Prediction of immunochemotherapy response for diffuse large B-cell lymphoma using artificial intelligence digital pathology.

Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous and prevalent subtype of aggressive non-Hodgkin lymphoma that poses diagnostic and prognostic challenges, particularly in predicting drug responsiveness. In this study, we used digital pathology and deep learning to predict responses to immunochemotherapy in patients with DLBCL. We retrospectively collected 251 slide images from 216 DLBCL patients treated with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP), with their immunochemotherapy response labels. The digital pathology images were processed using contrastive learning for feature extraction. A multi-modal prediction model was developed by integrating clinical data and pathology image features. Knowledge distillation was employed to mitigate overfitting on gigapixel histopathology images to create a model that predicts responses based solely on pathology images. Based on the importance derived from the attention mechanism of the model, we extracted histological features that were considered key textures associated with drug responsiveness. The multi-modal prediction model achieved an impressive area under the ROC curve of 0.856, demonstrating significant associations with clinical variables such as Ann Arbor stage, International Prognostic Index, and bulky disease. Survival analyses indicated their effectiveness in predicting relapse-free survival. External validation using TCGA datasets supported the model's ability to predict survival differences. Additionally, pathology-based predictions show promise as independent prognostic indicators. Histopathological analysis identified centroblastic and immunoblastic features to be associated with treatment response, aligning with previous morphological classifications and highlighting the objectivity and reproducibility of artificial intelligence-based diagnosis. This study introduces a novel approach that combines digital pathology and clinical data to predict the response to immunochemotherapy in patients with DLBCL. This model shows great promise as a diagnostic and prognostic tool for clinical management of DLBCL. Further research and genomic data integration hold the potential to enhance its impact on clinical practice, ultimately improving patient outcomes.

Isolation and characterization of novel 3,3'-iminodipropionitrile biodegrading Paracoccus communis, from an industrial wastewater treatment bioreactor.

Until now, bacteria able to degrade, 3,3'-iminodipropionitrile (IDPN), a neurotoxin that destroys vestibular hair cells, causing ototoxicity, culminating in irreversible movement disorders, had never been isolated. The aim of this study was to isolate a novel IDPN-biodegrading microorganism and characterize its metabolic pathway. Enrichment was performed by inoculating activated sludge from a wastewater treatment bioreactor that treated IDPN-contaminated wastewater in M9 salt medium, with IDPN as the sole carbon source. A bacterial strain with a spherical morphology that could grow at high concentrations was isolated on a solid medium. Growth of the isolated strain followed the Monod kinetic model. Based on the 16S rRNA gene, the isolate was Paracoccus communis. Whole-genome sequencing revealed that the isolated P. communis possessed the expected full metabolic pathway for IDPN biodegradation. Transcriptome analyses confirmed the overexpression of the gene encoding hydantoinase/oxoprolinase during the exponential growth phase under IDPN-fed conditions, suggesting that the enzyme involved in cleaving the imine bond of IDPN may promote IDPN biodegradation. Additionally, the newly discovered P. communis isolate seems to metabolize IDPN through cleavage of the imine bond in IDPN via nitrilase, nitrile hydratase, and amidase reactions. Overall, this study lays the foundation for the application of IDPN-metabolizing bacteria in the remediation of IDPN-contaminated environments.

Natural history of gastric leiomyoma.

BACKGROUND: Most gastric leiomyomas are asymptomatic and benign subepithelial tumors (SETs); however, some may increase in size or become symptomatic. Understanding their natural history is therefore important to their management. We investigated the natural history of histologically proven gastric leiomyomas. METHODS: We retrospectively reviewed histologically proven gastric leiomyoma cases at a tertiary center. The baseline characteristics of these cases were analyzed, and those with a follow-up period of at least 12 months without immediate resection were evaluated. The primary outcome was the frequency of size increase of more than 25% during the follow-up period, and the secondary outcome was the histopathologic results in cases that underwent resection. RESULTS: Among the 231 patients with histologically proven gastric leiomyomas, the most frequent location was the cardia (77.1%), and the median size was 3 cm (IQR 2-4 cm). Eighty-four cases were followed up over a median period of 50.8 months (IQR 27.2-91.3 months). During the follow-up period, tumor size increased in two cases (2.4%). Surgical results showed that one case was leiomyoma, and the other was leiomyosarcoma. Among the remaining cases without change in size, 15 underwent surgical resection (n = 10) or endoscopic resection (n = 5), and all cases were confirmed as leiomyoma. CONCLUSIONS: Most gastric leiomyomas are benign SETs, and an increase in size is not frequent, even in large-sized cases. Close monitoring with routine follow-up without resection may be sufficient in cases of histologically proven gastric leiomyoma. However, in cases of ulceration or size increase, resection may be beneficial.

