By: Winfred Shaw | June 04, 2016
A primer on different kinds of biomarkers and how these tools are enabling precision medicine.
Precision medicine is changing the way we think about and treat human disease. By taking into account individual variability at a molecular level, researchers and physicians are able to hone therapies and therapeutic decision making to address the needs of each patient more effectively. At the heart of this clinical evolution are the biomarkers that researchers and physicians use to identify patient or disease variability that is relevant to choice of therapy.
While everything that we measure in the clinical laboratory is a biomarker of some type, only a special subset is pertinent to precision medicine. Broadly speaking, and as reflected by the National Institutes of Health’s definition, biomarkers are characteristics that are objectively measured and evaluated as indicators of normal biologic processes, pathogenic processes, or biological responses to therapeutic intervention. Thinking about the universe of biomarkers in terms of usage or purpose can help to avoid confusion about a given biomarker’s relevance to precision medicine. Four categories of biomarker utility cover the gamut of biomarker applications:
Diagnostic. Diagnostic biomarkers indicate the presence or absence of a pathogenic process, enabling physicians and researchers to identify or confirm the nature or cause of disease. Examples include the enzymes aspartate aminotransferase (AST) and alanine aminotransferase (ALT), which, when found elevated in blood concurrently, indicate the presence of liver injury.
Prognostic. Prognostic biomarkers indicate future clinical course in the absence of a therapeutic intervention. In other words, what will happen to the patient if he or she receives no treatment? A well-known prognostic biomarker involves specific mutations in the BRCA1 and BRCA 2 genes, which indicate elevated risk of developing breast and ovarian cancers.
Predictive. Predictive biomarkers are used to identify patients who are most likely to be susceptible to a particular drug effect, whether salutary or harmful. Predictive biomarkers are the primary tool in precision medicine, because they enable researchers to identify specific sub-populations most likely to respond well to a treatment. Doing so can shorten drug development timelines by producing convincing signals of efficacy within smaller study populations. This in turn may enable the biopharmaceutical industry to bring more and better targeted therapies to market. Common examples of predictive biomarkers include KRAS mutations that predict poor response to anti-epidermal growth factor receptor (EGFR) therapies, and overexpression of estrogen in breast cancer patients, which predicts response to anti-endocrine therapies such as tamoxifen.
Pharmacodynamic. Pharmacodynamic (PD) biomarkers are molecular indicators of a drug’s effect. PD biomarkers are extremely valuable in disease treatment because they demonstrate a clear link between a specific drug regimen and a patient’s biological response, providing data indicating whether, or to what extent, a treatment and dose is working. For example, a reduction in protein kinase activity of BCR-ABL indicates imatinib used to treat chronic myelogenous leukemia (CML) is working at the prescribed dose. In the future, PD biomarkers may play a more prominent role in oncology, especially as techniques to examine circulating tumor cells (CTC) and cell-free DNA (cfDNA) released into the peripheral blood from dying tumor cells become more common means of characterizing clonal response to therapy.
It’s important to note that a given molecule may be suitable for use in more than one type of biomarker application. For example, the glycated hemoglobin (HbA1c) ratio may be used as a diagnostic biomarker (values greater than 6.5% indicate diabetes), a prognostic biomarker (higher values indicate greater hazard ratios for diabetic complications), or a PD biomarker (lowered levels indicate improved glycemic control with therapy). In the treatment of cancer, as more becomes known about the relationship between CTCs or cfDNA and patients’ solid tumors, molecules in the peripheral blood that are first used as PD biomarkers may become useful as predictive biomarkers as well.
We are entering an exciting era in our molecular-level understanding of disease biology and therapeutic interventions. A confluence of factors, including advances in computing power, data storage, data curation, and analytic sciences, underpin private and public research efforts to accelerate this understanding. The result will be an enhanced armamentarium of biomarkers useful for diagnosing disease, prognosticating clinical course, and predicting and assessing therapeutic response. Deployment of predictive biomarkers within the context of precision medicine holds a great deal of promise for improving human health.