Comparison of the efficacy and tolerability of elobixibat plus sodium picosulfate with magnesium citrate and split-dose 2-L polyethylene glycol with ascorbic acid for bowel preparation before outpatient colonoscopy: a study protocol for the multicentre, randomised, controlled E-PLUS trial.

BACKGROUND: Sodium picosulfate (SP)/magnesium citrate (MC) and polyethylene glycol (PEG) plus ascorbic acid are recommended by Western guidelines as laxative solutions for bowel preparation. Clinically, SP/MC has a slower post-dose defaecation response than PEG and is perceived as less cleansing; therefore, it is not currently used for major bowel cancer screening preparation. The standard formulation for bowel preparation is PEG; however, a large dose is required, and it has a distinctive flavour that is considered unpleasant. SP/MC requires a small dose and ensures fluid intake because it is administered in another beverage. Therefore, clinical trials have shown that SP/MC is superior to PEG in terms of acceptability. We aim to compare the novel bowel cleansing method (test group) comprising SP/MC with elobixibat hydrate and the standard bowel cleansing method comprising PEG plus ascorbic acid (standard group) for patients preparing for outpatient colonoscopy. METHODS: This phase III, multicentre, single-blind, noninferiority, randomised, controlled, trial has not yet been completed. Patients aged 40-69 years will be included as participants. Patients with a history of abdominal or pelvic surgery, constipation, inflammatory bowel disease, or severe organ dysfunction will be excluded. The target number of research participants is 540 (standard group, 270 cases; test group, 270 cases). The primary endpoint is the degree of bowel cleansing (Boston Bowel Preparation Scale [BBPS] score ≥ 6). The secondary endpoints are patient acceptability, adverse events, polyp/adenoma detection rate, number of polyps/adenomas detected, degree of bowel cleansing according to the BBPS (BBPS score ≥ 8), degree of bowel cleansing according to the Aronchik scale, and bowel cleansing time. DISCUSSION: This trial aims to develop a "patient-first" colon cleansing regimen without the risk of inadequate bowel preparation by using both elobixibat hydrate and SP/MC. TRIAL REGISTRATION: Japan Registry of Clinical Trials (jRCT; no. s041210067; 9 September 2021; https://jrct.niph.go.jp/ ), protocol version 1.5 (May 1, 2023).

Outcomes of Metastatic and Unresectable Small Bowel Adenocarcinoma in Japan According to the Treatment Strategy: A Nationwide Observational Study.

PURPOSE: Limited information is available regarding the characteristics and outcomes of stage IV small bowel adenocarcinoma (SBA) in Japan. This study examined the clinical and pathological characteristics and outcomes according to the treatment strategies in patients with stage IV SBA. METHODS: This retrospective observational study used the data of patients with jejunal or ileal adenocarcinoma collected by the Small Bowel Malignant Tumor Project of the Japanese Society for Cancer of the Colon and Rectum. Descriptive statistics were expressed as the mean (standard deviation) or median (range). Survival analysis was performed using Kaplan-Meier curves and pairwise log-rank tests. RESULTS: Data from 128 patients were analyzed. The treatment strategies were chemotherapy alone (26 of 128, 20.3%), surgery alone (including palliative surgery; 21 of 128, 16.4%), surgery + chemotherapy (74 of 128, 57.8%), and best supportive care (7 of 128, 5.5%). The median (range) overall survival was 16 (0-125) months overall, and 11 (1-38) months, 8 (0-80) months, 18 (0-125) months, and 0 (0-1) months for the chemotherapy, surgery, surgery + chemotherapy, and best supportive care groups, respectively. Three main categories of chemotherapeutic regimen were used: a combination of fluoropyrimidine and oxaliplatin (F + Ox), fluoropyrimidine and irinotecan (F + Iri), and single-agent fluoropyrimidine. Among patients treated with chemotherapy, the median (range) OS was 16 (1-106) months overall, and 17 (1-87) months, 29 (7-39) months, and 16 (1-106) months in patients treated with fluoropyrimidine, F + Iri, and F + Ox, respectively. CONCLUSION: Patients treated with surgery, chemotherapy, or both had a better prognosis than those who received best supportive care. Among patients who received chemotherapy, survival did not differ according to the chemotherapeutic regimen.

