Society: ASGE

Background: Using EsoCheck (EC), a non-endoscopic balloon device for sampling the distal esophagus, coupled with EsoGuard (EG), a DNA based testing screening for Barrett’s esophagus (BE), we have previously reported high sensitivity and specificity for detection of BE among patients with known BE compared with controls. We now utilized this screening method in the population at risk for developing BE and esophageal adenocarcinoma (ECA).
Aim: Determine the accuracy of EG/EC in patients eligible for BE screening.
Method: We recruited veterans who meet American College of Gastroenterology (ACG) Guideline criteria for endoscopic BE screening at our VA hospital. Study participants completed EC/EG followed by esophagogastroduodenoscopy (EGD). We compared the yield of guideline based screening with a theoretical strategy where patients have EC/EG, and only those with a positive result then have an EGD.
Results: We enrolled 47 participants (89.3% male, median age 60.9 (SD 11.2)). 46/47 patients successfully swallowed the EC. In total, 10.6% (5/47) had an endoscopic diagnosis of BE. EC/EG was positive for 14 patients, negative for 23, and 9 patients did not have sufficient material for EG analysis. We found that 10.8% (4/37) patients had BE diagnosed by EGD, compared with 28.6% (4/14). There were no cases of BE diagnosed by endoscopy among patients with a negative EC/EG test. Compared with EGD the sensitivity and specificity of the test was 100% (95% CI 39.8% to 100%) and 69.7% (95% CI 51.3% to 84.4%), respectively. The PPV, NPV, and accuracy were 28.12% (95% CI 18.91% to 39.63%), 100.00%, 72.91% (95% CI 55.81% to 86.16%).
Conclusions: The performance of the EC/EG test was promising in the screening population and this study sets the ground for future larger trials of an EC/EG intervention for BE detection in the wider screening population.

Background and aims: The early diagnosis and intervention of esophageal squamous cell carcinoma (ESCC) is particularly important due to poor prognosis and the lack of effective therapies. Comprehensive study on early ESCC at single cell level is rare due to the requirements of fresh and high-quality specimen obtained from ESD. This study aims to systematically describe the cellular atlas of human intramucosal ESCC.
Methods: Five paired samples of intramucosal ESCC, para-ESCC esophageal tissue from endoscopically resected specimen, and peripheral blood mononuclear cells were adopted for scRNA-seq analysis. A computational pipeline scMetabolism was applied to quantify the metabolic diversity of single cells.
Results: A total of 164,715 cells were profiled. The abundance of CD8⁺ TEXs , Tregs and PD1⁺CD4⁺T cells suggested an exhausted and suppressive immune microenvironment. Several genes in immune cells such as CXCL13, CXCR5 and PADI4 were identified as new biomarkers for poor prognosis. Epithelial cells exhibited high intra-tumor heterogeneity and two evolutional trajectories during ESCC tumorigenesis initiated from proliferative cells, and then through an intermediate state, to two different terminal states of normal differentiated epithelial cells or malignant cells, respectively. Intercellular interaction analysis based on ligand-receptor pairs revealed malignant cells interacting with CAFs via the MDK-NCL pathway, which was varified by IHC and coculture, is an important hallmark in the change of tumor microenvironment, and serves as a sign of CAF activation to stimulate downstream pathways for facilitating tumor invasion, and therefore suggesting a potential early biomarker of ESCC progression.The proliferation of CAFs was significantly promoted by the stimulation of tumor supernatant and the alteration could be rescued by the inhibitor of MDK (MDK inhibitor, iMDK)
Conclusion: This study demonstrates the changes of cell subsets and transcriptional levels in human intramucosal ESCC, providing unique insights into the development of novel biomarkers and potential intervention strategies.

<b>Figure 1 A,B. scRNA-Seq profiling of the human intramucosal ESCC microenvironment.</b><br /> <b>Figure 1 C~H. The single-cell transcriptomes of epithelial cells in human intramucosal ESCC.</b><br /> <b>Figure 1 I~L. </b><b>Exhausted T cells and suppressive immune microenvironment in human intramucosal ESCC.</b><br /> <b>Figure 1 M,N. Distinct fibroblast subpopulations in human intramucosal ESCC ecosystem. </b>

Figure 1 A,B. scRNA-Seq profiling of the human intramucosal ESCC microenvironment.
Figure 1 C~H. The single-cell transcriptomes of epithelial cells in human intramucosal ESCC.
Figure 1 I~L. Exhausted T cells and suppressive immune microenvironment in human intramucosal ESCC.
Figure 1 M,N. Distinct fibroblast subpopulations in human intramucosal ESCC ecosystem.

