Enabling large-scale screening of Barrett's esophagus using weakly supervised deep learning in histopathology.

Timely detection of Barrett's esophagus, the pre-malignant condition of esophageal adenocarcinoma, can improve patient survival rates. The Cytosponge-TFF3 test, a non-endoscopic minimally invasive procedure, has been used for diagnosing intestinal metaplasia in Barrett's. However, it depends on pathologist's assessment of two slides stained with H&E and the immunohistochemical biomarker TFF3. This resource-intensive clinical workflow limits large-scale screening in the at-risk population. To improve screening capacity, we propose a deep learning approach for detecting Barrett's from routinely stained H&E slides. The approach solely relies on diagnostic labels, eliminating the need for expensive localized expert annotations. We train and independently validate our approach on two clinical trial datasets, totaling 1866 patients. We achieve 91.4% and 87.3% AUROCs on discovery and external test datasets for the H&E model, comparable to the TFF3 model. Our proposed semi-automated clinical workflow can reduce pathologists' workload to 48% without sacrificing diagnostic performance, enabling pathologists to prioritize high risk cases.

Understanding the malignant potential of gastric metaplasia of the oesophagus and its relevance to Barrett's oesophagus surveillance: individual-level data analysis.

OBJECTIVE: Whether gastric metaplasia (GM) of the oesophagus should be considered as Barrett's oesophagus (BO) is controversial. Given concern intestinal metaplasia (IM) may be missed due to sampling, the UK guidelines include GM as a type of BO. Here, we investigated whether the risk of misdiagnosis and the malignant potential of GM warrant its place in the UK surveillance. DESIGN: We performed a thorough pathology and endoscopy review to follow clinical outcomes in a novel UK cohort of 244 patients, covering 1854 person years of follow-up. We complemented this with a comparative genomic analysis of 160 GM and IM specimens, focused on early molecular hallmarks of BO and oesophageal adenocarcinoma (OAC). RESULTS: We found that 58 of 77 short-segment (<3 cm) GM (SS-GM) cases (75%) continued to be observed as GM-only across a median of 4.4 years of follow-up. We observed that disease progression in GM-only cases and GM+IM cases (cases with reported GM on some occasions, IM on others) was significantly lower than in the IM-only cases (Kaplan-Meier, p=0.03). Genomic analysis revealed that the mutation burden in GM is significantly lower than in IM (p<0.01). Moreover, GM does not bear the mutational hallmarks of OAC, with an absence of associated signatures and driver gene mutations. Finally, we established that GM found adjacent to OAC is evolutionarily distant from cancer. CONCLUSION: SS-GM is a distinct entity from SS-IM and the malignant potential of GM is lower than IM. It is questionable whether SS-GM warrants inclusion in BO surveillance.

Adequacy of endoscopic recognition and surveillance of gastric intestinal metaplasia and atrophic gastritis: A multicentre retrospective study in low incidence countries.

BACKGROUND: Gastric atrophy (GA) and gastric intestinal metaplasia (GIM) are precursor conditions to gastric adenocarcinoma (GAC) and should be monitored endoscopically in selected individuals. However, little is known about adherence to recommendations in clinical practice in low-risk countries. OBJECTIVE: The aim of this study was to evaluate endoscopic recognition and adequacy of surveillance for GA and GIM in countries with low GAC prevalence. METHODS: We retrospectively analysed patients diagnosed with GIM or GA in three centers in The Netherlands and UK between 2012 and 2019. Cases with GIM and/or GA diagnosis at index endoscopy were retrieved through systematic search of pathology databases using 'gastric' and 'intestinal metaplasia' or 'atrophy' keywords. Endoscopy reports were analysed to ascertain accuracy of endoscopic diagnoses. Adequacy of surveillance was assessed following histological diagnosis at the index endoscopy based on ESGE guidelines published in 2012. RESULTS: We included 396 patients with a median follow-up of 57.2 months. Mean age was 66 years and the rates of antrum-predominant versus extensive GIM were comparable (37% vs 38%). Endoscopic recognition rates were 48.5% for GA and 16.3% for GIM. Surveillance was adequately carried out in 215 of 396 patients (54.3%). CONCLUSION: In countries with a low incidence of GAC, the rate of endoscopic recognition of gastric pre-cancerous lesions and adherence to surveillance recommendation are low. Substantial improvement is required in endoscopic training and awareness of guidelines recommendation in order to optimise detection and management of pre-malignant gastric conditions.