Clinicopathological features and prognosis of primary small bowel adenocarcinoma: a large multicenter analysis of the JSCCR database in Japan.

BACKGROUND: The clinicopathological features and prognosis of primary small bowel adenocarcinoma (PSBA), excluding duodenal cancer, remain undetermined due to its rarity in Japan. METHODS: We analyzed 354 patients with 358 PSBAs, between January 2008 and December 2017, at 44 institutions affiliated with the Japanese Society for Cancer of the Colon and Rectum. RESULTS: The median age was 67 years (218 males, 61.6%). The average tumor size was 49.9 (7-100) mm. PSBA sites consisted of jejunum (66.2%) and ileum (30.4%). A total of 219 patients (61.9%) underwent diagnostic small bowel endoscopy, including single-balloon endoscopy, double-balloon endoscopy, and capsule endoscopy before treatment. Nineteen patients (5.4%) had Lynch syndrome, and 272 patients (76.8%) had symptoms at the initial diagnosis. The rates for stages 0, I, II, III, and IV were 5.4%, 2.5%, 27.1%, 26.0%, and 35.6%, respectively. The 5-year overall survival rates at each stage were 92.3%, 60.0%, 75.9%, 61.4%, and 25.5%, respectively, and the 5-year disease-specific survival (DSS) rates were 100%, 75.0%, 84.1%, 59.3%, and 25.6%, respectively. Patients with the PSBA located in the jejunum, with symptoms at the initial diagnosis or advanced clinical stage had a worse prognosis. However, multivariate analysis using Cox-hazard model revealed that clinical stage was the only significant predictor of DSS for patients with PSBA. CONCLUSIONS: Of the patients with PSBA, 76.8% had symptoms at the initial diagnosis, which were often detected at an advanced stage. Detection during the early stages of PSBA is important to ensure a good prognosis.

Endoscopist-related factors affecting adenoma detection during colonoscopy: Data from the J-SCOUT study.

OBJECTIVES: Colonoscopy withdrawal times are associated with the adenoma detection rate (ADR). However, the relationship between ADR and cecal insertion time has been inadequately characterized. We aimed to evaluate endoscopist-related factors involved in the ADR, including the average individual colonoscopy insertion and withdrawal times. METHODS: This observational study used a colonoscopy database with pathology data from routine clinical practice in Japanese institutions. The odds ratios (OR) of endoscopist-related factors related to ADRs were examined using a generalized linear mixed model. RESULTS: Of the 186,293 colonoscopies performed during the study period, 47,705 colonoscopies by 189 endoscopists in four hospitals were analyzed for ADR. The overall ADR was 38.3% (95% confidence interval [CI] 37.8, 38.7). Compared to endoscopists with mean cecal insertion times of <5 min, the OR of ADR for those with mean cecal insertion times of 5-9, 10-14, and ≥15 min were 0.84 (95% CI 0.71, 0.99), 0.68 (95% CI 0.52, 0.90), and 0.45 (95% CI 0.25, 0.78), respectively. Compared to endoscopists with mean withdrawal times of <6 min, the OR of ADR for those with mean withdrawal times of 6-9, 10-14, and ≥15 min were 1.38 (95% CI 1.03, 1.85), 1.48 (95% CI 1.09, 2.02), and 1.68 (95% CI 1.04, 2.61), respectively. There were no significant differences in ADRs by endoscopist specialty, gender, or the total number of examinations performed. CONCLUSION: Individual mean colonoscopy insertion time was associated with ADR and might be considered as a colonoscopy quality indicator as well as withdrawal time.

Sleeping Beauty transposon mutagenesis identified genes and pathways involved in inflammation-associated colon tumor development.

