Drug development is a long, slow and expensive business. Costs of drug development are reported as high as $11 billion or $12 billion. Despite these rising levels of investment, the failure rate is still high--on average, 80% of drugs fail in the clinical trial process.
Some drugs fail because they only work in specific subsets of patients. Predictive biomarkers, which forecast how a patient will respond to a specific drug, can select those patients most likely to do well on a particular drug. This means that clinical trials can be shorter and involve fewer patients, with clearer endpoints and results that are easier to interpret. Biomarkers can make it clearer when a drug has potential to go straight into the next step of development, and they allow unsuccessful drugs to fail fast, so that R&D teams can cut their losses and move on to the next clinical candidate.
Biomarkers in clinical trials can cut costs. In an analysis of U.S. new drug approvals in breast cancer between 1998 and 2012, using a validated biomarker (HER2) cut the cost of clinical trials and reduced the risk of drug development, increasing the percentage of drugs that finally reached the market.
Biomarkers can also be used as surrogate endpoints, allowing researchers to track the efficacy of drugs in chronic disease. As an example, in Parkinson's disease, a mutant form of the protein alpha-synuclein can accumulate in the brain. This also appears in the blood, and could be used as a non-invasive marker of the effectiveness of treatment.
In multiple sclerosis, expression of a subunit of α4-integrin on peripheral blood mononuclear cells (PBMCs) falls with successful treatment with the antibody natalizumab (Tysabri), and this could be used as a surrogate marker of the effectiveness of treatment.