ARTICLE

Best Strategies for Biomarker Discovery

Introduction

Analytical evaluation and validation of biomarkers involves the use of assay technologies such as real-time polymerase chain reaction (RT-PCR) to determine gene expression or mutations, fluorescent in situ hybridization (FISH) for genetic alterations, and immunohistochemistry (IHC) or high content (HCS) analysis for protein expression and subcellular localization. This checklist will help you create a biomarker discovery process.

Biomarkers are analytically measurable and clinically validated substances or events (e.g., pathways) that represent diagnostic, predictive, and/or prognostic value within a normal or an abnormal biological condition. The significance and validity of the association of a biomarker with a physiological condition can be determined via discovery, characterization, and preclinical and clinical validation.

The analytical process involves a number of steps based on the type of biomarker, i.e., either DNA, RNA, protein, peptide, biomolecular modification(s), or biochemical pathway(s). This evaluation process may involve either one or a multiple combination of the biomolecules.

Analytical evaluation and validation of biomarkers involves use of assay technologies such as real-time polymerase chain reaction (RT-PCR) to determine gene expression or mutations, fluorescent in situ hybridization (FISH) for genetic alterations, and immunohistochemistry (IHC) or high content (HCS) analysis for protein expression and subcellular localization.


Biomarker discovery process

  1. Biomarker sample or study design
    1. Origin of the clinical sample or biospecimen: Critical factor that impacts the robustness of the study and reproducibility of the data. The study cohorts with appropriate annotations — as well as the methodologies used for sample preparation — impact the identification and subsequent characterization of biomarkers
  2. Molecular identification
    1. Published data on the biomarker(s) at the genetic level: Single or multiple genes that encode the protein biomarker(s) can be obtained from prior studies to establish the foundation for biomarker identification.
    2. Genetic analysis: If limited data is available at the genetic level, technologies such as RT-PCR and next generation sequencing (NGS) can be used to identify and characterize genes.
  3. Data management
    1. Big data generated from biomarker identification and characterization at the molecular and phenotypic level requires appropriate annotation and analysis tools as well as data warehousing or archiving capabilities.
  4. Biomarker characterization and validation
    1. Determination of the molecular size and structure of protein biomarkers via various analytical platforms such as chromatography and mass spectrometry, as well as detection and quantitation using technologies such as protein arrays, are key to biomarker discovery and characterization.
    2. Validation involves association of the biomarker to disease and/or physiological state. This can be determined by subcellular localization, using high-content analysis as well as via in-vivo studies in animal model systems.
  5. Clinical validation
    1. Preclinical and analytical validation of biomarkers above would need to correlate within the clinical environment, i.e., with phenotype within clinical samples. Emergence of cutting-edge technologies at the single-cell level, as well as transcript analysis without amplification, will further enhance the validity and relevance of biomarkers with clinical phenotype.


Challenges, bottlenecks, and common mistakes

Challenges are most commonly related to managing, storing, and analyzing big data. Bottlenecks are faced at almost every stage; therefore, a strong infrastructure to support scientists to manage data becomes critical. Furthermore, the analysis of this data can create its own bottleneck, placing greater premiums on robust data analysis platforms.

Leveraging external database storage capabilities is one option. A cloud environment is another platform to manage and store data. However, with proprietary data, particularly from industries, data security remains a major concern. Hence, leveraging cloud capabilities must ensure security and compliance to further open the doors to outsourced data management.


Optimal strategies

To overcome the challenges mentioned above, an optimal approach for both industry and academia would be to establish an internal framework supported by best practices via cross-functional teams that consider:

  • Standardized biomarker discovery approaches within preclinical settings.
  • Validation of the biomarker in clinical setting to demonstrate the phenotype prediction or diagnosis, or prognostic value during disease treatment.

The FDA continues to evaluate the significance and value of biomarkers in clinical decision-making and new therapeutics in precision medicine. To that, analytical discovery and development of biomarkers leading to its clinical relevance and validation become vital pillars to the transforming field of translational research and medicine.

Sources:

  1. Goossens, N., Nakagawa, S., Xiaochen, S., X., & Hoshida, Y. (2015). Cancer biomarker discovery and validation. Translational Cancer Research, 4 (3), 256-269. Retrieved from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4511498/