# Translating Gastric Cancer Genomics into Targeted Therapy: Mechanistic Insights from Animal Models and Patient-Derived Systems

## TL;DR

Many gastric cancer therapies that worked in animal models failed in humans, but HER2 and CLDN18.2 therapies succeeded due to accurate biomarkers and tumor dependency.

## Contribution

The paper highlights the importance of true tumor dependency and functional biomarkers for successful clinical translation in gastric cancer.

## Key findings

- HER2- and CLDN18.2-targeted therapies are rare successes due to precise biomarker-driven patient selection.
- Advanced patient-relevant models may reduce preclinical–clinical discrepancies in gastric cancer research.

## Abstract

What are the main findings?
Many gastric cancer-targeted therapies that showed efficacy in animal models (e.g., EGFR, MET, FGFR2, and PI3K inhibition) failed to improve survival in clinical trials, revealing major translational limitations.HER2- and CLDN18.2-targeted therapies represent rare successful examples of bench-to-bedside translation, supported by clear oncogenic dependency and precise biomarker-driven patient selection.

Many gastric cancer-targeted therapies that showed efficacy in animal models (e.g., EGFR, MET, FGFR2, and PI3K inhibition) failed to improve survival in clinical trials, revealing major translational limitations.

HER2- and CLDN18.2-targeted therapies represent rare successful examples of bench-to-bedside translation, supported by clear oncogenic dependency and precise biomarker-driven patient selection.

What are the implications of the main findings?
True tumor dependency and functional biomarker accuracy are critical determinants of successful clinical translation in gastric cancer.Advanced patient-relevant models, including PDX, organoids, and organ-on-a-chip platforms, may improve target validation and reduce preclinical–clinical discrepancies.

True tumor dependency and functional biomarker accuracy are critical determinants of successful clinical translation in gastric cancer.

Advanced patient-relevant models, including PDX, organoids, and organ-on-a-chip platforms, may improve target validation and reduce preclinical–clinical discrepancies.

Gastric cancer remains a leading cause of cancer-related mortality worldwide and is marked by pronounced molecular heterogeneity. Advances in genomic profiling have identified key genetic alterations, including oncogenes (HER2, PIK3CA, and MYC), tumor suppressor genes (TP53, CDH1, and ARID1A), and regulators of genome stability and cell architecture (MLH1, RHOA, and CLDN18), which have driven the development of targeted therapeutic strategies. Although genetically engineered mouse models and xenograft systems have been indispensable for functional validation and preclinical drug testing, many approaches that showed promising efficacy in animal models—such as inhibition of EGFR, MET, FGFR2, and the PI3K pathway—failed to translate into overall survival benefits in clinical trials, highlighting major translational limitations. In contrast, HER2- and CLDN18.2-targeted therapies represent rare but notable clinical successes, underscoring the importance of true oncogenic dependency, precise biomarker-driven patient selection, and robust preclinical validation. In this review, we systematically categorize gastric cancer-associated genes according to their biological functions, summarize representative animal models, and critically examine key successes and failures in clinical translation, emphasizing the need for biologically faithful models and precision-driven translational strategies.