A deep learning system for detection of early Barrett's neoplasia: a model development and validation study.

BACKGROUND: Computer-aided detection (CADe) systems could assist endoscopists in detecting early neoplasia in Barrett's oesophagus, which could be difficult to detect in endoscopic images. The aim of this study was to develop, test, and benchmark a CADe system for early neoplasia in Barrett's oesophagus. METHODS: The CADe system was first pretrained with ImageNet followed by domain-specific pretraining with GastroNet. We trained the CADe system on a dataset of 14 046 images (2506 patients) of confirmed Barrett's oesophagus neoplasia and non-dysplastic Barrett's oesophagus from 15 centres. Neoplasia was delineated by 14 Barrett's oesophagus experts for all datasets. We tested the performance of the CADe system on two independent test sets. The all-comers test set comprised 327 (73 patients) non-dysplastic Barrett's oesophagus images, 82 (46 patients) neoplastic images, 180 (66 of the same patients) non-dysplastic Barrett's oesophagus videos, and 71 (45 of the same patients) neoplastic videos. The benchmarking test set comprised 100 (50 patients) neoplastic images, 300 (125 patients) non-dysplastic images, 47 (47 of the same patients) neoplastic videos, and 141 (82 of the same patients) non-dysplastic videos, and was enriched with subtle neoplasia cases. The benchmarking test set was evaluated by 112 endoscopists from six countries (first without CADe and, after 6 weeks, with CADe) and by 28 external international Barrett's oesophagus experts. The primary outcome was the sensitivity of Barrett's neoplasia detection by general endoscopists without CADe assistance versus with CADe assistance on the benchmarking test set. We compared sensitivity using a mixed-effects logistic regression model with conditional odds ratios (ORs; likelihood profile 95% CIs). FINDINGS: Sensitivity for neoplasia detection among endoscopists increased from 74% to 88% with CADe assistance (OR 2·04; 95% CI 1·73-2·42; p<0·0001 for images and from 67% to 79% [2·35; 1·90-2·94; p<0·0001] for video) without compromising specificity (from 89% to 90% [1·07; 0·96-1·19; p=0·20] for images and from 96% to 94% [0·94; 0·79-1·11; ] for video; p=0·46). In the all-comers test set, CADe detected neoplastic lesions in 95% (88-98) of images and 97% (90-99) of videos. In the benchmarking test set, the CADe system was superior to endoscopists in detecting neoplasia (90% vs 74% [OR 3·75; 95% CI 1·93-8·05; p=0·0002] for images and 91% vs 67% [11·68; 3·85-47·53; p<0·0001] for video) and non-inferior to Barrett's oesophagus experts (90% vs 87% [OR 1·74; 95% CI 0·83-3·65] for images and 91% vs 86% [2·94; 0·99-11·40] for video). INTERPRETATION: CADe outperformed endoscopists in detecting Barrett's oesophagus neoplasia and, when used as an assistive tool, it improved their detection rate. CADe detected virtually all neoplasia in a test set of consecutive cases. FUNDING: Olympus.

[Rectal bleeding: easy to overcome or still a challenge in proctology?] / Rektale Blutungen: leicht zu beherrschen oder doch eine Herausforderung in der Proktologie?

Rectal bleeding is a frequent symptom in proctology. In most cases frequent causes, such as anal fissures and hemorrhoidal disease can be diagnosed and treated using a structured patient history and basic proctological diagnostic assessment; however, it is not uncommon for proctitis to be the reason for rectal bleeding, which necessitates interdisciplinary diagnostics and treatment. In addition to proctitis associated with chronic inflammatory bowel disease, prolapse-induced, radiogenic, ischemic, infectious types and proctitis associated with sexually transmitted diseases represent important differential diagnoses. Moreover, rectal cancer has to be excluded as the cause of rectal bleeding. Finally, with appropriate diligence most causes of rectal bleeding can be securely identified and effectively managed; however, special circumstances can necessitate interdisciplinary diagnostics and management, including conservative, topical, interventional and surgical treatment options.

An Interactive Web-Based Educational Tool Improves Detection and Delineation of Barrett's Esophagus-Related Neoplasia.