Confocal endomicroscopy diagnostic criteria for early signet-ring cell carcinoma in hereditary diffuse gastric cancer.

BACKGROUND: Recognition of early signet-ring cell carcinoma (SRCC) in patients with hereditary diffuse gastric cancer (HDGC) undergoing endoscopic surveillance is challenging. We hypothesized that probe-based confocal laser endomicroscopy (pCLE) might help diagnose early cancerous lesions in the context of HDGC. The aim of this study was to identify pCLE diagnostic criteria for early SRCC. METHODS: Patients with HDGC syndrome were prospectively recruited and pCLE assessment was performed on areas suspicious for early SRCC and control regions during an endoscopic surveillance procedure. Targeted biopsies were taken for gold standard histologic assessment. In Phase I two investigators assessed video sequences off-line to identify pCLE features related to SRCC. In Phase II pCLE diagnostic criteria were evaluated in an independent video set by the investigators blinded to the histologic diagnosis. Sensitivity, specificity, accuracy, and interobserver agreement were calculated. RESULTS: Forty-two video sequences from 16 HDGC patients were included in Phase I. Four pCLE patterns associated to SRCC histologic features were identified: (A) glands with attenuated margins, (B) glands with spiculated or irregular shape, (C) heterogenous granular stroma with sparse glands, (D) enlarged vessels with tortuous shape. In Phase II, 38 video sequences from 15 patients were assessed. Criteria A and B and C had the highest diagnostic accuracy, with a κ for interobserver agreement ranging from 0.153 to 0.565. A panel comprising these 3 criteria with a cut-off of at least one positive criterion had a sensitivity of 80.9% (95%CI:58.1-94.5%) and a specificity of 70.6% (95%CI:44.0-89.7%) for a diagnosis of SRCC. CONCLUSIONS: We have generated and validated off-line pCLE criteria for early SRCC. Future real-time validation of these criteria is required.

Single-Cell RNA Sequencing Unifies Developmental Programs of Esophageal and Gastric Intestinal Metaplasia.

Intestinal metaplasia in the esophagus (Barrett's esophagus IM, or BE-IM) and stomach (GIM) are considered precursors for esophageal and gastric adenocarcinoma, respectively. We hypothesize that BE-IM and GIM follow parallel developmental trajectories in response to differing inflammatory insults. Here, we construct a single-cell RNA-sequencing atlas, supported by protein expression studies, of the entire gastrointestinal tract spanning physiologically normal and pathologic states including gastric metaplasia in the esophagus (E-GM), BE-IM, atrophic gastritis, and GIM. We demonstrate that BE-IM and GIM share molecular features, and individual cells simultaneously possess transcriptional properties of gastric and intestinal epithelia, suggesting phenotypic mosaicism. Transcriptionally E-GM resembles atrophic gastritis; genetically, it is clonal and has a lower mutational burden than BE-IM. Finally, we show that GIM and BE-IM acquire a protumorigenic, activated fibroblast microenvironment. These findings suggest that BE-IM and GIM can be considered molecularly similar entities in adjacent organs, opening the path for shared detection and treatment strategies. SIGNIFICANCE: Our data capture the gradual molecular and phenotypic transition from a gastric to intestinal phenotype (IM) in the esophagus and stomach. Because BE-IM and GIM can predispose to cancer, this new understanding of a common developmental trajectory could pave the way for a more unified approach to detection and treatment. See related commentary by Stachler, p. 1291. This article is highlighted in the In This Issue feature, p. 1275.

The effect of procedural time on dysplasia detection rate during endoscopic surveillance of Barrett's esophagus.

