Integrated Multi-Omics Profiling Identifies Prognostic Biomarkers Predictive of Early Disease Progression and Therapeutic Response in Solid Malignancies: A Prospective Multicenter Cohort Study
Background Solid malignancies exhibit substantial molecular and clinical heterogeneity, limiting the prognostic accuracy of conventional clinicopathological variables and single-biomarker approaches. Integrated multi-omics profiling may provide a more comprehensive assessment of tumor biology by combining genomic, transcriptomic, proteomic, and metabolomic information. This study evaluated whether an integrated multi-omics model could identify prognostic biomarkers associated with early disease progression and therapeutic response across diverse solid tumors. Method This prospective multicenter cohort study enrolled 612 adults with histologically confirmed solid malignancies from five tertiary oncology centers in Armenia between January 2023 and December 2025. After prespecified exclusions, 588 patients were included in the final analysis. Pretreatment tumor tissue and peripheral blood samples underwent whole-exome sequencing, RNA sequencing, liquid chromatography–tandem mass spectrometry-based proteomics, and untargeted metabolomic profiling. Multi-omics datasets were normalized, corrected for batch effects, and integrated using Multi-Omics Factor Analysis and Similarity Network Fusion. Machine-learning models were developed to predict early disease progression within 12 months and therapeutic response. Clinical outcomes included objective response rate, progression-free survival, and overall survival. Results The median age of the cohort was 61 years, and 54.8% of patients were male. Lung, colorectal, and breast cancers were the most frequently represented malignancies. The objective response rate was 42.0%, and the disease control rate was 71.9%. During a median follow-up of 24.8 months, 39.6% of patients experienced disease progression and 29.9% died. Genomic profiling identified recurrent alterations in TP53, KRAS, PIK3CA, APC, EGFR, PTEN, BRAF, and ARID1A. Early disease progression was associated with increased expression of inflammatory, angiogenic, epithelial–mesenchymal transition, extracellular matrix remodeling, glycolytic, glutamine, and kynurenine-pathway signatures. Durable responders demonstrated enrichment of interferon signaling, antigen presentation, cytotoxic T-cell activity, and preserved mitochondrial metabolism. Integration of all molecular platforms identified a 32-marker prognostic signature comprising genomic, transcriptomic, proteomic, and metabolomic features. The integrated model achieved an area under the receiver operating characteristic curve of 0.93 and a Harrell concordance index of 0.91, outperforming clinical variables and each individual omics platform. High-risk classification was independently associated with shorter progression-free survival (adjusted hazard ratio, 2.84; 95% CI, 2.18–3.70; P < 0.001) and overall survival (adjusted hazard ratio, 2.61; 95% CI, 1.94–3.52; P < 0.001). Conclusion Integrated multi-omics profiling provided superior prognostic and predictive performance compared with conventional clinical factors or single-omics analyses. The identified 32-marker signature accurately stratified patients according to risk of early progression, therapeutic response, and survival across multiple solid malignancies. External validation and prospective clinical utility studies are required before routine implementation. Keywords: Multi-omics; solid malignancies; prognostic biomarkers; precision oncology; therapeutic response.