Chronic inflammation promotes development and progression of colorectal cancer (CRC). To comprehensively understand the molecular mechanisms underlying the development and progression of inflamed CRC, we perform in vivo screening and identify 142 genes that are frequently mutated in inflammation-associated colon tumors. These genes include senescence and TGFß-activin signaling genes. We find that TNFα can induce stemness and activate senescence signaling by enhancing cell plasticity in colonic epithelial cells, which could act as a selective pressure to mutate senescence-related genes in inflammation-associated colonic tumors. Furthermore, we show the efficacy of the Cdk4/6 inhibitor in vivo for inflammation-associated colonic tumors. Finally, we functionally validate that Arhgap5 and Mecom are tumor suppressor genes, providing possible therapeutic targets for CRC. Thus, we demonstrate the importance of the inactivation of senescence pathways in CRC development and progression in an inflammatory microenvironment, which can help progress toward precision medicine.

Effectiveness of endoscopic resection for colorectal neoplasms in ulcerative colitis: a multicenter registration study.

BACKGROUND AND AIMS: Patients with ulcerative colitis (UC) are at risk of developing colorectal cancer. The feasibility of endoscopic resection (ER) for UC-associated neoplasia has been suggested, but its efficacy and safety remain unclear. We aimed to assess the efficacy and safety of ER for colorectal neoplasms in patients with UC. METHODS: This was a retrospective, multicenter cohort study of patients with UC who initially underwent ER or surgery for colorectal neoplasms between April 2015 and March 2021. Patients who had prior colorectal neoplastic lesions were excluded. RESULTS: Among 213 men and 123 women analyzed, the mean age at UC onset was 41.6 years, and the mean age at neoplasia diagnosis was 56.1 years for 240 cases of total colitis, 59 cases of left-sided colitis, 31 cases of proctitis, and 6 cases of segmental colitis. EMR was performed for 142 lesions, and endoscopic submucosal dissection (ESD) was performed for 96 lesions. The perforation rate was 2.5% for all 238 lesions removed by ER and 6.3% for the 96 lesions removed by ESD. Among 146 ER lesions followed up with endoscopy, the local recurrence rate was 2.7%. The incidence of metachronous neoplasia after ER was 6.1%. All patients were followed a median of 34.7 months after initial treatment, and 5 died (all surgical cases). Overall survival was significantly higher in the ER group than in the surgery group (P = .0085). CONCLUSIONS: ER for colorectal neoplasms in UC may be acceptable in selected cases, although follow-up for metachronous lesions is necessary.

Difficult colorectal polypectomy: Technical tips and recent advances.

Colonoscopy has been shown to be an effective modality to prevent colorectal cancer (CRC) development. CRC reduction is achieved by detecting and removing adenomas, which are precursors of CRC. Most colorectal polyps are small and do not pose a significant challenge for trained and skilled endoscopists. However, up to 15% of polyps are considered "difficult", potentially causing life-threatening complications. A difficult polyp is defined as any polyp that is challenging for the endoscopist to remove owing to its size, shape, or location. Advanced polypectomy techniques and skills are required to resect difficult colorectal polyps. There were various polypectomy techniques for difficult polyps such as endoscopic mucosal resection (EMR), underwater EMR, Tip-in EMR, endoscopic submucosal dissection (ESD), or endoscopic full-thickness resection. The selection of the appropriate modality depends on the morphology and endoscopic diagnosis. Several technologies have been developed to aid endoscopists in performing safe and effective polypectomies, especially complex procedures such as ESD. These advances include video endoscopy system, equipment assisting in advanced polypectomy, and closure devices/techniques for complication management. Endoscopists should know how to use these devices and their availability in practice to enhance polypectomy performance. This review describes several useful strategies and tips for managing difficult colorectal polyps. We also propose the stepwise approach for difficult colorectal polyps.

The methylation level of a single cancer risk marker gene reflects methylation burden in gastric mucosa.

