Introduction

An important issue for unresectable advanced stage hepatocellular carcinoma (uHCC) patients with Child-Pugh class B liver cirrhosis (CP-B), is lack of safe treatment options that show therapeutic efficacy, thus it is considered to be an unmet need. Atezolizumab plus bevacizumab (At/Bv) and durvalumab plus tremelimumab (Dur/Tre) have each been used as first-line systemic immunotherapy for uHCC.

Aims & Methods

The present study aimed to elucidate clinical outcomes of uHCC patients with CP-B treated with those combination treatments. The records of 169 uHCC patients with CP-B in Japan, treated with At/Bv or Dur/Tre as first-line systemic therapy from 2020 to 2024 at participating institutions were examined (median age: 73 years, males: 143, Child-Pugh score 7:8:9 = 121:43:5, BCLC stage B:C = 65:104, At/Bv:Dur/Tre = 131:38). Each regimen was analyzed for therapeutic efficacy and safety, and the results compared, in a retrospective manner.

Results

Using RECIST, ver. 1.1, the objective response rate (ORR) was shown to be 25.9% in the At/Bv group, and 18.4% in the Dur/Tre group (p=0.08). Disease control rate (DCR) was significantly higher in the At/Bv group (64.9% vs. 39.5%, p=0.01). There were no significant differences between the groups for progression-free survival (PFS) [At/Bv vs. Dur/Tre=5.3 vs. 3.5 months (95% CI: 3.9–6.5 and 2.3–7.2, respectively), p=0.90], overall survival (OS) [10.5 vs. 12.5 months (95% CI: 9.1–13.3 and 7.9–NA, respectively), p=0.58], or post-progression survival (PPS) [4.9 vs. 7.5 months (95% CI: 3.7–7.5 and 2.2–NA, respectively), p=0.60]. The rate of incidence of immune-related adverse events (irAEs) of any grade was 11.5% in the At/Bv group and 23.7% in the Dur/Tre group (p=0.07), while irAEs of grade 3 or higher were noted in 3.1% and 15.8%, respectively (p=0.02) significantly greater in the Dur/Tre group).

Conclusion

The incidence rate of irAEs in each group was similar to that observed in clinical trials previously performed for these regimens. uHCC patients with CP-B, who received Dur/Tre had a greater rate of incidence of high-grade irAEs, while the At/Bv group had a better DCR, though there were no significant differences for PFS, OS, or PPS between the groups. Neither treatment was found to sufficiently improve the prognosis of uHCC patients with CP-B. Development of safe and effective treatment options for patients affected by uHCC, that can be administered even to those with CP-B remains necessary.

Disclosure

Atsushi Hiraoka: lecture fee Chugai, AstraZeneca, and Lilly.

Introduction

Hepatocellular carcinoma (HCC) is the sixth most common malignancy and the third leading cause of cancer-related deaths globally, with increasing incidence and mortality rates. Most patients are diagnosed at advanced stages, missing the opportunity for radical treatment. Sorafenib, a tyrosine kinase inhibitor, is the first-line treatment for advanced HCC; however, drug resistance typically develops within 6 months. Therefore, elucidating the mechanisms of sorafenib resistance and identifying new therapeutic targets are of significant importance. Our study focuses on the role and mechanism of Ribonucleotide Reductase M2 (RRM2) in sorafenib resistance in HCC.

Aims & Methods

The primary objective of this study is to investigate the role of RRM2 in sorafenib resistance in HCC and explore its underlying mechanisms. We aim to clarify whether RRM2 regulates aerobic glycolysis via the HIF-1α/HK2 pathway, thereby affecting sorafenib sensitivity. Our methods include:

• Analyzing RRM2 expression in HCC tissues and its correlation with clinical outcomes, such as disease-free survival (DFS) and overall survival (OS).
• Establishing sorafenib-resistant HCC cell lines (LM3-SR and SMMC7721-SR) and manipulating RRM2 expression using lentiviral vectors for knockdown and overexpression.
• Conducting in vitro experiments to assess the effects of RRM2 on cell proliferation, apoptosis, glycolysis, and sorafenib sensitivity using assays such as CCK8, flow cytometry, and Seahorse analysis.
• Performing in vivo studies using a nude mouse xenograft model to evaluate the impact of RRM2 knockdown on tumor growth and sorafenib efficacy.
• Utilizing RNA-seq, Western blot, PCR, and other techniques to investigate the molecular mechanisms, including the potential regulation of HIF-1α and HK2 by RRM2.

