October 1, 2020

In the era of modern medicine, healthcare providers have relied on unique ways of observing and measuring the function of the body. These measurements are critical to the processes of health assessment, diagnostic accuracy, and monitoring of diseases and their treatments, and biomarkers are one of our most effective tools for this.

What are Biomarkers?

Biomarkers are any substance or characteristic that when objectively measured indicates the presence of an abnormal condition within a patient. Biomarkers do not have to be “complex” and can fall into the category of symptoms. During this COVID-19 pandemic, many institutions have begun a program of measuring individual’s temperatures. They are also asking people to be aware of various symptoms such as sore throat, cough, fatigue, headache, etc. All of these indicators are biomarkers that the individual is ill and should take precautions. Beyond symptoms, however, biomarkers are often specific molecules. These molecules can include proteins, metabolites, or genes that are associated with a specific disease.

Classes of Biomarkers

Biomarkers are often categorized based on their clinical application. Diagnostic biomarkers are used to define the disease and often its stage. For example, the BCR-ABL gene is used to diagnose individuals with chronic myelogenous leukemia. Prognostic biomarkers are used to assess the patient’s overall outcome regardless of what treatment is selected. For example, women with human epidermal growth factor receptor 2 (HER-2) positive metastatic breast who also have mutations in the PIK3CA gene had poorer outcomes compared to women who had the PIK3CA wild-type gene. Biomarkers can also be predictive. Predictive biomarkers differ from prognostic biomarkers in that they give information based on a specific treatment. For example, a predictive biomarker is used to determine if a patient will achieve some symptomatic benefit, improved survival, or an adverse effect that results from being treated with a specific drug or medical device (e.g., radiation, etc.). An example of a predictive biomarker is the V600E mutation in the serine/threonine-protein kinase B-Raf protein. Patients with late-stage melanoma that express this mutation benefit from vemurafenib treatment.

The Discovery of Molecular Biomarkers

While non-specific symptoms such as pain and fever can be classified as biomarkers, identifying molecules that are uniquely specific to a certain condition are invaluable in absolutely diagnosing and treating patients. The molecular biomarkers can be genes, transcript, proteins, or metabolites. Identifying molecular biomarkers that are disease specific is a formidable task requiring carefully controlled experiments and years of data collection and analysis. In most cases, there is little indication as to the identity of the biomarker (or even if a useful one exists) requiring a broad survey of molecules within the biological samples that are going to be tested. Therefore, a purely discovery-driven strategy is often employed in which the aim is to identify and compare as many molecules as possible between samples taken from disease-affected and healthy individuals.

The molecular biomarkers can be genes, transcript, proteins, or metabolites.
The first thing that needs to be done is sample collection. Samples need to be collected from a group of patients that have been diagnosed with a disease of interest and healthy controls. The type of sample collected (i.e., serum, plasma, tissue, etc.) depends on the disease being investigated. Ideally a biomarker would be accessible in a non-invasively collected sample but trying to discover a biomarker that may be specific to a tissue disease is extremely challenging in a matrix as complicated as blood, for example.

In the next stage, the molecular profiles of the samples need to be acquired. The analytical platforms for doing these studies depend on the nature of the biomarker. For example, if the biomarker is suspected to be genetic, whole genome or exon sequencing may be the platform of choice. If the aim is to find a protein or metabolite biomarker, mass spectrometry or nuclear magnetic resonance would be a reasonable analytical choice. Since the identification and comparison of thousands of genes, proteins, or metabolites between samples requires significant resources, most biomarker discovery studies begin with only tens of samples from well curated sources.

After the molecular content of the disease-affected and healthy samples have been compared, the results typically show a large number of molecular differences between the two sample cohorts. The reasons for this large number of differences is the inherent difference in the molecular content of biological samples acquired from distinct individuals. The molecular content of individuals is very dynamic, changing based on nutrients, time, and genetic content. Therefore, if a specific biomarker does exist, it must be found within the large number of other molecules that also show some difference between individuals.

The next stage is to test all of the identified differences between the samples in the discovery phase. The theme of the experimental design switches from discovery-driven to hypothesis-driven. Specific assays must be designed to evaluate each potential biomarker within another set of samples acquired from additional cohorts of samples. The primary differences between these sample sets and those used in the discovery-driven phases is that a much greater number of samples need to be analyzed (i.e., hundreds versus tens). Conducting specific assays on targeted potential markers serves to reduce the number of potential biomarkers down to between 5-20 depending on the study designs selection criteria.

The final phase of biomarker discovery builds upon the previous stage; however, an even greater number (i.e., thousands) of samples are analyzed and samples from patients with closely related conditions to the disease of interest are included in the study. If everything works well, a molecule (or molecules) that is able to differentiate between healthy and disease-affected patients is identified. The entire study from design through to final discovery generally required three-five years and costs on the order of tens of millions of dollars.

The Value of Biomarkers

Biomarkers are an incredibly valuable tool in the diagnosis and treatment of patients. The use of a reverse-transcriptase polymerase chain reaction (RT-PCR) test to detect RNA-specific COVID-19 has enabled rapid identification of infected individuals. For cancer patients, identifying a biomarker that can specifically identify the type of cancer can make all of the difference in whether the patients survive or not.
One of the greatest uses of biomarkers is early disease detection. These biomarkers can identify disease conditions early, often enabling a treatment plan to apply quickly before the disease can progress. For example, many infants are tested for in-born errors of metabolism, a group of disorders caused by mutations in genes that encode proteins and enzymes involved in metabolism. Approximately 1 in 2500 infants have an inborn error in metabolism, which include such conditions as maple syrup urine disease, phenylketonuria, and galactosemia. The ability to recognize these conditions at birth have enabled nutritional plans to be implemented immediately in the patient.

Biomarkers are an incredibly valuable tool in the diagnosis and treatment of patients.
Oncology is arguably the one field that benefits most from novel biomarkers. One of the major factors that determine a cancer patient’s survival rate is the stage of diagnosis. The statistics are very clear; the earlier a cancer is diagnosed the greater the chance of survival. For example, individuals with invasive endothelial ovarian cancer diagnosed at stage 1 have a 90% survival rate. Those diagnosed with the same cancer at stage 4 have only a 17% survival rate. Scientists have many useful tools for treating cancer (e.g., surgery, chemotherapy, immunotherapy, surgery), however, applying these treatments as soon as possible increases their efficacy. It is generally accepted within the oncology community that metastasis accounts for 90% of all cancer deaths. It is not a difficult concept to grasp; it is easier to treat a small primary tumor into remission than a large tumor that has spread to the lungs, liver, and/or brain.

As progress continues in biomarkers research, our ability to diagnose and treat serious conditions will drastically improve. However, with this research comes important ethical questions that must be considered. Look for more on this topic in our next blog post.

Tim Veenstra, Ph.D. is an Associate Professor of Pharmacy Science in the Cedarville University School of Pharmacy. Prior to his position at Cedarville University, Dr. Veenstra worked for the National Cancer Institute where his laboratory focused on identification of molecules involved in cancer diagnostics and development.

Posted in: