Defining criteria for useful biomarkers

The following discussion cites many of the important characteristics of a useful molecular biomarker, many of which have been discussed previously57 and are discussed elsewhere in this volume. It may be unrealistic to expect that every biomarker will fulfill all the criteria; in that case, judicious use of several biomarkers that provide complementing and confirming data is a suitable alternative. This concept of using a suite of bio-markers was pioneered by Perera and colleagues,810 and it is often a beneficial approach. The strengths and/or deficiencies of each molecular biomarker should be carefully considered in experimental design.

A. Accuracy and reproducibility

A useful biomarker is highly accurate and reproducible, characteristics that may be thoroughly established during the development phase of the marker using statistical methods and laboratory studies. Internal controls are extremely valuable for ascertaining accuracy and reproducibility. Controls should also be used to characterize and quantify the dynamic range of the assay. Controls enhance data comparability between experiments in a single study, between distinct studies in a single lab, and between studies in different labs or with different models. Use of such controls can be built into the data management structure as well. Attention to accuracy and reproducibility is important as assays move from laboratory studies with model systems to clinical or population-based studies with large numbers of samples.

B. Specificity/sensitivity

Biomarkers are powerful tools because they can identify and quantify exposure, effect, or susceptibility in individual members of a population. Thus, a biomarker must be sufficiently sensitive to provide an accurate measurement in a sample of limited quantity from a single individual. Sensitivity and specificity are often inversely related to one another. Specificity must be sufficient to avoid a high false positive rate, and sensitivity must be sufficient to avoid a high false negative rate. Standards and controls can be used to analyze and optimize assay sensitivity and specificity.

C. Database retrieval and retrospective analysis

Data management and data comparability are extremely important in development and validation of biomarkers and for their use in population-based studies. Proper data management allows researchers to use data efficiently and to increase the data's value. Ideally, the database structure should allow the data to be reevaluated by third parties long after they are added to the database. For this reason, accessions to the database should include complete sets of raw data, and the database structure should be relational. Thus, researchers will be able to reexamine data when new insights are gained into disease processes (e.g., disease-related covariables are identified). Importantly, the statistical power of any study may increase significantly, when data from different studies are combined in a valid manner. With appropriate foresight, relational database structures can be optimized with these goals in mind.

D. Experimental models and biological plausibility

Biomarkers of exposure and biomarkers of effect distribute along the continuum from exposure to disease and in some cases, a biomarker can indicate both types of events (e.g., site-specific DNA adduct in the p53 gene and/or site-specific p53 mutation as marker of exposure to cigarette smoke and as early marker of lung cancer11,12). In most cases, the link between exposure and effect is ambiguous unless it is rigorously confirmed using animal and cellular models. Animal models allow researchers to understand the mode of action of environmental chemicals and the biological mechanism of their action. Such information leads to biologically plausible biomarkers of the disease process. In addition, cell and animal models are essential for developing and refining knowledge of dose-response and exposure kinetics. Thus, the biological basis and usefulness of a molecular biomarker can be enhanced by thorough use of cell and animal models.

E. Sampling requirements

Useful molecular biomarkers for population-based studies should rely on noninvasive sampling (i.e., buccal swab, urine, blood). This requirement reflects the need for high-throughput biomarkers and for large sample size in epidemiological studies. Sample banking is also important because it enhances access to experimental materials and facilitates high-quality research studies and retrospective analyses. Retrospective analyses can take advantage of newly developed technology or allow a new hypothesis to be tested without additional field work.

F. High-throughput/feasibility on population scale

Biomarker assays need high throughput capability for application in population-based studies. Current technological advances facilitate this requirement. DNA microarray, proteomics, immunodetection, and high-throughput sequencing or genotyping all have the potential for rapid analysis of many samples and endpoints using automated processing and data analysis.

G. Implementing biomarker criteria

It will be important to promote development of peer-reviewed guidelines that emphasize and/or require that a biomarker meets an accepted set of criteria and will be implemented in an ethical manner. These guidelines need to be implemented during the development phase of a biomarker as well as after the biomarker is implemented in a clinical or population-based setting. Mechanisms to fulfill this need have been presented previously7,13 and are discussed below. A recent report discussing guidelines for development of biomarkers of cancer13 recommended a phased program for biomarker development, akin to phases of drug development.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment