• info@icbcongress.com
• English
• Persian

• Login

Create Free Account

INTERNATIONAL CONGRESS OF

PharmacoGenetics

• Pharmacogenomics-Guided Therapy Selection in Oncology: 2024–2025 Evidence and Implementation

• Asal Naghipour-Kordlar,¹ Maryam Radmanfard,^2,*
  1. Faculty of Nursing, Tabriz University of Medical Sciences, Tabriz, Iran
  2. Department of Basic Sciences, Ta.C., Islamic Azad University, Tabriz, Iran
  
  
  
• Introduction: The integration of advanced computational approaches with molecular profiling has transformed the landscape of oncology, shifting from empirical treatment strategies toward precision medicine paradigms. Multimodal deep learning frameworks, which analyze genomic, clinical, imaging, and pathological data simultaneously, have demonstrated strong potential to improve therapeutic prediction and patient stratification beyond conventional methodologies (Yang et al., 2024) Concurrently, pharmacogenomics (PGx) has emerged as a critical pillar of precision oncology by enabling therapy selection based on germline and somatic genetic variation. Through predicting drug efficacy and toxicity, PGx enhances treatment safety and cost-effectiveness, particularly in oncology where adverse drug reactions can severely compromise outcomes. Nevertheless, the widespread adoption of PGx remains challenged by fragmented clinical evidence, infrastructural limitations, and the incomplete integration of genomic data into real-world care pathways (Marcu & Marcu, 2024). Recent advances provide strong support for the clinical implementation of PGx-guided therapy. The CODA-PGx framework, for example, leverages synthetic lethality principles and real-world evidence to identify genetically targeted treatment opportunities. This strategy successfully uncovered actionable vulnerabilities—such as heightened sensitivity to Carboplatin in STAG2-mutant tumors and Oxaliplatin in SMARCB1-mutant tumors—demonstrating the feasibility of tailoring therapeutic intensity to loss-of-function genetic contexts (Truesdell et al., 2024). Collectively, these developments emphasize the convergence of multimodal deep learning and pharmacogenomics as a transformative pathway for precision oncology. Their integration not only improves predictive accuracy and therapeutic outcomes but also addresses the translational gap between molecular discoveries and clinical practice, advancing the field toward the goal of fully personalized cancer care (Marcu & Marcu, 2024; Truesdell et al., 2024; Yang et al., 2024).
• Methods: A focused narrative review approach was applied. Relevant literature was identified through structured searches in biomedical databases using predefined keywords consistent with the study’s main focus, including pharmacogenomics, precision oncology, multimodal deep learning, genetically targeted therapy, and personalized cancer care. Publications were screened for relevance, and priority was given to peer-reviewed articles addressing pharmacogenomic biomarkers, therapeutic guidance, and emerging implementation strategies. Evidence was synthesized narratively to outline key concepts, innovations, and translational challenges in pharmacogenomics-guided oncology.
• Results: A comprehensive review of the selected literature revealed three major thematic developments in pharmacogenomics-guided therapy selection for oncology: 1. Multimodal Deep Learning Enhances Predictive Precision Multimodal deep learning models demonstrated superior capability in integrating diverse patient data—such as genomic, clinical, and imaging inputs—to improve therapeutic response predictions and enhance patient stratification compared to conventional approaches (Yang et al., 2024). 2. Big-Data Pharmacogenomics Supports Evidence-Based Decision-Making Pharmacogenomic analyses leveraging large-scale datasets emphasized the identification and implementation of germline biomarkers and highlighted the increasing utility of real-world data in shaping therapeutic strategies, although challenges remain in clinical translation (Marcu & Marcu, 2024). 3. Genetically Targeted Therapeutic Opportunities via CODA-PGx The CODA-PGx framework emerged as an innovative approach for genetically tailored therapy, utilizing synthetic lethality concepts and clinical outcome data. Notable associations included enhanced sensitivity to Carboplatin in STAG2-mutant tumors and sensitivity to Oxaliplatin in SMARCB1-mutant tumors, underscoring the potential for targeted treatment strategies in genetically defined contexts (Truesdell et al., 2024). 