BACKGROUND : Endoscopic surveillance of Barrett's esophagus (BE) with Seattle protocol biopsies is time-consuming and inadequately performed in routine practice. There is no recommended procedural time for BE surveillance. We investigated the duration of surveillance procedures with adequate tissue sampling and effect on dysplasia detection rate (DDR). METHODS : We performed post hoc analysis from the standard arm of a crossover randomized controlled trial recruiting patients with BE (≥C2 and/or ≥M3) and no clearly visible dysplastic lesions. After inspection with white-light imaging, targeted biopsies of subtle lesions and Seattle protocol biopsies were performed. Procedure duration and biopsy number were stratified by BE length. The effect of endoscopy-related variables on DDR was assessed by multivariable logistic regression. RESULTS : Of 142 patients recruited, 15 (10.6 %) had high grade dysplasia/intramucosal cancer and 15 (10.6 %) had low grade dysplasia. The median procedural time was 16.5 minutes (interquartile range 14.0-19.0). Endoscopy duration increased by 0.9 minutes for each additional 1 cm of BE length. Seattle protocol biopsies had higher sensitivity for dysplasia than targeted biopsies (86.7 % vs. 60.0 %; P = 0.045). Longer procedural time was associated with increased likelihood of dysplasia detection on quadrantic biopsies (odds ratio [OR] 1.10, 95 %CI 1.00-1.20, P = 0.04), and for patients with BE > 6 cm also on targeted biopsies (OR 1.21, 95 %CI 1.04-1.40; P = 0.01). CONCLUSIONS : In BE patients with no clearly visible dysplastic lesions, longer procedural time was associated with increased likelihood of dysplasia detection. Adequate time slots are required to perform good-quality surveillance and maximize dysplasia detection.

Endoscopic surveillance with systematic random biopsy for the early diagnosis of hereditary diffuse gastric cancer: a prospective 16-year longitudinal cohort study.

BACKGROUND: Hereditary diffuse gastric cancer, generally caused by germline pathogenic variants in CDH1, presents with early-onset signet ring cell carcinoma. Prophylactic total gastrectomy is the definitive treatment. Endoscopic surveillance can inform the timing of prophylactic total gastrectomy through detection of microscopic signet ring cell carcinoma foci. However, evidence is scarce about the optimal endoscopic sampling technique and characterisation of signet ring cell carcinoma foci in hereditary diffuse gastric cancer. We aimed to formally assess the diagnostic yield of different sampling strategies and to identify criteria for the characterisation of endoscopic lesions. METHODS: For this prospective longitudinal cohort study, we included individuals aged 18 years or older at the Cambridge University Hospitals National Health Service (NHS) Foundation Trust who fulfilled testing criteria for hereditary diffuse gastric cancer between June 1, 2005, and July 31, 2021. The primary outcome was detection of intramucosal signet ring cell carcinoma foci. We assessed the detection rate and anatomical location of signet ring cell carcinoma in random biopsy samples taken according to a systematic protocol compared with biopsies targeted to endoscopic findings. Endoscopic lesions were examined with white-light and narrow band imaging with magnification to assess the likelihood of cancerous foci. FINDINGS: 145 individuals were included, of whom 68 (47%) were male and 92 (63%) carried the CDH1 pathogenic variant. 58 (40%) patients were diagnosed with invasive signet ring cell carcinoma over a median follow-up time of 51 months (IQR 18-80). The first diagnosis of signet ring cell carcinoma was most commonly made from random biopsies (29 [50%] of 58 patients), rather than targeted biopsies (15 [26%] patients). The anatomical distribution of signet ring cell carcinoma foci detected by random biopsies more accurately reflected those identified in prophylactic total gastrectomy specimens than did targeted biopsies. Omitting random biopsies in our cohort would have led to an under-diagnosis rate of 42%. Using a novel panel of endoscopic criteria, gastric lesions containing signet ring cell carcinoma were predicted with a sensitivity of 67·3% and a specificity of 90·2%. INTERPRETATION: Random biopsies enhance the early detection of signet ring cell carcinoma and are complementary to targeted biopsies in surveillance of hereditary diffuse gastric cancer. This sampling method should be the standard of care when performing all surveillance endoscopies for individuals with hereditary diffuse gastric cancer. FUNDING: UK Medical Research Council.

Use of a non-endoscopic immunocytological device (Cytosponge™) for post chemoradiotherapy surveillance in patients with oesophageal cancer in the UK (CYTOFLOC): A multicentre feasibility study.

Background: Effective surveillance strategies are required for patients diagnosed with oesophageal squamous cell carcinoma (OSCC) or adenocarcinoma (OAC) for whom chemoradiotherapy (CRT) is used as a potentially-curative, organ-sparing, alternative to surgery. In this study, we evaluated the safety, acceptability and tolerability of a non-endoscopic immunocytological device (the Cytosponge™) to assess treatment response following CRT. Methods: This multicentre, single-arm feasibility trial took place in 10 tertiary cancer centres in the UK. Patients aged at least 16 years diagnosed with OSCC or OAC, and who were within 4-16 weeks of completing definitive or neo-adjuvant CRT, were included. Participants were required to have a Mellow-Pinkas dysphagia score of 0-2 and be able to swallow tablets. All patients underwent a single Cytosponge™ assessment in addition to standard of care (which included post-treatment endoscopic evaluation with biopsy for patients undergoing definitive CRT; surgery for those who received neo-adjuvant CRT). The primary outcome was the proportion of consented, evaluable patients who successfully underwent Cytosponge™ assessment. Secondary and tertiary outcomes included safety, study consent rate, acceptance rate, the suitability of obtained samples for biomarker analysis, and the comparative efficacy of Cytosponge™ to standard histology (endoscopy and biopsy or post-resection specimen) in assessing for residual disease. The trial is registered with ClinicalTrials.gov, NCT03529669. Findings: Between 18th April 2018 and 16th January 2020, 41 (42.7%; 95% confidence interval (CI) 32.7-53.2) of 96 potentially eligible patients consented to participate. Thirty-nine (95.1%, 95% CI 83.5-99.4) successfully carried out the Cytosponge™ procedure. Of these, 37 (95%) would be prepared to repeat the procedure. There were only two grade 1 adverse events attributed to use of the Cytosponge™. Thirty-five (90%) of the completed Cytosponge™ samples were suitable for biomarker analysis; 29 (83%) of these were concordant with endoscopic biopsies, three (9%) had findings suggestive of residual cancer on Cytosponge™ not found on endoscopic biopsies, and three (9%) had residual cancer on endoscopic biopsies not detected by Cytosponge™. Interpretation: Use of the CytospongeTM is safe, tolerable, and acceptable for the assessment of treatment response following CRT in OAC and OSCC. Further evaluation of Cytosponge™ in this setting is warranted. Funding: Cancer Research UK, National Institute for Health Research, Medical Research Council.

Quantification of TFF3 expression from a non-endoscopic device predicts clinically relevant Barrett's oesophagus by machine learning.

BACKGROUND: Intestinal metaplasia (IM) is pre-neoplastic with variable cancer risk. Cytosponge-TFF3 test can detect IM. We aimed to 1) assess whether quantitative TFF3 scores can distinguish clinically relevant Barrett's oesophagus (BO) (C≥1 or M≥3) from focal IM pathologies (C<1, M<3 or IM of gastro-oesophageal junction); 2) whether TFF3 counts can be automated to inform clinical practice. METHODS: Patients from the Barett's oEsophagus Screening Trial 2 (BEST2) case-control and BEST3 randomised trials were used. For aim 1, TFF3-positive glands were scored manually and correlated with clinical diagnosis. For aim 2, machine learning approach was used to obtain TFF3 count and logistic regression with cross-validation was trained on the BEST2 dataset (n = 529) and tested in the BEST3 dataset (n = 158). FINDINGS: Patients with clinically relevant BO had higher mean TFF3 gland count compared to focal IM pathologies (mean difference 4.14; 95% confidence interval, CI 2.76-5.52, p < 0.001). The mean class-balanced validation accuracy was 0.84 (95% CI 0.77-0.90), and precision of 0.95 (95% CI 0.87-1.00) for detecting clinically relevant BO. Applying this model on BEST3 showed precision of 0.91 (95% CI 0.85-0.97) for focal IM pathologies with a class-balanced accuracy of 0.77 (95% CI 0.69-0.84). Using this model, 55% of patients (87/158) in BEST3 would fall below the threshold for clinically relevant BO and could avoid gastroscopy, while only missing 5.1% of patients (8/158). INTERPRETATION: Automated Cytosponge-TFF3 gland quantification may enable thresholds to be set to trigger confirmatory gastroscopy to minimize overdiagnosis of focal IM pathologies with very low cancer-associated risk. FUNDING: Cancer Research UK (12088/16893 and C14478/A21047).

Kyoto international consensus report on anatomy, pathophysiology and clinical significance of the gastro-oesophageal junction.

