Early detection of early gastric cancer using image-enhanced endoscopy: Current trends

A novel color-aided system for diagnosis of early gastric cancer using magnifying endoscopy with narrow-band imaging

BACKGROUND

The Pink Zone Pattern (PP) sign is a typical color alteration of early gastric cancer (EGC) under magnifying endoscopic narrow-band imaging (ME-NBI). By integrating the color changes (PP sign) with the "vessel plus surface (VS)" classification system, we developed an innovative diagnostic system for EGC and named it "Pink Microsurface Microvascular (PSV)" system. Here, we aimed to elucidate the diagnostic performance of the PSV system for EGC.

METHODS

We conducted a single-center prospective clinical study (before-after design) consisting of 2 cross-sectional studies at 2 separate periods. In the before phase, 184 suspected lesions were evaluated using the VS system under ME-NBI; in the after phase, 183 suspected lesions were evaluated using the PSV system. We compared the diagnostic accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) between the VS group and the PSV group.

RESULTS

The accuracy, sensitivity, specificity, PPV, and NPV of the VS system for EGC were 84.6%, 87.0%, 83.6%, 67.8%, and 94.2%, respectively, and those for the PSV system were 93.0%, 92.0%, 93.4%, 85.2%, and 96.6%, respectively. The accuracy, specificity, and PPV of the PSV system were superior to those of the VS system. However, the sensitivity and NPV did not significantly differ between the VS system and the PSV system. The VS system was inconclusive for 22 lesions (12.0%) and the PSV system was inconclusive for 11 lesions (6.0%). The PSV system could identify more suspicious lesions than the VS system.

CONCLUSIONS

We propose a new PSV diagnostic system by combining the VS system and the PP sign. Compared with the VS system, the PSV system could identify more suspected lesions and improve the diagnostic performance of EGC.

A deep-learning based system using multi-modal data for diagnosing gastric neoplasms in real-time (with video)

BACKGROUND

White light (WL) and weak-magnifying (WM) endoscopy are both important methods for diagnosing gastric neoplasms. This study constructed a deep-learning system named ENDOANGEL-MM (multi-modal) aimed at real-time diagnosing gastric neoplasms using WL and WM data.

METHODS

WL and WM images of a same lesion were combined into image-pairs. A total of 4201 images, 7436 image-pairs, and 162 videos were used for model construction and validation. Models 1-5 including two single-modal models (WL, WM) and three multi-modal models (data fusion on task-level, feature-level, and input-level) were constructed. The models were tested on three levels including images, videos, and prospective patients. The best model was selected for constructing ENDOANGEL-MM. We compared the performance between the models and endoscopists and conducted a diagnostic study to explore the ENDOANGEL-MM's assistance ability.

RESULTS

Model 4 (ENDOANGEL-MM) showed the best performance among five models. Model 2 performed better in single-modal models. The accuracy of ENDOANGEL-MM was higher than that of Model 2 in still images, real-time videos, and prospective patients. (86.54 vs 78.85%, P = 0.134; 90.00 vs 85.00%, P = 0.179; 93.55 vs 70.97%, P < 0.001). Model 2 and ENDOANGEL-MM outperformed endoscopists on WM data (85.00 vs 71.67%, P = 0.002) and multi-modal data (90.00 vs 76.17%, P = 0.002), significantly. With the assistance of ENDOANGEL-MM, the accuracy of non-experts improved significantly (85.75 vs 70.75%, P = 0.020), and performed no significant difference from experts (85.75 vs 89.00%, P = 0.159).

CONCLUSIONS

The multi-modal model constructed by feature-level fusion showed the best performance. ENDOANGEL-MM identified gastric neoplasms with good accuracy and has a potential role in real-clinic.

Real-time use of artificial intelligence for diagnosing early gastric cancer by magnifying image-enhanced endoscopy: a multicenter diagnostic study (with videos)

BACKGROUND AND AIMS

Endoscopy is a pivotal method for detecting early gastric cancer (EGC). However, skill among endoscopists varies greatly. Here, we proposed a deep learning-based system named ENDOANGEL-ME to diagnose EGC in magnifying image-enhanced endoscopy (M-IEE).

METHODS

M-IEE images were retrospectively obtained from 6 hospitals in China, including 4667 images for training and validation, 1324 images for internal tests, and 4702 images for external tests. One hundred eighty-seven stored videos from 2 hospitals were used to evaluate the performance of ENDOANGEL-ME and endoscopists and to assess the effect of ENDOANGEL-ME on improving the performance of endoscopists. Prospective consecutive patients undergoing M-IEE were enrolled from August 17, 2020 to August 2, 2021 in Renmin Hospital of Wuhan University to assess the applicability of ENDOANGEL-ME in clinical practice.

