NeuroRx௡: The Journal of the American Society for Experimental NeuroTherapeutics Biomarkers: Potential Uses and Limitations Richard Mayeux Gertrude H. Sergievsky Center and the Taub Institute for Research on Alzheimer’s Disease and the Aging Brain, College of Physicians and Surgeons, Columbia University, New York, New York 10032 Summary: Biomarkers provide a dynamic and powerful ap- from the earliest manifestations to the terminal stages. This proach to understanding the spectrum of neurological disease brief review describes the major uses of biomarkers in clinical with applications in observational and analytic epidemiology, investigation. Careful assessment of the validity of biomarkers is randomized clinical trials, screening and diagnosis and prog- required with respect to the stage of disease. Causes of variability nosis. Defined as alterations in the constituents of tissues or in the measurement of biomarkers range from the individual to the body fluids, these markers offer the means for homogeneous laboratory. Issues that affect the analysis of biomarkers are classification of a disease and risk factors, and the can extend discussed along with recommendations on how to deal with our base information about the underlying pathogenesis of dis- bias and confounding. Key Words: Antecedent biomarkers, ease. Biomarkers can also reflect the entire spectrum of disease diagnostic biomarkers, variability, reliability, validity. INTRODUCTION fections, immunological and genetic disorders, and can- cer are well known.1,3 Their use in research has grown Biological markers (biomarkers) have been defined by out of the need to have a more direct measurement of Hulka and colleagues1 as “cellular, biochemical or mo- exposures in the causal pathway of disease that is free lecular alterations that are measurable in biological me- from recall bias, and that can also have the potential of dia such as human tissues, cells, or fluids.” More re- providing information on the absorption and metabolism cently, the definition has been broadened to include of the exposures.4 Neuroscientists have also relied on biological characteristics that can be objectively mea- biomarkers to assist in the diagnosis and treatment of sured and evaluated as an indicator of normal biological nervous system disorders and to investigate their cause. processes, pathogenic processes, or pharmacological re- Blood, brain, cerebrospinal fluid, muscle, nerve, skin, sponses to a therapeutic intervention.2 In practice, bi- and urine have been employed to gain information about omarkers include tools and technologies that can aid in the nervous system in both the healthy and diseased state. understanding the prediction, cause, diagnosis, progres- This paper focuses on biomarkers as defined by Hulka et sion, regression, or outcome of treatment of disease. For al.,1 i.e., direct measures of biological media, and other the nervous system there is a wide range of techniques papers in this issue will address brain imaging and other used to gain information about the brain in both the markers. healthy and diseased state. These may involve measure- The rapid growth of molecular biology and laboratory ments directly on biological media (e.g., blood or cere- technology has expanded to the point at which the ap- brospinal fluid) or measurements such as brain imaging plication of technically advanced biomarkers will soon which do not involve direct sampling of biological media become even more feasible.5–8 Molecular biomarkers but measure changes in the composition or function of will, in the hands of clinical investigators, provide a the nervous system. dynamic and powerful approach to understanding the Biomarkers of all types have been used by generations spectrum of neurological disease with obvious applica- of epidemiologists, physicians, and scientists to study tions in analytic epidemiology, clinical trials and disease human disease. The application of biomarkers in the prevention, diagnosis, and disease management. diagnosis and management of cardiovascular disease, in- TYPES OF BIOMARKERS Address correspondence and reprint requests to Dr. Richard Mayeux, Sergievsky Center, 630 West 168th Street, New York, New York Biomarkers have been classified by Perera and Wein- 10032. E-mail: [email protected]. stein3 based on the sequence of events from exposure to 182 Vol. 1, 182–188, April 2004 © The American Society for Experimental NeuroTherapeutics, Inc. BIOMARKERS: USES AND LIMITATIONS 183 FIG. 1. Disease pathway and potential impact of biomarkers. disease (FIG. 1). Though biomarkers readily lend them- of exposure, which are used in risk prediction, and bi- selves to epidemiological investigations, they are also omarkers of disease, which are used in screening and useful in the investigation