![[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, Bethesda, MD 20892
2 To whom correspondence should be addressed. E-mail: srivasts{at}mail.nih.gov.
 |
ABSTRACT |
|---|
 TOP ABSTRACT LITERATURE CITED |
|---|
KEY WORDS: • biomarkers • cancer • amino acids
Introduction
The biology of disease progression is a complex process that involves multiple sequential steps leading to cellular changes, altered signaling pathways, and metabolic events. These molecular events, which may serve as potential biomarkers, can be analyzed by laboratory methods and used to detect a disease or indicate the biological exposure to environmental substances including those encountered by dietary intake. Carcinogenesis involves initiation, progression, and metastasis. Identification of the genetic, molecular, and clinical events involved in the process enables the development of target-based therapeutic and preventive measures and the prediction of prognostic outcomes. In general, cancer biomarkers are classified based on their ability to detect the disease: early detection, high risk, prognostic, or predictive. A given marker may serve one or more of these functions and thus may fall into more than one category. For instance, prostate-specific antigen is used to detect and monitor prostate cancer in diagnosis and treatment (1). Cervical cancer–specific antigen-125 (CA-125) is commonly used for detecting cervical cancer (2). Although many of these markers help in detecting the presence of the disease, none has been validated for its usefulness in screening an asymptomatic population. Some genomic markers such as breast cancer–associated gene (BRCA)3-1 and BRCA-2 mutations identify the risk of breast cancer (3,4). Similarly, adenomatous polyposis syndrome, a genetic alteration that results in colonic polyps, helps evaluate the risk of colon cancer (5).
Nutritional biomarkers, in contrast, reflect the nutritional status with respect to the intake or metabolism of dietary constituents. A nutritional biomarker can be a biochemical, functional, or clinical index of status of a dietary constituent. Nutritional markers are broadly classified as short-, mid-, and long-term markers that predict the nutritional outcomes of the relevant changes in the biomarker levels to dietary changes (6). Like their diagnostic counterparts, nutritional biomarkers can also fall into more than one category depending on the intended purpose. For example, nutritional biomarkers can be used to validate dietary instruments such as dietary questionnaires. These markers may reflect a direct relationship to dietary intake and can be independently assessed; they are either the dietary nutrients themselves in the body fluids or direct end-products of the dietary substances. In some cases, biomarkers not only act as markers for dietary intake but also reflect a measure of nutritional status for a nutrient. Such biomarkers provide useful information regarding the metabolism of the nutrient and effects of disease processes.
Biomarkers and nutritional research
Biomarkers are influenced not only by disease but also by several normal physiological conditions such as age and genetic and environmental factors. This is even truer in the case of nutritional influence on the level of biomarkers. For a nutritional biomarker to be effective in nutritional research, it should reflect an absolute dietary intake or the metabolic event that follows the nutrient intake. However, the specificity of a nutritional biomarker is often impeded by a number of possible interactions that occur within the body. In addition to these biological uncertainties, the choice of specimen may be critical for the assessment of a nutritional biomarker, e.g., estimation of folate in plasma is subject to dietary changes, whereas erythrocyte folate more accurately reflects nutritional status (7,8). It is often intriguing to infer whether a given nutritional assessment represents the truer picture of a nutrient level, e.g., toenails and hair are considered ideal surrogates for selenium estimation; however, the validity of these markers has been questioned (6,9). In many instances, the robustness of an analytical technique is a critical factor. For example, analysis of vitamin C is a problematic issue for population studies, because the timely processing of the samples and the requirement of freshly prepared buffer for storage are of paramount importance (9).
It is unclear whether one biomarker could reflect nutritional status or whether a combination of biomarkers is needed, because it is often suggested that the actual protective effect of nutrients against diseases is the intake of a wide array of nutrients (10). Nutritional biomarker evaluation thus requires an extensive systematic approach to establish the status of the population, and the associated statistical and epidemiologic parameters are critical for nutritional research. In this article, we describe various factors that influence nutritional and cancer biomarker research and draw similarities between them. Moreover, we also discuss measures that have been adapted to validate cancer biomarkers that can potentially be applied to nutritional biomarker research.
Biological aspects.