OBJECTIVE: An international meeting was organised to develop consensus on (1) the landmarks to define the gastro-oesophageal junction (GOJ), (2) the occurrence and pathophysiological significance of the cardiac gland, (3) the definition of the gastro-oesophageal junctional zone (GOJZ) and (4) the causes of inflammation, metaplasia and neoplasia occurring in the GOJZ. DESIGN: Clinical questions relevant to the afore-mentioned major issues were drafted for which expert panels formulated relevant statements and textural explanations.A Delphi method using an anonymous system was employed to develop the consensus, the level of which was predefined as ≥80% of agreement. Two rounds of voting and amendments were completed before the meeting at which clinical questions and consensus were finalised. RESULTS: Twenty eight clinical questions and statements were finalised after extensive amendments. Critical consensus was achieved: (1) definition for the GOJ, (2) definition of the GOJZ spanning 1 cm proximal and distal to the GOJ as defined by the end of palisade vessels was accepted based on the anatomical distribution of cardiac type gland, (3) chemical and bacterial (Helicobacter pylori) factors as the primary causes of inflammation, metaplasia and neoplasia occurring in the GOJZ, (4) a new definition of Barrett's oesophagus (BO). CONCLUSIONS: This international consensus on the new definitions of BO, GOJ and the GOJZ will be instrumental in future studies aiming to resolve many issues on this important anatomic area and hopefully will lead to better classification and management of the diseases surrounding the GOJ.

The utility of P53 immunohistochemistry in the diagnosis of Barrett's oesophagus with indefinite for dysplasia.

AIMS: Barrett's oesophagus with indefinite for dysplasia (BE-IND) is a subjective diagnosis with a low interobserver agreement (IOA) among pathologists and uncertain clinical implications. This study aimed to assess the utility of p53 immunohistochemistry (p53-IHC) in assessing BE-IND specimens. METHODS AND RESULTS: Archive endoscopic biopsies with a BE-IND diagnosis from two academic centres were analysed. First, haematoxylin and eosin-stained slides (H&E) were reviewed by four expert gastrointestinal (GI) pathologists allocated into two groups (A and B). After a washout period of at least 8 weeks, H&E slides were reassessed side-to-side with p53-IHC available. We compared the rate of changed diagnosis and the IOA for all BE grades before and after p53-IHC. We included 216 BE-IND specimens from 185 patients, 44.0 and 32.9% of which were confirmed after H&E slide revision by groups A and B, respectively. More than half the cases were reclassified to a non-dysplastic BE (NDBE), while 5.6% of cases in group A and 7.4% in group B were reclassified to definite dysplasia. The IOA for NDBE, BE-IND, low-grade dysplasia (LGD) and high-grade dysplasia (HGD)/intramucosal cancer (IMC) was 0.31, 0.21, -0.03 and -0.02, respectively. Use of p53-IHC led to a >40% reduction in BE-IND diagnoses (P < 0.001) and increased IOA for all BE grades [κ = 0.46 (NDBE), 0.26 (BE-IND), 0.49 (LGD), 0.35 (HGD/IMC)]. An aberrant p53-IHC pattern significantly increased the likelihood of reclassifying BE-IND to definite dysplasia (odds ratio = 44.3, 95% confidence interval = 18.8-113.0). CONCLUSION: P53-IHC reduces the rate of BE-IND diagnoses and improves the IOA among pathologists when reporting BE with equivocal epithelial changes.

Image-Enhanced Endoscopy and Molecular Biomarkers Vs Seattle Protocol to Diagnose Dysplasia in Barrett's Esophagus.

BACKGROUND & AIMS: Dysplasia in Barrett's esophagus often is invisible on high-resolution white-light endoscopy (HRWLE). We compared the diagnostic accuracy for inconspicuous dysplasia of the combination of autofluorescence imaging (AFI)-guided probe-based confocal laser endomicroscopy (pCLE) and molecular biomarkers vs HRWLE with Seattle protocol biopsies. METHODS: Barrett's esophagus patients with no dysplastic lesions were block-randomized to standard endoscopy (HRWLE with the Seattle protocol) or AFI-guided pCLE with targeted biopsies for molecular biomarkers (p53 and cyclin A by immunohistochemistry; aneuploidy by image cytometry), with crossover to the other arm after 6 to 12 weeks. The primary end point was the histologic diagnosis from all study biopsies (trial histology). A sensitivity analysis was performed for overall histology, which included diagnoses within 12 months from the first study endoscopy. Endoscopists were blinded to the referral endoscopy and histology results. The primary outcome was diagnostic accuracy for dysplasia by real-time pCLE vs HRWLE biopsies. RESULTS: Of 154 patients recruited, 134 completed both arms. In the primary outcome analysis (trial histology analysis), AFI-guided pCLE had similar sensitivity for dysplasia compared with standard endoscopy (74.3%; 95% CI, 56.7-87.5 vs 80.0%; 95% CI, 63.1-91.6; P = .48). Multivariate logistic regression showed pCLE optical dysplasia, aberrant p53, and aneuploidy had the strongest correlation with dysplasia (secondary outcome). This 3-biomarker panel had higher sensitivity for any grade of dysplasia than the Seattle protocol (81.5% vs 51.9%; P < .001) in the overall histology analysis, but not in the trial histology analysis (91.4% vs 80.0%; P = .16), with an area under the receiver operating curve of 0.83. CONCLUSIONS: Seattle protocol biopsies miss dysplasia in approximately half of patients with inconspicuous neoplasia. AFI-guided pCLE has similar accuracy to the current gold standard. The addition of molecular biomarkers could improve diagnostic accuracy.