RESULTS

A total of 3099 patients undergoing M-IEE were enrolled in this study. The diagnostic accuracy of ENDOANGEL-ME for diagnosing EGC was 88.44% and 90.49% in internal and external images, respectively. In 93 internal videos, ENDOANGEL-ME achieved an accuracy of 90.32% for diagnosing EGC, significantly superior to that of senior endoscopists (70.16% ± 8.78%). In 94 external videos, with the assistance of ENDOANGEL-ME, endoscopists showed improved accuracy and sensitivity (85.64% vs 80.32% and 82.03% vs 67.19%, respectively). In 194 prospective consecutive patients with 251 lesions, ENDOANGEL-ME achieved a sensitivity of 92.59% (25/27) and an accuracy of 83.67% (210/251) in real clinical practice.

CONCLUSIONS

This multicenter diagnostic study showed that ENDOANGEL-ME can be well applied in the clinical setting. (Clinical trial registration number: ChiCTR2000035116.).

Deep learning system compared with expert endoscopists in predicting early gastric cancer and its invasion depth and differentiation status (with videos)

BACKGROUND AND AIMS

We aimed to develop and validate a deep learning-based system that covers various aspects of early gastric cancer (EGC) diagnosis, including detecting gastric neoplasm, identifying EGC, and predicting EGC invasion depth and differentiation status. Herein, we provide a state-of-the-art comparison of the system with endoscopists using real-time videos in a nationwide human-machine competition.

METHODS

This multicenter, prospective, real-time, competitive comparative, diagnostic study enrolled consecutive patients who received magnifying narrow-band imaging endoscopy at the Peking University Cancer Hospital from June 9, 2020 to November 17, 2020. The offline competition was conducted in Wuhan, China, and the endoscopists and the system simultaneously read patients' videos and made diagnoses. The primary outcomes were sensitivity in detecting neoplasms and diagnosing EGCs.

RESULTS

One hundred videos, including 37 EGCs and 63 noncancerous lesions, were enrolled; 46 endoscopists from 44 hospitals in 19 provinces in China participated in the competition. The sensitivity rates of the system for detecting neoplasms and diagnosing EGCs were 87.81% and 100%, respectively, significantly higher than those of endoscopists (83.51% [95% confidence interval [CI], 81.23-85.79] and 87.13% [95% CI, 83.75-90.51], respectively). Accuracy rates of the system for predicting EGC invasion depth and differentiation status were 78.57% and 71.43%, respectively, slightly higher than those of endoscopists (63.75% [95% CI, 61.12-66.39] and 64.41% [95% CI, 60.65-68.16], respectively).

CONCLUSIONS

The system outperformed endoscopists in identifying EGCs and was comparable with endoscopists in predicting EGC invasion depth and differentiation status in videos. This deep learning-based system could be a powerful tool to assist endoscopists in EGC diagnosis in clinical practice.

A comprehensive update on early gastric cancer: defining terms, etiology, and alarming risk factors

INTRODUCTION

Early gastric cancer (EGC) is a well-defined gastric malignancy that is limited to the mucosa or submucosa, irrespective of lymph node metastasis. At an early stage, gastric cancer often does not cause symptoms until it becomes advanced, and it is a heterogeneous disease and usually encountered in its late stages.

AREA COVERED

This comprehensive review will provide a novel insight into the evaluation of EGC epidemiology, defining terms, extensive etiology and risk factors, and timely diagnosis since prevention is an essential approach for controlling this cancer and reducing its morbidity and mortality.

EXPERT OPINION

The causative manner of EGC is complex and multifactorial. In recent years, researchers have made significant contributions to understanding the etiology and pathogenesis of EGC, and standardization in the evaluation of disease activity. Though the incidence of this cancer is steadily declining in some advanced societies owing to appropriate interventions, there remains a serious threat to health in developing nations. Early detection of resectable gastric cancer is crucial for better patient outcomes.