4. Guideline Integration Through Evidence-Based Recommendations Updated clinical guidelines now increasingly incorporate pharmacogenomic insights to inform therapeutic choices. These guidelines underscore the importance of systematically translating genetic evidence into standardized practice recommendations to enhance patient safety and care quality (Ornello et al., 2025). Collectively, these findings reveal a clear trajectory toward integrating advanced computational methods and pharmacogenomic evidence into oncology care—bridging discovery science with clinical applicability and guideline-informed practice. Table: Summary of Key Findings on Pharmacogenomics-Guided Therapy Selection in Oncology Theme Key Findings Multimodal Deep Learning Integration of genomic, clinical, and imaging data improves prediction of therapy response and patient stratification beyond conventional models. Big-Data Pharmacogenomics Large-scale analyses identify germline biomarkers and leverage real-world data to support therapy guidance, though translation barriers remain. CODA-PGx Framework Synthetic lethality approach reveals genotype-specific drug sensitivities (e.g., Carboplatin in STAG2-mutant, Oxaliplatin in SMARCB1-mutant tumors). Evidence-Based Guidelines Updated clinical recommendations increasingly integrate pharmacogenomic evidence to optimize therapy selection and patient safety. Discussion The findings summarized in this review highlight the rapidly evolving role of pharmacogenomics in precision oncology. Multimodal deep learning frameworks have demonstrated significant potential to improve therapeutic predictions by integrating diverse patient data types, offering a pathway to more individualized and effective cancer treatment strategies (Yang et al., 2024).These approaches enable the identification of complex interactions between genomic alterations, clinical features, and imaging phenotypes, facilitating more accurate patient stratification. Big-data pharmacogenomic analyses complement these computational approaches by systematically uncovering germline biomarkers associated with drug response and toxicity. Leveraging large-scale real-world datasets allows researchers to identify clinically actionable patterns and anticipate adverse reactions, although challenges persist in translating these findings into standardized clinical workflows (Marcu & Marcu, 2024). The CODA-PGx framework represents a key translational advance by combining synthetic lethality principles with real-world outcome data to inform genetically targeted therapeutic decisions. This method has successfully pinpointed genotype-specific vulnerabilities—such as enhanced sensitivity to Carboplatin in STAG2-mutant tumors and Oxaliplatin in SMARCB1-mutant tumors—underscoring the potential for tailored interventions based on molecular context (Truesdell et al., 2024). Furthermore, the increasing integration of pharmacogenomic evidence into clinical guidelines emphasizes the growing recognition of genetic factors in therapy selection. Evidence-based recommendations help bridge the gap between molecular discovery and clinical implementation, ultimately improving patient safety and therapeutic outcomes (Ornello et al., 2025). Overall, these converging lines of evidence support a future in which pharmacogenomics-guided therapy is routinely embedded in oncology practice. Challenges remain, including standardizing data integration, overcoming implementation barriers, and validating predictive models across diverse populations. Nevertheless, the combination of deep learning, large-scale genomic analysis, and guideline-based translation positions the field to advance toward truly personalized cancer care.
• Conclusion: Pharmacogenomics-guided therapy, supported by multimodal deep learning and large-scale genomic analyses, represents a transformative approach in oncology. Integrating genetic insights with clinical decision-making enables more precise patient stratification, optimization of drug efficacy, and reduction of adverse effects. Frameworks such as CODA-PGx demonstrate the potential for genotype-specific therapeutic targeting, while evidence-based guidelines ensure the safe translation of these findings into clinical practice. Continued advancements in data integration, model validation, and implementation strategies will be essential to realize the full potential of personalized cancer care and to establish pharmacogenomics as a standard component of oncology treatment pathways.
• Keywords: Pharmacogenomics, Precision Oncology, Multimodal Deep Learning, Genetically Targeted Therapy, Person