Prognostic model for lung adenocarcinoma based on experimental drug-resistant cell lines and clinical patients.

OBJECTIVEDespite advances in EGFR-TKIs for lung adenocarcinoma (LUAD), resistance remains a major hurdle. This study aimed to develop a prognostic model integrating immune microenvironment features and in vitro resistance mechanisms to predict outcomes and guide therapy.

MATERIALS AND METHODSerlotinib-, gefitinib-, and osimertinib-resistant HCC827 cell lines were established by exposing them to increasing EGFR-TKIs concentrations. RNA-sequencing was conducted on non-resistant HCC827 and erlotinib/gefitinibresistant cell lines. From the erlotinib-resistant, gefitinib-resistant cell lines and The Cancer Genome Atlas Program-Lung adenocarcinoma (TCGA-LUAD) data, a prognostic risk score model was constructed via Least Absolute Shrinkage and Selection Operator-Cox Proportional Hazards Model (LASSO-COX). Furthermore, immune infiltration was assessed using Gene Set Variation Analysis (GSVA), and single-cell RNA-seq (GSE241934) resolved expression patterns in EGFR-mutant vs. wild-type tumors. In vitro validation included RT-PCR in Osimertinib resistant (OR)-HCC827 cells.

RESULTSA 3-gene (PPP1R3G, CREG2, LYPD3) RiskScore were developed. The RiskScore predicted poor survival and resistance across all EGFR-TKI generations, with osimertinib-resistant HCC827 cells showing significant upregulation of signature genes. High-risk patients exhibited immune-suppressive microenvironments (enriched regulatory T cells, depleted mast cells) and distinct scRNA-seq profiles. A nomogram (C-index = 0.7) integrated RiskScore with clinical factors for personalized prognosis.

CONCLUSIONThis model bridges in vitro resistance mechanisms with clinical immune landscapes, offering a tool to stratify patients for EGFR-TKIs, immunotherapies, or combinatorial strategies.