Advancing Korean nationwide registry for hepatocellular carcinoma: a systematic sampling approach utilizing the Korea Central Cancer Registry database.

Hepatocellular carcinoma (HCC) presents a substantial public health challenge in South Korea as evidenced by 10,565 new cases annually (incidence rate of 30 per 100,000 individuals), in 2020. Cancer registries play a crucial role in gathering data on incidence, disease attributes, etiology, treatment modalities, outcomes, and informing health policies. The effectiveness of a registry depends on the completeness and accuracy of data. Established in 1999 by the Ministry of Health and Welfare, the Korea Central Cancer Registry (KCCR) is a comprehensive, legally mandated, nationwide registry that captures nearly all incidence and survival data for major cancers, including HCC, in Korea. However, detailed information on cancer staging, specific characteristics, and treatments is lacking. To address this gap, the KCCR, in partnership with the Korean Liver Cancer Association (KLCA), has implemented a systematic approach to collect detailed data on HCC since 2010. This involved random sampling of 10-15% of all new HCC cases diagnosed since 2003. The registry process encompassed four stages: random case selection, meticulous data extraction by trained personnel, expert validation, anonymization of personal data, and data dissemination for research purposes. This random sampling strategy mitigates the biases associated with voluntary reporting and aligns with stringent privacy regulations. This innovative approach positions the KCCR and KLCA as foundations for advancing cancer control and shaping health policies in South Korea.

Inhibition of sodium-glucose cotransporter-2 and liver-related complications in individuals with diabetes: a Mendelian randomization and population-based cohort study.

BACKGROUND AND AIMS: No medication has been found to reduce liver-related events. We evaluated the effect of sodium-glucose cotransporter-2 inhibitor (SGLT2i) on liver-related outcomes. APPROACH AND RESULTS: Single nucleotide polymorphisms associated with SGLT2 inhibition were identified, and a genetic risk score (GRS) was computed using the UK Biobank data (n=337,138). Two-sample Mendelian randomization (MR) was conducted using the FinnGen (n=218,792) database and the UK Biobank data. In parallel, a nationwide population-based study using the Korean National Health Insurance Service (NHIS) database was conducted. The development of liver-related complications (ie, hepatic decompensation, HCC, liver transplantation, and death) was compared between individuals with type 2 diabetes mellitus and steatotic liver diseases treated with SGLT2i (n=13,208) and propensity score-matched individuals treated with dipeptidyl peptidase-4 inhibitor (n=70,342). After computing GRS with 6 single nucleotide polymorphisms (rs4488457, rs80577326, rs11865835, rs9930811, rs34497199, and rs35445454), GRS-based MR showed that SGLT2 inhibition (per 1 SD increase of GRS, 0.1% lowering of HbA1c) was negatively associated with cirrhosis development (adjusted odds ratio=0.83, 95% CI=0.70-0.98, p =0.03) and this was consistent in the 2-sample MR (OR=0.73, 95% CI=0.60-0.90, p =0.003). In the Korean NHIS database, the risk of liver-related complications was significantly lower in the SGLT2i group than in the dipeptidyl peptidase-4 inhibitor group (adjusted hazard ratio=0.88, 95% CI=0.79-0.97, p =0.01), and this difference remained significant (adjusted hazard ratio=0.72-0.89, all p <0.05) across various sensitivity analyses. CONCLUSIONS: Both MRs using 2 European cohorts and a Korean nationwide population-based cohort study suggest that SGLT2 inhibition is associated with a lower risk of liver-related events.

Bioengineered amyloid peptide for rapid screening of inhibitors against main protease of SARS-CoV-2.

The coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) has evoked a worldwide pandemic. As the emergence of variants has hampered the neutralization capacity of currently available vaccines, developing effective antiviral therapeutics against SARS-CoV-2 and its variants becomes a significant challenge. The main protease (Mpro) of SARS-CoV-2 has received increased attention as an attractive pharmaceutical target because of its pivotal role in viral replication and proliferation. Here, we generated a de novo Mpro-inhibitor screening platform to evaluate the efficacies of Mpro inhibitors based on Mpro cleavage site-embedded amyloid peptide (MCAP)-coated gold nanoparticles (MCAP-AuNPs). We fabricated MCAPs comprising an amyloid-forming sequence and Mpro-cleavage sequence, mimicking in vivo viral replication process mediated by Mpro. By measuring the proteolytic activity of Mpro and the inhibitory efficacies of various drugs, we confirmed that the MCAP-AuNP-based platform was suitable for rapid screening potential of Mpro inhibitors. These results demonstrated that our MCAP-AuNP-based platform has great potential for discovering Mpro inhibitors and may accelerate the development of therapeutics against COVID-19.

High-dose proton pump inhibitor treatment is associated with a higher mortality in cirrhotic patients: A multicentre study.

BACKGROUND: Proton pump inhibitors (PPI) are frequently used in patients with cirrhosis. AIMS: This study aimed to determine whether PPI use is associated with the prognosis of cirrhotic patients. METHODS: We conducted a multicentre retrospective cohort study involving 1485 patients who had experienced hepatic encephalopathy (HE) from 7 referral centres in Korea. The primary outcome was overall survival and secondary outcomes included the development of cirrhotic complications, including recurrent HE, spontaneous bacterial peritonitis (SBP), hepatorenal syndrome (HRS), and gastrointestinal bleeding. Patients treated with PPI with a mean defined daily dose (mDDD) ≥0.5 (high-dose PPI group) were compared to those treated with PPI of an mDDD < 0.5 (No or low-dose PPI group) for each outcome. RESULTS: Among 1485 patients (median age, 61 years; male, 61%), 232 were assigned to the high-dose PPI group. High-dose PPI use was independently associated with a higher risk of death (adjusted HR [aHR] = 1.71, 95% confidence interval [CI] = 1.38-2.11, p < 0.001). This result was reproducible after propensity score-matching (PSM) (aHR = 1.90, 95% CI = 1.49-2.44, p < 0.001). High-dose PPI use was an independent risk factor of recurrent HE (before PSM: aHR = 2.04, 95% CI = 1.66-2.51, p < 0.001; after PSM: aHR = 2.16, 95% CI = 1.70-2.74, p < 0.001), SBP (before PSM: aHR = 1.87, 95% CI = 1.43-2.43, p < 0.001; after PSM: aHR = 1.76, 95% CI = 1.31-2.36, p = 0.002), HRS (before PSM: aHR = 1.48, 95% CI = 1.02-2.15, p = 0.04; after PSM: aHR = 1.47, 95% CI = 0.95-2.28, p = 0.09), and gastrointestinal bleeding (before PSM: aHR = 1.46, 95% CI = 1.12-1.90, p = 0.006; after PSM: aHR = 1.74, 95% CI = 1.28-2.37, p < 0.001). CONCLUSIONS: The use of high-dose PPI was independently associated with increased risks of mortality and cirrhotic complications.

Asian Pacific Association for the Study of the Liver clinical practice guidelines on liver transplantation.

Liver transplantation is a highly complex and challenging field of clinical practice. Although it was originally developed in western countries, it has been further advanced in Asian countries through the use of living donor liver transplantation. This method of transplantation is the only available option in many countries in the Asia-Pacific region due to the lack of deceased organ donation. As a result of this clinical situation, there is a growing need for guidelines that are specific to the Asia-Pacific region. These guidelines provide comprehensive recommendations for evidence-based management throughout the entire process of liver transplantation, covering both deceased and living donor liver transplantation. In addition, the development of these guidelines has been a collaborative effort between medical professionals from various countries in the region. This has allowed for the inclusion of diverse perspectives and experiences, leading to a more comprehensive and effective set of guidelines.

Endoscopic resection of gastric gastrointestinal stromal tumor using clip-and-cut endoscopic full-thickness resection: a single-center, retrospective cohort in Korea.

