Role of chromoendoscopy in colon cancer surveillance in inflammatory bowel disease

Application of artificial intelligence in gastrointestinal disease: a narrative review

Objective

We collected evidence on the application of artificial intelligence (AI) in gastroenterology field. The review was carried out from two aspects of endoscopic types and gastrointestinal diseases, and briefly summarized the challenges and future directions in this field.

Background

Due to the advancement of computational power and a surge of available data, a solid foundation has been laid for the growth of AI. Specifically, varied machine learning (ML) techniques have been emerging in endoscopic image analysis. To improve the accuracy and efficiency of clinicians, AI has been widely applied to gastrointestinal endoscopy.

Methods

PubMed electronic database was searched using the keywords containing “AI”, “ML”, “deep learning (DL)”, “convolution neural network”, “endoscopy (such as white light endoscopy (WLE), narrow band imaging (NBI) endoscopy, magnifying endoscopy with narrow band imaging (ME-NBI), chromoendoscopy, endocytoscopy (EC), and capsule endoscopy (CE))”. Search results were assessed for relevance and then used for detailed discussion.

Conclusions

This review described the basic knowledge of AI, ML, and DL, and summarizes the application of AI in various endoscopes and gastrointestinal diseases. Finally, the challenges and directions of AI in clinical application were discussed. At present, the application of AI has solved some clinical problems, but more still needs to be done.

Methylene blue MMX® tablets for chromoendoscopy. Bioavailability, colon staining and safety in healthy volunteers undergoing a full colonoscopy

Methylene blue-MMX® tablets are proposed as an aid for detection and visualisation of adenomas and carcinomas in patients undergoing colonoscopy, by improving their detection rate and highlighting the presence of the intestinal dysplastic lesions. Single total doses of 100 and 200 mg were administered to healthy volunteers undergoing a bowel cleansing preparation and a full colonoscopy to investigate the colonic staining. The pharmacokinetics of methylene blue and the safety after exposure to the tablets were also investigated. With 200 mg, the best staining, assessed as the sum of acceptable and good staining, was achieved in the ascending colon and rectosigmoid (75% subjects each), the transverse and the descending colon (approximately 63% each). Absence of staining or overstaining were reported for no colonic region of interest in any subject. Similar results were observed in the 100 mg dose group. Methylene blue blood concentrations reached a peak (C_max) in a median time (T_max) of 12 h with 100 mg and 16 h with 200 mg. AUC_0-t was 10.7 ± 6.7 μg/mLxh after 100 mg and 25.2 ± 7.4 μg/mLxh after 200 mg. Half-life ranged between 9 and 22 h after the lower dose and between 6 and 26 h after the higher dose. The cumulative urinary excretion was about 28% after 100 mg and about 39% after 200 mg up to 60 h post-dose. The overall frequency of adverse events after single dose of the test product administered along with a bowel cleansing preparation was 39%, but only one was related to the test product: abnormal transaminases. The most frequent adverse event was a transient polyuria (17%). One serious adverse event (gastrointestinal haemorrhage) led the subject to study discontinuation and hospitalisation and another subject withdrew the study due to one adverse event (haematemesis). Either event was not related to methylene blue.

Detection of colonic dysplasia in patients with ulcerative colitis using a targeted fluorescent peptide and confocal laser endomicroscopy: A pilot study

Aim

Targeted molecular probes have been used to detect sporadic colonic dysplasia during confocal laser endomicroscopy (CLE) with promising results. This is a feasibility pilot study aiming to assess the potential role of CLE combined with a fluorescent-labeled peptide to stain and detect dysplasia associated with Ulcerative Colitis.

Method

A phage-derived heptapeptide with predicted high binding affinity for dysplastic tissue, was synthesized and labeled with fluorescein. Eleven lesions with suspected dysplasia at endoscopy were excised from nine patients with long-standing ulcerative colitis. Specimens were sprayed with the peptide and examined by CLE. The CLE images were then compared to the corresponding histological sections.

Results

At definitive histology, 4 lesions were diagnosed as inflammatory polyps, 6 as dysplastic lesions and one as invasive cancer. In inflammatory polyps, the fluorescence signal came from peri-cryptal spaces and crypt lumen due to passive accumulation of the peptide in these areas. Dysplasia was associated with active binding of the peptide to dysplastic colonocytes.

Conclusion

Ex vivo staining of ulcerative colitis-associated dysplasia using a fluorescent labeled molecular probe and CLE is feasible. In vivo studies on larger populations are required to evaluate the safety and the effective contribution of molecular probes in cancer surveillance of ulcerative colitis.