Identification of functional methylated CpG loci in PD-L1 promoter as the novel epigenetic biomarkers for primary gastric cancer

Gastric cancer (GC) is considered one of the most lethal malignancies worldwide due to poor prognosis. Aberrant methylation has been demonstrated to be involved in PD-L1 dysregulated expression in human cancers and possesses a great value as a diagnostic biomarker. Given that, in this study, we investigated the methylation status of PD-L1 as a promising biomarker in primary gastric tumors and identified functional CpG loci undergoing aberrant methylation through tumorigenesis of GC. PD-L1 methylation was initially evaluated in-silico using TCGA-STAD dataset. Pearson's correlation analysis was further employed to identify the most significant functional methylated CpG loci of PD-L1 gene in TCGA-STAD patient cohort. Methylation status and its correlation with PD-L1 expression were also validated using q-MSP and qRT-PCR in a set of internal samples, including 25 paired primary gastric tumors and adjacent normal tissues. The obtained results from TCGA-STAD showed that PD-L1 is significantly hypermethylated through gastric tumorigenesis, mostly in two CpG loci overlapping with cg19724470 and cg15837913 probes. Besides, PD-L1 DNA methylation was negatively correlated with PD-L1 expression in tumor samples. Furthermore, hypermethylation of cg19724470 and cg15837913 regions was validated in primary gastric tumors compared to adjacent normal samples. Also, ROC curve analysis illustrated the high diagnostic value of PD-L1 methylation for early detection of GC (AUC = 0.8110). In conclusion, the findings of this study suggested that PD-L1 expression is regulated by methylation in functional CpG loci and its methylation could be considered as a valuable diagnostic target for GC.

Enhanced Image-Based Endoscopic Pathological Site Classification Using an Ensemble of Deep Learning Models

In vivo diseases such as colorectal cancer and gastric cancer are increasingly occurring in humans. These are two of the most common types of cancer that cause death worldwide. Therefore, the early detection and treatment of these types of cancer are crucial for saving lives. With the advances in technology and image processing techniques, computer-aided diagnosis (CAD) systems have been developed and applied in several medical systems to assist doctors in diagnosing diseases using imaging technology. In this study, we propose a CAD method to preclassify the in vivo endoscopic images into negative (images without evidence of a disease) and positive (images that possibly include pathological sites such as a polyp or suspected regions including complex vascular information) cases. The goal of our study is to assist doctors to focus on the positive frames of endoscopic sequence rather than the negative frames. Consequently, we can help in enhancing the performance and mitigating the efforts of doctors in the diagnosis procedure. Although previous studies were conducted to solve this problem, they were mostly based on a single classification model, thus limiting the classification performance. Thus, we propose the use of multiple classification models based on ensemble learning techniques to enhance the performance of pathological site classification. Through experiments with an open database, we confirmed that the ensemble of multiple deep learning-based models with different network architectures is more efficient for enhancing the performance of pathological site classification using a CAD system as compared to the state-of-the-art methods.

The Mexican consensus on the detection and treatment of early gastric cancer

Gastric cancer is one of the most frequent neoplasias in the digestive tract and is the result of premalignant lesion progression in the majority of cases. Opportune detection of those lesions is relevant, given that timely treatment offers the possibility of cure. There is no consensus in Mexico on the early detection of gastric cancer, and therefore, the Asociación Mexicana de Gastroenterología brought together a group of experts and produced the "Mexican consensus on the detection and treatment of early gastric cancer" to establish useful recommendations for the medical community. The Delphi methodology was employed, and 38 recommendations related to early gastric cancer were formulated. The consensus defines early gastric cancer as that which at diagnosis is limited to the mucosa and submucosa, irrespective of lymph node metástasis. In Mexico, as in other parts of the world, factors associated with early gastric cancer include Helicobacter pylori infection, a family history of the disease, smoking, and diet. Chromoendoscopy, magnification endoscopy, and equipment-based image-enhanced endoscopy are recommended for making the diagnosis, and accurate histopathologic diagnosis is invaluable for making therapeutic decisions. The endoscopic treatment of early gastric cancer, whether dissection or resection of the mucosa, should be preferred to surgical management, when similar oncologic cure results can be obtained. Endoscopic surveillance should be individualized.