BACKGROUND & AIMS: Endoscopic detection of early Barrett's esophagus-related neoplasia (BORN) is a challenge. We aimed to develop a web-based teaching tool for improving detection and delineation of BORN. METHODS: We made high-definition digital videos during endoscopies of patients with BORN and non-dysplastic Barrett's esophagus. Three experts superimposed their delineations of BORN lesions on the videos using special tools. In phase one, 68 general endoscopists from 4 countries assessed 4 batches of 20 videos. After each batch, mandatory feedback compared the assessors' interpretations with those from experts. These data informed the selection of 25 videos for the phase 2 module, which was completed by 121 new assessors from 5 countries. A 5-video test batch was completed before and after scoring of the four 5-video training batches. Mandatory feedback was as in phase 1. Outcome measures were scores for detection, delineation, agreement delineation, and relative delineation of BORN. RESULTS: A linear mixed-effect model showed significant sequential improvement for all 4 outcomes over successive training batches in both phases. In phase 2, median detection rates of BORN in the test batch increased by 30% (P < .001) after training. From baseline to the end of the study, there were relative increases in scores of 46% for detection, 129% for delineation, 105% for agreement delineation, and 106% for relative delineation (all, P < .001). Scores improved independent of assessors' country of origin or level of endoscopic experience. CONCLUSIONS: We developed a web-based teaching tool for endoscopic recognition of BORN that is easily accessible, efficient, and increases detection and delineation of neoplastic lesions. Widespread use of this tool might improve management of Barrett's esophagus by general endoscopists.

Early-stage adenocarcinoma of the esophagus with mid to deep submucosal invasion (pT1b sm2-3): the frequency of lymph-node metastasis depends on macroscopic and histological risk patterns.

The rate of lymph-node (LN) metastasis in early adenocarcinoma (EAC) of the esophagus with mid to deep submucosal invasion (pT1b sm2/3) has not yet been precisely defined. The aim of the this study was to evaluate the rate of LN metastasis in pT1b sm2/3 EAC depending on macroscopic and histological risk patterns to find out whether there may also be options for endoscopic therapy as in cancers limited to the mucosa and the upper third of the submucosa. A total of 1.718 pt with suspicion of EAC were referred for endoscopic treatment (ET) to the Dept. of Internal Medicine II at HSK Wiesbaden 1996-2010. In 230/1.718 pt, the suspicion (endoscopic ultrasound, EUS) or definitive diagnosis of pT1b EAC (ER/surgery) was made. Of these, 38 pt had sm2 lesions, and 69 sm3. Rate of LN metastasis was analyzed depending on risk patterns: histologically low-risk (hisLR): G1-2, L0, V0; histologically high-risk (hisHR): ≥1 criterion not fulfilled; macroscopically low-risk (macLR): gross tumor type I-II, tumor size ≤2 cm; macroscopically high-risk (macHR): ≥1 criterion not fulfilled; combined low-risk (combLR): hisLR+macLR; combined high-risk (combHR): at least 1 risk factor. LN rate was only evaluated in pt who had proven maximum invasion depth of sm2/sm3, and who in case of ET had a follow-up (FU) by EUS of at least 24 months. 23/38 pt with pT1b sm2 lesions and 39/69 pt with sm3 lesions fulfilled our inclusion criteria. In the pT1b sm2 group, rate of LN metastasis in the hisLR, hisHR, combLR, and combHR groups were 8.3% (1/12), 36.3% (4/11), 0% (0/5), and 27.8% (5/18). In the pT1b sm3 group, rate of LN metastasis in the hisLR, hisHR, combLR and combHR groups were 28.6% (2/7), 37.5% (12/32), 25% (1/4), and 37.1% (13/35). 30-day mortality of surgery was 1.7% (1/58 pt). In EAC with pT1b sm2/3 invasion, the frequency of LN metastasis depends on macroscopic and histological risk patterns. Surgery remains the standard treatment, because the rate of LN metastasis appears to be higher than the mortality risk of surgery. Whether a highly selected group of pT1b sm2 patients with a favourable risk pattern may be candidates for endoscopic therapy cannot be decided until the results of larger case volumes are available.

