![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
*
Cancer Biomarkers Research Group, Division of Cancer Prevention, National Cancer Institute and
Division of Extramural Activity, National Cancer Institute, Bethesda, MD 20892
2To whom correspondence should be addressed. E-mail: ss1a{at}nih.gov.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONDISCUSSIONLITERATURE CITED |
|---|
KEY WORDS: • biomarkers • cancer screening • early detection • public health • technology
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONDISCUSSIONLITERATURE CITED |
|---|
). To be useful clinically, tests for biomarkers must have high predictive accuracy and be easily measurable and reproducible, minimally invasive and acceptable to patients and physicians. Potential uses of biomarkers include the following: 1) monitoring patients with established cancer for recurrence, 2) early detection of asymptomatic patients, 3) aiding in the diagnosis of symptomatic patients, 4) surveillance of individuals known to be at high risk of cancer, and 5) surrogate endpoint markers for primary prevention strategies such as chemoprevention.
Several classes of biomarkers in cancer cells and bodily fluids have been studied, mostly in laboratories examining specific observations but also in limited clinical settings. Several biomarkers have shown only limited utility: e.g., CD44, telomerase, transforming growth factor-
(TGF-)3
, transforming growth factor-ß (TGF-ß), epidermal growth factor receptor erbB-2 (erbB-2), epidermal growth factor receptor erbB-3 (erbB-3), mucin 1 (MUC1), mucin 2 (MUC2) and cytokeratin 20 (CK20). Some are being used in clinical practice based on observational associations. Prostate specific antigen (PSA) and cancer antibody or tumor marker 125 (CA 125) fall under the latter category. However, reduction of mortality using these in the screening setting has not been shown (2
). To date only a small number of protein markers have been studied in depth. Fecal occult blood test (FOBT) is the only protein biomarker shown to decrease cause-specific mortality in cancer screens (3
). This underscores the need to develop new and novel protein-based markers that can be reproducibly detected in serum or directly detected in exfoliated cells at early stages.
Recent advances in fundamental and clinical cancer research have mostly focused on identifying mutations, translocations and other molecular genetic abnormalities in the DNA of patients with established cancers. Such work is of great importance in the delineation of the mechanisms of tumor induction and for the identification of individuals who have an inherited predisposition to cancer. However, if early detection is to be useful, then clearly it must be targeted toward detection of lesions early enough to intervene and change the outcome. Although the existence of mutations in some known oncogenes is quite high in some specific cancers, in others it is insufficient to be of benefit in routine clinical practice. On the contrary, RNA and protein-based diagnostic techniques may have distinct advantages over DNA-based techniques in that they give direct evidence of abnormal gene expression at the time of sampling in a given patient who may not be symptomatic. Therefore, both DNA- and protein-based techniques are complementary. Identification of DNA sequences that encode proteins can stimulate subsequent expression analysis in tissues or biological fluids (4
).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONDISCUSSIONLITERATURE CITED |
|---|
There has been a lack of consistent use of terminology related to biomarkers and their applications in early detection, prognosis and treatment. Recently, a National Institutes of Health (NIH) working group recommended definitions that have applications in health services (5
), including: biological marker (biomarker), a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes or pharmacological responses to a therapeutic intervention; clinical endpoint, a characteristic or variable that reflects how a patient feels or functions or how long a patient survives; surrogate endpoint (intermediate endpoint), a biomarker intended to substitute for a clinical endpoint. A clinical investigator uses epidemiologic, therapeutic, pathophysiologic or other scientific evidence to select a surrogate endpoint that is expected to predict clinical benefit or harm or lack of benefit or harm.
Biomarkers as health care devices
Biochemical tests are used in health care to screen for detection, in diagnosis and to grade and monitor the progression of disease. These tests vary not only in accuracy, precision and reliability, but they have performance characteristics, e.g., sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). Test sensitivity and specificity refer to the identification of patients with and without the disease, respectively. For a test to be useful, it must have high sensitivity and specificity. The PPV is the proportion of persons who tested positive who have the disease. The requirements for the performance characteristics of a test vary with the intended use. For diagnosis or monitoring following treatment, high sensitivity is important; whereas for screening, high specificity is of paramount importance because a test with low specificity will lead to a large number of false positive cases. This can lead to unnecessary diagnostic procedures, including biopsies, and can result in prohibitive costs. Therefore, the test properties must be estimated as accurately as possible in well designed studies before a test is used in health care.
