SPECIAL REPORT
Conference summary Navigating the Sea of Genomic Data, October 28-29, 2015 Bruce L. Pihlstrom, DDS, MS; Michael L. Barnett, DDS
T
he rapid pace of biomedical discoveries has substantially enhanced our ability to diagnose, treat, and prevent a wide variety of diseases. As a result, practitioners are faced with the challenge of assessing the scientific evidence that supports the use of new technologies and, in the case of commercial products, the validity of marketing claims. Sequencing of the human genome1 has made it possible to understand the etiology, pathogenesis, and risk of developing disease from a genetic perspective and has led to the development of genomic-based diagnostic and risk-assessment tests. To evaluate whether a genetic test may be useful in clinical practice, practitioners need to have a basic understanding of the state of genomic science, as well as its limitations. To assist oral health care professionals in assessing the science of genomics, the American Dental Association and the Task Force on Design and Analysis in Oral Health Research2,3 cosponsored a landmark conference, Navigating the Sea of Genomic Data, held October 28-29, 2015, at the American Dental Association headquarters building in Chicago, IL. The purpose of the conference was to review the basics of genomic science, promote sound design and analysis of genomic studies of oral diseases, and provide a basis or framework to guide practitioners in assessing new developments in genomics and genetic tests for oral diseases. This article summarizes key points and concepts presented by the speakers. Box 1 lists the speakers and the titles of their presentations. FUNDAMENTAL DEFINITIONS AND CONCEPTS
Because much of the terminology used at the conference may be unfamiliar to some readers of The Journal of the American Dental Association, Box 24 provides a glossary of commonly used terms. The term genomics refers to the
ABSTRACT Background. The rapid pace of biomedical discoveries in the past few years has resulted in substantial advances in our ability to diagnose, treat, and prevent a wide variety of diseases. The sequencing of the human genome offered the possibility of understanding the etiology, pathogenesis, and risk of developing disease from a genetic perspective and has resulted, for example, in the development of genomicbased diagnostic or risk-assessment tests for a number of medical and dental conditions. To assess the scientific evidence underlying such tests and determine whether they may be useful in clinical practice, practitioners need to have a basic understanding of the state-of-the-science of genomics and genetic testing. Objective. To assist practitioners in understanding the science of genomics, the American Dental Association and the Task Force on Design and Analysis in Oral Health Research co-sponsored a landmark conference, Navigating the Sea of Genomic Data, held October 28-29, 2015, at the American Dental Association headquarters building in Chicago, IL. The purpose of this conference was to review the basics of genomic science, promote sound design and analysis of genomic studies of oral diseases, and provide a basis or “framework” to guide practitioners in assessing new development in genomics and genetic tests for oral diseases. Overview. Presentations at this conference were made by 9 world-renowned scientists who discussed a wide range of topics involving genomic science, genetic testing for rare mendelian single gene disorders, and genetic testing for assessing the risk of experiencing common complex diseases. This article summarizes the key points and concepts presented by the speakers. Practical Implications. It is essential for oral health care professionals to have a fundamental understanding of genomic science so that they can evaluate new advances in this field and the use of genetic testing for the benefit of their patients. Key Words. Genetic testing; genomics; oral disease. JADA 2016:-(-):--http://dx.doi.org/10.1016/j.adaj.2015.12.016
Copyright ª 2016 American Dental Association. All rights reserved.
JADA
( )
- -
http://jada.ada.org
-
2016 1
SPECIAL REPORT
BOX 1
Conference presentations. Big Data, Genomics, and Reproducible Research (Keynote Speaker): John Ioannidis, MD, DSc, professor of health research and policy, Stanford University, Stanford, CA Emerging Concepts in Understanding Genomics; Genetic Testing in Dentistry: Thomas Hart, DDS, PhD, director, Dr. Anthony Volpe Research Center, American Dental Association Foundation, Gaithersburg, MD Lessons From GWAS Studies: Teri Manolio, MD, PhD, director, Division of Genomic Medicine, Research, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD Predictive and Not: Understanding the Mixed Messages From Our DNA: A. Cecile J.W. Janssens, MA, MSc, PhD, professor, Faculty of Epidemiology, Emory University, Atlanta, GA Statistical Issues in Data Analysis: John Barnard, PhD, head, Section of Statistical Genetics and Bioinformatics, The Cleveland Clinic, Cleveland, OH Epigenetics, RNA, and Missing Heritability: Michael Stitzel, PhD, assistant professor, The Jackson Laboratory of Genomic Medicine, Farmington, CT Future of Genomic and Genetic Testing in Healthcare: Robert Wildin, MD, chief, Genomic Healthcare Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD Genomic Education for Clinicians: Debra Regier, MD, PhD, director of genetic and genomic education, Children’s National Health System, Washington, DC A Strategy for Genetic Testing of Mendelian Dental Defects: James P. Simmer, DDS, PhD, professor of biologic and material sciences, School of Dentistry, University of Michigan, Ann Arbor, MI
entire genetic makeup contained in the chromosomes of an organism (that is, the genome), and the term genetics refers to the effects that genes have on an organism. The genome encompasses all of the genetic material in an organism as expressed in its chromosomes’ DNA. DNA consists of a double helix of repeated pairs of nucleotides or bases (adenine, thymine, guanine, and cytosine) (Figure 1).5 Single-locus variants in the sequences of these four DNA base chemicals are called single nucleotide polymorphisms (SNPs). For example, a SNP is present when some people have a cytosine and others have a thymine nucleotide in its place at a specific site on a chromosome. SNPs occur randomly at approximately every 100 to 300 locations along the sequence of the 3 billion base pairs in human DNA. Of the 20 to 30 million SNPs that have been identified, most appear to have no effect on cellular function, but a relatively small number are associated with disease or response to a drug. GENOMEWIDE ASSOCIATION STUDIES
The fundamental biological event in genetics is DNA recombination between chromosomes as a result of meiotic separation of genetic information passed through generations by means of sexual reproduction. The classic family-based study design involves use of genetic analysis to study families over a few generations and has been used for many years to investigate rare diseases or traits that are inherited as single-gene mendelian traits. However, family studies are less useful for studying complex common diseases in which multiple genes interact in combinations with one another and with the environment. For these diseases, genomewide association studies (GWASs) having larger samples of people are used to investigate DNA variants (SNPs) that have accumulated over many generations. GWASs most often have a case-control design that statistically compares the frequency of each genetic variant (SNP) in a sample of people drawn from a
2 JADA
( )
- -
http://jada.ada.org
-
2016
population having a phenotype (disease or trait) of interest (cases) with the corresponding SNPs in a matched sample of unaffected people (controls). Investigators typically use GWASs to compare hundreds of thousands of SNPs across the entire genome between case and control participants. Conceptually, the statistical analysis of case-control GWASs is relatively simple, involving multiple c2 tests. However, because the statistical tests are repeated hundreds of thousands of times, the likelihood of rejecting the null hypothesis and accepting falsepositive associations of some SNPs with the phenotype is high if a conventional criterion of statistical significance, such as P < .05 or P < .01, is used. Therefore, the level of statistical significance required in GWASs for each SNP’s test is set at P < 108. This stringent criterion for each test is used so that the likelihood of accepting any falsepositive association in a particular GWAS will be no more than P < .05. As in any valid case-control study, careful matching of case and control participants is essential, and it is important that the disease or trait (phenotype) is defined precisely and, for multicenter studies, standardized among enrollment sites. It also is critical that in a given study, the case and control participants have the same ethnic ancestry to avoid spurious associations that can result from the presence in different ethnic groups of SNPs that may be unrelated to the condition of interest. Before GWAS results can be accepted as applicable to the general population, they must be replicated in independent studies of populations having different ancestries. Many GWASs must involve thousands of people to find statistically significant associations between diseases or traits and SNPs. Meta-analyses and consortia of investigators often are used to pool data and achieve these
ABBREVIATION KEY. GWAS: Genomewide association study. SNP: Single nucleotide polymorphism.
SPECIAL REPORT
BOX 2
Glossary of terms.* Agnostic GWAS†: An agnostic GWAS is one in which no SNP‡ is thought, in advance (a priori), to have a higher probability of being associated with the phenotype of interest than any other SNP. Candidate gene: A gene located in a chromosome region suspected of being involved in a disease or disorder. Complex trait or Complex disorder: A trait or disorder that has a genetic component that does not follow strict mendelian inheritance but involves the interaction of multiple genes or gene-environment interactions. Epigenetics: The mechanisms that alter how genes are expressed without altering the underlying DNA nucleotide sequence. Genetics: The study of inheritance patterns of specific traits. Genome: All the genetic material in the chromosomes of a particular organism; its size is generally given as its total number of base pairs. Genome sequence: Order of nucleotides or bases within DNA molecules that make up an organism’s entire genome. The 4 bases are adenine, guanine, cytosine, and thymine, represented as A, G, C, and T, respectively. Genomewide association study (GWAS): A study in which the investigators evaluate the statistical association of genetic variation with outcomes or traits of interest by using 100,000 to 1,000,000 or more SNP markers across the genome. Genomics: The study all the genetic material in the chromosomes of a particular organism. Genotype: The genetic constitution of an organism as distinguished from its physical appearance (its phenotype). Phenotype: The physical characteristics of an organism. Single-gene disorder: Hereditary disorder caused by a mutant allele of a single gene (for example, Duchenne muscular dystrophy, retinoblastoma, or sickle cell disease). Single nucleotide polymorphism (SNP): DNA sequence variations that occur when a single nucleotide (adenine, guanine, cytosine, or thymine) in the genome sequence is altered. Trait: A phenotypic characteristic. Whole genome sequencing, complete genome sequencing, or entire genome sequencing: A laboratory process that determines the order of nucleotides or bases within DNA molecules of an entire organism’s genome. * Source: Some of the information and definitions have been derived in part from the US National Library of Medicine, National Institutes of Health, Department of Health and Human Services.4 † GWAS: Genomewide association study. ‡ SNP: Single nucleotide polymorphism.
large sample sizes.6 However, if data are not standardized properly, the usefulness of pooling data from many investigators may be limited by several factors, including the heterogeneity of the population samples, failure to define case phenotypes rigorously, and the use by investigators of different laboratory methods (that is, SNP chips or platforms) for assaying SNPs. Investigators typically present the results of GWASs as Manhattan plots in which the locations of genetic variants (SNPs) in the chromosomes are plotted on the horizontal axis against the P values of the SNP associations, with the phenotype of interest on the vertical axis, producing a display that resembles the skyline of Manhattan (Figure 2).7 It is important to emphasize that finding an association between a given SNP and a disease does not necessarily implicate the SNP in the causation of that disease but rather may suggest the possibility of an etiologic role for 1 or more genes that are located near the SNP in the strand of DNA. Previously, investigators thought that the occurrence of some common, complex diseases could be predicted using just a few so-called candidate genes encoding for proteins that were related biologically to disease pathogenesis. However, investigators in large agnostic GWASs have validated only approximately 1.2% of previously identified candidate genes. Moreover, GWAS investigators generally report small magnitudes of association (odds ratios) for most genetic variants (SNPs) associated with a given phenotype. GWASs having large
sample sizes (thousands of participants) increase the likelihood of finding associations of SNPs with common, complex diseases, but the strength of the associations found in such studies is often small. The expectation that a common genetic variant can affect one’s phenotype directly may be overly optimistic given the myriad regulatory or modifying processes in the cascade of events that occur from the start of gene transcription until final protein synthesis. In complex polygenic diseases, multiple genes act in combinations with one another and with the environment. The environment includes many endogenous and exogenous influences, such as antibiotics, drugs, bacteria, and viruses that could influence gene expression. Therefore, it is unrealistic to expect that GWASs necessarily will result in genetic tests that can be used to predict disease in a manner identical to that of well-known mendelian single gene diseases. Nevertheless, GWAS investigators have identified many SNPs that may be associated with complex diseases. In the 10 years since this technology became available, investigators in more than 2,300 articles have reported more than 17,000 statistically significant associations.8 Many of these SNPS are located in genes previously unsuspected of being related to the associated disease phenotypes, and nearly one-half are in regions of the genome containing no protein-encoding nucleotides. A catalog of statistically significant GWAS findings accessible online is provided jointly by the National
JADA
( )
- -
http://jada.ada.org
-
2016 3
SPECIAL REPORT
Figure 1. DNA double helix. Reproduced from the US National Library of Medicine, National Institutes of Health, Department of Health and Human Services.5
Human Genome Research Institute and the European Bioinformatics Institute.8 Many of the SNPs found in GWASs appear to be associated with multiple diseases, and their identification is shedding new light on the pathophysiology of complex diseases, as well as defining disease phenotypes better. GWAS investigators also are identifying promising targets for drug development. WHOLE GENOME SEQUENCING
In contrast to GWASs, which investigators use to assess common SNPs that occur once in approximately every 100 to 300 nucleotide base pairs in the genome, whole genome sequencing assesses the sequence of nucleotides in a person’s entire genome. Whole genome sequencing studies might be more effective than GWASs in helping to identify genetic markers of disease risk because investigators can use them to assess both common and rare genetic variants. In general, results of such studies have been variable. Occasionally, however, single markers or combinations of markers have been clinically useful, such as the BRCA gene mutations for predicting breast cancer risk. However, for most genetic variants detected, particularly as related to oral diseases, there has been considerable disagreement as to their clinical relevance and usefulness in routine practice. EPIGENETICS
Epigenetics refers to mechanisms that alter how genes are expressed without altering the underlying DNA nucleotide sequence.9 One can think of the DNA that codes for
4 JADA
( )
- -
http://jada.ada.org
-
2016
genetic traits as genomic hardware, and the epigenome is the software that controls the function of the hardware influencing the how, what, and when of the DNA genetic code expression. Moreover, environmental factors can modulate epigenomic chemicals and influence the phenotypic expression of DNA. An excellent primer of epigenetic mechanisms was published in The Scientist,10 illustrating how DNA is modified by methylation, in which a methyl group is added to DNA, or by histones, which are proteins that organize DNA into more complex structures called nucleosomes. Epigenetic modification does not change the DNA nucleotide sequence but rather modifies or regulates gene expression—for example, differentiation of stem cells into various cell types such as liver, brain, or gingival cells. Epigenetic modification also may regulate how the environment influences gene expression, such as activation of immune cells during an infection.11 Understanding how these “software instructions” or epigenetic mechanisms regulate our DNA “hardware” in interactions with ribonucleic acid, protein synthesis, cellular function, and the environment is being pursued by huge consortia of researchers, such as those involved in the International Human Genome Consortium.12 Epigenetic research also has focused on specific cells of oral tissues that are involved in inflammation and how these cells interact with environmental influences such as smoking, bacteria, nutrition, and oral hygiene. This focus is likely to be an exciting and productive future research area. TRANSLATION OF GENOMICS INTO CLINICAL PRACTICE
Translating genomics into clinical practice for assessing disease risk involves finding a genomic association that indicates increased disease risk, determining whether the finding is useful for separating patients who will benefit from an intervention from those who will not, and assessing whether patient outcomes improve as a result of using the genomic information. There are instances in which the addition of genomic information to disease risk assessment can be deceptive and of little value. For example, adding a genetic variant to a risk assessment that includes early clinical signs of a disease will do little to improve the ability to predict whether later stages of the disease will develop. Similarly, if the common clinical or behavioral risk factors for a disease are known and included in a risk assessment model, the addition of genetic information will do little to enhance predictive ability. A challenge in translating genomic findings to patient care is the ability to communicate accurate and reliable information to health care professionals and patients. Traditional teaching methods in education for health care professionals are not necessarily the most effective aid to learning in today’s environment. A concept in
SPECIAL REPORT
8
P = 4.8 × 10–7
–LOG10(P)
6
4
2
0 0
50,000
100,000
SINGLE NUCLEOTIDE POLYMORPHISMS Figure 2. Manhattan plot of statistical probability (P values) of a genomewide association scan for genes that affect the risk of developing age-related macular degeneration. Vertical dotted lines show chromosomal boundaries. The dotted horizontal line shows the cutoff for P ¼ .05 after correction for multiple comparisons. The arrow indicates the peak for SNP rs380390, the most significant association with age-related macular degeneration. log: Logarithm. Adapted with permission of the publisher from Klein and colleagues.7
education is that instructional methods are based best on less teaching and more on problem-based participatory learning. Because information gleaned during formal learning sessions about a topic such as genomics or genetic testing may not be retained for long periods, it is useful to have a means of obtaining information conveniently at the time it is actually needed. An example of a useful resource is the Genetic Home Reference website13 that can be accessed using a mobile device and provides a way of rapidly obtaining information during, for example, a clinic session when time is limited. GENETIC TESTING—AN OVERVIEW
The genetic basis of a given trait or disorder (phenotype) may be due to mendelian transmission of specific gene mutations, chromosomal anomalies, or complex interactions between genes and environmental factors. Many health care providers do not distinguish between single gene (mendelian) inheritance of rare diseases and
complex patterns of inheritance that involve many genes, variable gene expression, and interaction between the genes and the environment. Identification of specific gene and chromosomal variations has been used for many years to diagnose mendelian disorders such as amelogenesis imperfecta and chromosomal disorders such as Down syndrome. However, mendelian or chromosomal inheritance and methods of diagnosing these diseases or disorders must not be confused with use of genomics in the prediction, prevention, or treatment of diseases that have complex etiologies involving both genetic and environmental influences. The genetic component of polygenic diseases is not causative but rather is indicative of a small increase in the risk of developing disease. Whether a person will develop a disease is a function of the interaction between the genome and environmental factors. Diseases having mendelian patterns of inheritance involving 1 or a few genes affect a small fraction
JADA
( )
- -
http://jada.ada.org
-
2016 5
SPECIAL REPORT
of the population, have high heritability, and can be predicted with a high degree of accuracy. In contrast, diseases having low heritability, genetic complexity involving many genes, and strong environmental influences have poor predictability. Genetic tests are in use for diagnosis, disease prediction, determination of carrier status, prenatal and preimplantation testing, newborn testing, prediction of therapeutic response, and forensic analyses. Factors that health care professionals should consider when evaluating any genetic test include the following: - What is the type of genetic variation being tested: single gene mutation, multiple genes, a single genetic variant (SNP), or multiple SNPs? - Is the test for a rare disease caused by a single gene mutation and inherited in a mendelian pattern, or is the disease common and likely to have a complex etiology? - Does the test look for mendelian patterns of inheritance, or does it use GWASs to look for associations of genetic variants and common, complex diseases? - Is the test intended for research or clinical use? - Does the test have analytic validity—that is, are the laboratory assays precise, accurate, and reproducible? Can the test accurately detect whether a specific genetic variant is present or absent?13 - Does the test have clinical validity—that is, is the test clinically meaningful? How well does the genetic variant being tested relate to the presence, absence, or risk of developing a specific disease?13 - Does the test have clinical usefulness—that is, will having the genetic information change patient treatment or the clinical outcome? Does the test provide information about the diagnosis, treatment, management, or prevention of disease that will be helpful to the consumer?13 In summary, the purpose for which a genetic test is used must be defined clearly. Tests for mendelian disorders are used for the diagnosis or prediction of disease occurrence. These disorders are based on a single gene or a small subset of genes, and the test results are generally definitive and predictive. People with mutations in certain genes either will develop the disease (for example, Huntington’s disease) or will have a high probability of developing the disease (for example, breast cancer, colorectal cancer). Tests for common complex diseases involving multiple genes (polygenic) are not diagnostic but may be only indicative of small increases in the risk of developing a disease. To date, most genetic tests to assess the risk of developing complex common diseases, including periodontal disease and dental caries, have not been clinically useful. Genetic testing of rare mendelian dental disorders. Inherited dental disorders can be isolated and affect only the teeth (nonsyndromic) or can occur as a part of a syndrome with other diseases or tissue malformations.