[Gastro-Esophageal Reflux Disease and Malignant Progression - Equal Risk for Men and Women?] / Gastroösophageale Refluxerkrankung und maligne Transformation ­ für Männer und Frauen ein gleich großes Risiko?

Gastroesophageal reflux disease (GERD) is highly prevalent in the Western world. Patients with GERD have a 10-fold increased risk to develop a Barrett's esophagus. Patients with Barrett's esophagus have a higher risk for an esophageal adenocarcinoma. Men have more severe reflux with a higher grade of inflammation and acid reflux. This seems to be the reason why men develop a Barrett's esophagus more frequently - the risk is approximately 2-3-fold and the risk for an esophageal adenocarcinoma is even 3-6 times higher.

[Multiple smooth colon stenosis in 76-year-old female patient]. / Multiple glatte Kolonstenosen bei einer 76-jährigen Frau.

Differential diagnostic aspects of colon stenoses are discussed using the example case of a female patient presenting with multilocular colon metastases, who had lobular breast cancer 9 years previously. Typical is linitis plastica, which can indicate tumorous infiltration not only of the stomach, but also of the large intestine. Other endoscopic imaging and histological studies may, however, fail. The pathologist requires the anamnestic data relating to the breast cancer for exact assignment of the tumorous infiltration.

[Gastro-esophageal Reflux Disease and malignant progression--equal risk for men and women?]. / Gastroösophageale Refluxerkrankung und maligne Transformation--für Männer und Frauen ein gleich großes Risiko?

Gastroesophageal reflux disease (GERD) is highly prevalent in the Western world. Patients with GERD have a 10 fold increased risk to develop a Barrett's esophagus. Patients with Barrett's esophagus have a higher risk for an esophageal adenocarcinoma. Men have more severe reflux with a higher grade of inflammation and acid reflux. This seems to be the reason why men develop a Barrett's esophagus more frequently--the risk is approximately 2 to 3 fold and the risk for an esophageal adenocarcinoma is even 3 to 6 times higher.

[Detection of early neoplasia in Barrett's oesophagus: focus attention on index endoscopy in short-segment-Barrett's oesophagus with random biopsies]. / Detektion von frühen Neoplasien im Barrett-Ösophagus: Ein Plädoyer für die Wertigkeit der Indexendoskopie und 4-Quadranten-Biopsien im Short-Segment-Barrett-Ösophagus.

BACKGROUND: Detecting early neoplasias in Barrett's oesophagus (BE) is challenging. Recent publications have been focusing on improving the detection of such lesions during Barrett's surveillance. However in a recently published Danish register study calculating the risk for cancer-development in BE two-thirds of the diagnosed tumors were identified during the first examination or in the first year. This means that index endoscopy might be more effective than surveillance in detecting early neoplasia in BE. METHODS: In the period from January 2010 to April 2011, all patients who consecutively presented with a diagnosis of early neoplastic changes in BE were recorded prospectively. ANALYSIS: The analysis included data for 121 patients. In patients with short-segment BE (SSBE), neoplasia was only diagnosed in 6 % of cases in the surveillance examination, compared with 44 % of cases in long-segment BE (LSBE). The neoplastic lesion was identified visually in 43 patients (36 %) during the external EGD. Type II tumours were detected in 40 % (39/98) and were correctly assessed as neoplastic in 25 % of cases (24/98). CONCLUSIONS: 1. in patients with SSBE almost all early tumours are diagnosed by index endoscopy and not by Barrett's surveillance; 2. around 40 % of all early neoplasias are endoscopically invisible and are only diagnosed using four-quadrant biopsies; 3. the macroscopic tumour type has a substantial influence on the detection rate for neoplasia. If efforts to increase the detection rate for early neoplasia in BE are focused solely on the Barrett's surveillance method, then only a minority of patients - 20 % in the present group - will benefit from the measure. German clinical trials register, DRKS00 004 168.

[Family with marked hyperferritinemia as a result of hemochromatosis type 4 (ferroportin disease)]. / Familie mit ausgeprägter Hyperferritinämie als Folge einer Hämochromatose Typ 4 (Ferroportin-Krankheit).

