Molecular Subtyping of Lung Adenocarcinoma Reveals HMGA1 as Prognostic Biomarker

We are providing an unedited version of this manuscript to give early access to its findings. Before final publication, the manuscript will undergo further editing. Please note there may be errors present which affect the content, and all legal disclaimers apply. Lung adenocarcinoma (LUAD) shows marked heterogeneity, limiting the effectiveness of traditional classifications. We integrated bulk transcriptomic data from TCGA and multicenter cohorts with spatial transcriptomics for in situ validation. Analysis of glucocorticoid (GC) and mitophagy pathways identified 127 core genes, enabling a novel molecular subtyping and survival risk model that predicted prognosis and correlated with immune infiltration. High mobility group AT-hook 1 (HMGA1) emerged as a key regulator of tumor metabolism, immune evasion, and therapy resistance. Spatial transcriptomics with RCTD-based deconvolution showed a tumor-enriched and heterogeneous HMGA1 expression pattern and a positive spatial association with epithelial cell proportion in tumor sections. Single-cell RNA-seq analysis further resolved HMGA1 enrichment to tumor epithelial cells, and CIBERSORTx projection linked the HMGA1-high tumor epithelial cell state to poor prognosis and increased proliferation. Functional assays showed that high HMGA1 expression conferred sensitivity to metabolic inhibitors but resistance to proteasome inhibitors, while HMGA1 knockout suppressed tumor growth and enhanced anti–PD-1 efficacy. By integrating bulk modeling with spatial validation, this study establishes a GC–mitophagy–based classification and highlights HMGA1 as a promising biomarker for prognostic stratification and personalized immunotherapy in LUAD. This work was supported in part by grants from the National Natural Science Foundation of China [Grant Nos. 82573471 and 82272686], the Natural Science Foundation of Tianjin [Grant Nos. 23JCZDJC00200 and 25JCYBJC00270], and the Tianjin Key Medical Discipline Construction Project [Grant No. TJYXZDXK-3-003A]. These authors contributed equally: Junjie Yu, Yantao Jiang, Qingqing Xiong. Tianjin Medical University Cancer Institute and Hospital, National Clinical Research Center for Cancer; State Key Laboratory of Druggability Evaluation and Systematic Translational Medicine; Key Laboratory of Cancer Prevention and Therapy, Tianjin, PR China Junjie Yu, Yantao Jiang, Qingqing Xiong, Wei Luo, Luyao Tong, Yi Lu, Lianmin Zhang, Peng Chen & Tingting Qin Tianjin’s Clinical Research Center for Cancer; Department of Thoracic Oncology, Tianjin Lung Cancer Center, Tianjin Cancer Institute & Ho