Use of a Cytosponge biomarker panel to prioritise endoscopic Barrett's oesophagus surveillance: a cross-sectional study followed by a real-world prospective pilot.

BACKGROUND: Endoscopic surveillance is recommended for patients with Barrett's oesophagus because, although the progression risk is low, endoscopic intervention is highly effective for high-grade dysplasia and cancer. However, repeated endoscopy has associated harms and access has been limited during the COVID-19 pandemic. We aimed to evaluate the role of a non-endoscopic device (Cytosponge) coupled with laboratory biomarkers and clinical factors to prioritise endoscopy for Barrett's oesophagus. METHODS: We first conducted a retrospective, multicentre, cross-sectional study in patients older than 18 years who were having endoscopic surveillance for Barrett's oesophagus (with intestinal metaplasia confirmed by TFF3 and a minimum Barrett's segment length of 1 cm [circumferential or tongues by the Prague C and M criteria]). All patients had received the Cytosponge and confirmatory endoscopy during the BEST2 (ISRCTN12730505) and BEST3 (ISRCTN68382401) clinical trials, from July 7, 2011, to April 1, 2019 (UK Clinical Research Network Study Portfolio 9461). Participants were divided into training (n=557) and validation (n=334) cohorts to identify optimal risk groups. The biomarkers evaluated were overexpression of p53, cellular atypia, and 17 clinical demographic variables. Endoscopic biopsy diagnosis of high-grade dysplasia or cancer was the primary endpoint. Clinical feasibility of a decision tree for Cytosponge triage was evaluated in a real-world prospective cohort from Aug 27, 2020 (DELTA; ISRCTN91655550; n=223), in response to COVID-19 and the need to provide an alternative to endoscopic surveillance. FINDINGS: The prevalence of high-grade dysplasia or cancer determined by the current gold standard of endoscopic biopsy was 17% (92 of 557 patients) in the training cohort and 10% (35 of 344) in the validation cohort. From the new biomarker analysis, three risk groups were identified: high risk, defined as atypia or p53 overexpression or both on Cytosponge; moderate risk, defined by the presence of a clinical risk factor (age, sex, and segment length); and low risk, defined as Cytosponge-negative and no clinical risk factors. The risk of high-grade dysplasia or intramucosal cancer in the high-risk group was 52% (68 of 132 patients) in the training cohort and 41% (31 of 75) in the validation cohort, compared with 2% (five of 210) and 1% (two of 185) in the low-risk group, respectively. In the real-world setting, Cytosponge results prospectively identified 39 (17%) of 223 patients as high risk (atypia or p53 overexpression, or both) requiring endoscopy, among whom the positive predictive value was 31% (12 of 39 patients) for high-grade dysplasia or intramucosal cancer and 44% (17 of 39) for any grade of dysplasia. INTERPRETATION: Cytosponge atypia, p53 overexpression, and clinical risk factors (age, sex, and segment length) could be used to prioritise patients for endoscopy. Further investigation could validate their use in clinical practice and lead to a substantial reduction in endoscopy procedures compared with current surveillance pathways. FUNDING: Medical Research Council, Cancer Research UK, Innovate UK.

Data Set for Reporting Carcinoma of the Stomach in Gastrectomy.

