Integrated proteogenomic profiling to reveal prognostic subtypes and actionable targets in adenoid cystic carcinoma.

6067 Background: Adenoid cystic carcinoma (ACC) is a rare head and neck malignancy characterized by highly variable clinical outcomes and a lack of validated biomarkers to guide risk stratification or therapeutic decision-making. Despite a generally low tumor mutational burden, patients frequently experience late recurrence and distant metastasis, underscoring the need for biologically informed prognostic models. Existing genomic and transcriptomic classifiers have failed to adequately explain this heterogeneity. We therefore performed integrated multi-omic profiling to define biologically grounded prognostic subtypes and clinically actionable biomarkers in ACC. Methods: We established a multi-omics cohort of 55 surgically resected ACC tumors, including salivary gland–derived and pulmonary ACC, with matched adjacent normal tissues. Whole-exome sequencing, RNA sequencing, quantitative proteomics, and phosphoproteomics were performed. Cross-omics concordance and driver-anchored pathway effects were assessed. Transcriptomic and proteomic data were integrated using Similarity Network Fusion (SNF) to derive molecular subtypes. Prognostic proteins were evaluated using Cox regression. Targeted quantitative mass spectrometry was applied for orthogonal validation and for quantifying antibody–drug conjugate (ADC) targets and pathway stoichiometry. Results: ACC exhibited a low-mutational genomic landscape but profound proteomic and phosphoproteomic remodeling, with systematic activation of replication, cell-cycle, chromatin remodeling, and receptor tyrosine kinase–adhesion pathways, accompanied by suppression of immune-associated signaling. Driver effects were more consistently captured at the protein level than at the RNA level, indicating a proteome-driven tumor biology. SNF-based multi-omic integration identified three molecular subtypes with significantly different disease-free survival (DFS; log-rank P=0.0064), outperforming the conventional ACC-I/II classifier. These subtypes formed a biological continuum defined by metabolic robustness and cell-cycle competence. A compact three-protein signature independently stratified DFS. Quantitative profiling demonstrated broad and functionally relevant expression of ADC targets, including TROP2 and B7-H3, and identified ratio-based functional indices (e.g., IGF2/IGF1R and TP63/NOTCH1) associated with risk groups. Conclusions: ACC is a proteome-driven malignancy in which post-transcriptional regulation and signaling stoichiometry dominate clinical behavior. Integrated multi-omic analysis redefines ACC prognostic architecture and yields quantitative, translatable biomarkers that may inform precision risk stratification and therapeutic strategies in this rare head and neck cancer.