Deep learning assists detection of esophageal cancer and precursor lesions in a prospective, randomized controlled study.

Endoscopy is the primary modality for detecting asymptomatic esophageal squamous cell carcinoma (ESCC) and precancerous lesions. Improving detection rate remains challenging. We developed a system based on deep convolutional neural networks (CNNs) for detecting esophageal cancer and precancerous lesions [high-risk esophageal lesions (HrELs)] and validated its efficacy in improving HrEL detection rate in clinical practice (trial registration ChiCTR2100044126 at www.chictr.org.cn). Between April 2021 and March 2022, 3117 patients ≥50 years old were consecutively recruited from Taizhou Hospital, Zhejiang Province, and randomly assigned 1:1 to an experimental group (CNN-assisted endoscopy) or a control group (unassisted endoscopy) based on block randomization. The primary endpoint was the HrEL detection rate. In the intention-to-treat population, the HrEL detection rate [28 of 1556 (1.8%)] was significantly higher in the experimental group than in the control group [14 of 1561 (0.9%), P = 0.029], and the experimental group detection rate was twice that of the control group. Similar findings were observed between the experimental and control groups [28 of 1524 (1.9%) versus 13 of 1534 (0.9%), respectively; P = 0.021]. The system's sensitivity, specificity, and accuracy for detecting HrELs were 89.7, 98.5, and 98.2%, respectively. No adverse events occurred. The proposed system thus improved HrEL detection rate during endoscopy and was safe. Deep learning assistance may enhance early diagnosis and treatment of esophageal cancer and may become a useful tool for esophageal cancer screening.

Developing a Prognostic Model for Primary Biliary Cholangitis Based on a Random Survival Forest Model.

Background: Primary biliary cholangitis (PBC) is a rare autoimmune liver disease with few effective treatments and a poor prognosis, and its incidence is on the rise. There is an urgent need for more targeted treatment strategies to accurately identify high-risk patients. The use of stochastic survival forest models in machine learning is an innovative approach to constructing a prognostic model for PBC that can improve the prognosis by identifying high-risk patients for targeted treatment. Method: Based on the inclusion and exclusion criteria, the clinical data and follow-up data of patients diagnosed with PBC-associated cirrhosis between January 2011 and December 2021 at Taizhou Hospital of Zhejiang Province were retrospectively collected and analyzed. Data analyses and random survival forest model construction were based on the R language. Result: Through a Cox univariate regression analysis of 90 included samples and 46 variables, 17 variables with p-values <0.1 were selected for initial model construction. The out-of-bag (OOB) performance error was 0.2094, and K-fold cross-validation yielded an internal validation C-index of 0.8182. Through model selection, cholinesterase, bile acid, the white blood cell count, total bilirubin, and albumin were chosen for the final predictive model, with a final OOB performance error of 0.2002 and C-index of 0.7805. Using the final model, patients were stratified into high- and low-risk groups, which showed significant differences with a P value <0.0001. The area under the curve was used to evaluate the predictive ability for patients in the first, third, and fifth years, with respective results of 0.9595, 0.8898, and 0.9088. Conclusion: The present study constructed a prognostic model for PBC-associated cirrhosis patients using a random survival forest model, which accurately stratified patients into low- and high-risk groups. Treatment strategies can thus be more targeted, leading to improved outcomes for high-risk patients.

Combined use of extracorporeal membrane oxygenation with interventional surgery for acute pancreatitis with pulmonary embolism: A case report.

