.

BACKGROUND AND AIMS: Endoscopic resection (ER) is feasible for well-circumscribed tumors in patients with ulcerative colitis (UC); however, the specific manner for diagnosis of the tumor border is unclear. We evaluated the efficacy of magnifying endoscopy (ME) for the diagnosis of tumor borders in UC. METHODS: We analyzed endoscopically or surgically resected tumors in UC patients in whom both chromoendoscopy (CE) and ME were performed, retrospectively. We classified the tumors based on tumor border visibility and evaluated tumor's characteristics and ER outcomes. RESULTS: We examined 100 tumors from 76 UC patients (66 distinct and 34 indistinct on CE). In 22 (65%) indistinct tumors on CE, ME improved the tumor border visibility. Compared with distinct tumors on CE, nonpolypoid and large tumors were more common in indistinct tumors on CE. In indistinct tumors even on ME, flat or depressed morphologies and type V pit were more frequently than in other groups. Sixty-five distinct tumors on CE and 18 distinct tumors on ME alone were treated endoscopically, and their R0 resection rate were 91% and 95% (p > 0.99). CONCLUSIONS: ME can improve the tumor border visibility in UC, and ER is feasible for tumors whose border can be visualized on ME.

Potential roles of gastroesophageal reflux in patients with superficial esophageal squamous cell carcinoma without major causative risk factors.

BACKGROUND: Cigarette smoking, alcohol consumption, and Lugol-voiding lesions (LVLs) are the major causative risk factors of esophageal squamous cell carcinoma (ESCC); however, reports on ESCC cases unrelated to these risk factors are very limited. Here, we investigated the clinicopathological features and etiology of such cases. METHODS: We retrospectively analyzed 704 consecutive superficial ESCC tumors of 512 patients who were treated with endoscopic submucosal dissection. The enrolled patients were divided into two groups-the very low-risk (VLR)-group and risk (R)-group-based on the presence of the abovementioned risks. Clinical, endoscopic, and pathological characteristics and genetic findings were assessed in both groups. RESULTS: The VLR-group consisted of 21 (4.1%) patients, who were characteristically female. Patients in the VLR-group presented gastroesophageal reflux disease (GERD), hiatal hernia, and non-open-type atrophic gastritis, and were negative for Helicobacter pylori. We found unique endoscopic features-frequently observed in the posterior wall of the middle thoracic esophagus-with a linear shape that closely resembled the erosion-like form of GERD. Additionally, histopathological examination showed that these tumors presented atypical nuclei limited to the basal and parabasal layer, sequential to the surrounding changes that presented pathological chronic inflammation of esophagitis. Evaluation of somatic mutations in cancer-related genes using next-generation sequencing revealed that the positive carcinogenic potential (TP53 mutation) of the tumors was relatively frequent in the VLR-group. CONCLUSIONS: Our study suggests that ESCC without major causative factors is related to GERD, with no remarkable oncogenic difference.

The impact of tube replacement timing during LCIG therapy on PEG-J associated adverse events: a retrospective multicenter observational study.

BACKGROUND: Levodopa-carbidopa intestinal gel (LCIG) treatment, a unique drug delivery system for patients with advanced Parkinson's disease (PD), is covered by health insurance in Japan since September 2016. Various LCIG procedure/device-associated adverse events (AEs) have been reported; however, reports on their treatment have been limited. This is the first multicenter study to clarify the frequency and timing of device-related AEs. METHODS: Between September 2016 and December 2018, 104 patients introduced to the LCIG treatment for advanced PD in 11 hospitals were included. The patients' characteristics, AEs incidence, AEs time, and tube exchange time were investigated. RESULTS: The median follow-up period was 21.5 months. Minor AE cases were 29.4%, whereas major AE cases were 43.1%. Majority of major AEs (n = 55, 94.8%) were managed with endoscopic treatment, such as tube exchange. Few severe AEs required surgical treatment (n =3, 5.2%). The mean (range) exposure to percutaneous endoscopic gastrojejunostomy (PEG-J) was 14.7 (0-33) months. One year after the LCIG treatment introduction, 55 patients (54.0%) retained the original PEG-J tube. The mean PEG-J tube exchange time was 10.8 ± 7.0 months in all patients, 11.6 ± 4.7 and 10.5 ± 7.7 months in patients with scheduled exchange and who underwent exchange due to AEs, respectively. CONCLUSIONS: Some device-related AEs occurred during the LCIG treatment; however, only few were serious, most of which could be treated with simple procedures or tube replacement with endoscopy. Therefore, the LCIG treatment is feasible and safe and is a unique treatment option for PD, requiring endoscopists' understanding and cooperation.