Precision medicine in gastric cancer

Gastric cancer (GC) is a complex disease linked to a series of environmental factors and unhealthy lifestyle habits, and especially to genetic alterations. GC represents the second leading cause of cancer-related deaths worldwide. Its onset is subtle, and the majority of patients are diagnosed once the cancer is already advanced. In recent years, there have been innovations in the management of advanced GC including the introduction of new classifications based on its molecular characteristics. Thanks to new technologies such as next-generation sequencing and microarray, the Cancer Genome Atlas and Asian Cancer Research Group classifications have also paved the way for precision medicine in GC, making it possible to integrate diagnostic and therapeutic methods. Among the objectives of the subdivision of GC into subtypes is to select patients in whom molecular targeted drugs can achieve the best results; many lines of research have been initiated to this end. After phase III clinical trials, trastuzumab, anti-Erb-B2 receptor tyrosine kinase 2 (commonly known as ERBB2) and ramucirumab, anti-vascular endothelial growth factor receptor 2 (commonly known as VEGFR2) monoclonal antibodies, were approved and introduced into first- and second-line therapies for patients with advanced/metastatic GC. However, the heterogeneity of this neoplasia makes the practical application of such approaches difficult. Unfortunately, scientific progress has not been matched by progress in clinical practice in terms of significant improvements in prognosis. Survival continues to be low in contrast to the reduction in deaths from many common cancers such as colorectal, lung, breast, and prostate cancers. Although several target molecules have been identified on which targeted drugs can act and novel products have been introduced into experimental therapeutic protocols, the overall approach to treating advanced stage GC has not substantially changed. Currently, surgical resection with adjuvant or neoadjuvant radiotherapy and chemotherapy are the most effective treatments for this disease. Future research should not underestimate the heterogeneity of GC when developing diagnostic and therapeutic strategies aimed toward improving patient survival.

CEACAM6 is upregulated by Helicobacter pylori CagA and is a biomarker for early gastric cancer

Early detection of gastric cancers saves lives, but remains a diagnostic challenge. In this study, we aimed to identify cell-surface biomarkers of early gastric cancer. We hypothesized that a subset of plasma membrane proteins induced by the Helicobacter pylori oncoprotein CagA will be retained in early gastric cancers through non-oncogene addiction. An inducible system for expression of CagA was used to identify differentially upregulated membrane protein transcripts in vitro. The top hits were then analyzed in gene expression datasets comparing transcriptome of gastric cancer with normal tissue, to focus on markers retained in cancer. Among the transcripts enriched upon CagA induction in vitro, a significant elevation of CEACAM6 was noted in gene expression datasets of gastric cancer. We used quantitative digital immunohistochemistry to measure CEACAM6 protein levels in tissue microarrays of gastric cancer. We demonstrate an increase in CEACAM6 in early gastric cancers, when compared to matched normal tissue, with an AUC of 0.83 for diagnostic validity. Finally, we show that a fluorescently conjugated CEACAM6 antibody binds avidly to freshly resected gastric cancer xenograft samples and can be detected by endoscopy in real time. Together, these results suggest that CEACAM6 upregulation is a cell surface response to H. pylori CagA, and is retained in early gastric cancers. They highlight a novel link between CEACAM6 expression and CagA in gastric cancer, and suggest CEACAM6 to be a promising biomarker to aid with the fluorescent endoscopic diagnosis of early neoplastic lesions in the stomach.

Efficacy Evaluation of SAVE for the Diagnosis of Superficial Neoplastic Lesion

The detection of non-polypoid superficial neoplastic lesions using current standard of white light endoscopy surveillance and random biopsy is associated with high miss rate. The subtle changes in mucosa caused by the flat and depressed neoplasms often go undetected and do not qualify for further investigation, e.g., biopsy and resection, thus increasing the risk of cancer advancement. This paper presents a screening tool named the saliency-aided visual enhancement (SAVE) method, with an objective of highlighting abnormalities in endoscopic images to detect early lesions. SAVE is a hybrid system combining image enhancement and saliency detection. The method provides both qualitative enhancement and quantitative suspicion index for endoscopic image regions. A study to evaluate the efficacy of SAVE to localize superficial neoplastic lesion was performed. Experimental results for average overlap index >0.7 indicated that SAVE was successful to localize the lesion areas. The area under the receiver-operating characteristic curve obtained for SAVE was 94.91%. A very high sensitivity (100%) was achieved with a moderate specificity (65.45%). Visual inspection showed a comparable performance of SAVE with chromoendoscopy to highlight mucosal irregularities. This paper suggests that SAVE could be a potential screening tool that can substitute the application of burdensome chromoendoscopy technique. SAVE method, as a simple, easy-to-use, highly sensitive, and consistent red flag technology, will be useful for early detection of neoplasm in clinical applications.

Efficacy of small-volume simethicone given at least 30 min before gastroscopy

AIM

To evaluate the efficacy of 5 mL simethicone solution in decreasing gastric foam if given at least 30 min before gastroscopy.