<b>Figure 2 A,B. Distinct fibroblast subpopulations in human intramucosal ESCC ecosystem. </b><br /> <b>Figure 2 C~G. Altered crosstalk between subclusters in human intramucosal ESCC. </b><br /> <b>Figure 2 H. Table showing Pathological characteristics of patients and percentage of E7 cell percentages</b>

Figure 2 A,B. Distinct fibroblast subpopulations in human intramucosal ESCC ecosystem.
Figure 2 C~G. Altered crosstalk between subclusters in human intramucosal ESCC.
Figure 2 H. Table showing Pathological characteristics of patients and percentage of E7 cell percentages

Background and Aim: Computer Aided Detection (CADe) systems have the potential to improve endoscopic detection of early neoplasia in Barrett’s Esophagus (BE) patients. We aimed to test a recently developed CADe system during live endoscopic procedures.
Method: The CADe system was developed using a large and heterogeneous BE training dataset originating from 15 international endoscopy centers, including 6.237 neoplastic (1.304 patients) and 7.595 non-dysplastic images (1.103 patients). Subsequently, it underwent rigorous external validation by means of multiple ex-vivo benchmarking studies. The CADe system displayed robust performance and significantly increased the neoplasia detection rate of endoscopists. In this pilot study, the CADe system was evaluated during endoscopic procedures of BE patients with a neoplastic lesion or with non-dysplastic Barrett’s esophagus (NDBE) in two tertiary hospitals. The protocol comprised a sequence of white light endoscopy videos obtained by a BE expert endoscopist with real-time evaluation and feedback by the CADe system. First, the Barrett’s segment was completely visualized with a standardized pullback video, starting at the gastric folds up to the maximum extent of the Barrett’s segment. Thereafter, every 2 centimeters of the Barrett’s segment, a 10 second overview video was recorded, starting in retroflexed position. Ground truth (the presence or absence of visible abnormalities requiring targeted biopsy) was established by the endoscopist before starting the protocol, followed by post-hoc histopathological confirmation (by targeted biopsies/endoscopic resection or acquisition of random biopsies). Outcome measure was the stand-alone performance of the CADe system in terms of sensitivity and specificity per patient.
Results: A total of 15 neoplastic and 15 NDBE patients were enrolled in the study. The CADe system correctly detected all neoplastic lesions on a per patient basis, resulting in a sensitivity of 100%. 14 out of 15 visible lesions were correctly diagnosed in the pullback videos. The missed neoplastic lesion was subsequently detected in the level videos. The CADe system incorrectly predicted neoplasia in 8 NDBE patients. Extrapolated to clinical practice, this would result in one additional targeted biopsy per patient whereas the mean Barrett length in this study would dictate 16 random biopsies. Histopathological examination confirmed neoplasia in 13 neoplastic cases (11x adenocarcinoma, 2x high-grade dysplasia), 2 cases did not contain dysplasia. In 4 cases in the non-dysplastic group, low-grade dysplasia was found in the random biopsy protocol.
Conclusion: This study is one of the first to evaluate a CADe system for real-time BE neoplasia detection in the endoscopy suite. The system correctly diagnosed all neoplastic lesions against the background of an acceptable number of false positive detections.

Figure 1) Visible abnormalities detected by the CADe system. Histopathological results (left to right): high-grade dysplasia, adenocarcinoma, high-grade dysplasia).

Figure 1) Visible abnormalities detected by the CADe system. Histopathological results (left to right): high-grade dysplasia, adenocarcinoma, high-grade dysplasia).