BACKGROUND: Acute pancreatitis (AP) is an acute inflammatory process of the pancreas characterized by self-digestion of pancreatic tissue, which can trigger a systemic inflammatory response. Venous thrombosis, resulting from a hypercoagulable state, is a vascular complication of AP. AP complicated by pulmonary embolism (PE) is very rare, and the combined use of extracorporeal membrane oxygenation (ECMO) with a vascular interventional procedure for AP complicated by PE is even rarer. CASE SUMMARY: A 32-year-old man with a history of obesity developed rapidly worsening AP secondary to hypertriglyceridemia. During treatment, the patient developed chest tightness, shortness of breath, and cardiac arrest. Computed tomography (CT) scans of his upper abdomen were consistent with pancreatitis. PE was identified by chest CT angiography involving the right main pulmonary artery and multiple lobar pulmonary arteries. The patient's D-dimer level was significantly elevated (> 20 mg/L). The patient received high-frequency oxygen inhalation, continuous renal replacement therapies, anti-infective therapy, inhibition of pancreatic secretion, emergent endotracheal intubation, and advanced cardiac life support with cardiopulmonary resuscitation. Following both ECMO and a vascular interventional procedure, the patient recovered and was discharged. CONCLUSION: PE is a rare but potentially lethal complication of AP. The early diagnosis of PE is important because an accurate diagnosis and timely interventional procedures can reduce mortality. The combined use of ECMO with a vascular interventional procedure for AP complicated by PE can be considered a feasible treatment method. A collaborative effort between multiple teams is also vital.

Cost-effective low-coverage whole-genome sequencing assay for the risk stratification of gastric cancer.

BACKGROUND: Gastric cancer (GC), a multifactorial disease, is caused by pathogens, such as Helicobacter pylori (H. pylori) and Epstein-Barr virus (EBV), and genetic components. AIM: To investigate microbiomes and host genome instability by cost-effective, low-coverage whole-genome sequencing, as biomarkers for GC subtyping. METHODS: Samples from 40 GC patients were collected from Taizhou Hospital, Zhejiang Province, affiliated with Wenzhou Medical University. DNA from the samples was subjected to low-coverage whole-genome sequencing with a median genome coverage of 1.86 × (range: 1.03 × to 3.17 ×) by Illumina × 10, followed by copy number analyses using a customized bioinformatics workflow ultrasensitive chromosomal aneuploidy detector. RESULTS: Of the 40 GC samples, 20 (50%) were found to be enriched with microbiomes. EBV DNA was detected in 5 GC patients (12.5%). H. pylori DNA was found in 15 (37.5%) patients. The other 20 (50%) patients were found to have relatively higher genomic instability. Copy number amplifications of the oncogenes, ERBB2 and KRAS, were found in 9 (22.5%) and 7 (17.5%) of the GC samples, respectively. EBV enrichment was found to be associated with tumors in the gastric cardia and fundus. H. pylori enrichment was found to be associated with tumors in the pylorus and antrum. Tumors with elevated genomic instability showed no localization and could be observed in any location. Additionally, H. pylori-enriched GC was found to be associated with the Borrmann type II/III and gastritis history. EBV-enriched GC was not associated with gastritis. No statistically significant correlation was observed between genomic instability and gastritis. Furthermore, these three different molecular subtypes showed distinct survival outcomes (P = 0.019). EBV-positive tumors had the best prognosis, whereas patients with high genomic instability (CIN+) showed the worst survival. Patients with H. pylori infection showed intermediate prognosis compared with the other two subtypes. CONCLUSION: Thus, using low-coverage whole-genome sequencing, GC can be classified into three categories based on disease etiology; this classification may prove useful for GC diagnosis and precision medicine.

Ninety-four thousand-case retrospective study on antibacterial drug resistance of Helicobacter pylori.

BACKGROUND: The resistance rate to antibacterial drugs is the key inhibitor of Helicobacter pylori (H. pylori) eradication treatment. AIM: To evaluate the prevalence and drug resistance of H. pylori based on big data. METHODS: Gastric mucosal specimens were collected from naive patients undergoing upper gastrointestinal endoscopy for H. pylori culture and antimicrobial susceptibility testing (AST), including clarithromycin, levofloxacin, metronidazole and amoxicillin. Every 10 years of age was grouped as an age group. The H. pylori infection and resistance were explored based on the age group and gender. RESULTS: The number of H. pylori-positive specimen was 94509 in 283823 gastric mucosal specimens, with an infection rate of 33.30%. The infection rate increased with age, and males had a higher infection rate than females. The average resistance rate of H. pylori to amoxicillin and metronidazole was 0.21% and 93.72%, which remained stable. The average resistance rate to clarithromycin was 23.99% with an increasing trend from 14.43% to 38.24%. The average resistance rate to levofloxacin was 30.29%, which increased from 17.07% to 39.42% and mostly stabilized after 2017. The resistance rate of H. pylori increased with age, except amoxicillin. H. pylori in females are at higher risk of resistance to metronidazole, but not to amoxicillin, regardless of the age group. Meanwhile, H. pylori in females are at higher risk of resistance to levofloxacin and clarithromycin in the 21-50 age group. The single, dual, triple and quadruple-drug resistance rate was 54.59%, 29.03%, 11.71% and 0.11%, respectively. CONCLUSION: The resistance of H. pylori in Taizhou city is serious. Guided by the consensus report, individualized treatment based on AST is recommended.

