Introduction

Several studies have shown that image enhancement technologies like NBI and BLI in expert hands can meet the PIVI criteria [1-6]. This has led to various calls for implementation of this technique in clinical practice. However, attempts at more widespread generalisation of optical diagnosis have failed due to the inability of general endoscopists to meet the established standards.
CAD-EYE (Fujifilm, Japan) is a novel AI driven optical diagnosis system (CADx) which has shown encouraging preliminary results in expert hands or in single centre studies.

Aims & Methods

To evaluate the diagnostic performance of AI (CAD-EYE) based optical diagnosis of colorectal polyps in the hands of both experts and non-experts and to explore the impact of AI based prediction on the learning curve of non-expert endoscopists.
A multicenter prospective observational study which compares the performance of CAD-EYE with that of the endoscopists, combined with monthly review meetings by all endoscopists involved in the study with the aim of group learning and improving the skills of optical diagnosis for non-expert endoscopists.

Results

Optical diagnosis (OD) was performed with high confidence by the endoscopists in 674/782 (86.2%) of polyps. High confidence OD was achieved in 91.5% polyps by expert endoscopists compared to 85.1% by non-expert endoscopists [p<0.05].
The overall sensitivity, NPV and accuracy by all endoscopists were 84.4%, 74.1% and 86.2% respectively. CAD-EYE diagnosis was recorded in 672/782 (85.9%) of polyps. The overall sensitivity, NPV and accuracy were 83.0%, 72.8% and 85.9% respectively [p>0.05].
Optical diagnosis was performed with high confidence by the endoscopist in 416/590 (84.9%) of diminutive polyps (≤5mm). The overall sensitivity, NPV and accuracy were 84.6%, 74.5% and 84.9% respectively. CAD-EYE diagnosis was recorded in 413/590 (84.2%) of diminutive polyps. The overall sensitivity, NPV and accuracy was 83.4%, 73.2% and 84.2% respectively [p>0.05].
273/590 diminutive polyps were located in the rectosigmoid. Endoscopist’s sensitivity, NPV and accuracy was 93.1%, 92.6% and 91.9% respectively, compared to CAD-EYE whose sensitivity, NPV and accuracy were 93.7%, 92.4% and 92.9% respectively [p>0.05]
The endoscopist’s optical diagnosis-based post-polypectomy surveillance interval was calculated in all patients and compared with histology based surveillance and it achieved a 90.7% [CI 84.6-93.8] and 92.9% [CI 89.4-95.0 ] concordance with BSG and ESGE guidelines respectively; achieving the ASGE PIVI criterion. The post-polypectomy surveillance interval based on CAD-EYE diagnosis was calculated and compared with histology based surveillance and it achieved 91.9% [CI 88.7%–93.2 %] and 93.4 % [CI 90.2 %–96.8%] concordance with the BSG & ESGE guidelines; also achieving the ASGE PIVI criterion.
An improvement in the performance of non-experts during the study was noted with an improvement in accuracy from 63% for the first 50 polyps to 85% for the last 50 polyps, with an associated increase in sensitivity (65% and 82%), specificity (50% and 91%), NPV (55% and 79%) and also degree of confidence (65% and 90%) [p<0.05 for all KPIs]

Conclusion

Our data proves that CAD-EYE based diagnosis can meet both PIVI criterion and help introduce both resect & discard and leave in-situ strategies even in the hands of non-experts.
Real time Optical diagnosis of polyps has the potential to transform clinical practice by reducing pathology & polypectomy costs, setting up surveillance intervals without waiting for histology, reducing morbidity and procedure time.
We have also demonstrated an improvement in performance of non-expert endoscopists over the course of the study. This is a novel finding and has not been shown in any study before and we feel that this is an important finding which can help inform the policy makers to invest in CADx.

References

1. Abu Dayyeh BK, Thosani N, Konda V et al. ASGE technology committee systematic review and meta-analysis assessing the ASGE PIVI thresholds for adopting real-time endoscopic assessment of the histology of diminutive colorectal polyps. Vol. 81, Gastrointestinal Endoscopy. 2015. p. 502–16.
2. Rees C, Rajashekar P, Wilson A et al: Narrow band imaging optical diagnosis of small colorectal polyps in routine clinical practice: the Detect Inspect Characterise Resect and Discard 2 (DISCARD 2) study. Gut 2017;66:887–895.
3. Neumann H, Vieth M, Bisscops R et al. Leaving colorectal polyps in place can be achieved with high accuracy using blue light imaging (BLI) United European Gastroenterol J. 2018 Aug
4. Rondonotti E, Paggi S, Amato A et al. Blue-light imaging compared with high-definition white light for real-time histology prediction of colorectal polyps less than 1 centimeter: a prospective randomized study.
5. Vleugels JLA, Hazewinkel Y, Dijkgraa MGW. Optical diagnosis expanded to small polyps: post-hoc analysis of diagnostic performance in a prospective multicenter study. Endoscopy 2019; 51: 244-252)
6. Rondonotti E, Hassan C, Andrealli A et al. Clinical validation of BASIC classification for the resect and discard strategy for diminutive colorectal polyps. Clin Gastroenterol Hepatol 2020; 18: 2357–2365

