Optimal second-line (2L) and third-line (3L) systemic therapies after first-line (1L) atezolizumab plus bevacizumab (AtezoBev) for advanced hepatocellular carcinoma (HCC) remain undefined. This prospective multicenter study evaluated the efficacy and safety of various 2L and 3L systemic therapies following AtezoBev. We enrolled 94 patients with advanced HCC who were intolerant of or progressed to 1L AtezoBev treatment from four hospitals in Korea. Five patients who did not receive 2L therapy were excluded, and the remaining 89 patients were analyzed. Survival outcomes were analyzed using Kaplan-Meier survival curves and Cox proportional hazards models. Among the 89 analyzed patients, median progression-free survival (PFS) was significantly longer in those who received 2L-lenvatinib compared to 2L-sorafenib (5.5 months vs 3.2 months: hazard ratio [HR], 0.43; 95% confidence interval [CI], 0.24 to 0.77; p=0.005) or 2L-regorafenib (4.7 months), while median overall survival (OS) did not differ significantly. Among 48 patients receiving 3L therapy, regorafenib showed a trend toward improved OS over sorafenib (HR, 0.33; 95% CI, 0.11 to 1.02; p=0.051), supported by restricted mean survival time analysis (p=0.028). Adverse events differed across therapies, with skin reactions and proteinuria notably linked to regorafenib and lenvatinib. Sequential 2L-lenvatinib followed by 3L-sorafenib yielded significantly better PFS than 2L-sorafenib followed by 3L-regorafenib (5.49 months vs 2.99 months: HR, 0.26; 95% CI, 0.10 to 0.65; p=0.003). Longer AtezoBev exposure prior to 2L treatment was associated with improved OS (HR, 1.05; 95% CI, 1.00 to 1.11; p=0.041). In advanced HCC following 1L AtezoBev failure, 2L lenvatinib prolonged PFS, and 3L regorafenib improved OS. These findings underscore the need to optimize sequential systemic therapies to enhance survival.

Severe alcoholic hepatitis (SAH) is a life-threatening form of alcoholic liver disease resulting in high short-term mortality. Mesenchymal stem cells (MSCs) have potent immunomodulatory effects and have been evaluated in various clinical trials for the treatment of chronic liver diseases. However, clinical evidence in patients with alcoholic hepatitis remains scarce, and the underlying mechanisms of MSCs in this population are not yet fully understood. An integrative meta-analysis identified conserved transcriptomic signatures of alcoholic hepatitis. These signatures were validated in an ethanol-induced murine model. A mouse model of SAH was induced via subacute ethanol exposure (5 g/kg) combined with thioacetamide injection. MSCs were administered at two concentrations (5×105 or 1×106 cells), depending on the treatment group. In the animal model, MSCs treatment visibly alleviated liver injury induced by thioacetamide and ethanol. Significant reductions in tumor necrosis factor-α (p<0.05) and α-smooth muscle actin (p<0.01) levels were observed, accompanied by notable changes in inducible nitric oxide synthase, interleukin-1β, and transforming growth factor-β1 levels. From the meta-analysis, seven upregulated and 17 downregulated genes were identified. Subsequent quantitative polymerase chain reaction and Western blot analyses consistently validated four upregulated genes that demonstrated overlapping expression patterns across both the meta-analysis and in vivo experiments. MSCs therapy significantly attenuates liver injury, inflammation, and fibrosis in SAH model mice. The observed messenger RNA-protein expression mismatches highlight the complexity of molecular regulation in acute hepatitis and underscore the importance of multilevel analysis in evaluating stem cell therapy. These results provide valuable insights into the mechanisms of MSC-mediated liver repair and suggest key targets for MSC therapy and response assessment in SAH.

Malignant gastric outlet obstruction (MGOO) occurs in 2% to 25% of patients with pancreatic ductal adenocarcinoma (PDAC) who do not undergo surgical intervention. Over the past decade, duodenal self-expandable metal stent (D-SEMS) has been widely used for MGOO and has demonstrated high technical (89.1% to 100%) and clinical (85.7% to 94.3%) success rates. Endoscopic ultrasonography-guided gastroenterostomy (EUS-GE) has emerged as a promising alternative with comparable technical (89.3% to 98.9%) and clinical (89.0% to 100%) success rates. Notably, EUS-GE reduces the 6-month reintervention rate by approximately 25% relative to D-SEMS in patients with MGOO, thus making it an increasingly popular treatment option. Despite its advantages, EUS-GE may not be suitable for patients with massive ascites, extensive peritoneal carcinomatosis, gastric linitis plastica, or an inaccessible small bowel. Moreover, EUS-GE requires a longer procedure time and incurs higher overall costs, even when considering the increased reintervention rate associated with D-SEMS. While the risk of adverse events is similar between EUS-GE and D-SEMS, EUS-GE requires a higher level of operator expertise to ensure safety and may also require deeper levels of anesthesia than D-SEMS. Advancements in cancer therapy have prolonged survival in PDAC patients, but most do not require reintervention for MGOO, unlike those with other malignancies. Thus, EUS-GE may be suitable for patients with a longer life span who are willing to undergo deep anesthesia at a high-volume center with an experienced specialist. Given these considerations, more comprehensive studies are needed before EUS-GE can be recommended as a standard replacement for D-SEMS in the treatment of MGOO.