Instruction: Esophageal chromoendoscopy with Lugol's iodine staining (LIS) is known to improve the endoscopic visualization of esophageal squamous dysplasia and cancer. However, the use of LIS often causes mucosal irritation leading to severe thoracoabdominal discomforts after operation and possesses potential risk of allergic reaction. We have developed a novel staining with PVP-I and iodized salt as the main components to reduce the toxic mucosal damages, on the premise of ensuring the dyeing effect. The clinical usefulness of this novel staining for detecting suspected lesions and easing mucosal irritation were evaluated in this study.
Methods: We collected human esophageal ESD postoperative specimens and sprayed the novel staining on these specimens to observe the dyeing effect. After elution by Vitamin C solution, the specimens were sprayed with LIS. Pictures above were collected and image analysis software Image J was used to analyze the shape of lesions outlined. Differences in “lesion area” outlined between the novel staining and LIS were examined by the intraclass correlation coefficient (ICC) and One-Way Repeated Measures ANOVA when required. We have also completed animal experiments in order to verify the safety and irritation of the novel staining. 18 KM mice were randomly divided into three groups, and esophageal perfusion was performed with physiological saline, 1% novel staining and 1% LIS respectively. 2 mice in each group were killed, then completely removed esophagus and stomach for anatomy according to the time (2 hours, 1 day, 3 days after esophageal perfusion). The mucosal damage index (UI) was calculated according to Guth method.
Results: A total of 51 esophageal ESD specimens were stained in endoscopy center of West China hospital. The dyeing effect of novel staining is similar to that of LIS in terms of the color boundary definition of the esophagus [Figure 1]. The results showed that ICC=0.995 (>0.75), t=0.092 (>0.05), indicating that there was no difference between the two stainings in the “lesion area” outlined. In animal experiments, there was no accidental death, no obvious abnormality in activity, eating and spirit of mice in three groups after esophageal perfusion. In the acute inflammation period (1 day), the degree of redness and swelling of the gastric mucosa of mice reached the highest, the average UI value of the novel staining group was lower than LIS group, and gradually relieved after 3 days [Figure 2]. The above results indicate that the novel staining is safe and less irritating to the esophageal gastric mucosa than LIS.
Conclusion: This novel staining showed certain effectiveness, safety and feasibility compared with routine Lugol’s staining in the detection of early esophageal cancer. For safety reasons, we have not conducted experiments in humans, further clinical researchs will be carried out in the future.

Figure 1. Staining effect of human esophageal ESD specimens.<br /> A. Before staining<br /> B. Novel staining<br /> C. Elution by Vitamin C<br /> D. Lugol's iodine staining

Figure 1. Staining effect of human esophageal ESD specimens.
A. Before staining
B. Novel staining
C. Elution by Vitamin C
D. Lugol's iodine staining

Figure 2. Characteristics of esophageal and gastric mucosa in KM mice after esophageal perfusion. <br /> A. Physiological saline (2h)<br /> B. Physiological saline (1d)<br /> C. Physiological saline (3d)<br /> D. 1% novel staining (2h)<br /> E. 1% novel staining (1d)<br /> F. 1% novel staining (3d)<br /> G. 1% Lugol's iodine staining (2h)<br /> H. 1% Lugol's iodine staining (1d)<br /> I. 1% Lugol's iodine staining (3d)

Figure 2. Characteristics of esophageal and gastric mucosa in KM mice after esophageal perfusion.
A. Physiological saline (2h)
B. Physiological saline (1d)
C. Physiological saline (3d)
D. 1% novel staining (2h)
E. 1% novel staining (1d)
F. 1% novel staining (3d)
G. 1% Lugol's iodine staining (2h)
H. 1% Lugol's iodine staining (1d)
I. 1% Lugol's iodine staining (3d)