A biomarker can be considered a device to effect rational decision making on the health care of cancer patients in a cost-effective manner. Biomarker measurement can support key decisions throughout the drug development process, especially for cancer. However, the basis for desirable performance characteristics for clinical decision making should be a careful process and, whenever possible, the receiver operating characteristics (ROC) plots must be generated to help clarify the test performance. ROC curves are used to measure the test accuracy in regard to sensitivity and specificity (Fig. 1
). The overall performance of the test is measured by the position of the ROC line. The line for a perfect test will rise rapidly and reach close to the top left-hand corner, where both the sensitivity and specificity are located. A test with poor performance will have a line close to the rising diagonal.
|
–9
) are techniques commonly used for differential diagnosis in bladder cancer. With the possible exception of cytologic examination, these methods are not suitable for screening. When used as screening tests and coupled with therapy, none of these methods has been proven to lead to reduced mortality. Not only are they expensive and time consuming but they also are hampered by technical considerations. Flow cytometry, for instance, requires irrigation of the bladder, necessitating bladder catheterization rather than voided urine samples (9
). Standard cytology of the urine has inherent limitations in detecting low-grade tumors in the bladder, particularly as the individual cells of a Grade I transitional cell carcinoma often are morphologically indistinguishable from normal cells (10
). Endoscopic surveillance may fail to identify malignant changes in the bladder epithelium, especially in the early in situ stages. Tests for hematuria (11
) may be more promising but have not been tested in prospective studies.
Imaging and molecular technologies are developing rapidly but are not necessarily focused on the detection of early cancer. The potential for new imaging and molecular techniques to significantly improve the detection of localized lung cancer provides an unprecedented opportunity to understand the biology, improve diagnosis, enhance treatment and reduce mortality. These strategies have just begun to explore the utility of spiral computed tomography (CT), fluorescence bronchoscopy, PET imaging [see Gohagan et al. (12
) for review] and proteomic and genomic analysis of tumors and other specimens (4
). These approaches (and in particular the application of spiral CT) have the potential to identify biomarkers for small and early lesions that have not been readily accessible in clinical practice through more conventional detection methods. Molecular profiling may assist in identifying high-risk populations and offers a unique opportunity to study early carcinogenesis and potentially to reduce cancer mortality through the available effective treatment modalities amenable to early cancer.
Biomarkers as a competing device (technology) in health care
Technology in health care services (medical technology) is defined as "any discrete and identifiable regimen or modality used to diagnose and treat illness, prevent disease, maintain patient well being, or facilitate the provision of health services." In this article, we have focused our discussion on biomarkers for diagnosing illness. Technology such as laboratory tests (biomarkers), X-rays, CT scans and other imaging techniques, genetic and biochemical analyses are included in this discussion. In medical settings, there are three types of technologies: competing technologies include technologies that are equally effective but have different costs associated with their applications; cost-saving technologies include technologies that, in comparison with other technologies, are intended to reduce the cost of the diagnosis and prevention of disease both for individuals and for large groups of people; and intermediate technologies include technologies that increase productivity or improve performance, often at a lower cost, or generate additional costs while accomplishing something not possible previously. At present, biomarkers could possibly be considered half-way technology until they are rigorously tested, validated and shown to be effective; i.e., to identify a clinical endpoint with greater certainty (13
).
The translation of emerging technology into a market place (health care) is a complex process that involves the integration of scientific rationale and the regulatory process. The U.S. Food and Drug Administration (FDA) considers there to be a preclinical and four clinical phases for technology evaluation. In the preclinical phase, a new technology is conceptualized and developed. The four clinical phases are as follows: 1) phase 1, to perform an initial evaluation of a developing technology in a human population; 2) phase 2, to perform clinical studies involving limited numbers of human subjects to gather preliminary evidence regarding effectiveness and safety; 3) phase 3, to conduct controlled clinical studies to provide a reasonable assurance of safety and effectiveness in defined populations; and 4) phase 4, to address long-term safety and to better characterize the performance of the technology within a larger population. Phase 4 usually is conducted once a technology has been approved for marketing (14
).
In a recently published report, Houn and colleagues (15
) developed a hierarchical model with five levels by which to assess the efficacy of a given for technology. These are listed in order of increasing strength: 1) technical performance measuring the analytical sensitivity of a test; 2) performance accuracy in the success of the reader and the technology in disease detection or diagnosis; 3) patient management technology resulting in improvements in patient care over current standards; and 4) patient outcome efficacy demonstrating that the technology in question has long-term safety and a proven record in improving the quality of life within a larger population; and 5) social efficacy and cost effectiveness (Fig. 2
).