CONTEXT.­: A standardized detailed surgical pathology report is the cornerstone of gastric cancer management. OBJECTIVE.­: To guide management and prognostication for patients with gastric carcinomas globally, the International Collaboration on Cancer Reporting aimed to produce an evidence-based international pathology reporting data set with a panel of globally recognized expert pathologists and clinicians. DESIGN.­: Based on published guidelines/data sets for gastric carcinomas, a working draft was developed by the chair of the expert panel of pathologists and clinicians. The draft was then circulated to the panel and discussed in a series of teleconferences and email communications until consensus was achieved. The draft data set was uploaded on the International Collaboration on Cancer Reporting Web site for public comment. The data set was reviewed in consideration of the feedback, and a final version was approved by the panel. RESULTS.­: This data set was developed for gastrectomy specimens for primary gastric carcinomas, including neuroendocrine carcinomas and mixed neuroendocrine-nonneuroendocrine neoplasms. Well-differentiated neuroendocrine tumors, nonepithelial malignancies, and secondary tumors were excluded from this data set. The final data set contains 15 core (required) elements and 8 noncore (recommended) elements. A commentary is provided for each element. CONCLUSIONS.­: The International Collaboration on Cancer Reporting has published freely available, evidence-based data sets for gastric cancer reporting. Standardized reporting has been shown to improve patient care and facilitates data exchange and analysis for quality assurance, cancer epidemiology, and clinical and basic research.

Thyroid diagnostic modalities (fine needle aspiration and core needle biopsy) with histology correlation: a tertiary centre experience.

AIMS: To determine the proportion of thyroid fine needle aspiration (FNA) and core needle biopsy (CNB) cases reported at a single institute into each UK Royal College of Pathologists (RCPath) Thy1-5 and local T category, respectively. Where subsequent histology was available, malignancy rates, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and diagnostic accuracy were compared for both procedures. METHODS: 1591 FNAs (2010-2018) and 514 CNBs (2013-2018) cases were identified, together with paired histology excision specimens. RESULTS: The FNA samples were classified as: Thy1: 45.3%, Thy2/Thy2c: 22.1%, Thy3a/Thy3f: 28%, Thy4: 1.6% and Thy5: 3%; while the CNB were classified as: T1: 7.2%, T2: 22.4%, T3 59.3%, T4: 1% and T5: 10.1%. Comparison of FNA and CNB classified as Thy5/T5 showed a 100% risk of malignancy (ROM), sensitivity (98% vs 100%), specificity (14.1% vs 12.1%), PPV (29.4% vs 29.4%), NPV (94.9% vs 100%) and accuracy (36.5% vs 35.6%), respectively, for a diagnosis of malignancy. ROMs for other categories were: Thy1/T1 (9% vs 6.7%), Thy2/T2 (5.1% vs 0%), Thy3/T3 (17.5% vs 18.4%) and Thy4/T4 (73.3% vs 100%). CONCLUSIONS: The proportion of cases in each RCPath Thy category has remained relatively stable during the 9-year study period, with the exception of the Thy3a category, which has increased over time. This finding is in line with other more recent reports in the literature and the proportion of T3 cases in the CNB group. The proportion of Thy2/Thy2c cases has also reduced over time, reflecting a local change in the triaging protocol for probable benign lesions. Both FNA and CNB showed comparable performance in our study.

Acceptability and Adequacy of a Non-endoscopic Cell Collection Device for Diagnosis of Barrett's Esophagus: Lessons Learned.

BACKGROUND: Endoscopic screening for Barrett's esophagus (BE) is common, costly, and underperformed in at-risk people. A non-endoscopic cell collection device can be used to collect esophageal cells, enabling BE screening. AIMS: This study assessed the acceptability and adequacy of a commercial non-endoscopic cell collection device in a US population. METHODS: Six sites enrolled patients with confirmed BE or heartburn/regurgitation for ≥ 6 months. Patients underwent administration of the device, consisting of a sponge encapsulated in a capsule. The capsule dwelled in the stomach for 7.5 min and was retracted via an attached suture. An adequate sample was ≥ 1 columnar cell by H&E staining. Sample quality was rated using a 0-5 scale, with 0 = no columnar cells and 5 = plentiful groups. Trefoil Factor 3 (TFF3) staining was performed. Accuracy was assessed using esophagogastroduodenoscopy (EGD)/biopsy as the gold standard. RESULTS: Of 191 patients, 99.5% successfully swallowed the device. Overall sample adequacy was 91% (171/188), with 84% (158/188) high quality. The detachment rate was 2/190 (1%). Overall sensitivity, specificity, and accuracy of the assay with TFF3 staining were 76%, 77%, and 76%. Sensitivity, specificity, and accuracy for ≥ 3 cm BE were 86%, 77%, and 82%. Asked if willing to repeat the procedure, 93% would, and 65% indicated a preference for the device over EGD. CONCLUSIONS: This study demonstrated a high rate of sample adequacy and promising acceptability of this non-endoscopic sampling device in a US population. Diagnostic characteristics suggest that non-endoscopic assessment of BE deserves further development as an alternative to endoscopy.