Introduction

Currently, the only management option for patients with celiac disease (CeD) is lifelong adherence to a gluten-free diet (GFD). Here, we report the results of a phase 2 study of the computationally designed glutenase TAK-062 (zamaglutenase) for the treatment of CeD in patients on a GFD (NCT05353985).

Aims & Methods

Eligible patients were aged >18 years with ongoing CeD-related gastrointestinal (GI) symptoms and villous height to crypt depth ratio (Vh:Cd) <2.5, despite maintenance of a GFD for ≥1 year. Study phases were: (1) a 4-week screening period to confirm symptoms and perform endoscopy; (2) a 4-week single-blind placebo run-in, including 2 weeks receiving placebo simulated inadvertent gluten exposure (SIGE) bars (Weeks −4 to −2) and 2 weeks receiving SIGE bars containing approximately 500 mg of gluten 3× per week (TIW; Week −2 to Day 1); and (3) a 24-week treatment period with patients randomized to receive TAK-062 600 mg or placebo with meals 3× per day (TID), plus gluten-containing SIGE TIW, followed by an endoscopy. Symptoms were assessed daily using the 5-item Celiac Disease Symptom Diary v2.1 (CDSD), that evaluates the severity of CeD symptoms on a 5-point Likert scale.

The primary endpoint was change from baseline to Week 12 in mean weekly CDSD GI symptom severity scores, and the secondary endpoint was change from baseline to Week 24 in Vh:Cd. Safety was assessed throughout the study. Differences between groups were assessed using a mixed model for repeated measures (primary endpoint) and ANCOVA (secondary endpoint) at a 2-sided significance level of 0.05.

Results

A total of 153 patients were randomized (placebo: n = 76; TAK-062: n = 77); one patient in the TAK-062 arm did not receive treatment. Mean age was 46.0 years, 75.2% of patients were female, mean time since diagnosis was 8.7 years, and 24.2% were seropositive. Mean baseline Vh:Cd was 1.78 and mean baseline weekly CDSD GI symptom severity score was 0.93. There were no appreciable differences between groups in baseline characteristics.

There was no significant improvement in weekly CDSD GI symptom severity score from baseline to Week 12 in either arm, and no significant difference between groups; non-significant improvements of approximately 0.1 were observed in both arms (Table). Vh:Cd remained stable to Week 24 in the placebo arm, but significantly decreased by 0.34 (95% confidence interval: −0.53, −0.14) in the TAK-062 arm, with a significant difference between groups (Table). Similar results were observed in all subgroups assessed.

TAK-062 was clinically well tolerated despite the significant decrease in Vh:Cd, with comparable proportions of patients in both arms experiencing an adverse event (AE) (placebo: 64.5%; TAK-062: 73.7%) or serious AE (placebo: 5.3%; TAK-062: 1.3%). The most frequent AEs were GI-related and occurred at a similar frequency in both arms, with the exception of diarrhea (placebo: 2.6%; TAK-062: 17.1%).

Conclusion

TAK-062 did not demonstrate efficacy for improving symptoms or histology in patients with ongoing active CeD on a GFD. Vh:Cd significantly worsened in the TAK-062 arm, with a significant difference between groups; possible reasons for this are under investigation. Stable Vh:Cd in the placebo arm indicates that SIGE successfully mitigates trial-related disease improvements previously observed in similar cohorts.