|
Biomarkers in the competing market place
The health benefits of emerging technologies may result from the encouraging findings in other fields. Early detection of cancer, for example, will enhance treatment if there is an effective treatment available. Early detection of illness and deviation from normal functioning will be accomplished through regular inexpensive biomarker-based assays. Subtypes of disease will be diagnosed more accurately. Customized treatment will be tailored to the patient’s risk factors. Disease management will utilize a wide range of monitoring tools such as new generations of molecular and genetic biomarkers. For example, stool screening for colorectal cancer is vigorously pursued by investigators in search of biomarkers and provides important advantages over other screening methods, such as colonoscopy and flexible sigmoidoscopy. The latter methods are uncomfortable for patients and are expensive. Therefore, better screening tools are needed that would exhibit the combined features of high sensitivity, specificity, affordability and broad acceptability for the general population, health care providers and third-party payers. FOBTs have been used to screen colorectal cancer for > 3 decades and continue to be the most frequently used stool-based screening tool in the United States. Several types of FOBTs are available currently. The most frequently used is the guaiac-impregnated Hemoccult card that detects blood in the fecal smear. However, FOBTs are inherently insensitive (26%) and nonspecific (88–98%). Stool screening is noninvasive, requires no unpleasant cathartic preparation, can be performed using a mailed-in specimen card from a physician’s office and, unlike sigmoidoscopy, reflects the full length of the colorectum (3
).
Favorable cost-effective analysis of clinical tests, including the assays for biomarkers, will pave the way for the acceptance of biomarker-based assays in the market place. Recent advances in the knowledge of cancer biology and genetics have raised the prospect that genomic- and proteomic-based assays could soon be commercialized. Two factors will dominate the cost effectiveness of newer technologies: the unit cost of the test and the life-year benefit resulting from the diagnostic intervention (16
). It is likely that the unit cost of the test will decrease dramatically over time because of technical innovations with high-throughput assays. The life-year benefit data will have to be generated for each test as it develops to convince providers to support the use of such tests.
Future directions
Biomarkers are making their way into investigation and are producing an enormous amount of information. With the rapid advances in genomic and proteomic technology, the discovery of useful biomarkers has arisen to unravel the unique signatures of cancer cells. These signatures are enabling investigators to pose scientific questions and address problems that, until recently, were inconceivable. For instance, the focus on biomarker-based disease detection has now shifted from one biomarker to a panel of biomarkers in distinguishing normal cells from precancerous or cancerous cells. Statistical methodologies are being developed to analyze multivariate data in reference to disease outcome. Sophisticated bioinformatic tools and computation algorithms are being developed to extract knowledge from data being generated by high-throughput technology in genomics and proteomics.
The National Cancer Institute’s Early Detection Research Network (http://edrn.nci.nih.gov) is supporting the integration of discovery, evaluation and validation of biomarkers to meet the stringent criteria articulated by the FDA for use as replacements for standard-of-care technologies, as adjuncts or as complementary criteria. The EDRN has three main components: Biomarkers Developmental Laboratories, Biomarkers Validation Laboratories and Clinical/Epidemiology Centers. The Biomarkers Developmental Laboratories conduct studies for the development and characterization of new, or refinement of existing, biomarkers; the Biomarkers Validation Laboratories serve as a EDRN resource for clinical and laboratory validation of biomarkers, to include technological development, standardization of assay methods and refinement; and the Clinical/Epidemiology Centers conduct clinical and epidemiological research on the application of biomarkers. Statistical, logistics and informatics support are provided through an auxiliary Data Management and Coordinating Center (DMCC). The DMCC also develops the theoretical statistical approaches to the simultaneous pattern analysis of multiple markers.
In summary, biomarkers as competing technologies for cancer screening will be judged on their effectiveness in reducing cancer mortality by the early detection of precancerous lesions that are amenable to surgical removal or treatment by chemotherapy and chemoprevention.
|
FOOTNOTES |
|---|
3 Abbreviations used: CA 125, cancer antibody or tumor marker 125; CK20, cytokeratin 20; CT, computed tomography; DMCC, Data Management and Coordinating Center; erbB-2, epidermal growth factor receptor erbB-2; erbB-3, epidermal growth factor receptor erbB-3; FDA, Food and Drug Administration; FDG-PET, fluoro-2-deoxyglucose-positron emission tomography; FOBT, fecal occult blood test; MUC1, mucin 1; MUC2, mucin 2; NIH, National Institutes of Health; NPV, negative predictive value; PPV, positive predictive value; PSA, prostate-specific antigen; ROC, receiver operating characteristics; TGF-, transforming growth factor-; TGF-ß, transforming growth factor-ß.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONDISCUSSIONLITERATURE CITED |
|---|
1. Srinivas, P. R., Kramer, B. S. & Srivastava, S. (2001) Trends in biomarker research for cancer detection. Lancet Oncology 2:698-704.[Medline]