The biological relevance of a nutritional marker can only be ascertained after confounding factors are controlled, which requires a comprehensive understanding of the biology and analytical measurements. A nutrient of interest may be in a mixture of compounds with similar biological action, e.g., endocrine disrupters or antioxidants. In such cases, the total content of these compounds rather than a single nutrient should be measured (11). Similarly, there are 
50 types of carotenoids consumed in the human diet, whereas only four or five compounds are found in serum in measurable amounts, and ß-carotene is the only one measured (12,13). In addition, the biomarker of interest may be the metabolite of a dietary compound that is responsible for the biological effect, which makes the biomarker research more complex. Adding to the complexity, the bioavailability of a nutritional biomarker can be affected by other dietary factors such as fat, carotenoids, or host factors, which may modify the correlation between the level of nutrient intake and the biomarker (11,14,15). Many dietary substances such as amino acids, when consumed in larger quantities, were found to have adverse effects on experimental animals (16). However, it is difficult to establish such toxic effects in humans because of the lack of quantitative data. One of the major problems associated with using biomarkers is that biomarkers vary with age, lifestyle, and environmental factors. For this reason it is essential to develop methods that can selectively detect changes in the biomarker level that are related to exposure and not to other environmental or physiological factors, especially for a wide range of study designs including aggregate (ecological) and analytical observational study designs and randomized control trials (17).
There is variation among individuals in physiology and nutrient metabolism. The absorption of a nutrient by the body depends on the nutritional status of the tissues. Low calcium or iron levels would prompt efficient absorption across the lumen (18,19). Similarly, the presence of certain substances in the diet may interfere with absorption, as in the case of phytic or oxalic acids, which interfere with divalent ion absorption (18). Absorption of vitamin B-12, for example, is largely influenced by the presence of intestinal intrinsic factor, and the actual deficiency of the vitamin is attributed to the absence of this factor (20). The presence of certain compounds in the diet as both contaminants and additives may determine the etiology of the disease and thus may need to be evaluated. Food processing may incorporate several toxic or beneficial bioactive substances that need to be continuously monitored (8).
Corollary between diagnostic and nutritional biomarkers. Biomarkers for cancer detection and biomarkers for nutritional research share several common attributes and performance characteristics, although the significance may vary with the final outcome. In the case of diagnostic markers, the final outcome is always the incidence of disease; however, in the nutritional setting, the outcome/use varies, e.g., level of nutrients, level of intake, etc. (Table 1). These performance characteristics must be addressed before they are applied to population studies.
|
Specificity. Specificity describes the true-negative results (tested individuals without disease who test negative for that disease) expressed as a percentage of all tested individuals who do not have that disease (the total of true negatives and false positives). When applied to cancer, specificity is the chance that a person without cancer will have a negative test.
Positive predictive value. The positive predictive value is the proportion of individuals who test positive and also have the disease. The requirements for the performance characteristics of a biomarker vary with its intended use. For diagnosis or monitoring, high sensitivity is important, whereas for screening, specificity is of paramount importance. However, in some cases, for example, in hepatitis C viral infection, sensitivity is more important for screening (21).
Sensitivity and specificity have different interpretations in the context of nutritional research. For example, in the context of nutritional uptake, sensitivity refers to the ability of an assay to measure the "true" concentrations of a nutrient of interest independent of confounding variables such as absorption and metabolic rates, age, sex, etc., whereas specificity refers to the ability of an assay to distinguish the nutrient of interest from other nutrients. If the purpose of the biomarker is to stratify subjects based on their basal level from an altered level of a nutrient, then the sensitivity and specificity have requirements similar to their diagnostic counterparts. Here the objective is to minimize classification error rates in reference to a particular nutrient. The ability to measure the concentration of a nutrient in body fluids may not reflect much about the ability to measure the nutrient intake. To alleviate the problems associated with a biomarker of exposure such as toxicity related to amino acid intake, either the sensitivity of the assay must increase, or a large study population is required to detect differences between the physiological basal level and an increased intake of the amino acids.