Disclosure

LE has nothing to disclose. ML reports receiving consulting fees from Anokion, EllaOla, and Takeda, and research funding from Moderna (institution) and Takeda (institution). PR has nothing to disclose. CC has nothing to disclose. JS reports receiving consulting fees from Chugai, GIG, and Takeda, and research funding from Takeda. JAM reports grants from ImmusanT, National Institutes of Health, Department of Defense, Immunogenix, Takeda Pharmaceuticals, Allakos, Provention Bio, Oberkotter Foundation, and 9 Meters Biopharma; a contract (to institution) from Kanyos Bio (a wholly owned subsidiary of Anokion); consultancy fees from Johnson and Johnson, Bristol Myers Squibb, Intrexon Corporation, Dren Bio, Neoleukin Therapeutics, Reistone Biopharma, Immunic Therapeutics, Senda Biosciences, Brightseed Bio, Chugai Pharma, Alimentiv, Equillium, Interlude Biopharma, Ukko Inc., argenx; has received royalties from Torax Medical and Evelo Biosciences; and holds shares in Entero Healthcare and Minnesota Medical Technologies. DSS reports receiving honoraria and research funding from Takeda. HL, JRM, MZ, SG, and DAL are employees and shareholders of Takeda.

Introduction

Accurate endoscopic diagnosis of colorectal polyps is essential for determining appropriate treatment strategies. The Japan NBI Expert Team (JNET) classification and pit pattern diagnosis are used for diagnosis, but their comparative diagnostic performance in real-world clinical settings remains unclear, particularly between experts and non-experts.

Aims & Methods

This post-hoc analysis aimed to evaluate the diagnostic accuracy of both classification across different levels of expertise using data from a prospective multicenter trial. We conducted a post-hoc analysis of the deTXIon trial¹ , a randomized controlled trial conducted at eight Japanese institutions between March and October 2023. From the per-protocol set, we analyzed colorectal lesions that had both JNET classification and pit pattern diagnosis with confirmed histopathology. JNET classification categorizes lesions into Type 1 (corresponding to hyperplastic polyps [HPs] and sessile serrated lesions [SSLs]), Type 2A (low-grade adenomas), Type 2B (high-grade adenomas or superficial submucosal invasive cancer), and Type 3 (deep submucosal invasive cancer). Pit pattern classification included Type II ( HPs/SSLs ), Type III_S/III_L/IV (low-grade adenomas), Type V_I low-grade (high-grade adenoma), and Type V_I high-grade/V_N (submucosal invasive cancer). We calculated sensitivity (Sen), specificity (Spe), positive predictive value (PPV), and negative predictive value (NPV) for each classification stratified by endoscopist expertise and lesion size.

Results

A total of 1,617 colorectal lesions with a median size (IQR) of 2(1-4) mm were included in this analysis. Histopathological examination revealed that 71.5% (1,156/1,617) were low-grade adenomas, 1.0% (16/1,617) were high-grade adenomas or cancer, 19.3% were HPs (312/1,617), and 8.2% (133/1,617) were SSLs. For JNET Type 1 lesions, diagnostic accuracy was comparable between experts (Sen: 82.8%, Spe: 95.9%, PPV: 88.3%, NPV: 93.7%) and non-experts (Sen: 81.7%, Spe: 93.4%, PPV: 83.1%, NPV: 92.8%). Similarly, for JNET Type 2A, both groups showed high diagnostic performance (experts: Sen 93.8%, Spe 81.0%, PPV: 92.6%, NPV: 83.8%; non-experts: Sen 91.5%, Spe 78.3%, PPV: 90.3%, NPV: 80.6%).
In contrast, pit pattern diagnosis experts showed higher performances, particularly for Type II (experts: Sen 80.9%, Spe 96.3%, PPV: 89.0%, NPV: 93.1%; non-experts: Sen 57.9%, Spe 92.8%, PPV: 76.0%, NPV: 84.8%). The diagnostic performance for small lesions (≤5mm) was consistently high with JNET classification (Type 1: Sen 79.5%, Spe 94.3%; Type 2A: Sen 93.7%, Spe 79.2%), while pit pattern diagnosis showed lower performance (Type II: Sen 70.8%, Spe 94.5%; Type III_S/III_L/IV: Sen 93.6%, Spe 79.5%), for these lesions among non-experts.

Conclusion

JNET classification demonstrates comparable diagnostic performance between expert and non-expert endoscopists, likely due to its standardized and intuitive visual criteria. Pit pattern diagnosis shows experience-dependent performance differences, suggesting that NBI-based assessment may be more suitable for routine clinical practice, especially for diminutive lesions.

References

1. Toyoshima N, Mizuguchi Y, Takamaru H,et al. The Efficacy of Texture and Color Enhancement Imaging Observation in the Detection of Colorectal Lesions: A Multicenter, Randomized Controlled Trial (deTXIon Study). Gastroenterology. 2025 in press

Disclosure

These authors disclose the following: Yutaka Saito, Masashi Misawa, Shin-ei Kudo, and Toshio Uraoka have received lecture fees from Olympus
Corporation. The remaining authors disclose no conflicts.

