Colorectal carcinoma is currently treated with a combination of chemotherapeutic drugs supplemented with targeted drugs. Since there is an eminent need to improved therapeutic success we employed the phenotypically-driven therapeutically guided multidrug optimization (TGMO) technology to identify personalized optimal drug combinations (ODCs). Using this technology we obtained low dose synergistic and selective ODCs for a panel of human colorectal carcinoma cells that remained active in 3-dimensional heterotypic co-culture models. From RNA sequencing and phosphoproteomics analysis we noticed considerable overlap in the top 20 most active kinases in all cell lines and that the mechanism converges around MAP kinase signaling and cell cycle inhibition despite differential cell mutation status, transcript expression levels and protein kinase activity. RNA sequencing revealed differentially expressed genes that confirmed these observations. These combinations were subsequently translated to in vivo models. Interestingly, the optimized drug mixtures reduced tumor volume with 80% and significantly outperformed the standard chemotherapy (FOLFOX) in these models, while preventing toxicity, observed after FOLFOX treatment. Pharmacokinetics studies demonstrated that the combinations supported significantly enhanced drug bioavailability.