A view point of cancer proteomics looking back over its short history

Abstract

Biomarkers are essential tools to achieve correct diagnosis, and the development of cancer biomarkers has been one of the most major topics in cancer prtoteomics. Many efforts and research budgets were devoted to biomarker studies in the last decade. However, there is no biomarker, which was developed by proteomics and proven to be clinically useful. The limitations of modalities and validation have been considered factors to hinder the successful biomarker study. However, the fundamental problem of biomarker studies in the laboratory was not well considered. In the most of biomarker studies, almost equal number of cases and controls are compared to identify the biomarker candidates. When high sensitivity and specificity are obtained by the biomarkers created in that condition, those biomarkers are considered to be candidates to be promising. However, the number of cancer patients in the ordinary population is extremely lower than non-cancerous persons, and the sensitivity and specificity calculated in the laboratory conditions may not make sense in the clinical setting. We need to employ the research theme in which the clinical situation can be reproduced in the laboratory. Such theme may include the biomarkers for differential diagnosis, and prediction of response to treatments or prognosis. Proper stratification of samples based on the statistical thought also improves the study for early detection. In conclusions, to obtain the clinically useful biomarkers by cancer proteomics, we need to consider the realistic utility of biomarkers in the clinical setting.