The Proteomic Search for Diagnostic Biomarkers: Lost in Translation?

Abstract

The search for biomarkers based on proteomic discovery strategies is being conducted in different scientific constellations. One major group is organized around governmentsponsored programs involving a network of academic institutions focusing on biomarkers systematically. An example of such a multicenter initiative for the United States is the “Directors Challenge Program,” which aims to characterize the molecular basis and related changes in protein expression of major cancers (www.cancerdiagnosis.nci.nhi.gov/challenge). Simultaneously there are major research and development programs in the pharmaceutical industry also focused on the identification of biomarkers but for different reasons (1). Various alliances and partnerships have also been formed between academic institutions and between industry and academia. Focusing throughout on diagnostic applications of biomarkers and ignoring the even larger efforts to identify proteins that can serve as drug targets, the goal of industry-driven efforts is not primarily the identification of diagnostically relevant proteins itself, but clear economic imperatives call for the eventual introduction of those novel biomarkers into the market. So far, financial analysts estimate that close to one billion dollars have been spent by or invested into companies with business models based on proteomic strategies for the identification of drug targets and/or biomarkers, including diagnostic biomarkers. However, there are few, if any, new biomarkers that have resulted in a commercial product by completing the proteomic process chain of identification, validation in clinical trials, and approval by regulatory agencies. On the contrary, the number of new protein targets approved by the registration agencies has been declining over the past decade, a trend not reversed or even stopped by any of the ongoing activities in proteomics. Moreover a survey of commercially available products used as diagnostic tools reveals that an amazingly small number of proteins in the human proteome have been identified as diagnostically relevant targets, but instead entire assay families have been developed that target the very same proteins in different configurations (2). Furthermore the scientific community as well as the general public has been led to believe that major diagnostic breakthroughs have already been achieved as a result of activities in proteomics and that consequently a series of novel diagnostic biomarkers are in place to fill the diagnostic gaps in a reliable and beneficial way. This misleading information policy has already proven to be counterproductive because expectations have been raised that have ultimately not held up to scrutiny. There have been high investments on one side but a low success rate on the other. Given this discrepancy, one feels challenged to start analyzing some of the potential hurdles in the processes leading from the discovery phase to the validation phase to product development. The rationale for these sequential steps has already been outlined (1, 3, 4), but several aspects should be revisited from different angles. Some of the considerations below were presented at the 21st Asilomar Conference on Mass Spectrometry in October 2005.