Introduction

Colorectal cancer (CRC) remains a leading cause of cancer-related deaths in the U.S., with persistent disparities across racial and ethnic groups. Current uniform screening guidelines overlook population differences in risk. Polygenic risk scores (PRS) offer a pathway to personalized, risk-stratified screening, but their cost-effectiveness across diverse populations is unclear.

Aims & Methods

In this study, we estimated the cost-effectiveness of PRS-based, risk-stratified screening across four major U.S. racial and ethnic groups (White, Black, Asian and Hispanic), modelling lifetime costs, quality-adjusted life-years gained (QALYGs), and colonoscopy use to assess both overall and subgroup-specific outcomes.
PRS data from the Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO) were used to stratify individuals into risk 40 groups. The microsimulation model MISCAN-Colon was used to determine the optimal colonoscopy screening strategy for each risk group within the racial and ethnic groups. Investigated strategies varied in terms of starting age (40–60 years), stopping age (70–85 years), and screening interval (5–15 years). Within each racial and ethnic group, the cost-effectiveness of the optimal strategies of the 40 risk groups was compared with uniform colonoscopy screening (ages 45–75, every 10 years).

Results

Risk-stratified screening produced different recommendations across risk levels and populations. For low-risk groups, strategies ranged from no screening in Whites and Blacks to a one-time colonoscopy at age 55 for Asians and at age 60 for Hispanics. In contrast, for high-risk groups, all populations benefited most from intensive screening with colonoscopy every 5 years from ages 40–85. Across all racial and ethnic groups, PRS-guided screening outperformed uniform strategies, preventing more CRC cases and deaths but at the cost of higher colonoscopy use and costs. In constrained analyses, when QALYGs were fixed to the level of uniform screening, risk-stratified screening reduced costs by 6–27% and when costs were fixed to the level of uniform screening QALYGs increased by 8–29% These efficiency gains were most pronounced in Hispanics (–27% costs; +29% QALYGs) and Asians (–14% costs; +13% QALYGs), with smaller but consistent improvements in Whites (-9% costs; +12% QALYGs) and Blacks (-6% costs; +8%QALYGs).

Table 1. Lifetime Effects and Costs per 1,000 40-Year-Old Individuals for No Screening, Uniform Screening, and PRS-Guided Risk-Stratified Screening at a $100,000/QALYG Threshold, With QALYGs- and Cost-Constrained Analyses, Stratified by Racial and Ethnic Group (Results only for Hispanics and Asians)
Asians

Conclusion

PRS-guided screening may offer a pathway to more tailored CRC prevention, with modest overall gains but potentially greater cost savings and health benefits in Hispanics and Asians.

References

1. Carethers JM. Racial and ethnic disparities in colorectal cancer incidence and mortality. Adv Cancer Res. 2021;151:197-229.
2. May FP, Yang L, Corona E, Glenn BA, Bastani R. Disparities in Colorectal Cancer Screening in the United States Before and After Implementation of the Affordable Care Act. Clin Gastroenterol Hepatol. 2020;18(8):1796-804 e2.
3. Dixon P, Keeney E, Taylor JC, Wordsworth S, Martin RM. Can polygenic risk scores contribute to cost-effective cancer screening? A systematic review. Genet Med. 2022;24(8):1604-17.
4. Tamlander M, Jermy B, Seppälä TT, Färkkilä M, FinnGen, Widén E, et al. Genome-wide polygenic risk scores for colorectal cancer have implications for risk-based screening. Br J Cancer. 2024;130(4):651-9.

Disclosure

No conflicts of interest.

Introduction

Post endoscopy upper gastrointestinal cancers (PEUGIC) are often missed opportunities to diagnose earlier or even prevent cancer. PEUGIC in patients undergoing surveillance for pre-malignant conditions are a potentially important surveillance quality indicator. A national process for identifying all PEUGIC in the English National Health Service and performing root cause analysis was developed.

Aims & Methods

Potential PEUGIC between 2017 and 2023 occurring 3-36 months after an index endoscopy which did not result in a cancer diagnosis were identified from National Cancer Registration Dataset and Hospital Episode Statistics held by the National Cancer Registration and Analysis Service. Data from local root cause analysis were uploaded onto a secure portal for national analysis. Only PEUGIC following surveillance endoscopy were included in this analysis. Logistic regression analysis was used to identify associations.