Safety and efficacy of water pressure endoscopic submucosal dissection for colorectal tumors with submucosal fibrosis (with video).

BACKGROUND AND AIMS: Colorectal neoplasms with submucosal fibrosis are the most challenging targets of endoscopic resection. Water pressure endoscopic submucosal dissection (WP-ESD) is a recently introduced procedure that has several advantages over conventional endoscopic submucosal dissection (C-ESD). This study aimed to assess the efficacy and safety of WP-ESD for fibrotic colorectal neoplasms. METHODS: This retrospective observational study investigated 133 colorectal neoplasms expected to have submucosal fibrosis that were resected by WP-ESD or C-ESD between April 2012 and April 2020. Eighty-seven lesions after endoscopic or surgical treatment, 18 with biopsy scar with fold convergence and 28 in patients with ulcerative colitis, were included. The differences in treatment outcomes, including procedure time and adverse event proportions, between the WP-ESD and C-ESD groups were analyzed. The clinical course after perforation using WP-ESD was also evaluated, including postprocedural multidetector CT findings obtained immediately after WP-ESD. RESULTS: Severe submucosal fibrosis was observed in 96 lesions (72.2%). The median procedure time was significantly shorter in the WP-ESD group than in the C-ESD group (43.5 minutes [interquartile range {IQR}, 32.8-73] vs 72 minutes [IQR, 45-105]; P = .0041). The multivariate analysis revealed WP-ESD as an independent factor for a short procedure time (odds ratio, 2.90; 95% confidence interval, 1.28-6.55). The proportions of post-ESD electrocoagulation syndrome (11.6% vs 13.1%) and perforation (20.4% vs 22.8%) were similar between the groups. Four of 11 patients with perforation who underwent WP-ESD showed fluid collection on postprocedural multidetector CT images. CONCLUSIONS: WP-ESD can shorten procedure time for treating fibrotic colorectal neoplasms.

Superficial neoplasia involving the Ileocecal valve: Clinical outcomes of endoscopic submucosal dissection.

BACKGROUND: Endoscopic submucosal dissection (ESD) is the treatment of choice for colorectal superficial neoplasia, but certain anatomical locations are challenging even for skilled endoscopists. Ileocecal valve (ICV) is considered a technically challenging site for ESD. OBJECTIVE: Aim of this study was to analyze efficacy and safety of Endoscopic Submucosal Dissection in the treatment of colorectal neoplasia involving the ileocecal valve (ICV) DESIGN: Retrospective study. PATIENTS: We retrospectively evaluated 1507 consecutive patients undergoing ESD at two tertiary referral centres for ESD (Italy and Japan) from January 2008 to March 2020. MAIN OUTCOME MEASURES: Demographic, clinical, procedural, and follow-up data was collected, analysed, and compared between patients with ileocecal valve lesions (ICVL) and patients with non-ICVL. RESULTS: Overall, 1507 patients were enrolled (872 M, 57.8%), of these 53 patients had lesions involving the ICV. Mean age was 70.2 years (range, 53-83 years). En-bloc resection was achieved in 52 (98%) patients. The median specimen size of ICVL was 36.4 mm (range, 8-80 mm), significantly smaller than non-ICVL (p = 0.005). Procedure time was significantly longer in the ICVL group, (71.3 vs. 58.9 min; p = 0.03). Non Granular Type Laterally Spreading Tumors (LST-NG) were significantly more frequent in the ICVL group compared to rectum (52.8% vs. 25.7; p = 0.0001). En-bloc resection rate in the ileocecal region did not differ significantly between groups (p = 0.20). Complications such as perforation and postoperative occurred respectively in 3/53 (5.7%) and 1/53 (2%) patient, and were treated conservatively. At first surveillance colonoscopy performed at 6 months, recurrent adenoma was detected in 2/53 patients (3.9%). CONCLUSIONS: ESD is safe and effective for the treatment of colorectal neoplasia involving the ileocecal valve if performed by expert endoscopist in referral centres.

Small intestinal thrombotic microangiopathy following kidney transplantation diagnosed by balloon-assisted enteroscopy.