Methodological aspects. Although a variety of specimens can be selected for identifying biomarkers, specimens collected by noninvasive means are more acceptable and easier to handle and store. A selected specimen should contain ranges of biomarkers that are detectable by laboratory methods and be physiologically relevant. Such specimens could include blood, urine, stool, and sputum. However, the choice of specimen depends on the purpose of the study. For example, hair, nails, and breast milk may serve as better specimens for analyzing some biomarkers such as trace elements (9). Certain specimens such as fecal samples may only be amenable for small-scale studies, whereas some of the exhaustive sample-collection procedures such as 24-h urine-sample collection for dietary nitrogen may be required for some studies. Systemic biomarkers such as hormones and other regulators of metabolism may predict overall health status pertaining to that nutrient, whereas more specific tissue-related biomarkers such as bone-density to predict the onset of osteoporosis might provide information on organ development and response to nutritional status (22).
Laboratory issues. Proper sample handling and storage are of utmost importance when analyzing biomarkers. Urine, for example, may need to be acidified to improve the shelflife of the samples. Some biomarkers may be unstable when exposed to light, heat, oxygen, or other factors or may simply decay over time. Care must be taken to avoid exposure of the samples to a variety of laboratory conditions, and the amount of time elapsed between the sample collection and analysis may be critical. The assay to detect a biomarker should be relatively simple and cost effective and should have the capacity for high-throughput scalability. Appropriate quality-control ranges must be present whenever possible. Success of an analytical method relies on the sensitivity of the method to detect the biomarker in samples at minimum levels or in repeated analysis. For this reason, the validity, reliability, and repeatability of the results of the biomarker assay are of great importance. Repeatability refers to the sensitivity of the analytical method to detect the biomarker in question with repeated measurements. Reliability refers to the correlation of measurements for a biomarker in repeated assays. Validity of the biomarker refers to the ability to identify a highly repeatable characteristic of a biological sample. It is usually described as the correspondence of a biomarker with the actual exposure. Some samples may render themselves as informative tools for studying certain nutritional issues but may not be applicable to larger studies. For example, fecal components can be important for measuring fibers or bile acids that are relevant to colon cancer and other diseases; however, they can only be applied to small-scale studies. Urine samples may be valuable to study water-soluble nutrients but are limited by the nutrient saturation of the tissues and dietary intake (6). Some samples like breast milk may be easily accessible but may not be relevant to many diseases with epidemiological importance. Knowledge of the presence of the biomarker in a given serologic component is essential so that the appropriate component can be collected without collecting all the fractions for the analysis.
Technological aspects. Current advances in genomics, epigenomics, proteomics, and metabolomics have greatly accelerated biomarker discovery for nutritional research. By determining molecular patterns under various conditions, it is possible to develop a molecular signature for a given disease or physiological status. Also, these profiles could be used to measure the influence of dietary components. These novel approaches largely complement the conventional biochemical or imaging methods in identifying more effective biomarkers.
Approaches to biomarker discovery
Several approaches specific to molecular events have emerged for biomarker discovery. These are briefly summarized here.
Nutrigenomics. Determination of molecular patterns under various nutritional conditions can yield interesting results related to the genes that are turned "on" or "off" in pathways related to nutrient metabolism. The techniques of transcriptomics include DNA-based large-scale microarrays, differential display, and serial analysis of gene expression to study gene-expression patterns in response to external stimuli or nutrients (23). Recently it was shown that amino acids can modify the gene expression of target genes (24). Amino acid depletion could trigger many molecular events leading to the activation of upregulation and translation of the cationic amino acid transporter gene (25). In rat hepatocytes, addition of amino acids leads to swelling, which implies regulation of gene expression (26,27).
Nutritional proteomics. Although gene-expression patterns provide substantial information about the physiological response to external nutrients, they do not always correlate with protein concentrations. Moreover, proteins are subject to proteolysis or post-translational modifications such as glycosylation or phosphorylation that are not evident from genomic-based biomarkers. Biomarker discovery strategies that target protein expression are becoming popular because proteomic approaches enable the characterization of proteins involved in disease progression or physiological conditions. Various proteomic approaches include two-dimensional electrophoresis, isotope-coded affinity tags, surface-enhanced laser desorption/ionization (SELDI), matrix-assisted laser desorption/ionization (MALDI), liquid chromatography coupled to mass/mass spectroscopy, and MALDI–time of flight; these methods give rise to protein fingerprints, tissue-based arrays, protein microarrays, and antibody-based arrays. Although the field is still evolving, there is great promise for novel biomarker development. In a recent study, Sironi et al. (28) used proteomic approaches to show the effects of salt loading on the protein patterns in body fluids. Van Eijk and Deutz (29) studied the plasma protein–synthesis patterns using proteomic approaches. Using a targeted proteomic approach, Brooks et al. (17) showed that the angiotensin II type 1a receptor plays an important role in regulation of Na+ transporter and channel proteins.