Diagnostic yield of endomicroscopy for dysplasia in primary sclerosing cholangitis associated inflammatory bowel disease: a feasibility study

BACKGROUND AND STUDY AIMS

Primary sclerosing cholangitis associated inflammatory bowel disease (PSC-IBD) is characterized by a high risk of colorectal dysplasia. Surveillance colonoscopies with random biopsies have doubtful power for dysplasia detection. Our aim was to prospectively investigate the feasibility and efficacy of pCLE in surveillance colonoscopies in patients with PSC-IBD.

PATIENTS AND METHODS

Sixty-nine patients with PSC-IBD underwent colonoscopy in 2 steps. On the way from rectum to cecum, the mucosa was inspected with high definition endoscopy (HDE) and random biopsies were taken according to the standard routine. On the way from cecum to rectum, fluorescein-enhanced pCLE and chromoendoscopy were performed. Regions where random biopsies had been taken, as well as visible lesions, were examined with pCLE and targeted biopsies were taken of lesions suspicious for dysplasia. Two investigators, blinded to histology and endoscopy results, analyzed all pCLE videos off-line.

RESULTS

Nineteen biopsies obtained in 13 patients (17 targeted biopsies, 2 random biopsies) revealed the presence of low-grade dysplasia. Thirteen lesions with dysplasia were endoscopically visible but by using pCLE-targeted biopsies, additional endoscopically invisible dysplasias in 4 biopsies obtained from 3 patients were detected. The sensitivity, specificity, and accuracy of pCLE in predicting dysplasia were respectively 89 % (95 % CI: 65 - 98), 96 % (95 % CI: 94 - 97), and 96 % (95 % CI: 94 - 97). pCLE showed a good performance for differentiating neoplastic from non-neoplastic mucosa with negative predictive value of 99 %.

CONCLUSIONS

pCLE in PSC-IBD surveillance is feasible and may be a good complement to HDE. Future research should aim at elucidating whether real-time pCLE is applicable in PSC-IBD surveillance.

Chromocolonoscopy

Chromoendoscopy techniques improve the visualization of mucosal structures. This article reviews and summarizes key studies addressing the impact of chromoendoscopy on colonic neoplasia detection and differentiation of neoplastic from non-neoplastic polyps in average and high-risk populations, including patients with colonic inflammatory bowel disease (IBD). In this context, there are convincing data that chromoendoscopy differentiates neoplastic from non-neoplastic polyps in average-risk populations with high accuracy. Moreover, dye-based chromoendoscopy improves neoplasia detection in colonic IBD surveillance.

Toward a consensus on endoscopic surveillance of patients with colonic inflammatory bowel disease

Surveillance colonoscopy in patients with inflammatory bowel disease (IBD) with colonic involvement is recommended by multiple national and international gastrointestinal societies. Recommendations differ on the timing of initial screening colonoscopy, recommended surveillance intervals, optimal technique for dysplasia detection, and management of endoscopically visible and nonvisible dysplasia. This article reviews current society guidelines, highlighting similarities and differences, in an attempt to summarize areas of consensus on surveillance protocols in IBD, while drawing attention to controversial areas in need of further research.

Meta-analysis for evaluating the accuracy of endoscopy with narrow band imaging in detecting colorectal adenomas

BACKGROUND AND AIM

The aim of this study was to determine whether the use of the narrow band imaging (NBI) system could enhance the accuracy of adenoma detection during an endoscopic examination of the colon and rectum.

METHODS

MEDLINE, EMBASE, and the Cochrane Library databases were searched along with a hand search of abstracts from relevant conferences up to June 2011. The rates of adenoma and flat adenoma detection, and withdrawal time were analyzed using Review Manager 4.2.

RESULTS

A total of 3049 subjects in eight trials were included. Meta-analysis revealed that there was no statistically significant difference in the rates of adenoma detection between the NBI group and the white light colonoscopy group (pooled relative risk [RR]: 1.09, 95% confidence interval [CI]: 1.00-1.19, P = 0.05). However, after exclusion of high-definition television modalities, the rate of adenoma detection by NBI was significantly higher than that by white light, particularly for patients with one adenoma (pooled RR 1.36, 95%CI 1.07-1.71, P = 0.02). Endoscopy with the NBI system significantly increased the rate of flat adenoma detection (pooled RR 1.96, 95%CI 1.09-3.52, P = 0.02). However, endoscopy with NBI had longer withdrawal time than that with white light (pooled weighted mean difference: 0.90, 95%CI: 0.38-1.42, P = 0.0006).