Results

• Clinical Correlation: RRM2 expression was significantly higher in HCC tissues compared to adjacent normal tissues. Higher RRM2 levels were associated with shorter DFS and OS in HCC patients.
• In Vitro Findings: RRM2 knockdown in sorafenib-resistant HCC cells (LM3-SR and SMMC7721-SR) increased sensitivity to sorafenib, as evidenced by reduced cell viability and increased apoptosis. Additionally, RRM2 knockdown decreased glycolysis, as indicated by reduced lactate production and glucose uptake. Conversely, RRM2 overexpression in HCC cells promoted resistance to sorafenib and enhanced glycolysis.
• Molecular Mechanisms: RRM2 regulated the expression of glycolytic enzymes, particularly HK2, through the HIF-1α pathway. Knockdown of RRM2 reduced HIF-1α and HK2 expression, while overexpression of HIF-1α or HK2 could reverse the effects of RRM2 knockdown on glycolysis and sorafenib sensitivity.
• In Vivo Validation: In a nude mouse xenograft model, RRM2 knockdown enhanced the inhibitory effects of sorafenib on tumor growth, as demonstrated by reduced tumor volume and increased apoptosis.

Conclusion

Our study demonstrates that RRM2 plays a crucial role in sorafenib resistance in HCC by regulating aerobic glycolysis via the HIF-1α/HK2 pathway. Targeting RRM2 or its downstream pathways may represent a promising strategy to overcome sorafenib resistance and improve therapeutic outcomes in HCC patients. Future work will focus on further elucidating the molecular mechanisms and exploring the potential clinical applications of RRM2 as a therapeutic target.

Disclosure

The authors declare that they have no conflict of interest.

Introduction

Gastroenterologists face dual demands from clinical practice and academic research, necessitating efficient literature processing. While large language models (LLMs) show promise as healthcare tools, their variable performance in literature interpretation remains understudied, particularly regarding content completeness and reliability. This study systematically evaluates four prominent LLMs (ChatGPT-4o, Claude-3.5sonnet, Doubao-pro, Deepseek-V3) in explaining five gastroenterology literature types (original articles, case reports, reviews, editorials, guidelines), providing evidence-based recommendations for practical applications.

Aims & Methods

We tasked four LLMs with interpreting 125 gastroenterology publications (25 per literature type, 5 interpretations per article) spanning artificial intelligence, endoscopy, gastrointestinal cancers, and inflammatory bowel disease. Five gastroenterologists assessed interpretations using five criteria:
1. Accuracy: Correct identification of study design.
2. Format compliance: Adherence to predefined structure.
3. Content completeness: Key concept coverage (quantified via standardized rating scales).
4. Conciseness: Absence of extraneous content.
5. Reliability: Absence of hallucinated content.
Domain experts developed literature-specific completeness scales. Statistical analysis employed Pearson's chi-square tests for categorical variables (accuracy, format compliance, conciseness, reliability) and Kruskal-Wallis tests with Dunn's post-hoc comparisons for completeness scores. We applied Bonferroni correction for multiple comparisons (R version 4.4.2), with α=0.05 defining statistical significance.

Results

The 625 evaluations revealed significant inter-model differences across all metrics (p<0.001). ChatGPT-4o and Deepseek-V3 demonstrated superior accuracy (99.2% vs 100.0%) and format compliance (97.6% vs 99.2%) compared to Claude-3.5sonnet and Doubao-pro. Claude-3.5sonnet produced the most concise outputs (99.4%), while Deepseek showed the lowest concision (92.8%). Doubao exhibited perfect reliability (100.0%) versus 93.6%-97.0% in other models. Median completeness scores ranked: ChatGPT-4o (84.5%) > Claude-3.5sonnet (80.0%) = Doubao-pro (80.0%) > Deepseek-V3 (79.3%), with ChatGPT-4o significantly worse than others (Dunn's test, p<0.05).

Conclusion

Gastroenterologists can select LLMs based on the application priority of literature processing: ChatGPT-4o and Deepseek-V3 excel in accuracy and format compliance; Claude-3.5sonnet offers superior conciseness; while Doubao-pro demonstrates optimal reliability.