Iron overload in MR-imaging with decreased signal intensity in T2 weighting of liver, spleen, adrenal gland and pituitary gland in combination with an extremely elevated ferritin level of 9859  ng/mL and a positive family history of hyperferritinaemia led to the diagnosis of the rare hemochromatosis type 4 (synonym: ferroportin disease) in the case of a 62-year-old patient. The autosomal dominant disease was confirmed by analysis of the SLC40A1-gene. Histologically, a liver cirrhosis was detected. This was neither detectable in the case of the two similarly aged cousins (ferritin about 4750  ng/mL, transferrin saturation normal), nor in the case of the 82-year-old mother (ferritin 7860  ng/dL, transferrin saturation 58 %). Hemochromatosis type 4 with worldwide less than 200 described cases is caused by a disorder of the hepcidin ferroportin metabolism, which regulates the iron export from the cells. A hepatocellular carcinoma may occur even without cirrhosis. Therefore, surveillance of these patients is necessary. Treatment options are therapeutic phlebotomies and alternatively iron-chelating drugs (Deferoxamin, Deferasirox) if the patient develops anaemia.

Quality of life after surgical treatment of early Barrett's cancer: a prospective comparison of the Ivor-Lewis resection versus the modified Merendino resection.

INTRODUCTION: The Merendino (MER) procedure has been evaluated as an alternative to transthoracic esophageal resection (TER) for early stage Barrett's carcinoma. Apart from reducing morbidity and mortality, improvements concerning postoperative health-related quality of life (HRQL) have been postulated. The aim of our study was to compare HRQL between these procedures. MATERIALS AND METHODS: Between July 2000 and July 2007, 117 patients with early Barrett's carcinoma underwent surgery. Patients with tumor recurrence were excluded from the study. HRQL was assessed 1 and 2 years after surgery using the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Core Questionnaire (EORTC-QLQ-C30) and the QLQ-OES18 module. Patients recently diagnosed with early Barrett's carcinoma served as controls. Symptoms that showed a difference of more than ten between the control and the study groups were considered clinically relevant and were tested for significant differences between the study groups using the Mann-Whitney U test (p < 0.05). RESULTS: The response rates for the questionnaires ranged between 70 and 93 %. In the MER group, more items reflected a clinical relevant impairment of HRQL than in the TER group. Significant complaints in the MER group included nausea/vomiting, appetite loss, local pain, difficulties with social eating, and choking. Moreover, we found a significant restriction concerning global health and emotional and social functioning in this group 1 year after surgery. 2 years postoperatively, hardly any differences between the operative techniques could be detected. The only symptom in favor of the MER procedure was a better dysphagia score postoperatively. CONCLUSION: Our study suggests that MER procedure is not superior to subtotal esophagectomy with regard to HRQL.

Predictive factors for initial treatment response after circumferential radiofrequency ablation for Barrett's esophagus with early neoplasia: a prospective multicenter study.

BACKGROUND AND STUDY AIMS: Radiofrequency ablation (RFA) is safe and effective for the eradication of neoplastic Barrett's esophagus; however, occasionally there is minimal regression after initial circumferential balloon-based RFA (c-RFA). This study aimed to identify predictive factors for a poor response 3 months after c-RFA, and to relate the percentage regression at 3 months to the final treatment outcome. METHODS: We included consecutive patients from 14 centers who underwent c-RFA for high grade dysplasia at worst. Patient and treatment characteristics were registered prospectively. "Poor initial response" was defined as < 50 % regression of the Barrett's esophagus 3 months after c-RFA, graded by two expert endoscopists using endoscopic images. Predictors of initial response were identified through logistic regression analysis. RESULTS: There were 278 patients included (median Barrett's segment C4M6). In poor initial responders (n = 36; 13 %), complete response for neoplasia (CR-neoplasia) was ultimately achieved in 86 % (vs. 98 % in good responders; P < 0.01) and complete response for intestinal metaplasia (CR-IM) in 66 % (vs. 95 %; P < 0.01). Poor responders required 13 months treatment (vs. 7 months; P < 0.01) for a median of four RFA sessions (vs. three; P < 0.01). We identified four independent baseline predictors of poor response: active reflux esophagitis (odds ratio [OR] 37.4; 95 % confidence interval [CI] 3.2 - 433.2); endoscopic resection scar regeneration with Barrett's epithelium (OR 4.7; 95 %CI 1.1 - 20.0); esophageal narrowing pre-RFA (OR 3.9; 95 %CI 1.0 - 15.1); and years of neoplasia pre-RFA (OR 1.2; 95 %CI 1.0 - 1.4). CONCLUSIONS: Patients with a poor initial response to c-RFA have a lower ultimate success rate for CR-neoplasia/CR-IM, require more treatment sessions, and a longer treatment period. A poor initial response to c-RFA occurs more frequently in patients who regenerate their endoscopic resection scar with Barrett's epithelium, and those with ongoing reflux esophagitis, neoplasia in Barrett's esophagus for a longer time, or a narrow esophagus.