CONCLUSIONS

Endoscopy with NBI seems to improve the detection of flat adenomas, particularly with high-definition technology, but prolongs the withdrawal time. These results indicate that endoscopy routinely using the NBI system for the surveillance of adenomas may be recommended after the technique is further modified.

The diagnostic accuracy of chromoendoscopy for dysplasia in ulcerative colitis: meta-analysis of six randomized controlled trials

AIM

The diagnostic accuracy of chromoendoscopy for dysplasia in ulcerative colitis (UC) was systematically evaluated.

METHOD

  Original studies in any language were searched from PubMed and Embase. Meta-analysis of prospective studies that compared chromoendoscopy with histological diagnosis was carried out. Sensitivity, specificity, and diagnostic odds ratio (DOR) were calculated for each study and pooled together; summary receiver operating characteristic (ROC) and subgroup analyses were performed, while the quality of the study and heterogeneity were assessed.

RESULTS

Six randomized controlled trials were included, which used methylene blue or indigo carmine dye spray. The meta-analysis demonstrated a pooled sensitivity of 83.3% (95% confidence interval (CI), 35.9-99.6%), specificity of 91.3% (95% CI, 43.8-100%), and DOR of 17.544 (95% CI, 1.245-247.14). Subgroup analyses revealed that both the methylene blue dye spray subgroup and the unspecified endoscopist subgroup include the same studies, and their pooled sensitivity and specificity were 0.737 and 0.917, respectively. The other subgroup, which used indigo carmine dye spray, had overall higher sensitivity (0.930) and lower specificity (0.910).

CONCLUSION

Chromoendoscopy has medium to high sensitivity and a high diagnostic accuracy for dysplastic lesions in UC.

Methylene blue MMX tablets for chromoendoscopy. Safety tolerability and bioavailability in healthy volunteers

Methylene blue-MMX tablets are proposed as colonic diagnostic staining. Methylene blue taken prior to colonoscopy is expected to provide an effective staining of colonic and rectal mucosa leaving unstained the dysplastic or polypoid areas. The present single dose, open-label study investigated the safety of methylene blue after single oral doses of 200 and 400mg in healthy volunteers. The absolute bioavailability was also investigated after the intake of 2L of bowel cleansing preparation in 2h and by comparing the dose of 200mg with a single iv dose of 100mg in the same subjects. Only non-serious adverse events occurred. Related events occurred to 8/22 subjects. Most of the events were mild and transient. Abnormal transaminases, gastrointestinal disorders and dysuria frequency were 13.6%. After intake of the laxative and the oral dose of 200mg, systemic exposure to methylene blue was shown in all subjects with concentrations increasing for 12h. The peak was reached in a median of 16 h. Peak blood concentration did not increase proportionally with the dose. AUC(0-t) was 32.94 μg/mL × h after 200mg and 38.08 μg/mL × h after 400mg. Half life ranged between 14 and 27 h after the lower dose and between 6 and 26 h after the higher dose. The cumulative excretion was about 40% of the injected dose, 39.67% after 200mg and 23.48% after 400mg. Absolute bioavailability of methylene blue calculated as ratio between AUC(0-t) oral/iv corrected for the dose was on average F(abs)=139.19 ± 52.00%.

Relationship between pit patterns of colorectal polypoid lesions classified by magnifying chromoendoscopy and expression of hMLH1 and hMSH2 proteins in colorectal mucosa

AIM: To explore the relationship between pit patterns of colorectal polypoid lesions classified by magnifying chromoendoscopy and expression of human MutL homolog 1/2 (hMLH1/2) proteins in colorectal mucosa.

METHODS: Colorectal lesions in 146 patients were classified as type I to V pit patterns by magnifying chromoendoscopy using the Kudo criteria. All lesions were pathologically confirmed as nonneoplastic, adenomatous or cancerous lesions. Colorectal mucosal biopsy specimens were used to detect the expression of hMLH1 and hMSH2 proteins by immunohistochemistry.