Background/Aims: To overcome the technical limitations of classic endoscopic resection for gastric gastrointestinal stromal tumors (GISTs), various methods have been developed. In this study, we examined the role and feasibility of clip-and-cut procedures (clip-and-cut endoscopic full-thickness resection [cc-EFTR]) for gastric GISTs. Methods: Medical records of 83 patients diagnosed with GISTs after endoscopic resection between 2005 and 2021 were retrospectively reviewed. Moreover, clinical characteristics and outcomes were analyzed. Results: Endoscopic submucosal dissection (ESD) and cc-EFTR were performed in 51 and 32 patients, respectively. The GISTs were detected in the upper third of the stomach for ESD (52.9%) and cc-EFTR (90.6%). Within the cc-EFTR group, a majority of GISTs were located in the deep muscularis propria or serosal layer, accounting for 96.9%, as opposed to those in the ESD group (45.1%). The R0 resection rates were 51.0% and 84.4% in the ESD and cc-EFTR groups, respectively. Seven (8.4%) patients required surgical treatment (six patients underwent ESD and one underwent cc-EFTR,) due to residual tumor (n=5) and post-procedure adverse events (n=2). Patients undergoing R0 or R1 resection experienced recurrence during a median 14-month follow-up period, except for one patient in the ESD group. Conclusions: Cc-EFTR displayed high R0 resection rates as the procedure safely and successfully removed small gastric GISTs.

Evaluating the Diagnostic Potential of Connected Speech for Benign Laryngeal Disease Using Deep Learning Analysis.

OBJECTIVES: This study aimed to evaluate the performance of artificial intelligence (AI) models using connected speech and vowel sounds in detecting benign laryngeal diseases. STUDY DESIGN: Retrospective. METHODS: Voice samples from 772 patients, including 502 with normal voices and 270 with vocal cord polyps, cysts, or nodules, were analyzed. We employed deep learning architectures, including convolutional neural networks (CNNs) and time series models, to process the speech data. The primary endpoint was the area under the receiver's operating characteristic curve for binary classification. RESULTS: CNN models analyzing speech segments significantly outperformed those using vowel sounds in distinguishing patients with and without benign laryngeal diseases. The best-performing CNN model achieved areas under the receiver operating characteristic curve of 0.895 and 0.845 for speech and vowel sounds, respectively. Correlations between AI-generated disease probabilities and perceptual assessments were more pronounced in the connected-speech analyses. However, the time series models performed worse than the CNNs. CONCLUSION: Connected speech analysis is more effective than traditional vowel sound analysis for the diagnosis of laryngeal voice disorders. This study highlights the potential of AI technologies in enhancing the diagnostic capabilities of speech, advocating further exploration, and validation in this field.

Rapid deep learning-assisted predictive diagnostics for point-of-care testing.

Prominent techniques such as real-time polymerase chain reaction (RT-PCR), enzyme-linked immunosorbent assay (ELISA), and rapid kits are currently being explored to both enhance sensitivity and reduce assay time for diagnostic tests. Existing commercial molecular methods typically take several hours, while immunoassays can range from several hours to tens of minutes. Rapid diagnostics are crucial in Point-of-Care Testing (POCT). We propose an approach that integrates a time-series deep learning architecture and AI-based verification, for the enhanced result analysis of lateral flow assays. This approach is applicable to both infectious diseases and non-infectious biomarkers. In blind tests using clinical samples, our method achieved diagnostic times as short as 2 minutes, exceeding the accuracy of human analysis at 15 minutes. Furthermore, our technique significantly reduces assay time to just 1-2 minutes in the POCT setting. This advancement has the potential to greatly enhance POCT diagnostics, enabling both healthcare professionals and non-experts to make rapid, accurate decisions.

Enhancing Loop-Mediated Isothermal Amplification through Nonpowered Nanoelectric Preconcentration.

The global threat posed by the COVID-19 pandemic has catalyzed the development of point-of-care (POC) molecular diagnostics. While loop-mediated isothermal amplification (LAMP) stands out as a promising technique among FDA-approved methods, it is occasionally susceptible to a high risk of false positives due to nonspecific amplification of a primer dimer. In this work, we report an enhancing LAMP technique in terms of assay sensitivity and reliability through streamlined integration with a nonpowered nanoelectric preconcentration (NPP). The NPP, serving as a sample preparation tool, enriched the virus concentration in samples prior to the subsequent LAMP assay. This enrichment enabled not only to achieve more sensitive assay but also to shorten the assay time for all tested clinical samples by ∼10 min compared to the conventional LAMP. The shortened assay time suppresses the occurrence of nonspecific amplification by not providing the necessary incubation time, effectively suppressing misidentification by false positives. Utilizing this technique, we also developed a prototype of the POC NPP-LAMP kit. This kit offers a streamlined diagnostic process for nontrained individuals, from the sample enrichment, transfer of the enriched sample to LAMP assays, which facilitates on-site/on-demand diagnosis of SARS-CoV-2. This development holds the potential to contribute toward preventing not only the current outbreak but also future occurrences of pandemic viruses.