[S3 Guideline on "sedation in gastrointestinal endoscopy": how much does the new guideline cost in everyday hospital work? A calculation model and analysis of implementation in 2011 among ALGK members]. / S3-Leitlinie "Sedierung in der gastrointestinalen Endoskopie": Was kostet die neue Leitlinie im stationären Alltag? Ein Rechenmodell und eine Analyse der Umsetzung 2011 unter den Mitgliedern der ALGK.

BACKGROUND: The introduction of the S3 guideline on "Sedation in Gastrointestinal Endoscopy" in 2008 led to substantial organisational, structural, staffing-related, financial, and legal changes in the everyday work of departments of gastrointestinal endoscopy. This study examines the economic effects of this and the change in the legal situation. In addition, the extent to which the guideline has been implemented was assessed using an electronic questionnaire circulated to the members of the Working Group of Head Gastroenterologists in Hospitals (ALGK). METHODS: The increased financial burden in the area of staff costs resulting from correct compliance with the sedation guideline was calculated using research figures from an endoscopy department in a maximum-care non-university hospital, dating from 2009. Interpretation of the legal implications of the guideline was provided by lawyers familiar with medical legislation. The question of the extent to which the guideline has been implemented in everyday clinical practice in endoscopy departments in 2011 was investigated using an evaluation questionnaire sent to the members of the ALGK. RESULTS: Implementation of the S3 sedation guideline leads to a substantial increase in the financial burden in the area of staff costs. Assuming 8000 in-patient endoscopy procedures, a conservative estimate indicates extra costs amounting to € 257 462.- per year (gross costs for the employer). The analysis of the questionnaire sent to ALGK members on the implementation of the S3 guideline 3 years after its publication showed that its major points, particularly the deployment of a third staff member for sedation, have not been acted on. CONCLUSIONS: The S3 guideline on "Sedation in Gastrointestinal Endoscopy" leads to a substantial increase in the financial burden in the area of staff costs by requiring the presence of a third person exclusively concerned with sedation. This recommendation was issued by the authors of the guideline without any evidence being available. In addition, it leads to a clear change in the legal situation, which in case of claims arising is associated with substantial implications for the physician responsible. The questionnaire evaluation among the members of the ALGK showed that the guideline has not so far been implemented in in-patient gastroenterology.

How is early Barrett's cancer currently diagnosed and treated in Western Europe? Results of a survey at 52 university hospitals in eight Western European countries.

BACKGROUND AND STUDY AIMS: Endoscopic treatment is increasingly being accepted for early Barrett's cancer (EBC), as it is associated with few complications, excellent long-term results, and almost no mortality. This study investigated current standards and treatment strategies for EBC in eight countries in Western Europe. METHODS: A standardized questionnaire with questions on the endoscopic diagnosis, staging, and treatment of EBC was developed and sent to 107 university gastroenterology departments. The data were analyzed anonymously. RESULTS: The response rate was 49 % (52/107). For work-up of early Barrett's neoplasia, 67 % of hospitals use high-resolution endoscopes, with routine four-quadrant and targeted biopsies of visible lesions in 94 % of the cases. Narrow-band imaging and chromoendoscopy are used in 67 % of the cases, and other advanced imaging and staining techniques in 65 %. Before treatment, 63 % of the respondents recommended conventional endosonography, 6 % miniprobe endosonography, and 19 % both. Endoscopic resection is carried out at 98 % of the hospitals. Argon plasma coagulation is used for ablation in 52 % of the cases and radiofrequency ablation in 27 %. An 80-year-old patient with localized mucosal EBC would be treated endoscopically in all of the hospitals. Endoscopic therapy was recommended for 50-year-old patients with mucosal EBC by 87 % of the hospitals; esophageal resection was recommended for multifocal EBC by 15 % in 80-year-old patients, by 63 % in 50-year-old patients and by 44 % in patients with incipient submucosal infiltration. CONCLUSIONS: About two-thirds of the university hospitals use high-resolution endoscopy and advanced imaging. Endoscopic therapy is the accepted standard for treating localized mucosal Barrett's cancer in Western Europe; esophageal resection is recommended by the majority (63 %) for a young patient with multifocal EBC.