BACKGROUND: Gastric cancer risk can be accurately predicted by measuring the methylation level of a single marker gene in gastric mucosa. However, the mechanism is still uncertain. We hypothesized that the methylation level measured reflects methylation alterations in the entire genome (methylation burden), induced by Helicobacter pylori (H. pylori) infection, and thus cancer risk. METHODS: Gastric mucosa of 15 healthy volunteers without H. pylori infection (G1), 98 people with atrophic gastritis (G2), and 133 patients with gastric cancer (G3) after H. pylori eradication were collected. Methylation burden of an individual was obtained by microarray analysis as an inverse of the correlation coefficient between the methylation levels of 265,552 genomic regions in the person's gastric mucosa and those in an entirely healthy mucosa. RESULTS: The methylation burden significantly increased in the order of G1 (n = 4), G2 (n = 18), and G3 (n = 19) and was well correlated with the methylation level of a single marker gene (r = 0.91 for miR124a-3). The average methylation levels of nine driver genes tended to increase according to the risk levels (P = 0.08 between G2 vs G3) and was also correlated with the methylation level of a single marker gene (r = 0.94). Analysis of more samples (14 G1, 97 G2, and 131 G3 samples) yielded significant increases of the average methylation levels between risk groups. CONCLUSIONS: The methylation level of a single marker gene reflects the methylation burden, which includes driver gene methylation, and thus accurately predicts cancer risk.

Factors associated with increased duration of endoscopic submucosal dissection for rectal tumors: a 22-year retrospective analysis.

BACKGROUND AND AIMS: Colorectal endoscopic submucosal dissection (ESD) is widely used in several countries. However, it is associated with technical difficulties. Rectal ESD is considered an intermediate step in colorectal ESD training. Nevertheless, some rectal lesions require a longer procedure time than usual, and the reason for the prolonged time taken for these specific lesions remains unclear. Therefore, this study aimed to clarify the factors associated with prolonged rectal ESD. METHODS: In total, 483 rectal lesions resected using ESD from February 1998 to June 2021 were investigated. Prolonged ESD procedure time was defined as the time from the first submucosal injection to lesion removal exceeding 120 minutes, whereas other procedures were defined as average ESD procedure time. Clinicopathologic and endoscopic findings were compared between the 2 groups using univariate and multivariate analyses. RESULTS: One hundred forty-four lesions were resected using a prolonged ESD procedure time of 202.9 ± 92.3 minutes, whereas 339 lesions were resected using an average ESD procedure time of 77.8 ± 29.4 minutes. Multivariate analysis revealed that tumors involving the dentate line (P = .026), resection size ≥50 mm (P < .001), invasion depth ≥T1b (P = .006), and circumferential range ≥2/3 (P = .001) were independent risk factors for prolonged-duration ESDs, regardless of whether the procedure was performed by an expert or not. CONCLUSIONS: The results of the present study suggest that the location of a lesion involving the dentate line, resection size ≥50 mm, circumferential range ≥2/3, and invasion depth ≥T1b are the independent risk factors for prolonged ESD procedure time.

Management of colorectal high-grade dysplasia or cancer resected by cold snare polypectomy: a multicenter exploratory study.

BACKGROUND: The clinical course and surveillance strategy for patients who undergo cold snare polypectomy (CSP) for high-grade dysplasia (HGD) or cancer is unclear. We investigated the management of colorectal HGDs and cancers following CSP. METHODS: This Japanese nationwide multicenter exploratory study was retrospectively conducted on patients who had undergone CSP for colorectal HGDs or cancers and follow-up colonoscopy at least once from 2014 to 2020. We investigated the detection rate of CSP scars, local recurrence rate (LRR), risk factors for local recurrence, and follow-up strategy. This study was registered with the University Hospital Medical Information Network (UMIN) Clinical Trials Registry (UMIN000043670). RESULTS: We included 155 patients with 156 lesions. CSP scars were identified in 22 (31.4%), 41 (54.7%), and 10 (90.9%) patients with curative, borderline, and non-curative resection, respectively. Among them, residual tumors were observed in one (4.5%), six (14.6%), and three (30.0%) cases, respectively. The total LRR was 13.7% (95% confidence interval: 6.8-23.8). R1 resection cases (either horizontal or vertical margins positive for tumors) were associated with local recurrence (p = 0.031). Salvage endoscopic and surgical resections were performed on 21 and 10 patients, respectively. Among them, the proportion of endoscopically suspected residual tumors was significantly higher (p < 0.001) in the residual tumor-positive group (100%) than in the residual tumor-negative group (28.6%). CONCLUSIONS: LRR after CSP for HGDs or cancers was 13.7% based on scar-identified cases. Salvage endoscopic or surgical resection should be performed according to the curability of the lesion and endoscopic findings during colonoscopic surveillance.

Cost-effectiveness analysis of computer-aided detection systems for colonoscopy in Japan.

OBJECTIVES: The usefulness of computer-aided detection systems (CADe) for colonoscopy has been increasingly reported. In many countries, however, data on the cost-effectiveness of their use are lacking; consequently, CADe for colonoscopy has not been covered by health insurance. We aimed to evaluate the cost-effectiveness of colonoscopy using CADe in Japan. METHODS: We conducted a simulation model analysis using Japanese data to examine the cost-effectiveness of colonoscopy with and without CADe for a population aged 40-74 years who received colorectal cancer (CRC) screening with a fecal immunochemical test (FIT). The rates of receiving FIT screening and colonoscopy following a positive FIT were set as 40% and 70%, respectively. The sensitivities of FIT for advanced adenomas and CRC Dukes' A-D were 26.5% and 52.8-78.3%, respectively. CADe colonoscopy was judged to be cost-effective when its incremental cost-effectiveness ratio (ICER) was below JPY 5,000,000 per quality-adjusted life-years (QALYs) gained. RESULTS: Compared to conventional colonoscopy, CADe colonoscopy showed a higher QALY (20.4098 vs. 20.4088) and lower CRC incidence (2373 vs. 2415 per 100,000) and mortality (561 vs. 569 per 100,000). When the CADe cost was set at JPY 1000-6000, the ICER per QALY gained for CADe colonoscopy was lower than JPY 5,000,000 (JPY 796,328-4,971,274). The CADe cost threshold at which the ICER for CADe colonoscopy exceeded JPY 5,000,000 was JPY 6040. CONCLUSIONS: Computer-aided detection systems for colonoscopy has the potential to be cost-effective when the CADe cost is up to JPY 6000. These results suggest that the insurance reimbursement of CADe for colonoscopy is reasonable.

Detection of colorectal adenomas with texture and color enhancement imaging: Multicenter observational study.

OBJECTIVES: We aimed to evaluate the efficacy of texture and color enhancement imaging (TXI), which allows the acquisition of brighter images with enhanced color and surface structure in colorectal polyp detection compared to white light imaging. METHODS: Patients who underwent colonoscopy with repeated ascending colon observation using TXI and white light imaging between August 2020 and January 2021 were identified in three institutions. The outcomes included the mean number of adenomas detected per procedure (MAP), adenoma detection rate (ADR), and ascending colonic adenoma miss rate (Ac-AMR). Logistic regression was used to determine the effects of the variables on the outcomes. RESULTS: We included 1043 lesions from 470 patients in the analysis. The MAP, ADR, flat polyp detection rate, and Ac-AMR in TXI and white light imaging were 1.5% (95% confidence interval 1.3-1.6%) vs. 1.0% (0.9-1.1%), 58.2% (51.7-64.6%) vs. 46.8% (40.2-53.4%), 66.2% (59.8-72.2%) vs. 49.8% (43.2-56.4%), and 17.9% (12.1-25.2%) vs. 28.2% (20.0-37.6%), respectively. TXI, age, withdrawal time, and endoscopy type were identified as significant factors affecting the MAP and the ADR using multivariate regression analysis. CONCLUSIONS: Our study indicates that TXI improve the detection of colorectal neoplastic lesions. However, prospective randomized trials are required to confirm these findings.

