What Is A Biomarker? Definitions, Types, And Research Applications Explained

The key message here is that biomarkers help not only in disease diagnosis (diagnostic biomarkers) but also in identifying potential treatments and tracking disease progression, regression, and outcome after an intervention.

Biomarkers include biomolecules like carbohydrates, proteins, lipids to genes, DNA, RNA, platelets, enzymes, hormones, etc. Anything that helps in the identification of a disease can serve as a biomarker be it a metabolite, a change in biological structure or biological processes, or a characteristic feature. Biomarkers can be classified in many ways.

Drug discovery and development is a tricky business. Failure rates of new drugs (meaning they fail at some point during development and clinical testing) are as high as 90% and can reach even higher for certain conditions like Alzheimer’s disease. 

Identifying reliable biomarkers can aid drug discovery and potentially increase success rates. Let’s take a deeper look at why biomarkers are important in drug discovery.

Biomarkers play a key role in the initial phases of drug discovery and development, as they can help identify potential therapeutic targets through identifying molecular pathways of disease. In addition, biomarkers can aid our understanding of the mechanisms of candidate therapeutics, which can further help assess the validity of potential therapeutics.

Biomarkers are also vital in later stages of drug development and are used in clinical research and clinical trials to assess response to treatments.

There are numerous ways to classify biomarkers. Read on to discover the various ways biomarkers are classified and how this relates to disease identification, progression, and clinical outcomes.

One note of caution is that biomarkers can fall into multiple types. For example, BRCA1 expression is both a predictive biomarker and a prognostic biomarker for the response to chemotherapy in sporadic epithelial ovarian cancer. (11)

According to Genetics and Molecular Biology methods, biomarkers can be classified as follows. (2)

An example of a type 0 biomarker is the measurement of serum creatinine to measure kidney function or monitor injury to kidneys. (6)

These are drug activity biomarkers and they indicate the effect of drug intervention. They may be further divided into (3)

Examples of type 1 biomarkers include blood glucose levels (efficacy biomarker) to monitor the effects of insulin treatment, (6) and cytokines (mechanism biomarker) in autoimmune diseases such as rheumatoid arthritis. (7) 

These are surrogate markers (also known as surrogate endpoints) that serve as a substitute for a clinical outcome of a disease and also help predict the effect of a therapeutic intervention. Such markers may correlate with clinical endpoints but may not have a guaranteed relationship.

Cholesterol in heart disease is one example of a Type 2 biomarker – where elevated levels are correlated with increased risk of heart disease but the relationship is not always present, as there are cases of heart disease with low cholesterol levels and some patients do not develop heart disease despite high levels of cholesterol.

Prognostic (Greek) means “fore-knowing or foreseeing.” Prognostic biomarkers are the ones that suggest the likely outcome of a disease in an untreated individual.

An example of a prognostic biomarker is the mutation status of PIK3CA in HER-2 positive metastatic breast cancer where individuals with mutated PIK3CA have been found to have lower rates of disease-free survival. (8)

Predictive biomarkers are used to identify those patients who are likely to have a positive clinical outcome in response to a given treatment. Thus, with the help of predictive biomarkers, it is possible to give a particular therapy to the patients for which it is most likely to be effective. 

Erlotinib maintenance treatment is a therapeutic intervention used to treat advanced non–small-cell lung cancer. For patients with an EGFR mutation in the tumor, progression-free survival is significantly lower following erlotinib treatment than for those without an EGFR mutation receiving the same treatment. This makes EGFR mutation status a predictive biomarker for response to erlotinib treatment.

These biomarkers help in determining the pharmacological effects of a drug and can inform if the treatment is working as expected. 

PI3K inhibitors are used to treat a variety of cancers. The PI3K signaling pathway includes multiple downstream targets including AKT. On activation of the PI3K pathway, AKT becomes phosphorylated and therefore phosphorylated AKT (pAKT) can act as a pharmacodynamic biomarker to confirm that Pi3K inhibitor treatment is indeed inhibiting the PI3K pathway. (9) If pAKT levels drop, the inhibitor is working. 

Surrogate biomarkers are identical in all but name to type 2 biomarkers, so see the section above for more detail. Another example of a surrogate biomarker is blood pressure for cardiac disease. 

To be useful, surrogate markers should be an effective substitute for the clinical outcome, and effects on the substitute should predict clinical outcomes.

However, there is a possibility, because the biomarker is just a surrogate, that the effect of pharmacological therapeutics on the surrogate may not affect the outcome, limiting its usefulness. (10)

The final way to classify biomarkers is by dividing them into two broad types: (5)

These can also be called antecedent biomarkers and are used in risk prediction.

According to the US Food and Drug Administration (FDA),(2,3) an ideal biomarker should be:

Different diseases have different biomarkers associated with them that are specific to that particular disease. 

Hopefully, we’ve answered the question “what is a biomarker?”, provided details of the various types of biomarkers along with useful examples including several cancer biomarkers, and highlighted their importance in drug discovery and development. 

Do you use biomarkers in your research? Let us know in the comments below!