Endoscopic resection of large (≥ 4 cm) upper gastrointestinal subepithelial tumors originating from the muscularis propria layer: a single-center study of 101 cases (with video).

BACKGROUND AND AIMS: Although endoscopic resection (ER) is already established as a minimally invasive technique for small (< 4.0 cm) upper gastrointestinal subepithelial tumors originating from the muscularis propria layer (MP-SETs), published data of ER for large (≥ 4.0 cm) upper gastrointestinal MP-SETs are extremely rare and limited to case reports. This retrospective study aimed to evaluate the feasibility and safety of ER for large (≥ 4.0 cm) upper gastrointestinal MP-SETs in a large case series. METHODS: Between June 2012 and December 2018, 101 patients with large (≥ 4 cm) upper gastrointestinal MP-SETs were enrolled in this study. The main outcome measures included complete resection, total complications, and local residual or recurrent tumor. RESULTS: The rate of complete resection was 86.1%. Thirteen patients (12.9%) experienced complications including gas-related complications (6/101, 5.9%), localized peritonitis (4/101, 4.0%), esophageal/cardiac mucosal laceration (2/101, 2.0%), and delayed bleeding (1/101, 1.0%). These 13 patients recovered after endoscopic and conservative treatment. The independent risk factor for incomplete resection was tumor size (P = 0.005), and the independent risk factors for total complications were tumor size (P = 0.011) and tumor extraluminal growth (P = 0.037). During the median follow-up of 36 months, local residual tumor was detected in 1 patient. No local recurrence occurred in any patient. CONCLUSIONS: Despite being associated with a relatively low complete resection rate, ER is an alternative therapeutic method for large (≥ 4.0 cm) upper gastrointestinal MP-SETs when performed by an experienced endoscopist. This method is especially valuable for patients who are unwilling to undergo surgery.

Diagnostic significance of soluble human leukocyte antigen-G for gastric cancer.

Human leukocyte antigen-G (HLA-G) is a novel tumor marker. Increased level of soluble HLA-G (sHLA-G) in various tumor types has been reported. However, the potential diagnostic value of sHLA-G with other tumor markers in gastric cancer (GC) diagnosis is yet to be explored. In this study, plasma level of sHLA-G was measured in 81 GC patients, 53 benign gastric disease patients and 77 normal controls by ELISA. The serum levels of alpha fetoprotein (AFP), carcinoembryonic antigen (CEA), cancer antigen 125 (CA125), cancer antigen 19-9 (CA19-9) and cancer antigen 72-4 (CA72-4) were also determined. Data showed that plasma level of sHLA-G in GC was dramatically increased compared with normal controls and benign gastric disease patients (both p<0.001). The AUC for sHLA-G was 0.730 (p<0.001), superior to serum AFP, CEA, CA125, CA19-9 and CA72-4. After evaluating three cut-offs of sHLA-G, we concluded sHLA-G (cut-off at 128U/ml) plus CA125 in two-biomarker panel test and CA125 plus CA199 plus sHLA-G or CA125 plus CA724 plus sHLA-G in three-biomarker panel test were better choices for GC discrimination. Our findings indicated that sHLA-G was a potential biomarker for GC diagnosis and the combination of sHLA-G with CA125, CA19-9 and CA72-4 can improve the clinical screening and diagnosis for GC.