Abstract P103: Large-Scale Integrative Analysis Identifies Expression-Based Predictors of EGFR-TKI Response

Abstract

Lung adenocarcinomas patients harboring classical mutations in the epidermal growth factor receptor (EGFR), notably Exon 19 deletions and L858R insertions, are typically treated with tyrosine kinase inhibitors (TKIs) as first-line therapy. While these tumours initially respond well, resistance frequently arises within 12 to 18 months, presenting a major clinical challenge. Current clinical practice relies on mutational profiling to guide EGFR-TKI therapy. Yet, this approach has critical limitations. Mutation status alone fails to capture the full complexity of tumor biology, including transcriptomic dynamics, downstream signaling adaptations, and phenotypic plasticity. Some EGFR-wildtype tumors exhibit EGFR-dependent transcriptional behavior and may respond to EGFR-TKIs despite lacking classical mutations, highlighting the need for broader molecular assessments. To address this, we performed a large-scale integrative analysis of transcriptomic data from 1,380 treatment-naïve lung adenocarcinomas across 10 public and internal cohorts. This diverse cohort includes both EGFR-mutant and wild-type tumours, enabling systematic investigation of gene expression programs associated with EGFR activity and TKI response. Our analysis revealed that EGFR-mutant tumours display upregulation of pathways related to small molecule transport and glutamine metabolism, involving targets such as SLC1A5 and GLS. These patterns align with known roles of EGFR in nutrient regulation and with studies demonstrating synergy between EGFR and glutaminase inhibition. Conversely, EGFR-wildtype tumours were enriched for innate immune-related gene expression, consistent with previous reports of enhanced immune infiltration in these tumours. These findings suggest that transcriptomic profiling captures both metabolic and immunological characteristics associated with EGFR dependency. For clinicians, this offers the potential to improve patient stratification and identify TKI-sensitive tumours beyond mutation status. For researchers, the results provide a foundation for integrative models that combine genomic and transcriptomic data to refine therapeutic decision-making in precision oncology.