METHODS

This was a randomized, placebo controlled, endoscopist blinded study performed at Changi General Hospital. Patients were at least 21 years old, had no prior surgical resection of the upper gastrointestinal tract, and scheduled for elective diagnostic gastroscopies. The primary outcome was the total mucosal visibility score (TMVS) which was evaluated using McNally score. The sample size was calculated to be 24 per group (SD 2.4, 80% power, P < 0.05, 2-sample t test).

RESULTS

Fifty-four patients were randomised to receive either simethicone [1 mL liquid simethicone (100 mg) in 5 mL of water] or placebo (5 mL of water) at least 30 min before their gastroscopy. Six accredited consultants conducted the gastroscopy, and the interobserver agreement of scoring TMVS was good with a Kappa statistic of 0.73. The simethicone group had significantly better mean TMVS compared to placebo (5.78 ± SD 1.65 vs 8.89 ± SD 1.97, P < 0.001). The improvement was statistically significant for the duodenum and the gastric antrum, angularis, body, and fundus. Percent 51.9 of patients in the simethicone group had a TMVS of 4 (no bubbles at all) to 5 (only 1 area with minimal bubbles), while in the placebo group 3.7% of patients had TMVS of 4 or 5. The number needed to treat was 2.1 to avoid a TMVS of 6 and more. The simethicone group also had a significantly shorter procedure time with less volume of additional flushes required during gastroscopy to clear away obscuring gastric foam.

CONCLUSION

With a premedication time of at least 30 min, 5 mL simethicone can significantly decrease gastric foam, decrease the volume of additional flushes, and shorten gastroscopy time.

Diagnostic efficacy of magnifying endoscopy with narrow-band imaging for gastric neoplasms: a meta-analysis

Background

Magnifying endoscopy with narrow-band imaging (ME-NBI) is a novel, image-enhanced endoscopic technique for differentiating gastrointestinal neoplasms and potentially enabling pathological diagnosis.

Objectives

The aim of this analysis was to assess the diagnostic performance of ME-NBI for gastric neoplasms.

Methods

We performed a systematic search of the PubMed, EMbase, Web of Science, and Cochrane Library databases for relevant studies. Meta-DiSc (version 1.4) and STATA (version 11.0) software were used for the data analysis. Random effects models were used to assess diagnostic efficacy. Heterogeneity was tested by the Q statistic and I² statistic. Meta-regression was used to analyze the sources of heterogeneity.

Results

A total of 10 studies, with 2151 lesions, were included. The pooled characteristics of these studies were as follows: sensitivity 0.85 (95% confidence interval [CI]: 0.81–0.89), specificity 0.96 (95% confidence interval [CI]: 0.95–0.97), and area under the curve (AUC) 0.9647. In the subgroup analysis, which compared the diagnostic efficacy of ME-NBI and white light imaging (WLI), the pooled sensitivity and specificity of ME-NBI were 0.87 (95% CI: 0.80–0.92) and 0.93 (95% CI: 0.90–0.95), respectively, and the area under the curve (AUC) was 0.9556. In contrast, the pooled sensitivity and specificity of WLI were 0.61 (95% CI: 0.53–0.69) and 0.65 (95% CI: 0.60–0.69), respectively, and the area under the curve (AUC) was 0.6772.

Conclusions

ME-NBI presents a high diagnostic value for gastric neoplasms and has a high specificity.

Optical imaging of hemoglobin oxygen saturation using a small number of spectral images for endoscopic application

Tissue hypoxia is associated with tumor and inflammatory diseases, and detection of hypoxia is potentially useful for their detailed diagnosis. An endoscope system that can optically observe hemoglobin oxygen saturation (StO2) would enable minimally invasive, real-time detection of lesion hypoxia in vivo. Currently, point measurement of tissue StO2 via endoscopy is possible using the commercial fiber-optic oximeter T-Stat, which is based on visible light spectroscopy at many wavelengths. For clinical use, however, imaging of StO2 is desirable to assess the distribution of tissue oxygenation around a lesion. Here, we describe our StO2 imaging technique based on a small number of wavelength ranges in the visible range. By assuming a homogeneous tissue, we demonstrated that tissue StO2 can be obtained independently from the scattering property and blood concentration of tissue using four spectral bands. We developed a prototype endoscope system and used it to observe tissue-simulating phantoms. The StO2 (%) values obtained using our technique agreed with those from the T-Stat within 10%. We also showed that tissue StO2 can be derived using three spectral band if the scattering property is fixed at preliminarily measured values.