BACKGROUND: Immunogenic cell death (ICD) can be triggered by various therapies to induce anti-tumor immune responses, significantly enhancing treatment effectiveness, and is widely utilized in tumor immunotherapy. METHODS: LUAD data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) validated ICD-related molecular subtypes via consensus clustering. Clinical features, ICD genes, driver genes, mutations, tumor microenvironment, immune checkpoints, and drug sensitivity were compared. RT-qPCR, Western blot, immunofluorescence, ELISA, flow cytometry, and tube formation assays validated findings. RESULTS: Differential expression of 33 ICD genes was observed between tumor and normal tissues. These genes were clustered into two groups via consensus clustering and validated with GEO data. Prognostic analysis indicated superior outcomes in cluster 2 across TCGA and GEO cohorts. Significant disparities in clinicopathological characteristics like stage, gender, and age were noted between subtypes. Cluster 2 exhibited heightened expression of ICD-related genes, driver genes, immune checkpoints, and immune cells. Cluster 2 also showed increased sensitivity to chemotherapy drugs. IFNG overexpression in A549 and H1299 cells induced CRT exposure, HMGB1 release, and ATP secretion, thereby promoting dendritic cell maturation and enhancing CD8+ T cell function. Additionally, IFNG boosted tumor angiogenesis via HMGB1 pathways, which could be mitigated by HMGB1 inhibition. CONCLUSION: Identification of novel ICD-related molecular subtypes holds promise for guiding personalized therapies, assessing prognosis, and predicting immunotherapy efficacy in LUAD. IFNG emerges as a potential prognostic biomarker and therapeutic target, influencing both the tumor microenvironment and angiogenesis. These findings offer new insights into therapeutic strategies targeting IFNG-mediated pathways in LUAD.