6 JADA
( )
- -
http://jada.ada.org
-
2016
Syndromic conditions may occur simultaneously with or subsequent to the clinical appearance of dental malformations. Genetic diagnosis of these disorders is essential to determine whether the cause is a gene mutation that affects only the teeth or whether it is a mutation associated with a syndrome that involves other tissues as well. Determination of the genetic basis allows health care professionals and patients to minimize or deal with the potentially serious syndromic manifestations that may accompany mendelian tooth disorders. For example, with tooth agenesis (hypodontia), dentinal and enamel defects can be isolated inherited defects or also can be part of a syndrome. A mutation of the AXIN2 gene causes severe oligodontia, but mutations of this gene also can be associated with colorectal cancer or precancerous lesions of various types.14 Identification of this gene mutation in people who have many congenitally missing teeth could provide an early warning of predilection for colorectal cancer and the need for frequent colorectal cancer screening. Moreover, autosomal forms of DSPP and COLIA1 or COLIA2 gene mutations are present in most patients having dentinal disorders. Testing for the presence of these gene mutations is available at some academic institutions. People having a mutation in the DSPP gene have clinical manifestations that are limited to defects in dentin, but those having mutations in COLIA1 or COLIA2 have defects in other collagen tissues that cause bone fragility, as in osteogenesis imperfecta. It is impossible to tell the differences between these 2 conditions by examining the teeth; genetic testing is important to establish whether the dentin defect is caused by mutations in the DSPP gene or by genes that cause collagen tissue disorders in other organs. Genetic testing for common complex oral diseases. Rapid advances in genetic science have led to unrealistic expectations among health care professionals, patients, and the public, fostered by companies that offer genetic testing under the pretext of precision medicine and personalized care. Because a person’s DNA can be sequenced, it does not necessarily mean that it can be used to predict, prevent, diagnose, and guide treatment for common, complex diseases. Genetic tests offered to the public that claim to predict the risk of developing complex diseases often have limitations. For example, results from commercially available genetic tests involving saliva may indicate an elevated risk of developing disease on the basis of small increases in odds ratios considerably lower than the 2.0 often considered to be the threshold for clinically significant increased disease risk. Moreover, some genetic tests marketed to the public for predicting disease risk have been based on findings obtained from populations having a specific ethnic ancestry. Use of such a test would be of questionable value to people from other ethnic populations.
SPECIAL REPORT
Some companies offer genetic tests for common, complex oral diseases. However, there is no genetic test yet that can be used reliably to predict or assess risk of developing any complex oral disease, such as dental caries or periodontitis. Moreover, given the complexity of the human genome and its interaction with the environment, it is unlikely that genomic studies of association will lead to a meaningful predictive genetic test for the occurrence or progression of these diseases. CONCLUDING REMARKS
At the conclusion of the conference, co-chair Dr. Michael Glick shared key points that are important for practicing clinicians: - In common, complex diseases such as dental caries and periodontal disease, multiple genes interact with one another and with the environment in such a way that the genetic component of these diseases is not causative but may be indicative of a small increase in the risk of developing the disease. - The clinical usefulness or purpose of any genetic test must be considered, not only whether it is a valid and reproducible laboratory test. - The usefulness of genetic and genomic tests for diagnosing disease is different than the usefulness of such tests for predicting disease risk or disease progression. - Health care providers who use genetic and genomic tests need to be able to understand the meaning of the tests and discuss the findings, clinical implications, and test limitations with patients. - There are few genetic markers for common, complex diseases that have been validated by means of GWASs or replicated studies. - Finding SNPs that are associated with a disease or a disease biomarker does not necessarily imply causation or progression of the