Risk of hepatocellular carcinoma after curative treatment when switching from tenofovir disoproxil fumarate or entecavir to tenofovir alafenamide: A real-world multicenter cohort study.

AIM: Antiviral treatment reduces the risk of developing hepatocellular carcinoma (HCC) in patients with chronic hepatitis B. However, there is a lack of high-quality evidence regarding the preventive effects of tenofovir alafenamide (TAF) on HCC. We evaluated the impact of TAF use after curative treatment on HCC recurrence. METHODS: Patients who underwent surgery or radiofrequency ablation as a curative treatment for HCC were selected. Those patients who continued antiviral treatment with nucleos(t)ide analogs (NAs; entecavir [ETV] or tenofovir disoproxil fumarate [TDF]) or switched to TAF were included. The primary outcome was HCC recurrence, and the time-varying effect of NA use on HCC recurrence was analyzed using various statistical methods. RESULTS: Among 2794 consecutive patients with chronic hepatitis B who received curative treatment for HCC, 199 subsequently switched from ETV or TDF to TAF. After a median of 3.0 years, 1303 patients (46.6%) experienced HCC recurrence. After propensity score matching (ratio 1:10), switching to TAF was not associated with an increased HCC recurrence (HR 1.00, 95% CI 0.68-1.47; p = 1.00) by time-varying Cox analysis. Switching to TAF was not associated with HCC recurrence in subgroups of NA (HR 1.06, 95% CI 0.67-1.67; p = 0.81 for TDF, and HR 1.09, 95% CI 0.51-2.33; p = 0.82 for ETV). Kaplan-Meier analysis showed comparable HCC recurrence-free survival between patients who switched to TAF and those who continued with their NA (p = 0.08). Time-varying Cox analyses in various subgroups confirmed the primary findings. CONCLUSIONS: TAF is as effective as TDF and ETV in preventing HCC recurrence after curative treatment.

Extrahepatic malignancies in metabolic dysfunction-associated fatty liver disease: A nationwide cohort study.

BACKGROUND AND AIMS: Metabolic dysfunction-associated fatty liver disease (MAFLD) encompasses heterogeneous fatty liver diseases associated with metabolic disorders. We aimed to evaluate the association between MAFLD and extrahepatic malignancies based on MAFLD subtypes. METHODS: This nationwide cohort study included 9 298 497 patients who participated in a health-screening programme of the National Health Insurance Service of Korea in 2009. Patients were further classified into four subgroups: non-MAFLD, diabetes mellitus (DM)-MAFLD, overweight/obese-MAFLD and lean-MAFLD. The primary outcome was the development of any primary extrahepatic malignancy, while death, decompensated liver cirrhosis and liver transplantation were considered competing events. The secondary outcomes included all-cause and extrahepatic malignancy-related mortality. RESULTS: In total, 2 500 080 patients were diagnosed with MAFLD. During a median follow-up of 10.3 years, 447 880 patients (6.0%) with extrahepatic malignancies were identified. The DM-MAFLD (adjusted subdistribution hazard ratio [aSHR] = 1.13; 95% confidence interval [CI] = 1.11-1.14; p < .001) and the lean-MAFLD (aSHR = 1.12; 95% CI = 1.10-1.14; p < .001) groups were associated with higher risks of extrahepatic malignancy than the non-MAFLD group. However, the overweight/obese-MAFLD group exhibited a similar risk of extrahepatic malignancy compared to the non-MAFLD group (aSHR = 1.00; 95% CI = .99-1.00; p = .42). These findings were reproduced in several sensitivity analyses. The DM-MAFLD was an independent risk factor for all-cause mortality (adjusted hazard ratio [aHR] = 1.41; 95% CI = 1.40-1.43; p < .001) and extrahepatic malignancy-related mortality (aHR = 1.20; 95% CI = 1.17-1.23; p < .001). CONCLUSION: The diabetic or lean subtype of MAFLD was associated with a higher risk of extrahepatic malignancy than non-MAFLD. As MAFLD comprises a heterogeneous population, appropriate risk stratification and management based on the MAFLD subtypes are required.

