[Proteomics for biomarker development toward personalized medicine]

Cancer is a genetically and clinically diverse disease, and the therapeutic strategies should be optimized for individual cases. Recent advances in pharmacogenomics have generated molecular targeting anti-cancer drugs, and their optimized uses are an emergent challenge. Biomarkers are key tools to assess the malignant potential of tumor cells and to establish risk-stratified therapies. The proteome is a functional translation of the genome and it directly regulates the malignant phenotypes of tumor cells; thus, a proteomic approach could make significant contributions to biomarker development. In the past decade, proteomics technologies have progressed extensively, and biomarkers to predict the response to treatments were developed using clinical materials in various types of malignancies. For example, proteomics identified a strong biomarker candidate, pfetin, as a novel prognostic biomarker in gastrointestinal stromal tumor, where the anticancer drug became available to reduce the risk of post-operative metastasis. The prognostic utility of pfetin was immunohistochemically established by multi-institutional validation studies, and we expect that in the near future we will be able to select patients who may need adjuvant therapy by measuring the expression of pfetin in surgical specimens. These observations suggested the utility of proteomics for biomarker development. Other than the application of advanced technologies, the key points in biomarker studies are the use of an adequate number of clinical materials with problem-oriented experimental designs. Collaborations between basic researchers and clinicians are critical for the effective approach towards realistic biomarkers by proteomics.