Prediction of tissue-of-origin of early stage cancers using serum miRNomes.

BACKGROUND: Noninvasive detection of early stage cancers with accurate prediction of tumor tissue-of-origin could improve patient prognosis. Because miRNA profiles differ between organs, circulating miRNomics represent a promising method for early detection of cancers, but this has not been shown conclusively. METHODS: A serum miRNA profile (miRNomes)-based classifier was evaluated for its ability to discriminate cancer types using advanced machine learning. The training set comprised 7931 serum samples from patients with 13 types of solid cancers and 5013 noncancer samples. The validation set consisted of 1990 cancer and 1256 noncancer samples. The contribution of each miRNA to the cancer-type classification was evaluated, and those with a high contribution were identified. RESULTS: Cancer type was predicted with an accuracy of 0.88 (95% confidence interval [CI] = 0.87 to 0.90) in all stages and an accuracy of 0.90 (95% CI = 0.88 to 0.91) in resectable stages (stages 0-II). The F1 score for the discrimination of the 13 cancer types was 0.93. Optimal classification performance was achieved with at least 100 miRNAs that contributed the strongest to accurate prediction of cancer type. Assessment of tissue expression patterns of these miRNAs suggested that miRNAs secreted from the tumor environment could be used to establish cancer type-specific serum miRNomes. CONCLUSIONS: This study demonstrates that large-scale serum miRNomics in combination with machine learning could lead to the development of a blood-based cancer classification system. Further investigations of the regulating mechanisms of the miRNAs that contributed strongly to accurate prediction of cancer type could pave the way for the clinical use of circulating miRNA diagnostics.

Unexpected adhesive bowel obstruction after endoscopic submucosal dissection of early sigmoid colon cancer.

Various complications of colorectal endoscopic submucosal dissection (ESD) have been reported, including bleeding, penetration, perforation, and coagulation syndrome. However, the occurrence of bowel obstruction after ESD is rare. We report a case of adhesive bowel obstruction after ESD for a laterally spreading tumor in the sigmoid colon. The 35-mm tumor was successfully removed by ESD without intraoperative complications. The patient had a fever, lower abdominal pain, and a small amount of bloody stool the day after ESD. Endoscopy revealed minor bleeding from the ESD scar, which was treated by hemostatic clips. Pathological analysis showed adenocarcinoma was exposed to the vertical margin; therefore, the resection was non-curative. At 39 days after ESD and 36 days after discharge, the patient had abdominal pain and nausea. She was readmitted with a diagnosis of adhesive bowel obstruction. Conservative treatment was ineffective; therefore, she underwent sigmoidectomy combined with partial resection of the small intestine because of small intestinal stenosis caused by inflammation. The pathological examination showed localized peritonitis around the sigmoid colon where ESD was performed. There was more fibrosis along the serous surface of the small intestine than on the sigmoid colon. We concluded that there was a micro-perforation that could not be detected by endoscopy or physical examination. This case indicates that adhesive bowel obstruction may occur as a complication of ESD.

"Endoscopic" adenoma detection rate as a quality indicator of colonoscopy: First report from the J-SCOUT study.