Thrombotic microangiopathy (TMA) is a serious complication following kidney transplantation. Although intestinal TMA is a major organ injury and causes abdominal pain, diarrhea and bloody stools, the clinical and endoscopic characteristics of small intestinal TMA remain unclear. Here, we report a drug-induced small intestinal TMA, which did not meet the laboratory-defined TMA criteria but was diagnosed by balloon-assisted enteroscopy (BAE). A 32-year-old woman who underwent kidney transplantation at the age of 10 years complained of abdominal pain, diarrhea and bloody stools one month after starting everolimus (EVE) as an immunosuppressant. Although she did not meet the diagnostic criteria for TMA serologically, BAE revealed a circumferential ulcer in the jejunum, and the pathological findings of a biopsy specimen showed microvascular thrombi, compatible with intestinal TMA. Her symptoms improved upon the discontinuation of EVE, demonstrating that EVE can cause drug-induced intestinal TMA. The present case suggests that BAE should be performed when abdominal pain, diarrhea, and bloody stools occur in patients receiving immunosuppressive medication following kidney transplantation, even if there is no evidence of TMA according to the laboratory definition.

Factors influencing interruption of colorectal endoscopic submucosal dissection.

BACKGROUND AND AIMS: Although colorectal endoscopic submucosal dissection (ESD) has become a standardized procedure worldwide, the difficulty of the procedure is well known. However, there have been no studies assessing the causes of treatment interruption. The present study aimed to evaluate the factors involved in the interruption of colorectal ESD. METHODS: We retrospectively analyzed 1116 consecutive superficial colorectal neoplasms of 1012 patients who were treated with ESD between August 2008 and September 2018. The clinicopathological characteristics and treatment outcomes were analyzed. RESULTS: Interrupted ESD was reported in 14 lesions (1.3%) of the total study population. Univariate analysis of clinical characteristics indicated that age, 0-I macroscopic-type tumor, and tumor location on the left side colon were risk factors for interruption. Multivariate analysis revealed that 0-I macroscopic-type tumor was the sole preoperative independent risk factor for interruption. Univariate analysis revealed that the presence of muscle-retracting sign (MRS), deep submucosal tumor invasion, and intermediate invasive growth pattern represented the etiology of interruption. Multivariate analysis indicated that MRS can be a sole key sign for the interruption. Additionally, the resectability and curability of 0-I type tumors were significantly inferior to those of predominantly lateral spreading tumors. Observations of 0-I macroscopic-type tumors, MRS, and submucosal deep invasion were significantly more frequent in interrupted cases. Conventional endoscopic images without magnification endoscopy were more associated with interruption than irregular surfaces or Vi pit patterns in cases with 0-I type tumors. CONCLUSION: ESD of 0-I type tumors is highly disruptive, and undiagnosable submucosal infiltration can reduce the curability.

Incidence of metachronous gastric cancer after endoscopic submucosal dissection associated with eradication status of Helicobacter pylori.

OBJECTIVE: This study compared the clinicopathological features and treatment outcomes of patients with primary early gastric cancers (EGCs) who had undergone Helicobacter pylori eradication and endoscopic submucosal dissection (ESD) with those of patients who were H. pylori-positive and had undergone ESD. Additionally, we investigated the incidence of metachronous cancer in these patients. METHODS: We retrospectively analyzed 1849 EGCs in 1407 patients who underwent ESD whom 201 primary EGCs were detected after H. pylori eradication (eradication group) and 1648 primary EGCs were detected in patients infected with H. pylori (infection group). We evaluated the clinicopathological features and treatment outcomes of the first ESD. We next divided 938 patients whose follow-up periods were >1 year into three groups, an eradication group (n = 61), an infection group (n = 562), and an eradication after ESD group (n = 315). The groups' cumulative metachronous occurrence rates were determined. RESULTS: The eradication group's median tumor size was significantly smaller, and the tumors were significantly more likely to be flat/depressed than those in the infection group. The groups did not differ regarding the treatment outcomes. The cumulative incidence of metachronous cancer was significantly higher in the eradication group than in the eradication after ESD group (P = 0.0454) and in the infection group than in the eradication after ESD group (P = 0.0233). CONCLUSION: The treatment outcomes for EGC in the eradication group were favorable. The higher incidence of metachronous cancer in the eradication group suggests that careful endoscopic follow-up examinations are required.

Appropriate endoscopic treatment selection and surveillance for superficial non-ampullary duodenal epithelial tumors.

OBJECTIVES: Superficial nonampullary duodenal epithelial tumors (SNADETs) have become frequently detected and referred for endoscopic resection (ER). However, optimal treatment methods and long-term outcomes after ER of SNADETs have not been fully elucidated. We aimed to clarify them by analyzing our large cohort of patients with SNADETs. MATERIALS AND METHODS: We enrolled 190 consecutive tumors from 189 patients undergoing ER between January 2004 and September 2019. Cases were stratified into endoscopic submucosal dissection (ESD), conventional endoscopic mucosal resection, (CEMR) and underwater endoscopic mucosal resection (UEMR). Baseline characteristics and short-term outcomes were compared between the groups. Long-term outcomes were also investigated with a median follow-up of 36 months. RESULTS: ESD significantly exceeded CEMR (96.4% vs. 52.9%; p = .0026) and UEMR (96.4% vs. 50.0%; p = .0008) in complete resection rates for 11- to 20-mm lesions; the differences were not significant for lesions ≤10 mm. Local recurrence only occurred in patients with an incomplete resection. Only patients with submucosal invasion died from the primary neoplasms. The 3- and 5-year disease-free survivals were 91.3% and 83.5%. CONCLUSIONS: While tumors ≤10 mm seem to be good indications for endoscopic mucosal resection, ESD should be considered for larger tumors to better achieve complete resection. Patients with submucosal invasive carcinomas have a great risk of cancer death. Therefore, a close follow-up and an additional treatment are desirable.