2. Srivastava, S. Henson, D. J. Gazdar, A. eds. Molecular Pathology of Early Cancer 1998 IOS Press Amsterdam. .
3. Ahlquist, A. L., Skoletsky, J. E., Boynton, K. A., Harrington, J. J., Mahoney, D. W., Pierceall, W. E., Thibodeau, S. N. & Shuber, A. P. (2000) Colorectal cancer screening by detection of altered human DNA in stool: feasibility of multitarget assay panel. Gastroenterology 119:1219-1227.[Medline]
4. Verma, M., Wright, G. L., Hanash, S. M., Gopal-Srivastava, R. & Srivastava, S. (2001) Proteomic approaches within the NCI Early Detection Research Network for the discovery and identification of cancer biomarkers. Ann. N. Y. Acad. Sci. 945:103-115.[Medline]
5. De Gruttola, V. G., Clax, P., DeMets, D. L., Downing, G. J., Ellenberg, S. S., Friedman, L., Gail, M. H., Prentice, R., Wittes, J. & Zeger, S. L. (2001) Considerations in the evaluation of surrogate endpoints in clinical trials. Summary of a National Institutes of Health Workshop. Control. Clin. Trials 22:485-502.[Medline]
6. Ross, R. K., Paganini-Hill, A. & Henderson, H. E. (1988) Epidemiology of bladder cancer. Skinner, D. G. Lieskovsky, G. eds. Diagnosis and Management of Genitourinary Cancer 1988:23-32 W. B. Saunders Company Philadelphia, PA. .
7. Piccoli, C. W. & Rifkin, M. D. (1990) Magnetic resonance imaging of the prostate and bladder. Top. Magn. Reson. Imaging 2:51-66.[Medline]
8. Melamed, M. R. (1990) Flow cytometry detection and evaluation of bladder tumors. J. Occup. Med. 32:829-833.[Medline]
9. Stein, B. S. (1989) Advances in the diagnosis and treatment of bladder cancer. R. I. Med. J. 72:289-293.
10. Macfarlane, M. T., Figlin, R. A. & de Kernion, J. B. (1993) Neoplasm of the bladder. Holland, J. F. Freii, E., III Bast, R. C. Kufe, D. F. Morton, D. L. Weichselbaum, R.R. eds. Cancer Medicine 1993:1546-1561 Lea & Febiger Malvern, PA. .
11. Messing, E. M. & Vaillancourt, M. A. (1990) Hematuria screening for bladder cancer. J. Occup. Med. 32:838-845.[Medline]
12. Gohagan, J. K., Srivastava, S., Ross, S. C. & Black, W. C. (1999) New screening technologies in cancer screening. Kramer, B. S. Gohagan, J. K. Prorok, P. C. eds. Cancer Screening Theory and Practice 1999:559-594 Marcel Dekker, Inc. New York. .
13. Health services technology in managing health services organizations. Rakich, J. S. Longest, B. B. Darr, K. eds. Managing Health Services Organizations 3rd ed. :179-203 Health Professions Press Baltimore, MD. .
14. Maxim, P. E. & Gutman, S. I. (1999) Regulation of medical devices by the Food and Drug Administration with emphasis on in vitro diagnostic devices in cancer screening. Kramer, B. S. Gohagan, J. K. Prorok, P. C. eds. Cancer Screening Theory and Practice 1999:77-88 Marcel Dekker, Inc. New York. .
15. Houn, F., Bright, R. A., Bushar, H. F., Croft, B. Y., Finder, C. A., Gohagan, J. K., Jennings, R. J., Keegan, P., Kessler, L. G. & Kramer, B. S., et al (2000) Study design in the evaluation of breast cancer imaging technologies. Acad. Radiol. 7:684-689.[Medline]
16. Brown, M. L. & Kessler, L. G. (1995) The use of gene tests to detect hereditary predisposition to cancer: economic considerations. J. Natl. Cancer Inst. 87:1131-1136.
This article has been cited by other articles:
|
|
 |
|
  R. S. Vasan Biomarkers of Cardiovascular Disease: Molecular Basis and Practical Considerations Circulation, May 16, 2006; 113(19): 2335 - 2362. [Full Text] [PDF]
|
|
|
|
 |
|
 C. A Hobbs, M. A Cleves, S. Melnyk, W. Zhao, and S J. James Congenital heart defects and abnormal maternal biomarkers of methionine and homocysteine metabolism Am. J. Clinical Nutrition, January 1, 2005; 81(1): 147 - 153. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