Data on outcomes of superficial nonampullary duodenal tumors (SNADETs) are limited. We evaluated clinical features, outcomes, and complications of endoscopic resection (ER) in a large single-center cohort. Patients who underwent ER for SNADETs between 2008 and 2023 were retrospectively reviewed. Clinicopathologic features, resection outcomes, adverse events, and recurrence were analyzed. A total of 176 SNADETs from 163 patients (mean age, 57±13 years; 56.4% male) were included. The median follow-up duration was 22 months. The mean lesion size was 9.8±5.4 mm. ER methods were cold snare polypectomy (CSP) (n=29), cold endoscopic mucosal resection (EMR) (n=42), hot EMR (n=83), endoscopic submucosal dissection (ESD) (n=13), and underwater EMR (n=8). En bloc and histologic curative resection rates were 85.7% and 91.5%, respectively. Bleeding occurred in nine patients (5.1%), and perforation occurred in three patients (1.7%); both were mostly managed endoscopically. Bleeding was more common in the second portion of duodenum, and with hot EMR and ESD, whereas no bleeding occurred after CSP or underwater EMR. CSP and underwater EMR showed the shortest procedure times, whereas cold EMR and hot EMR had longer procedure times; ESD had the longest procedure time. Local recurrence occurred after en bloc curative ESD in one patient (0.6%). ER is effective and safe for SNADETs, achieving high curative resection rates, low recurrence, and acceptable complication risks. Selection of resection method should be individualized, with cold EMR and underwater EMR offering safe and efficient alternatives for the treatment of smaller lesions, while ESD may be reserved for the treatment of larger lesions or complex cases.

Background/Aims: Current guidelines recommend biannual ultrasound for hepatocellular carcinoma (HCC) surveillance in chronic hepatitis B (CHB) patients. However, computed tomography (CT) or magnetic resonance imaging (MRI) may be used when ultrasound is inadequate. The clinical impact of these alternative modalities remains unclear. Methods: CHB patients undergoing regular HCC surveillance were classified into two groups: ultrasound-only and alternative surveillance (CT/MRI). Patients were stratified into high- and low-risk groups using the Risk Estimation for HCC in CHB (REACH-B) score. Outcomes included 10-year overall survival (OS), HCC tumor size, Barcelona Clinic Liver Cancer (BCLC) stage at diagnosis, and OS after HCC diagnosis. Propensity score matching was applied to balance baseline characteristics. Results: A total of 2,024 patients were included after propensity score matching to ensure balanced baseline characteristics, with 1,012 patients in each group. OS was similar (ultrasound-only 96.0% vs alternative 96.8%; p=0.379). HCC occurred in 66 patients in each group. Alternative surveillance detected smaller HCC tumors (1.6 cm vs 2.1 cm; p<0.001) and more BCLC stage 0 cases (alternative 71.2% vs ultrasound-only 42.4%; p=0.003). The OS after HCC diagnosis was higher with alternative surveillance (83.0% vs 67.0%; p=0.025). In high-risk patients (n=970), alternative surveillance increased the OS (97.3% vs 93.6%; p=0.029) and the OS after HCC diagnosis (83.0% vs 60.6%; p=0.010). No significant differences were observed in low-risk patients. Conclusions: CT/MRI-based alternative surveillance led to earlier HCC detection and improved post-diagnosis survival, particularly in high-risk CHB patients, supporting its potential role as an alternative to ultrasound in selected populations.

Background/AimsThis study aimed to evaluate the effectiveness of fibrin glue application in reducing bleeding after endoscopic submucosal dissection (ESD) in the general patient population.MethodsThis randomized controlled trial enrolled 262 patients who underwent ESD for gastric neoplasms. The participants were randomized into the fibrin glue group (n=133) or the control group (n=129). The primary outcome was the post-ESD bleeding rate, which was compared between the two groups. Additionally, exploratory subgroup analyses were conducted for high-risk patients, which included patients taking antithrombotic agents (ATAs) and patients with resected specimens measuring 4 cm or greater.ResultsIn the modified intention-to-treat population (n=252), the overall bleeding rate occurred in 10.7% of cases, including 10.3% of patients in the fibrin glue group and 11.1% of patients in the control group (p=0.839). In the per-protocol population (n=248), the bleeding rates were 10.3% in the fibrin glue group and 10.7% in the control group (p=1.000). Among ATA users, the overall bleeding rate was 18.0%, including 9.5% of patients in the fibrin glue group and 24.1% of patients in the control group (p=0.271). Among ATA users with specimens measuring 4 cm or larger, the overall bleeding rate was 18.4%, including 12.5% of patients in the fibrin glue group and 22.7% of patients in the control group (p=0.675).ConclusionsRoutine application of fibrin glue after ESD did not significantly reduce postoperative bleeding. Although the results of subgroup analyses suggested a potential reduction in early bleeding among high-risk patients, these findings warrant further investigation.

