Surgery and adjuvant chemotherapy can save the lives of many patients with early-stage colorectal cancer. On the basis of individual patient data from 18 trials and more than 20,800 patients, a current study has demonstrated substantial cure rates. After 8 years of surgery plus chemotherapy for selected stage II, and III patients, recurrence rates were minimal. However, 34% of these patients recur, particularly in the early period of 2 years after treatment and 38% die within 8 years [1].

How can this substantial recurrence and mortality rates be improved? Targeted agents and molecular genetic markers to predict response represent today the two major tools to improve survival. But despite progress, the benefit of treatment with monoclonal antibodies (cetuximab, panitumumab, bevacizumab) is limited to a few only patients. There is skepticisms that these targeted agents simply delay recurrence but not prevent recurrence and thus they cannot improve cure. In the field of response prediction, KRAS and perhaps BRAF mutation status could be used as biomarkers. But these markers can identify only a small proportion of responder patients.

In the era of quantitative genetics, personal genomics, systems biology and in silico models, how could current large amount of experimental networking biology-based data and signaling pathways network research be translated into innovative complex combinatorial drugs and predictive markers? How closed we are in the personalized, biopredictive cancer medicine? This article evaluates new technologies-based research that has revolutionized scientific though, discovery drugs direction, and the challenges toward clinical implications.