Results

3151 potential PEUGIC were examined by local reviewers in 144 English hospitals. 635 (20%) were excluded and data were available for 2516 PEUGIC. 618 PEUGIC (24.6%) were undergoing surveillance for: Barrett’s oesophagus (83%); chronic atrophic gastritis (7%) or other conditions (10%). PEUGIC sites were oesophagus 85%, stomach 13% and duodenum 1%. 5% of PEUGIC were not related to the pre-malignant condition under surveillance.
Table 1 summarises the PEUGIC root cause analysis results based on the World Endoscopy Organisation categorisation. Inadequate endoscopy or decision making occurred in 30% of PEUGIC (categories B/D) and was associated with patient harm OR 2.61 (95% CI 1.74 -3.92).
Seattle biopsy protocol compliance was only 58% in Barrett’s surveillance PEUGIC. Additional focal or cancer associated lesions were present at index endoscopy in the PEUGIC segment in 23%.
Oesophageal and gastric PEUGIC often presented with stage 2 cancers or greater (48% and 43% respectively). 45.2% of oesophageal and 47.0% of gastric PEUGIC were considered at least possibly avoidable. Treatment was adversely affected with palliative treatment in 22% of oesophageal and 30% of gastric PEUGIC. 6% led to premature death.
Endoscopy practices associated with stage ≥2 cancer were: sedation without opioids OR 1.75 (95% CI 1.03-2.97); lack of image enhancement OR 1.78 (95% CI 1.18-2.70) and lack of dedicated surveillance sessions OR 2.08 (95% CI 1.46-2.98).

Table 1 Root causes of all surveillance PEUGIC

Conclusion

Inadequate assessment or decision making was noted in 30% of surveillance PEUGIC, which adversely affected treatment in 20% of oesophageal and 30% of gastric PEUGIC. 48% of oesophageal and 43% of gastric PEUGIC patients presented with stage 2 cancer or greater suggesting potential failures of surveillance. Optimal sedation practice, image enhancement and dedicated surveillance sessions were associated with stage 1 PEUGIC and better outcomes.

Introduction

Colorectal cancer is one of the most prevalent cancers with an estimated 1.8 million newly developed cases and 0.9 million cancer-related deaths in 2020.Resection of adenomatous polyps decreases the incidence and mortality of colorectal cancer. Narrow-band imaging (NBI) and linked color imaging (LCI) are the most predominantly conducted image-enhanced endoscopy using short-wavelength light but their usefulness in screening and surveillance colonoscopy remains controversial.

Aims & Methods

Our aim in this study was to investigate the superiority of LCI over NBI for ADR.
This is an open-label, two-arm, parallel, 1:1, multicenter, randomized, controlled, right-sided colonoscopy study and included patients who underwent colonoscopy at six endoscopy centers in Japan from November 2020 through November 2023. The patients were randomly assigned to undergo a proximal-sided colonoscopy (cecum and ascending colon) with LCI or 2G-NBI followed by WLI. The primary outcome included the ADR and secondary outcomes involved polyp detection rate (PDR), polyp miss rate, and adenoma miss rate. This study was registered on November 4, 2020, in the Japan Registry of Clinical Trials (jRCT1012200007).

Results

During the study period, 1,655 patients were randomly assigned to the LCI group (828 patients, primary LCI followed by secondary WLI) and the 2G-NBI group (827 patients, primary 2G-NBI followed by secondary WLI), and 776 patients in the LCI group and 752 patients in the 2G-NBI groups were finally analyzed. The ADRs were 26.7% and 23.4% and the PDRs were 32.6% and 28.9% in primary LCI and 2G-NBI, respectively, with no statistically significant differences in ADR and PDR between groups (P = 0.14 and P = 0.12). However, multivariate regression analysis revealed primary LCI observation as an independent predictor for reducing PDR in the secondary WLI (Odds ratio [OR]: 0.72, 95% confidence interval [CI]: 0.53−0.97, P = 0.029). Moreover, polyp analysis demonstrated that miss rates of adenomas and polyps were significantly lower in the LCI group than in the NBI group (OR: 0.54, 95% CI: 0.39−0.76 in adenoma; OR: 0.59, 95% CI: 0.44−0.78 in polyp).

Conclusion

LCI is not superior to 2G-NBI in ADR and PDR in a clinical right-sided colonoscopy setting. However, LCI significantly reduced miss rates of adenomas and polyps.

Disclosure

Naohisa Yoshida has received lecture and/or consulting fees from Fujifilm and Olympus; Shoko Ono has received lecture and/or consulting fees from Fujifilm and Olympus; Momoko Tsuda has received lecture and/or consulting fees from Fujifilm; Nobuaki Yagi have received lecture and/or consulting fees from Fujifilm and Olympus; Mototsugu Kato has received lecture and/or consulting fees from Fujifilm. The remaining authors disclose no conflicts.