Utility and Cost-Effectiveness of a Nonendoscopic Approach to Barrett's Esophagus Surveillance After Endoscopic Therapy.

BACKGROUND & AIMS: A non-endoscopic approach to Barrett's esophagus (BE) surveillance after radiofrequency ablation (RFA) would offer a less invasive method for monitoring. We assessed the test characteristics and cost-effectiveness of the Cytosponge (Medtronic, Minneapolis, MN) in post-RFA patients. METHODS: We performed a multicenter study of dysplastic BE patients after at least one round of RFA. A positive Cytosponge before endoscopy was defined as intestinal metaplasia (IM) on cytological assessment and/or TFF3 immunohistochemistry. Sensitivity, specificity, and receiver operator characteristic (ROC) curves were calculated. Multivariable regression was used to estimate the odds of a positive Cytosponge in BE. A microsimulation cost-effectiveness model was performed to assess outcomes of various surveillance strategies: endoscopy-only, Cytosponge-only, and alternating endoscopy/Cytosponge. RESULTS: Of 234 patients, Cytosponge adequately sampled the distal esophagus in 175 (75%). Of the 142 with both endoscopic and histologic data, 19 (13%) had residual/recurrent BE. For detecting any residual Barrett's, Cytosponge had a sensitivity of 74%, specificity of 85%, accuracy of 84%, and ROC curve showed an area under the curve of 0.74. The adjusted odds of a positive Cytosponge in BE were 17.1 (95% CI, 5.2-55.9). Cytosponge-only surveillance dominated all the surveillance strategies, being both less costly and more effective. Cytosponge-only surveillance required <1/4th the endoscopies, resulting in only 0.69 additional EAC cases/1000 patients, and no increase in EAC deaths when compared to currently-practiced endoscopy-only surveillance. CONCLUSIONS: A positive Cytosponge test was strongly associated with residual BE after ablation. While the assay needs further refinement in this context, it could serve as a cost-effective surveillance examination.

Cytosponge-TFF3 Testing can Detect Precancerous Mucosal Changes of the Stomach.

Gastric intestinal metaplasia (GIM) and gastric atrophy (GA) are associated with increased risk of gastric cancer and are indications for endoscopic surveillance when affecting the proximal stomach.1 Endoscopic screening is not cost-effective in areas with low-moderate incidence of gastric cancer2; noninvasive methods to detect GIM/GA are currently lacking.3.

Molecular phenotyping reveals the identity of Barrett's esophagus and its malignant transition.

The origin of human metaplastic states and their propensity for cancer is poorly understood. Barrett's esophagus is a common metaplastic condition that increases the risk for esophageal adenocarcinoma, and its cellular origin is enigmatic. To address this, we harvested tissues spanning the gastroesophageal junction from healthy and diseased donors, including isolation of esophageal submucosal glands. A combination of single-cell transcriptomic profiling, in silico lineage tracing from methylation, open chromatin and somatic mutation analyses, and functional studies in organoid models showed that Barrett's esophagus originates from gastric cardia through c-MYC and HNF4A-driven transcriptional programs. Furthermore, our data indicate that esophageal adenocarcinoma likely arises from undifferentiated Barrett's esophagus cell types even in the absence of a pathologically identifiable metaplastic precursor, illuminating early detection strategies.

Triage-driven diagnosis of Barrett's esophagus for early detection of esophageal adenocarcinoma using deep learning.

Deep learning methods have been shown to achieve excellent performance on diagnostic tasks, but how to optimally combine them with expert knowledge and existing clinical decision pathways is still an open challenge. This question is particularly important for the early detection of cancer, where high-volume workflows may benefit from (semi-)automated analysis. Here we present a deep learning framework to analyze samples of the Cytosponge-TFF3 test, a minimally invasive alternative to endoscopy, for detecting Barrett's esophagus, which is the main precursor of esophageal adenocarcinoma. We trained and independently validated the framework on data from two clinical trials, analyzing a combined total of 4,662 pathology slides from 2,331 patients. Our approach exploits decision patterns of gastrointestinal pathologists to define eight triage classes of varying priority for manual expert review. By substituting manual review with automated review in low-priority classes, we can reduce pathologist workload by 57% while matching the diagnostic performance of experienced pathologists.