An endoscopic treatment strategy for superficial tumors in patients with ulcerative colitis.

BACKGROUND AND AIM: Endoscopic resection is feasible for superficial tumors in patients with ulcerative colitis; however, endoscopic resection options have not been evaluated comprehensively. We evaluated the efficacy and safety of endoscopic submucosal dissection and endoscopic mucosal resection, and decision making regarding endoscopic resection options for patients with ulcerative colitis. METHODS: Endoscopically treated tumors from patients with ulcerative colitis were analyzed retrospectively. We evaluated en bloc and R0 resection, adverse events, local tumor recurrence, and metachronous lesion occurrence rates. RESULTS: We examined 102 tumors (mean size, 12 mm; non-polypoid, 55 tumors) from 74 patients with ulcerative colitis, of whom, 39 and 63 underwent endoscopic submucosal dissection and endoscopic mucosal resection, respectively. The R0 resection rate was significantly higher for endoscopic submucosal dissection (97%) than for endoscopic mucosal resection (80%) (P = 0.0015). For 11-20-mm tumors, the R0 resection rate was significantly higher for endoscopic submucosal dissection (94%) than for endoscopic mucosal resection (55%) (P = 0.0027); the endoscopic submucosal dissection and endoscopic mucosal resection R0 rates did not differ for ≤ 10-mm tumors. The non-polypoid tumor R0 resection rates were significantly higher for endoscopic submucosal dissection (100%) than for endoscopic mucosal resection (65%) (P < 0.001) and did not differ regarding the polypoid tumor R0 resection rates (75% vs 86%, P = 0.49). Four patients experienced intraoperative perforation during endoscopic submucosal dissection. No local recurrences occurred. Metachronous high-grade dysplasia occurred in three patients during surveillance. CONCLUSIONS: In patients with ulcerative colitis, endoscopic submucosal dissection is suitable for ≥ 11-mm and non-polypoid tumors, whereas endoscopic mucosal resection is acceptable for ≤ 10-mm tumors.

Mixed histology poses a greater risk for noncurative endoscopic resection in early gastric cancers regardless of the predominant histologic types.

OBJECTIVES: Clinicopathologic characteristics and treatment outcomes of mixed-histological-type (MT) early gastric cancers (EGCs) treated with endoscopic submucosal dissection (ESD) have not been sufficiently elucidated. We aimed to clarify them in comparison with pure-histological-type EGCs. METHODS: We used 3022 consecutive EGCs in 2281 patients treated with ESD from our prospectively maintained database. Cases were stratified into four groups according to the final diagnosis of the resected specimen are as follows: 2780 pure differentiated-type (DT), 127 DT-predominant MT (D-MT), 87 pure undifferentiated-type (UDT), and 28 UDT-predominant MT (U-MT). Clinicopathologic characteristics and treatment outcome were compared between pure DT and D-MT, and between pure UDT and U-MT separately. Risk factors for deep submucosal invasion, lymphovascular invasion, and a final diagnosis of MT were identified using multivariate analysis. RESULTS: Both D-MT (41.7 vs. 92.0%; P < 0.0001) and U-MT (35.7 vs. 75.9%; P = 0.0002) showed a significantly lower curative resection rate than their pure histologic counterparts. Multivariate analysis revealed that MT was an independent risk factor for deep submucosal (OR 6.55; 95% CI, 4.18-10.14) and lymphovascular (OR 4.74; 95% CI, 2.72-8.29) invasion. Preoperative biopsy results that did not show well-differentiated tubular adenocarcinoma (OR 28.2; 95% CI, 18.9-42.9) were an independent risk factor for a final diagnosis of MT. CONCLUSIONS: MT poses a greater risk for noncurative resection regardless of the predominant histologic types, reflecting more aggressive malignant potential. Although a biopsy examination rarely shows MT, clinicians should consider the possibility of MT when a biopsy examination does not show well-differentiated tubular adenocarcinoma.