of the natural history and diagnosis and monitoring of disease progression. prognosis of a disease. Schulte9 has outlined the capa- Biomarkers used in risk prediction, in screening, and as bilities of biomarkers (Table 1). In addition to delineat- diagnostic tests are well established, and they offer dis- ing the events between exposure and disease, biomarkers tinct and obvious advantages. The classification of many have the potential to identify the earliest events in the neurological diseases is based on either standardized natural history, reducing the degree of misclassification clinical criteria or histological diagnoses. Biomarkers of both disease and exposure, opening a window to po- also have the potential to identify neurological disease at tential mechanisms related to the disease pathogenesis, an early stage, to provide a method for homogeneous accounting for some of the variability and effect modi- classification of a disease, and to extend our knowledge- fication of risk prediction. Biomarkers can also provide base concerning the underlying disease pathogenesis. insight into disease progression, prognosis, and response These advantages have direct application to all types of to therapy. clinical investigation, from clinical trials to observational There are two major types of biomarkers: biomarkers studies in epidemiology. In epidemiological (or quasi-experimental) investiga- TABLE 1. Contributions of Valid Biomarkers to Clinical tions, biomarkers improve validity while reducing bias in Research the measurement of exposures (or risk factors) for neu- Capabilities of Biomarkers rological disease. Rather than relying on a history of exposure to a putative risk factor, direct measurement of ● Delineation of events between exposure and disease the level of exposure or the chromosomal alteration re- ● Establishment of dose–response sulting from the exposure lessens the possibility of mis- ● Identification of early events in the natural history ● Identification of mechanisms by which exposure and classification of exposure. Such misclassifications not disease are related only produce inaccurate and deceptive results but also ● Reduction in misclassification of exposures or risk fac- reduce the power of studies to detect health effects. Thus, tors and disease the use of biomarkers improves the sensitivity and speci- ● Establishment of variability and effect modification ficity of the measurement of the exposures or risk factors. ● Enhanced individual and group risk assessments Molecular biomarkers have the additional potential to 6 Modified from Schulte.9 identify individuals susceptible to disease. Molecular NeuroRx௡, Vol. 1, No. 2, 2004 184 RICHARD MAYEUX genetics have already had an impact on neurological useful in the cross-sectional investigation of acute dis- practice, leading to improved diagnosis. Classification of ease because of the pharmacologic properties of the populations in terms of the degree of susceptibility on the chemical or toxin. It is very difficult to find biomarkers basis of such biomarkers produces greater accuracy than for exposures that are stable over the long periods re- relying on historical definitions of susceptibility.7,10,11 quired for prospective studies of chronic neurological For example, a biomarker will allow the stratification of diseases such as Alzheimer’s disease. Banked serum or a population on the basis of a specific “genotype” asso- plasma may be of value in some instances depending on ciated with a disease rather than relying on a report of the the disorder being investigated and the pharmacologic “family history” of the disease. The ability to quantify characteristics of the biomarker. Issues of timing, persis- “susceptibility” in this way can be an extremely impor- tence, dose, and storage site all must be considered for tant method for estimating disease risk among various this class of biomarker. populations. Genetic susceptibility Epidemiologic analyses can examine familial aggrega- BIOMARKERS OF EXPOSURE OR tion and assess genetic and environmental contributions ANTECEDENT BIOMARKERS to a disease by using life table methods and recurrence risk. Mutations in genes that result in Mendelian forms of Environmental exposures, effect modifiers, or risk disease are typically deterministic. Variant alleles in factors genes or polymorphisms may be related to susceptibility When a disease is suspected of resulting from a toxic but are not deterministic. Most adult-onset degenerative exposure, researchers naturally wish to measure the de- diseases of the nervous system are likely to be a com- gree of exposure. External exposure is the measured posite of related characteristics, heritable and environ- concentration of the toxin in an individual’s immediate mental. The correlated