phenotypic expression of the disease. - GWASs having larger sample sizes may not be clinically informative because they likely will find additional associations of disease and genetic markers that have small effect sizes. - Samples of oral tissue may be better than blood for the genetic and genomic study of common and complex oral diseases. - The expression of our DNA code involves a cascade of many regulatory and modifying events during the process of gene expression as the phenotype. - There may be a need for a policy or professional guidelines for the clinical use of genetic and genomic information. - There is a need to explore the ethical issues of using genomic information for disease prediction, prevention, diagnosis, and therapy. n
Dr. Pihlstrom is a professor emeritus, School of Dentistry, University of Minnesota, Minneapolis, MN, and the associate editor for research, The Journal of the American Dental Association, Chicago, IL. Address correspondence to Dr. Pihlstrom at 4801 Fairmont Ave., Suite 902, Bethesda, MD 20814, e-mail [email protected]. Dr. Barnett is a clinical professor, School of Dental Medicine, University at Buffalo, State University of New York, Buffalo, NY. Disclosure. Dr. Pihlstrom and Dr. Barnett are members of the Task Force on Design and Analysis in Oral Health Research, which provided financial support for the conference. The conference was co-chaired by Dr. Sebastian Ciancio, executive director of the Task Force on Design and Analysis in Oral Health Research; Dr. Daniel Meyer, who was the chief science officer, American Dental Association, when the conference was held; and Dr. Michael Glick, editor of The Journal of the American Dental Association. Any errors of fact or interpretation of conference proceedings are the full responsibility of the authors. However, this summary would not have been possible without the encouragement, constructive criticism, and enthusiastic support of the following people who attended the conference and spent many hours reviewing drafts of this summary: Al Best, PhD; Sebastian Ciancio, DDS; Deborah Dawson, PhD; Michael Glick, DMD; Peter Imrey, PhD; Bryan Michalowicz, DDS, MS; and Malcom Snead, DDS, PhD. 1. Venter JC, Adams MD, Myers EW, et al. The sequence of the human genome. Science. 2001;291(5507):1304-1351. 2. Kingman A, Imrey PB, Pihlstrom BL, Zimmerman SO. Chilton, Fertig, Fleiss, and the Task Force on Design and Analysis in Dental and Oral Research. J Dent Res. 1997;76(6):1239-1243. 3. The Task Force on Design and Analysis in Oral Health Research. Task Force website. Available at: http://taskforceondesign.org/. Accessed November 12, 2015. 4. US National Library of Medicine, National Institutes of Health, Department of Health and Human Services. Genetics home reference. Available at: http://ghr.nlm.nih.gov/glossary=humanhenomeproject. Accessed January 15, 2016. 5. US National Library of Medicine, National Institutes of Health, Department of Health and Human Services. GeneEd: Genetics, Education, Discovery. Structure of the double helix. Available at: http://geneed.nlm. nih.gov/topic_subtopic.php?tid=15&sid=16. Accessed January 13, 2016. 6. Evangelou E, Ioannidis JP. Meta-analysis methods for genomewide association studies and beyond. Nat Rev Genet. 2013;14(6): 379-389. 7. Klein RJ, Zeiss C, Chew EY, et al. Complement factor H polymorphism in age-related macular degeneration. Science. 2005;308(5720): 385-389. 8. Burdett T (EBI), Hall PN (NHGRI), Hasting E (EBI), Hindorff LA (NHGRI), Junkins HA (NHGRI), Klemm AK (NHGRI), MacArthur J (EBI), Manolio TA (NHGRI), Morales J (EBI), Parkinson H (EBI) and Welter D (EBI). The NHGRI-EBI Catalog of published genome-wide association studies. Available at: www.ebi.ac.uk/gwas. Accessed January 15, 2016. 9. National Human Genome Research Institute. Epigenomics. Available at: https://www.genome.gov/27532724. Accessed November 6, 2015. 10. Kubicek S. Epigenetics: a primer. Available at: http://images.thescientist.com/content/images/articles/58007/epigenetics_primer.jpg. Accessed January 3, 2016. 11. Obata Y, Furusawa Y, Hase K. Epigenetic modifications of the immune system in health and disease. Immunol Cell Biol. 2015;93(3):226-232. 12. National Institutes of Health. International Human Epigenome Consortium. Available at: http://ihec-epigenomes.org. Accessed November 7, 2015. 13. US National Library of Medicine, National Institutes of Health, Department of Health and Human Services. Genetics home reference. Available at: http://ghr.nlm.nih.gov/. Accessed December 15, 2015. 14. Lammi L, Arte S, Somer M, et al. Mutations in AXIN2 cause familial tooth agenesis and predispose to colorectal cancer. Am J Hum Genet. 2004; 74(5):1043-1050.