Nutritional metabolomics.
A general limitation of expression studies is the correlation between the changes in gene expression to the functional assignment of specific genes. Metabolic changes influence nutritional requirements and utilization. Examining metabolomics or changes in metabolic profiles can be an important part of an integrative approach for assessing physiological responses to external agents. A comprehensive approach is important in identifying a nutritional biomarker, because the biomarkers of nutritional status relate to changes in a wide variety of trace elements, vitamins, and other nonorganic or small organic molecules. Recent metabolomic advances include nuclear magnetic resonance spectroscopy, mass spectrometry, chromatographic analysis, and metabolic network analysis models to predict metabolic fluxes. Xu et al. (30) showed that peroxisome proliferator-activated receptor-
influences substrate utilization for glucose production in hepatocytes. Rosiglitazone has been shown to induce varied effects on tissue metabolism and the lipid metabolome (31). Although systemic metabolomic signatures can be more informative, organ-specific metabolomic approaches would reduce the complexity and yield helpful information about dietary factors (32).
Nutritional epigenomics. Certain environmental factors influence the selective activation or inactivation of a gene or a set of genes by epigenetic changes such as methylation of DNA patterns, mostly in the promoter region, or acetylation of histones (33). In many events these patterns are inherited, which makes them vital targets for disease progression. It was shown that dietary substances bring about many of these epigenetic changes by affecting DNA methylation through changing the availability of methylating compounds such asS-adenosylmethionine or influencing methylating pathways (34). For this reason, exploration of epigenetic nutritional biomarkers may give rise to helpful information about the roles of dietary substances in relation to diseases such as cancer.
Challenges with nutritional biomarkers
Biomarkers offer a tremendous amount of information about health and disease and therefore hold great promise for nutritional epidemiology. However, the biomarker evaluation is often impeded by a lack of ideal endpoints and suitable laboratory methods to detect these conditions. Although significant progress can be achieved in this field through animal models, unfortunately, good models are still lacking for the study of various nutritional conditions. Other issues that pertain to nutritional biomarkers as discovery pathways include a lack of knowledge of the dietary constituents that may play a significant role in health and disease and uncertainties due to the dynamic nature of dietary changes and the difficulties in quantifying them effectively. Clearly, there is a demand for nutritional biomarker research that can address these issues.
Technological advances add another confounding factor to these uncertainties. Although our knowledge of biomarkers has expanded over the years, assays and analyses based on genomic and proteomic approaches have become increasingly challenging.
From a practical viewpoint, these assays need to be robust and thereby lend themselves as powerful tools to biomarker research. However, such approaches are successful only if the procedures are simple, inexpensive, and minimally invasive. Development of high-throughput assays is highly desirable to allow us to screen larger populations in less time. Biomarkers based on one technology may not necessarily provide all the needed information regarding physiological status. Similarly, DNA-sequence data are limited by a lack of information on post-translational modifications or biological consequences of altered gene expression due to environmental factors such as diet. We still lack effective computer models to simulate various physiological conditions and thereby provide a wealth of information without the need for expensive, long-term clinical trials and experimental research. These simulation studies could provide information on the natural history of the nutritional status and the biomarker, treatment effects on the nutritional status in relation to disease, cost information, and population behavior (35–37). Bioinformatics must evolve to enable the analysis and integration of the high-dimensional, high-throughput data that are generated on such multivariate platforms. Many technologies such as proteomics and metabolomics are still in their infancy and therefore necessitate a systematic iterative approach for biomarker research.
Cancer biomarker research suffers from a lack of validated biomarkers that can successfully predict risk and detect the disease at an early stage and thereby improve the chances of therapeutic intervention. Although many efforts have been focused on biomarker discovery leading to a wide range of promising biomarkers, none has been validated. Unfortunately there are no recommended criteria for validating a biomarker, although some biomarker approaches (such as Pap-smear testing to screen for cervical cancer) have gained wider acceptance on their own over time (38,39).
