Genomic Alterations and Molecular Mechanisms of Therapeutic Response and resistance in Gastric Cancer, Hepatocellular Carcinoma, and Gastrointestinal Stromal Tumors: A Bioinformatics-Based Approach

다국어 초록 (Multilingual Abstract)

Gastrointestinal cancers remain a major health burden, and targeted therapies have improved outcomes for selected patient groups. However, responses vary widely, and resistance frequently develops. This dissertation aimed to investigate the genomic and transcriptional factors associated with treatment response and resistance in three gastrointestinal malignancies: HER2-positive gastric cancer, hepatocellular carcinoma (HCC), and gastrointestinal stromal tumors (GISTs). In HER2-positive advanced gastric cancer, targeted sequencing of 151 patients treated with trastuzumab-based chemotherapy identified key genomic features linked to clinical outcomes. Non-focal ERBB2 amplification was associated with longer progression-free survival, whereas alterations in the ERBB2 tyrosine kinase domain and NOTCH3 were associated with poorer prognosis. These findings suggest that quantitative ERBB2 amplified segment length and specific co-alterations may influence the degree of HER2 dependency and response to targeted therapy. For HCC, multi-omics profiling of 231 tumors was performed to explore synthetic lethality (SL) as a therapeutic approach. Although several SL candidates involving CDKN2A were identified through a DAISY-based pipeline, functional validation did not confirm true synthetic lethality. A complementary network-based strategy highlighted NEK2 as a potential vulnerability in TP53-mutant HCC, and pharmacologic inhibition showed greater cytotoxicity compared with sorafenib in vitro. These findings support NEK2 as a candidate therapeutic target in a molecularly defined subset of HCC. In GIST, transcriptomic analysis of 14 tumor samples before and after imatinib treatment revealed marked transcriptional changes in resistant tumors. While SPP1 was frequently upregulated, functional testing suggested they were adaptive rather than causative. HAND1 emerged as a potential regulator of resistance, warranting further mechanistic investigation. Overall, this work demonstrates how integrating genomic profiling with functional analyses can help identify biomarkers and therapeutic vulnerabilities. The findings may contribute to the development of more personalized treatment strategies for gastrointestinal cancers.

목차 (Table of Contents)

• 1. Introduction 1
• 2. Methods 3
• 2. 1. Gastric cancer study 3
• 2. 1. 1. HER2-positive advanced gastric cancers 3
• 2. 1. 2. Targeted panel sequencing 3
• 2. 1. 3. Genomic alteration analysis 3
• 2. 1. 4. Statistical analysis 4
• 2. 2. Hepatocellular carcinoma study 4
• 2. 2. 1. Clinical Samples 4
• 2. 2. 2. Multi-omics data including Whole exome sequencing, CNV and RNA sequencing 4
• 2. 2. 3. Design SL-SDL detection algorithm 4
• 2. 2. 4. Utilizing external databases 6
• 2. 2. 5. RNA-i based functional examination 6
• 2. 3. Gastrointestinal Stromal Tumor study 6
• 2. 3. 1. Patient grouping and next-generation sequencing 6
• 2. 3. 2. Cell culture 6
• 2. 3. 3. RNA Isolation and Quantitative Real-Time PCR (qRT-PCR) 7
• 2. 3. 4. Cell Viability Assay and IC₅₀ Determination 7
• 2. 3. 5. Mutation and expression level profile 7
• 3. Results 8
• 3. 1. Gastric Cancer (HER2-positive, n = 151; Targeted Panel Sequencing) 8
• 3. 1. 1. Patient Characteristics 8
• 3. 1. 2. Genomic Profile of HER2-Positive Tumors Treated with Trastuzumab 11
• 3. 1. 3. Clinical Relevance of ERBB2 Mutations, Particularly Within the Tyrosine Kinase Domain 15
• 3. 1. 4. Prognostic Implications of NOTCH3 Alterations 18
• 3. 1. 5. Prognostic Value of ERBB2-Amplified Segment Length 22
• 3. 2. Hepatocellular Carcinoma (231 patients; WES, RNA-seq, and CNV analysis) 25
• 3. 2. 1. Clinical samples and genomic landscape 25
• 3. 2. 2. Identification of candidate SL and SDL gene pairs for HCCs using restructured algorithm 28
• 3. 2. 3. Expanded approaches for detection of SL and SDL gene pairs 32
• 3. 2. 4. Identification of NEK2 as a SL pair of inactivated TP53 and a therapeutic target 32
• 3. 3. Gastrointestinal Stromal Tumor (GIST) 36
• 3. 3. 1. Sample group 36
• 3. 3. 2. Transcriptomic profiling after Imatinib exposure 36
• 3. 3. 3. Validation of candidate overexpressed genes (SPP1 and HAND1) 41
• 4. Discussion 44
• 4. 1. Findings from HER2-Positive Gastric Cancer 44
• 4. 2. Insights from Synthetic Lethality Analysis in HCC 45
• 4. 3. Understanding Imatinib Resistance in GIST 46
• 5. Conclusion 47