Comparison of atezolizumab plus bevacizumab and lenvatinib for hepatocellular carcinoma with portal vein tumor thrombosis.

BACKGROUND/AIM: Atezolizumab plus bevacizumab and lenvatinib are currently available as first-line therapy for the treatment of unresectable hepatocellular carcinoma (HCC). However, comparative efficacy studies are still limited. This study aimed to investigate the effectiveness of these treatments in HCC patients with portal vein tumor thrombosis (PVTT). METHODS: We retrospectively included patients who received either atezolizumab plus bevacizumab or lenvatinib as first-line systemic therapy for HCC with PVTT. Primary endpoint was overall survival (OS), and secondary endpoints included progressionfree survival (PFS) and disease control rate (DCR) determined by response evaluation criteria in solid tumors, version 1.1. RESULTS: A total of 52 patients were included: 30 received atezolizumab plus bevacizumab and 22 received lenvatinib. The median follow-up duration was 6.4 months (interquartile range, 3.9-9.8). The median OS was 10.8 months (95% confidence interval [CI], 5.7 to not estimated) with atezolizumab plus bevacizumab and 5.8 months (95% CI, 4.8 to not estimated) with lenvatinib (P=0.26 by log-rank test). There was no statistically significant difference in OS (adjusted hazard ratio [aHR], 0.71; 95% CI, 0.34-1.49; P=0.37). The median PFS was similar (P=0.63 by log-rank test), with 4.1 months (95% CI, 3.3-7.7) for atezolizumab plus bevacizumab and 4.3 months (95% CI, 2.6-5.8) for lenvatinib (aHR, 0.93; 95% CI, 0.51-1.69; P=0.80). HRs were similar after inverse probability treatment weighting. The DCRs were 23.3% and 18.2% in patients receiving atezolizumab plus bevacizumab and lenvatinib, respectively (P=0.74). CONCLUSION: The effectiveness of atezolizumab plus bevacizumab and lenvatinib was comparable for the treatment of HCC with PVTT.

Dysphagia associated with esophageal wall thickening in patients with nonspecific high-resolution manometry findings: Understanding motility beyond the Chicago classification version 4.0.

BACKGROUND: Previous studies have demonstrated that 50% of patients with normal high-resolution manometry (HRM) findings or ineffective esophageal motility (IEM) may have abnormal functional luminal imaging probe (FLIP) results. However, the specific HRM findings associated with abnormal FLIP results are unknown. Herein, we investigated the relationship between nonspecific manometry findings and abnormal FLIP results. METHODS: We retrospectively analyzed 684 patients who underwent HRM at a tertiary care center in Seoul, Korea, based on the Chicago Classification version 4.0 protocol. KEY RESULTS: Among the 684 patients, 398 had normal HRM findings or IEM. Of these 398 patients, eight showed esophageal wall thickening on endoscopic ultrasonography or computed tomography; however, no abnormalities were seen during esophagogastroduodenoscopy. Among these eight patients, seven showed repetitive simultaneous contractions (RSCs) in at least one of the two positions: 61% (±29%) in 10 swallows in the supine position and 51% (±30%) in five swallows in the upright position. Four patients who underwent FLIP had a significantly decreased esophagogastric junction distensibility index (1.0 ± 0.5 m m 2 mmHg - 1 at 60 mL). Two of these patients underwent per-oral endoscopic myotomy (POEM) due to a lack of response to medication. Esophageal muscle biopsy revealed hypertrophic muscle with marginal eosinophil infiltration. CONCLUSIONS & INFERENCES: A subset of patients (2%) with normal HRM findings or IEM and RSCs experienced dysphagia associated with poor distensibility of the thickened esophageal wall. FLIP assessment or combined HRM and impedance protocols may help better define these patients who may respond well to POEM.

Ensemble Deep Learning Model to Predict Lymphovascular Invasion in Gastric Cancer.