In an effort to accelerate biomarker research, the National Cancer Institute's Early Detection Research Network (EDRN) has adapted a five-phase approach to biomarker development and evaluation (40). The purpose of the EDRN is to coordinate research among biomarker development laboratories, biomarker validation laboratories, clinical repositories, and population-screening programs. By coordinating these efforts, EDRN hopes to facilitate collaboration and promote quality and timely efforts in cancer biomarker development. Currently there are 18 biomarker development laboratories (BDLs), 2 biomarker validation laboratories (BVLs), and 8 clinical and epidemiological centers (CECs) that are constantly monitored by the Data Management and Coordination Center, which supports the infrastructure of EDRN (Fig. 1).
|
These phases are being successfully applied at EDRN to decide which biomarker is worthy of clinical validation using a triage system as depicted in Figure 2. Since its conception in 1999, many biomarkers in the pipeline are being validated for potential application in cancer detection. For example, microsatellite-instability assay is under validation for its potential application in bladder cancer diagnosis. SELDI proteomic signatures are being evaluated for detection of prostate cancer. A similar approach can also be adapted to develop nutritional biomarkers for epidemiological studies wherever applicable. The actual priorities can be modified based on the specific objectives of the nutritional status being studied. Nutritional biomarkers must be thoroughly evaluated for efficacy and properties. Establishment of scientific consortia such as EDRN for nutritional biomarker development may facilitate coordination among researchers for biomarker development and validation and for promotion of the process of nutritional biomarker development. In addition, smaller, well-defined, controlled studies should be conducted with newer markers before larger population studies are addressed. Although no single method of biomarker evaluation is perfect, a combination of efforts is required to validate the usefulness of nutritional biomarkers.
|
Biomarkers in health and disease could provide useful information for nutritional studies. However, it is often difficult to establish the causality between health states and specific dietary constituents in the first place. To successfully interpret nutritional exposure and outcome, it is necessary to assess conditions with a near-perfect biomarker that can evaluate nutritional status and external factors effectively. It may be possible to address the issue in an iterative manner wherein biomarkers may help to reveal more about mechanisms, which in turn may assist in the development of other biomarkers. The iterative process allows for successful accumulation of information on biomarkers. Potential success of a biomarker depends essentially on its sensitivity, specificity, predictive value, ease of use, applicability, and relevance. Ultimately the true value of a biomarker depends on how much it can be used to measure the impact of diet on health. The link between diet and health, which can be objectively measured through continuous efforts in biomarker research, will open new doors to disease prevention and cure.
|
FOOTNOTES |
|---|
3 Abbreviations used: BDL, biomarker development laboratory; BRCA, breast cancer–associated gene; BVL, biomarker validation laboratory; CEC, clinical and epidemiological center; EDRN, Early Detection Research Network; MALDI, matrix-assisted laser desorption/ionization; SELDI, surface-enhanced laser desorption/ionization.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTLITERATURE CITED |
|---|
1. Lein, M., Kwiatkowski, M., Semjonow, A., Luboldt, H. J., Hammerer, P., Stephan, C., Klevecka, V., Taymoorian, K., Schnorr, D., et al. (2003) A multicenter clinical trial on the use of complexed prostate specific antigen in low prostate specific antigen concentrations. J. Urol. 170: 1175–1179.[Medline]
2. Bast, R. C., Jr., Xu, F. J., Yu, Y. H., Barnhill, S., Zhang, Z. & Mills, G. B. (1998) CA 125: the past and the future. Int. J. Biol. Markers 13: 179–187.[Medline]
3. Nicoletto, M. O., Donach, M., De Nicolo, A., Artioli, G., Banna, G. & Monfardini, S. (2001) BRCA-1 and BRCA-2 mutations as prognostic factors in clinical practice and genetic counselling. Cancer Treat. Rev. 27: 295–304.[Medline]
4. Scott, R. J. & Sobol, H. H. (1999) Prognostic implications of cancer susceptibility genes: any news? Recent Results Cancer Res. 151: 71–84.[Medline]
5. Bertario, L., Russo, A., Sala, P., Varesco, L., Giarola, M., Mondini, P., Pierotti, M., Spinelli, P. & Radice, P. (2003) Multiple approach to the exploration of genotype-phenotype correlations in familial adenomatous polyposis. J. Clin. Oncol. 21: 1698–1707.