Quantitative fluorescence endoscopy (QFE) enables real-time in vivo visualization and quantification of targeted fluorescence tracers and can thereby highlights tumor tissue. QFE has proven its potential in detection of residual disease but is limited to the mucosal side only. Ultrasound guided needle biopsy (USNB) reflectance spectroscopy potentially enables assessment of lymph nodes and all esophageal wall layers. This pilot study evaluates the clinical impact of USNB fluorescence spectroscopy using bevacizumab-800CW for the identification of residual tumor in patients with esophageal cancer. Patients diagnosed with locally advanced esophageal adenocarcinoma (cT1b-4a N0-3 M0) scheduled for neoadjuvant chemoradiotherapy (CROSS) followed by surgery were included. Patients were intravenously injected with either 4.5, 10 or 25 mg bevacizumab-800CW 2-3 days prior to the QFE to establish the optimal dose. FME collected real-time white light and fluorescence images of the tumorbed and normal esophageal tissue. Quantitative fluorescent measurements were acquired luminal using multi-diameter single fiber reflectance / single fiber fluorescence (MDSFR/SFF) and inside the esophageal wall and within lymph nodes using USNB fluorescence spectroscopy. Additionally, optical properties like the reduced scattering coefficient and absorption coefficient were determined from the MDSFR/SFF data. Fifteen patients were included in the dose-escalation, where 25mg had the best differentiation between tumorbed and normal tissue (TBR: 2.63±0.60, p=0.0002). USNB was evaluated in 6/8 patients from the 25mg cohort. The tumorbed (p=0.059) and lymph nodes (p=0.041) showed a higher fluorescence intensity compared to normal tissue, as expected. Fluorescence measurements of normal tissue were comparable between the luminal MDSFR/SFF and USNB measurements. In 33% of all patients residual tumor was found only below a tumor-free mucosa. Last, the optical properties, the reduced scattering coefficient and absorption coefficient at 800nm, both showed a trend between normal tissue and amount of residual tumor. This trend was decreasing for the scattering and increasing for the absorption. In the present study we showed the potential of bevacizumab-800CW and QFE for response monitoring in esophageal cancer. QFE allows for both superficial and deeper tissue measurements to determine both the targeted fluorescence tracers and optical properties in vivo. A phase II study should establish the sensitivity and specificity of this bevacizumab-800CW and QFE and its additional value for response monitoring for individual patients.

Ultrasound-guided needle biopsy (USNB) fluorescence spectroscopy. (A) in vivo measurements of the esophageal wall. (B) The measured intrinsic fluorescence in normal tissue, the primary tumor site and within lymph nodes.

Ultrasound-guided needle biopsy (USNB) fluorescence spectroscopy. (A) in vivo measurements of the esophageal wall. (B) The measured intrinsic fluorescence in normal tissue, the primary tumor site and within lymph nodes.

Introduction: Barrett’s esophagus (BE) can progress from non-dysplastic BE (NDBE) to low-grade dysplasia, high-grade dysplasia (HGD), and esophageal adenocarcinoma (EAC). Early detection of BE patients at high risk for progression is essential for optimizing patient outcomes, as effective treatments are available to prevent progression to EAC. However, current management practices have limited ability to risk-stratify patients, and there are patients diagnosed as NDBE who harbor missed prevalent HGD/EAC or progress to HGD/EAC during guideline-recommended surveillance intervals. A tissue systems pathology test (TissueCypher, TSP-9) has been validated to provide a more accurate and objective method to risk-stratify patients with BE. This study aimed to evaluate the clinical experience with TSP-9 in an expert gastrointestinal and foregut surgery center, and to report the impact of the test in guiding risk-aligned management of patients with NDBE.

Methods: The TSP-9 test was ordered for 12 patients who had an endoscopy in the previous 42 days with biopsies diagnosed by a gastrointestinal subspecialist pathologist as NDBE. The test was ordered to enable risk stratification of NDBE to guide decision-making regarding use of endoscopic eradication therapy (EET) versus surveillance, and to guide surveillance intervals. The TSP-9 test reports a risk class (low, intermediate, or high) for progression to HGD/EAC within 5 years. The impact of the test results on BE management decisions was assessed.

Results: Five patients were male (41.7%), 7 were female (58.3%), and age ranged from 55-80 years (mean, 68.8+7.4). Ten patients had short-segment and 2 had long-segment BE. Ten patients were Medicare-eligible (Table 1). Nine patients received a low-risk result, 1 scored intermediate-risk, and 2 scored high-risk result by the TSP-9 test (Fig. 1A). Fifty percent of patients had management decisions impacted by the test results (Fig. 1B). Following shared decision-making, all 3 patients who scored high or intermediate risk, and had short-segment BE, received EET. With support from the TSP-9 results, the EET was covered by Medicare. In 3 of 9 patients who scored low risk, the TSP-9 test results supported downstaging of management from surveillance in 1-2 years to surveillance in 3 years.