Background/AimsThe efficacy of proton pump inhibitors (PPIs) and potassium-competitive acid blockers (P-CABs) in the treatment of eosinophilic esophagitis (EoE) has been well established. This study aimed to clarify the impact of PPIs/P-CABs on esophageal wall thickness and clinical symptoms in EoE patients.MethodsPatients who were consecutively diagnosed with asymptomatic esophageal eosinophilia and EoE and treated with PPIs/P-CABs were assessed in this study. Esophageal wall thickness before and after treatment was evaluated using endoscopic ultrasonography.ResultsThirteen patients were asymptomatic, while 20 presented with gastrointestinal symptoms at baseline. Treatment led to significant decreases in symptom scores, the EoE Endoscopic Reference Scores, and the EoE Histologic System Scores compared with those at baseline. Following treatment, significant reductions were observed in the total esophageal wall thickness (TWT) and thickness from the surface to the muscular layer (TSM) across the upper, middle, and lower esophagus compared with baseline values (median TWT, 2.0 mm vs 1.7 mm, p=0.005; 2.3 mm vs 1.9 mm, p=0.004; 2.9 mm vs 2.3 mm, p<0.001; median TSM, 1.1 mm vs 0.9 mm, p=0.001; 1.3 mm vs 1.1 mm, p<0.001; 1.8 mm vs 1.4 mm, p<0.001, respectively). Similar trends were observed in the lower esophagus of patients with asymptomatic esophageal eosinophilia, with the TWT and TSM values significantly lower after treatment (median TWT, 2.7 mm vs 2.5 mm, p=0.045; median TSM, 1.7 mm vs 1.5 mm, p=0.008, respectively). These findings were consistent in patients treated with either PPIs (p=0.027 and p=0.018, respectively) or P-CABs (p<0.001 and p<0.001, respectively).ConclusionsPPIs/P-CABs reduce esophageal wall thickening, particularly in the mucosal and submucosal layers.

Background/AimsSilymarin has been reported to be hepatoprotective and to improve liver function; however, its clinical effectiveness in specific liver diseases remains unclear. This study aimed to evaluate the impact of Legalon, which contains silymarin as its active ingredient, on changes in liver function test values and to assess its potential use as a practical treatment option for liver diseases.MethodsThis multicenter retrospective cohort study used the Common Data Model. Data were collected from adult patients with liver disease who were first prescribed Legalon between January 1, 2013, and December 31, 2022, across 10 medical institutions in South Korea. Changes in liver function test values at follow-up time points were compared with baseline values.ResultsPatients who were prescribed Legalon for at least 6 months showed a statistically significant decrease in liver function test values (aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase) compared with baseline values. At 3 and 6 months, aspartate aminotransferase decreased by approximately 23.18% and 24.54%, alanine aminotransferase decreased by 20.24% and 25.12%, and alkaline phosphatase decreased by 3.02% and 5.90%, respectively. All three parameters showed a sustained decline.ConclusionsOur findings indicate that silymarin (Legalon) induces a significant reduction in liver function test values, thus suggesting that this drug exerts medium to long-term hepatoprotective benefits. Moreover, the synergistic effects of silymarin with standard treatments highlight its potential as a complementary therapy for liver diseases.

Background/AimsThe aim of this study was to evaluate the efficacy of serum pepsinogen (sPG) tests for gastritis, gastric adenoma (GA), and gastric cancer (GC) using the enzyme-linked immunosorbent assay-based GastroPanel kit and to investigate the correlation between GastroPanel- and Latex-enhanced Turbidimetric Immunoassay (L-TIA)-derived sPG results.MethodssPG I and II levels and PG I/II ratios were measured using both kits in 2,204 participants, including 1,109 controls, 316 GA, and 779 GC patients.ResultsThe GastroPanel- and L-TIA-derived sPG results showed high concordance. An sPG I concentration of 70 ng/mL and a PG I/II ratio of 3 measured with the L-TIA kit corresponded to 100 ng/mL and 5.3 with the GastroPanel kit. sPG I decreased in the GA and GC groups, whereas sPG II was lower in the GA group, but higher in the GC group than that in control group. The PG I/II ratios significantly decreased in the GA and GC groups, especially for the intestinal type. The sensitivity and specificity of PG I/II ratio ≤5.3 using the GastroPanel kit for the detection of GA or GC were 51%–59% and 61%–66%, respectively, which were slightly higher than 51%–58% and 58%–63% using the L-TIA kit. The group with a PG I/II ratio ≤5.3 and Helicobacterpylori-negative status had the highest risk with an adjusted odds ratio of 3.36 for GA and 2.25 for GC, with more prominent increase in diffuse-type compared to intestinal-type.ConclusionsThe GastroPanel kit showed non-inferiority compared to the L-TIA kit.