[Endoscopic therapy in early gastric cancer]. / Endoskopische Therapie von Magenfrühkarzinomen.

Endoscopic treatment of early gastric cancer has been shown to be effective and safe. It is a minimally invasive and organ-preserving treatment approach that can safely be used as an alternative to surgical resection. A prerequisite of any endoscopic treatment with a curative intent is a very low risk of lymph node metastasis of the lesion intended to be endoscopically resected. As in high-volume surgical centers all endoscopic procedures with a curative intent should also be carried out in centers with a high expertise in the different endoscopic resection (ER) techniques available, the major techniques being suck-and-cut ER and endoscopic submucosal dissection (ESD). Because of the risk of secondary neoplasia after curative endoscopic therapy, a standardized follow-up protocol is required. The majority of secondary neoplastic lesions can again be treated endoscopically. In the present article an overview of initial staging procedures, techniques, indications, as well as follow-up strategies after endoscopic therapy for early gastric cancer is given.

[Early esophageal cancer: pro-endoscopic resection]. / Frühes Ösophaguskarzinom: Pro endoskopische Resektion.

In several publications including more than 1,000 endoscopically treated patients it could be demonstrated that endoscopic treatment of early esophageal cancer is effective and safe. A combination of endoscopic resection (ER) of early Barrett's neoplasia followed by ablation of the remaining non-dysplastic Barrett's mucosa was able to significantly reduce the rate of recurrence and metachronous neoplasia. Long-term complete remission rates of more than 95% can be achieved. Compared to esophagectomy, endoscopic treatment has only a very low complication rate and almost no mortality. Therefore, ER is the treatment of choice and recommended by several international guidelines. Esophageal resection only plays a minor role in patients with early esophageal cancer.

Accuracy of endoscopic ultrasound in preoperative staging of esophageal cancer: results from a referral center for early esophageal cancer.

BACKGROUND AND STUDY AIM: Endoscopic ultrasound (EUS) has been regarded as the most accurate staging tool in esophageal cancer. Staging results have a strong impact on the decision as to whether a patient should undergo endoscopic treatment, surgery alone, or neoadjuvant therapy. This retrospective study was conducted to analyze the accuracy of esophageal cancer staging using EUS. METHODS: All patients who received EUS for staging of esophageal cancer before esophagectomy from February 2003 to December 2007 at a high volume academic tertiary care center were included. RESULTS: 179 consecutive patients (mean age 64.4 +/- 9.5 years; 142 men) underwent esophageal resection for Barrett's adenocarcinoma (n = 134) and squamous cell cancer (n = 45). Postoperatively, 99 patients were staged as having T1 cancers (55 %), 30 patients T2 (17%), 46 patients T3 (26%), and four patients T4 (2%). The sensitivity and specificity of EUS relative to the T stage were 82% and 91%, respectively, for T1; 43% and 85% for T2; and 83% and 86% for T3. The overall accuracy for EUS in identifying the correct T stage was 74% (95%CI 66-80). Positive lymph nodes were diagnosed histologically in 68 patients (38%). The sensitivity, specificity and accuracy of EUS for the diagnosis of N1 were 71%, 74% and 73% (95%CI 65-79), respectively. CONCLUSIONS: The diagnostic accuracy of EUS in patients with esophageal cancer is still unsatisfactory. T2 cancers in particular are frequently overstaged, with a significant effect on the subsequent treatment strategy.