OBJECTIVES: To examine whether reasonable detection rate of endoscopically diagnosed lesions as adenoma ("endoscopic" adenoma detection rate [ADR]) could be calculated with a database generated from colonoscopy reports and whether it could be used as a surrogate colonoscopy quality indicator of "pathological" ADR. METHODS: A lesion-by-lesion database of colonoscopies performed between 2010 and 2020 at eight Japanese endoscopy centers and corresponding pathology database were integrated. Differences in numbers of detected polyps, "endoscopic" and "pathological" adenomas, and what these differences could be attributed to were examined. Polyp detection rate (PDR), "endoscopic" and "pathological" ADRs, and correlation coefficients between "pathological" ADR and PDR or "endoscopic" ADR by each endoscopist were calculated. RESULTS: Overall, 129,065 colonoscopy reports were analyzed. Among a total of 146,854 polyps, more "endoscopic" adenomas (n = 117,359) were observed than "pathological" adenomas (n = 70,076), primarily because adenomas were not resected on site, rather than because of a misdiagnosis. In all patients analyzed, PDR, "endoscopic" and "pathological" ADRs were 56.4% (95% confidence interval [CI] 56.2-56.7), 48.0% (95% CI 47.7-48.3), and 32.7% (95% CI 32.5-33.0), respectively. "Endoscopic" and "pathological" ADRs from each endoscopist showed a high correlation in hospitals where adenomas were usually resected at the time of examination. CONCLUSIONS: By appropriately describing endoscopically diagnosed lesions as "adenomas" in endoscopy reports, "endoscopic" ADR might be used as a surrogate colonoscopy quality indicator of "pathological" ADR (UMIN000040690).

Robust automated prediction of the revised Vienna Classification in colonoscopy using deep learning: development and initial external validation.

BACKGROUND: Improved optical diagnostic technology is needed that can be used by also outside expert centers. Hence, we developed an artificial intelligence (AI) system that automatically and robustly predicts the pathological diagnosis based on the revised Vienna Classification using standard colonoscopy images. METHODS: We prepared deep learning algorithms and colonoscopy images containing pathologically proven lesions (56,872 images, 6775 lesions). Four classifications were adopted: revised Vienna Classification category 1, 3, and 4/5 and normal images. The best algorithm-ResNet152-in the independent internal validation (14,048 images, 1718 lesions) was used for external validation (255 images, 128 lesions) based on neoplastic and non-neoplastic classification. Diagnostic performance of endoscopists was compared using a computer-assisted interpreting test. RESULTS: In the internal validation, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy for adenoma (category 3) of 84.6% (95% CI 83.5-85.6%), 99.7% (99.5-99.8%), 90.8% (89.9-91.7%), 89.2% (88.5-99.0%), and 89.8% (89.3-90.4%), respectively. In the external validation, ResNet152's sensitivity, specificity, PPV, NPV, and accuracy for neoplastic lesions were 88.3% (82.6-94.1%), 90.3% (83.0-97.7%), 94.6% (90.5-98.8%), 80.0% (70.6-89.4%), and 89.0% (84.5-93.6%), respectively. This diagnostic performance was superior to that of expert endoscopists. Area under the receiver-operating characteristic curve was 0.903 (0.860-0.946). CONCLUSIONS: The developed AI system can help non-expert endoscopists make differential diagnoses of colorectal neoplasia on par with expert endoscopists during colonoscopy. (229/250 words).

Conditional inference tree models to perceive depth of invasion in T1 colorectal cancer.

BACKGROUND AND AIM: Accurate diagnosis of invasion depth for T1 colorectal cancer is of critical importance as it decides optimal resection technique. Few reports have previously covered the effects of endoscopic morphology on depth assessment. We developed and validated a novel diagnostic algorithm that accurately predicts the depth of early colorectal cancer. METHODS: We examined large pathological and endoscopic databases compiled between Jan 2015 and Dec 2018. Training and validation data cohorts were derived and real-world diagnostic performance of two conditional interference tree algorithms (Models 1 and 2) was evaluated against that of the Japan NBI-Expert Team (JNET) classification used by both expert and non-expert endoscopists. RESULTS: Model 1 had higher sensitivity in deep submucosal invasion than that of JNET alone in both training (45.1% vs. 28.6%, p < 0.01) and validation sets (52.3% vs. 40.0%, p < 0.01). Model 2 demonstrated higher sensitivity than Model 1 (66.2% vs. 52.3%, p < 0.01) in excluding deeper invasion of suspected Tis/T1a lesions. CONCLUSION: We discovered that machine-learning classifiers, including JNET and macroscopic features, provide the best non-invasive screen to exclude deeper invasion for suspected Tis/T1 lesions. Adding this algorithm improves depth diagnosis of T1 colorectal lesions for both expert and non-expert endoscopists.