6. Potischman, N. (2003) Biologic and methodologic issues for nutritional biomarkers. J. Nutr. 133 (suppl. 3): 875S–880S.[Medline]
7. O'Keefe, C. A., Bailey, L. B., Thomas, E. A., Hofler, S. A., Davis, B. A., Cerda, J. J. & Gregory, J. F., 3rd. (1995) Controlled dietary folate affects folate status in nonpregnant women. J. Nutr. 125: 2717–2725.
8. Crews, H., Alink, G., Andersen, R., Braesco, V., Holst, B., Maiani, G., Ovesen, L., Scotter, M., Solfrizzo, M., et al. (2001) A critical assessment of some biomarker approaches linked with dietary intake. Br. J. Nutr. 86 (suppl. 1): S5–S35.
9. Bates, C., Thurnham, D., Bingham, S., Margetts, B. & Nelson, M. (1991) Biochemical markers of nutritional intake. In: Design Concepts in Nutritional Epidemiology (Margetts, B. & Nelson, M., eds.). Oxford University Press, New York, NY.
10. Schatzkin, A., Lanza, E., Freedman, L. S., Tangrea, J., Cooper, M. R., Marshall, J. R., Murphy, P. A., Selby, J. V., Shike, M., et al. (1996) The polyp prevention trial I: rationale, design, recruitment, and baseline participant characteristics. Cancer Epidemiol. Biomarkers Prev. 5: 375–383.
11. Wild, C. P., Andersson, C., O'Brien, N. M., Wilson, L. & Woods, J. A. (2001) A critical evaluation of the application of biomarkers in epidemiological studies on diet and health. Br. J. Nutr. 86 (suppl. 1): S37–S53.
12. Krinsky, N. I. (1993) Actions of carotenoids in biological systems. Annu. Rev. Nutr. 13: 561–587.[Medline]
13. Michaud, D. S., Giovannucci, E. L., Ascherio, A., Rimm, E. B., Forman, M. R., Sampson, L. & Willett, W. C. (1998) Associations of plasma carotenoid concentrations and dietary intake of specific carotenoids in samples of two prospective cohort studies using a new carotenoid database. Cancer Epidemiol. Biomarkers Prev. 7: 283–290.[Abstract]
14. Stahl, W. & Sies, H. (1992) Uptake of lycopene and its geometrical isomers is greater from heat-processed than from unprocessed tomato juice in humans. J. Nutr. 122: 2161–2166.
15. Beecher, G. R. & Khachik, F. (1992) Qualitative relationship of dietary and plasma carotenoids in human beings. Ann. N. Y. Acad. Sci. 669: 320–321.[Medline]
16. Hayashi, Y. (2003) Application of the concepts of risk assessment to the study of amino acid supplements. J. Nutr. 133: 2021S–2024S.
17. Brooks, H. L., Allred, A. J., Beutler, K. T., Coffman, T. M. & Knepper, M. A. (2002) Targeted proteomic profiling of renal Na+ transporter and channel abundances in angiotensin II type 1a receptor knockout mice. Hypertension 39: 470–473.
18. Allen, L. & Wood, R. J. (1994) Calcium and phosphorus. In: Modern Nutrition in Health and Disease, 8th ed. ed. Lea and Febiger, Philadelphia, PA.
19. Tapiero, H., Gate, L. & Tew, K. D. (2001) Iron: deficiencies and requirements. Biomed. Pharmacother. 55: 324–332.[Medline]
20. Sauberlich, H. E., Dowdy, R. P. & Skala, J. H. (1973) Laboratory tests for the assessment of nutritional status. CRC Crit. Rev. Clin. Lab. Sci. 4: 215–340.[Medline]
21. Gupta, S., Bent, S. & Kohlwes, J. (2003) Test characteristics of alpha-fetoprotein for detecting hepatocellular carcinoma in patients with hepatitis C. A systematic review and critical analysis. Ann. Intern. Med. 139: 46–50.