Conclusions: The TSP-9 test results demonstrated significant clinical utility in NDBE patients by enabling upstaging of care to EET for patients with short-segment BE who scored high/intermediate risk, and by supporting extension of surveillance intervals to 3 years in patients who scored low risk. At-risk NDBE patients can be missed if only clinicopathologic variables are considered, but individualized risk assessment by TSP-9, used in conjunction with clinicopathologic data, can enable risk-aligned care for NDBE in a manner consistent with improved health outcomes.

<b>Table 1.</b> Characteristics of Patients with Non-Dysplastic Barrett’s Esophagus (NDBE)

Table 1. Characteristics of Patients with Non-Dysplastic Barrett’s Esophagus (NDBE)

Background: High-resolution microendscopy (HRME) is a low-cost, portable endoscopic probe capable of providing real-time “optical biopsy” images of the esophagus. Artificial intelligence (AI) algorithms can assist with interpretation of the microscopic and nuclear features of these histopathology-like images. Integration of patient-level risk factors may allow further optimization of these AI algorithms for comprehensive neoplasia assessment. Our aim was to compare the predictive ability of a HRME deep learning algorithm alone for detection of esophageal squamous cell neoplasia (ESCN) to that of comprehensive risk models combining the algorithm with demographic and clinical risk factors.

Methods: We used data from a multicenter clinical trial comprising of one hospital in Houston, Texas and two hospitals in China of patients undergoing ESCN screening and surveillance endoscopy with HRME imaging. HRME images were interpreted using a multi-task convolutional neural network using nuclear segmentation and classification based on Y-Net architecture. For patients with multiple imaged sites, we used the mean of the deep learning algorithm scores for the per-patient analysis. Discrimination of the deep learning algorithm in detecting ESCN was compared to that of comprehensive logistic regression risk models which incorporated demographic and clinical risk factors into the algorithm. Sensitivity analysis was also performed using machine learning Random Forest models.

Results: A total of 233 patients yielded 266 esophageal sites (176 with neoplasia) imaged with HRME. Our deep learning algorithm showed strong discrimination for detecting ESCN (area under the receiver operator characteristics [AUROC] 0.864, 95% confidence interval [CI] 0.818-0.911 per-site; AUROC 0.851, 95% CI 0.796-0.907 per-patient). Without the deep learning algorithm, patient-level risk factors alone (race, age, personal history of esophageal neoplasia, family history of esophageal neoplasia, smoking, alcohol use) had AUROC 0.848 (95% CI 0.796-0.920) for predicting ESCN. When we combined the deep learning algorithm with select patient-level risk factors (personal history of esophageal neoplasia, family history of esophageal neoplasia, smoking, alcohol use), the model improved further (AUROC 0.921, 95% CI 0.879-0.963; Figure 1). The robust discrimination of the combined model was also confirmed using Random Forest analysis (AUROC 0.881). The deep learning algorithm performed better in ESCN screening (AUROC 0.956) than surveillance (AUROC 0.741; Figure 2).

Conclusion: We found that the predictive ability of a deep learning algorithm in detecting ESCN on HRME images improved further with addition of patient-level demographic and clinical risk factors. These risk factors could be incorporated into AI algorithms used in endoscopic imaging technologies to optimize cancer detection in real-time.

Comparison of discriminatory ability of the deep learning algorithm alone and of the deep learning algorithm combined with patient-level risk factors in detecting esophageal squamous cell neoplasia in the entire study cohort (233 patients of whom 166 had esophageal neoplasia).

Comparison of discriminatory ability of the deep learning algorithm alone and of the deep learning algorithm combined with patient-level risk factors in detecting esophageal squamous cell neoplasia in the entire study cohort (233 patients of whom 166 had esophageal neoplasia).

Comparison of discriminatory ability of the deep learning algorithm alone and of the deep learning algorithm combined with patient-level risk factors among 58 patients who underwent esophageal squamous cell neoplasia (ESCN) screening and 175 patients who underwent ESCN surveillance.

Comparison of discriminatory ability of the deep learning algorithm alone and of the deep learning algorithm combined with patient-level risk factors among 58 patients who underwent esophageal squamous cell neoplasia (ESCN) screening and 175 patients who underwent ESCN surveillance.

Sharmila Anandasabapathy

Appearances