RESULTS: A total of 256 polypoid lesions were found in 146 patients by magnifying chromoendoscopy. The rates of loss of hMLH1 and hMSH2 protein expression increased gradually from type I to type V pit patterns in 256 polypoid lesions [hMLH1: 0.0% (0/11), 1.61% (1/62), 19.68% (25/127), 33.33% (1/3), 32.26% (10/31), 36.36%(8/22); hMSH2: 0.00% (0/11), 3.22% (2/62), 16.53% (21/127), 33.33% (1/3), 35.48% (11/31), 40.90% (9/22); all P < 0.01]. The rates of loss of hMLH1 and hMSH2 expression were 2.70% and 4.05% in nonneoplastic lesions, 23.07% and 16.92% in adenoma lesions, and 25% and 30.76% in cancerous lesions. The rates of loss of hMLH1 and hMSH2 protein expression were significantly higher in cancerous lesions than in adenomatous and nonneoplastic lesions (both P < 0.01). There was no significant difference between the rate of loss of hMLH1 and that of hMSH2 protein expression in all lesions (all P > 0.05).

CONCLUSION: The rates of loss of hMLH1 and hMSH2 protein expression gradually increased from type I to type V pit patterns, suggesting that mutation or functional deficiency of DNA mismatch repair genes is an early event in colorectal carcinogenesis. Lesion classification by magnifying chromoendoscopy or detection of the loss of hMLH1 and hMSH2 protein expression can help identify precancerous and colorectal lesions from colorectal lesions.

Feasibility of confocal endomicroscopy in the diagnosis of pediatric gastrointestinal disorders

AIM: To evaluate the feasibility and utility of confocal laser endomicroscopy (CLE) in the description of normal gastrointestinal (GI) mucosa and in the diagnosis of GI disorders in children, in comparison to histology.

METHODS: Forty-four patients (19 female) median age 10.9 years (range 0.7-16.6 years) with suspected or known GI pathology underwent esophago-gastro-duodenoscopy (OGD) (n = 36) and/or ileocolonoscopy (IC) (n = 31) with CLE using sodium fluorescein and acriflavine as contrast agents. Histological sections were compared with same site confocal images by two experienced pediatric and GI histopathologists and endoscopists, respectively.

RESULTS: Duodenum and ileum were intubated in all but one patient undergoing OGD and IC. The median procedure time was 16.4 min (range 7-25 min) for OGD and 27.9 min (range 15-45 min) for IC. A total of 4798 confocal images were compared with 153 biopsies from the upper GI tract from 36 procedures, and 4661 confocal images were compared with 188 biopsies from the ileocolon from 31 procedures. Confocal images were comparable to conventional histology both in normal and in pathological conditions such as esophagitis, Helicobacter pylori gastritis, celiac disease, inflammatory bowel disease, colonic heterotopia, and graft versus host disease.

CONCLUSION: CLE offers the prospect of targeting biopsies to abnormal mucosa, thereby increasing diagnostic yield, reducing the number of biopsies, decreasing the burden on the histopathological services, and reducing costs.

Endoscopic mucosal resection of colonic lesions: current applications and future prospects

The introduction of submucosal fluid injection has remarkably extended the range of endoscopically resectable polyps. The limiting factor for endoscopic resection is not polyp size, but polyp depth. Endoscopic ultrasound is a useful adjunctive diagnostic tool to assess the depth of invasion. The success of are section ultimately depends on pathologic confirmation of a benign nature of this lesion or of a cancer limited to the mucosa. Selected well-differentiated cancers without lymphovascular invasion of the superficial submucosa can be successfully resected endoscopically.

Surveillance of patients at increased risk of colon cancer: inflammatory bowel disease and other conditions

Colorectal cancer (CRC) is the second most common cause of cancer-related mortality in the United States. Colonoscopic screening with removal of adenomatous polyps in individuals at average risk is known to decrease the incidence and associated mortality from colon cancer. Certain conditions, notably inflammatory bowel disease involving the colon, a family history of polyps or cancer, a personal history of colon cancer or polyps, and other conditions such as acromegaly, ureterosigmoidostomy, and Streptococcus bovis bacteremia are associated with an increased risk of colonic neoplasia. This article reviews the CRC risks associated with these conditions and the currently recommended surveillance strategies.

Diagnostics of inflammatory bowel disease

The diagnosis of inflammatory bowel disease (IBD) with its 2 main subforms, Crohn's disease and ulcerative colitis, is based on clinical, endoscopic, radiologic, and histologic criteria. This paradigm remains unchanged despite the advent of new molecular technologies for the examination of serum proteins and genetic sequences, respectively. The main innovations in diagnostic technologies include the development of more sophisticated endoscopic and noninvasive imaging techniques with the aim of improving the identification of complications, in particular malignant diseases associated with IBD. The future will see further progress in the identification of genetic susceptibility factors and of protein biomarkers and their use to describe the molecular epidemiology of IBD. It can be expected that future diagnostic algorithms will include molecular parameters to detect early disease or guide therapies by predicting the individual course of disease.