JADA
( )
- -
http://jada.ada.org
-
2016 7
Conference summary Navigating the Sea of Genomic Data, October 28-29, 2015 Bruce L. Pihlstrom, DDS, MS; Michael L. Barnett, DDS
T
he rapid pace of biomedical discoveries has substantially enhanced our ability to diagnose, treat, and prevent a wide variety of diseases. As a result, practitioners are faced with the challenge of assessing the scientific evidence that supports the use of new technologies and, in the case of commercial products, the validity of marketing claims. Sequencing of the human genome1 has made it possible to understand the etiology, pathogenesis, and risk of developing disease from a genetic perspective and has led to the development of genomic-based diagnostic and risk-assessment tests. To evaluate whether a genetic test may be useful in clinical practice, practitioners need to have a basic understanding of the state of genomic science, as well as its limitations. To assist oral health care professionals in assessing the science of genomics, the American Dental Association and the Task Force on Design and Analysis in Oral Health Research2,3 cosponsored a landmark conference, Navigating the Sea of Genomic Data, held October 28-29, 2015, at the American Dental Association headquarters building in Chicago, IL. The purpose of the conference was to review the basics of genomic science, promote sound design and analysis of genomic studies of oral diseases, and provide a basis or framework to guide practitioners in assessing new developments in genomics and genetic tests for oral diseases. This article summarizes key points and concepts presented by the speakers. Box 1 lists the speakers and the titles of their presentations. FUNDAMENTAL DEFINITIONS AND CONCEPTS
Because much of the terminology used at the conference may be unfamiliar to some readers of The Journal of the American Dental Association, Box 24 provides a glossary of commonly used terms. The term genomics refers to the
ABSTRACT Background. The rapid pace of biomedical discoveries in the past few years has resulted in substantial advances in our ability to diagnose, treat, and prevent a wide variety of diseases. The sequencing of the human genome offered the possibility of understanding the etiology, pathogenesis, and risk of developing disease from a genetic perspective and has resulted, for example, in the development of genomicbased diagnostic or risk-assessment tests for a number of medical and dental conditions. To assess the scientific evidence underlying such tests and determine whether they may be useful in clinical practice, practitioners need to have a basic understanding of the state-of-the-science of genomics and genetic testing. Objective. To assist practitioners in understanding the science of genomics, the American Dental Association and the Task Force on Design and Analysis in Oral Health Research co-sponsored a landmark conference, Navigating the Sea of Genomic Data, held October 28-29, 2015, at the American Dental Association headquarters building in Chicago, IL. The purpose of this conference was to review the basics of genomic science, promote sound design and analysis of genomic studies of oral diseases, and provide a basis or “framework” to guide practitioners in assessing new development in genomics and genetic tests for oral diseases. Overview. Presentations at this conference were made by 9 world-renowned scientists who discussed a wide range of topics involving genomic science, genetic testing for rare mendelian single gene disorders, and genetic testing for assessing the risk of experiencing common complex diseases. This article summarizes the key points and concepts presented by the speakers. Practical Implications. It is essential for oral health care professionals to have a fundamental understanding of genomic science so that they can evaluate new advances in this field and the use of genetic testing for the benefit of their patients. Key Words. Genetic testing; genomics; oral disease. JADA 2016:-(-):--http://dx.doi.org/10.1016/j.adaj.2015.12.016
Copyright ª 2016 American Dental Association. All rights reserved.
JADA
( )
- -
http://jada.ada.org
-
2016 1
SPECIAL REPORT
BOX 1
Conference presentations. Big Data, Genomics, and Reproducible Research (Keynote Speaker): John Ioannidis, MD, DSc, professor of health research and policy, Stanford University, Stanford, CA Emerging Concepts in Understanding Genomics; Genetic Testing in Dentistry: Thomas Hart, DDS, PhD, director, Dr. Anthony Volpe Research Center, American Dental Association Foundation, Gaithersburg, MD Lessons From GWAS Studies: Teri Manolio, MD, PhD, director, Division of Genomic Medicine, Research, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD Predictive and Not: Understanding the Mixed Messages From Our DNA: A. Cecile J.W. Janssens, MA, MSc, PhD, professor, Faculty of Epidemiology, Emory University, Atlanta, GA Statistical Issues in Data Analysis: John Barnard, PhD, head, Section of Statistical Genetics and Bioinformatics, The Cleveland Clinic, Cleveland, OH Epigenetics, RNA, and Missing Heritability: Michael Stitzel, PhD, assistant professor, The Jackson Laboratory of Genomic Medicine, Farmington, CT Future of Genomic and Genetic Testing in Healthcare: Robert Wildin, MD, chief, Genomic Healthcare Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD Genomic Education for Clinicians: Debra Regier, MD, PhD, director of genetic and genomic education, Children’s National Health System, Washington, DC A Strategy for Genetic Testing of Mendelian Dental Defects: James P. Simmer, DDS, PhD, professor of biologic and material sciences, School of Dentistry, University of Michigan, Ann Arbor, MI
entire genetic makeup contained in the chromosomes of an organism (that is, the genome), and the term genetics refers to the effects that genes have on an organism. The genome encompasses all of the genetic material in an organism as expressed in its chromosomes’ DNA. DNA consists of a double helix of repeated pairs of nucleotides or bases (adenine, thymine, guanine, and cytosine) (Figure 1).5 Single-locus variants in the sequences of these four DNA base chemicals are called single nucleotide polymorphisms (SNPs). For example, a SNP is present when some people have a cytosine and others have a thymine nucleotide in its place at a specific site on a chromosome. SNPs occur randomly at approximately every 100 to 300 locations along the sequence of the 3 billion base pairs in human DNA. Of the 20 to 30 million SNPs that have been identified, most appear to have no effect on cellular function, but a relatively small number are associated with disease or response to a drug. GENOMEWIDE ASSOCIATION STUDIES