Lymphovascular invasion (LVI) is one of the most important prognostic factors in gastric cancer as it indicates a higher likelihood of lymph node metastasis and poorer overall outcome for the patient. Despite its importance, the detection of LVI(+) in histopathology specimens of gastric cancer can be a challenging task for pathologists as invasion can be subtle and difficult to discern. Herein, we propose a deep learning-based LVI(+) detection method using H&E-stained whole-slide images. The ConViT model showed the best performance in terms of both AUROC and AURPC among the classification models (AUROC: 0.9796; AUPRC: 0.9648). The AUROC and AUPRC of YOLOX computed based on the augmented patch-level confidence score were slightly lower (AUROC: -0.0094; AUPRC: -0.0225) than those of the ConViT classification model. With weighted averaging of the patch-level confidence scores, the ensemble model exhibited the best AUROC, AUPRC, and F1 scores of 0.9880, 0.9769, and 0.9280, respectively. The proposed model is expected to contribute to precision medicine by potentially saving examination-related time and labor and reducing disagreements among pathologists.

Clinical outcomes of argon plasma coagulation for the treatment of gastric low grade dysplasia.

BACKGROUND & AIMS: Argon plasma coagulation (APC) could be considered a treatment modality for small gastric low grade dysplasia (LGD) instead of endoscopic resection (ER). Our study investigated the clinical outcomes of APC for treating gastric LGD and associated variables with local recurrence. METHODS: This study included 911 patients who underwent APC for gastric neoplasms at the tertiary hospital from July 2007 to March 2022 with a minimal follow-up of 12 months. 112 subjects without any information about H. pylori infection status, 164 subjects who underwent APC for salvage therapy, 5 subjects with high grade dysplasia, and 12 subjects with cancer were excluded. Through a retrospective review of medical data, the clinical outcomes and variables associated with the local recurrence were analyzed. RESULTS: A total of 618 patients with LGD (median age of 64 years old) were followed up for a median of 30 months and local recurrence has happened in 21 patients (3.4%). Multivariate analysis showed lesion size (hazard ratio 1.06, 95% confidential interval 1.01-1.12) was associated with the local recurrence. Among 557 lesions smaller than 10 mm, local recurrence was found in 14 cases (2.6%) and local recurrence was in 7 cases (9.5%) of 109 tumors larger than 10 mm (p <0.004). CONCLUSIONS: In gastric LGD smaller than 10 mm without scars, APC is a good treatment modality in place of ER. However, when a lesion is larger, APC should be selected carefully with close monitoring.

Reappraisal of Portal Vein Tumor Thrombosis as a Prognostic Factor for Patients with Hepatocellular Carcinoma.

Background/Aims: : This study aimed to assess whether hepatocellular carcinoma (HCC) patients with portal vein tumor thrombosis (PVTT) could have favorable prognoses with proper treatment under selective conditions. Methods: : This retrospective, single-center study involved 1,168 patients diagnosed with HCC between January 2005 and December 2006, before the introduction of sorafenib. Overall survival (OS) was estimated using the Kaplan-Meier method, and the Cox proportional hazards model was used to identify and adjust the variables associated with OS. Results: : In nodular-type HCC, the OS differed significantly according to the presence of PVTT (log-rank p<0.001), and the level of PVTT, not only its presence, was a major independent factor affecting OS. PVTT at the Vp1-3 branch was associated with significantly longer OS than was PVTT at the Vp4 level (hazard ratio [HR], 1.82; 95% confidence interval [CI], 1.04 to 3.21). In multivariate analysis, the OS was further stratified according to the PVTT level and tumor type, representing that nodular HCC without PVTT exhibited the best OS, whereas nodular HCC with Vp4 PVTT (adjusted HR, 2.59; 95% CI, 1.57 to 4.28) showed a poor prognosis similar to that of infiltrative HCC. The PVTT level was consistently correlated with OS in patients treated with transarterial chemoembolization. Nodular HCC without PVTT showed the best prognosis, while nodular HCC with Vp1-3 PVTT also exhibited a favorable OS, although inferior to that without PVTT (adjusted HR, 1.47, 95% CI, 0.92 to 2.36). Conclusions: : Active treatment such as transarterial chemoembolization can be considered for selected PVTT cases. The level of PVTT and type of HCC were independent prognostic factors.