22. Branca, F., Hanley, A. B., Pool-Zobel, B. & Verhagen, H. (2001) Biomarkers in disease and health. Br. J. Nutr. 86 (suppl. 1): S55–S92.
23. Wagener, C. (1997) Molecular diagnostics. J. Mol. Med. 75: 728–744.[Medline]
24. Meijer, A. J. (2003) Amino acids as regulators and components of nonproteinogenic pathways. J. Nutr. 133: 2057S–2062S.
25. Straus, D. S. (1994) Nutritional regulation of hormones and growth factors that control mammalian growth. FASEB (Fed. Am. Soc. Exp. Biol.) J. 8: 6–12.
26. Haussinger, D. (1996) The role of cellular hydration in the regulation of cell function. Biochem. J. 313: 697–710.
27. van Sluijters, D. A., Dubbelhuis, P. F., Blommaart, E. F. & Meijer, A. J. (2000) Amino-acid-dependent signal transduction. Biochem. J. 351: 545–550.
28. Sironi, L., Tremoli, E., Miller, I., Guerrini, U., Calvio, A. M., Eberini, I., Gemeiner, M., Asdente, M., Paoletti, R. & Gianazza, E. (2001) Acute-phase proteins before cerebral ischemia in stroke-prone rats: identification by proteomics. Stroke 32: 753–760.
29. van Eijk, H. M. & Deutz, N. E. (2003) Plasma protein synthesis measurements using a proteomics strategy. J. Nutr. 133: 2084S–2089S.
30. Xu, J., Xiao, G., Trujillo, C., Chang, V., Blanco, L., Joseph, S. B., Bassilian, S., Saad, M. F., Tontonoz, P., et al. (2002) Peroxisome proliferator-activated receptor alpha (PPARalpha) influences substrate utilization for hepatic glucose production. J. Biol. Chem. 277: 50237–50244.
31. Watkins, S. M., Reifsnyder, P. R., Pan, H. J., German, J. B. & Leiter, E. H. (2002) Lipid metabolome-wide effects of the PPARgamma agonist rosiglitazone. J. Lipid Res. 43: 1809–1817.
32. Reo, N. V. (2002) NMR-based metabolomics. Drug Chem. Toxicol. 25: 375–382.[Medline]
33. Jones, P. A. (2003) Epigenetics in carcinogenesis and cancer prevention. Ann. N. Y. Acad. Sci. 983: 213–219.[Medline]
34. Ross, S. A. (2003) Diet and DNA methylation interactions in cancer prevention. Ann. N. Y. Acad. Sci. 983: 197–207.[Medline]
35. Ross, K. S., Carter, H. B., Pearson, J. D. & Guess, H. A. (2000) Comparative efficiency of prostate-specific antigen screening strategies for prostate cancer detection. J. Am. Med. Assoc. 284: 1399–1405.
36. Urban, N., Drescher, C., Etzioni, R. & Colby, C. (1997) Use of a stochastic simulation model to identify an efficient protocol for ovarian cancer screening. Control Clin. Trials 18: 251–270.[Medline]
37. Etzioni, R., Cha, R. & Cowen, M. E. (1999) Serial prostate specific antigen screening for prostate cancer: a computer model evaluates competing strategies. J. Urol. 162: 741–748.[Medline]
38. Coste, J., Cochand-Priollet, B., de Cremoux, P., Le Gales, C., Cartier, I., Molinie, V., Labbe, S., Vacher-Lavenu, M. C. & Vielh, P. (2003) Cross sectional study of conventional cervical smear, monolayer cytology, and human papillomavirus DNA testing for cervical cancer screening. Br. Med. J. 326: 733–737.
39. Kinney, W., Sawaya, G. F., Sung, H. Y., Kearney, K. A., Miller, M. & Hiatt, R. A. (2003) Stage at diagnosis and mortality in patients with adenocarcinoma and adenosquamous carcinoma of the uterine cervix diagnosed as a consequence of cytologic screening. Acta Cytol. 47: 167–171.[Medline]
40. Sullivan Pepe, M., Etzioni, R., Feng, Z., Potter, J. D., Thompson, M. L., Thornquist, M., Winget, M. & Yasui, Y. (2001) Phases of biomarker development for early detection of cancer. J. Natl. Cancer Inst. 93: 1054–1061.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