The fundamental biological event in genetics is DNA recombination between chromosomes as a result of meiotic separation of genetic information passed through generations by means of sexual reproduction. The classic family-based study design involves use of genetic analysis to study families over a few generations and has been used for many years to investigate rare diseases or traits that are inherited as single-gene mendelian traits. However, family studies are less useful for studying complex common diseases in which multiple genes interact in combinations with one another and with the environment. For these diseases, genomewide association studies (GWASs) having larger samples of people are used to investigate DNA variants (SNPs) that have accumulated over many generations. GWASs most often have a case-control design that statistically compares the frequency of each genetic variant (SNP) in a sample of people drawn from a
2 JADA
( )
- -
http://jada.ada.org
-
2016
population having a phenotype (disease or trait) of interest (cases) with the corresponding SNPs in a matched sample of unaffected people (controls). Investigators typically use GWASs to compare hundreds of thousands of SNPs across the entire genome between case and control participants. Conceptually, the statistical analysis of case-control GWASs is relatively simple, involving multiple c2 tests. However, because the statistical tests are repeated hundreds of thousands of times, the likelihood of rejecting the null hypothesis and accepting falsepositive associations of some SNPs with the phenotype is high if a conventional criterion of statistical significance, such as P < .05 or P < .01, is used. Therefore, the level of statistical significance required in GWASs for each SNP’s test is set at P < 108. This stringent criterion for each test is used so that the likelihood of accepting any falsepositive association in a particular GWAS will be no more than P < .05. As in any valid case-control study, careful matching of case and control participants is essential, and it is important that the disease or trait (phenotype) is defined precisely and, for multicenter studies, standardized among enrollment sites. It also is critical that in a given study, the case and control participants have the same ethnic ancestry to avoid spurious associations that can result from the presence in different ethnic groups of SNPs that may be unrelated to the condition of interest. Before GWAS results can be accepted as applicable to the general population, they must be replicated in independent studies of populations having different ancestries. Many GWASs must involve thousands of people to find statistically significant associations between diseases or traits and SNPs. Meta-analyses and consortia of investigators often are used to pool data and achieve these
ABBREVIATION KEY. GWAS: Genomewide association study. SNP: Single nucleotide polymorphism.
SPECIAL REPORT
BOX 2
Glossary of terms.* Agnostic GWAS†: An agnostic GWAS is one in which no SNP‡ is thought, in advance (a priori), to have a higher probability of being associated with the phenotype of interest than any other SNP. Candidate gene: A gene located in a chromosome region suspected of being involved in a disease or disorder. Complex trait or Complex disorder: A trait or disorder that has a genetic component that does not follow strict mendelian inheritance but involves the interaction of multiple genes or gene-environment interactions. Epigenetics: The mechanisms that alter how genes are expressed without altering the underlying DNA nucleotide sequence. Genetics: The study of inheritance patterns of specific traits. Genome: All the genetic material in the chromosomes of a particular organism; its size is generally given as its total number of base pairs. Genome sequence: Order of nucleotides or bases within DNA molecules that make up an organism’s entire genome. The 4 bases are adenine, guanine, cytosine, and thymine, represented as A, G, C, and T, respectively. Genomewide association study (GWAS): A study in which the investigators evaluate the statistical association of genetic variation with outcomes or traits of interest by using 100,000 to 1,000,000 or more SNP markers across the genome. Genomics: The study all the genetic material in the chromosomes of a particular organism. Genotype: The genetic constitution of an organism as distinguished from its physical appearance (its phenotype). Phenotype: The physical characteristics of an organism. Single-gene disorder: Hereditary disorder caused by a mutant allele of a single gene (for example, Duchenne muscular dystrophy, retinoblastoma, or sickle cell disease). Single nucleotide polymorphism (SNP): DNA sequence variations that occur when a single nucleotide (adenine, guanine, cytosine, or thymine) in the genome sequence is altered. Trait: A phenotypic characteristic. Whole genome sequencing, complete genome sequencing, or entire genome sequencing: A laboratory process that determines the order of nucleotides or bases within DNA molecules of an entire organism’s genome. * Source: Some of the information and definitions have been derived in part from the US National Library of Medicine, National Institutes of Health, Department of Health and Human Services.4 † GWAS: Genomewide association study. ‡ SNP: Single nucleotide polymorphism.
large sample sizes.6 However, if data are not standardized properly, the usefulness of pooling data from many investigators may be limited by several factors, including the heterogeneity of the population samples, failure to define case phenotypes rigorously, and the use by investigators of different laboratory methods (that is, SNP chips or platforms) for assaying SNPs. Investigators typically present the results of GWASs as Manhattan plots in which the locations of genetic variants (SNPs) in the chromosomes are plotted on the horizontal axis against the P values of the SNP associations, with the phenotype of interest on the vertical axis, producing a display that resembles the skyline of Manhattan (Figure 2).7 It is important to emphasize that finding an association between a given SNP and a disease does not necessarily implicate the SNP in the causation of that disease but rather may suggest the possibility of an etiologic role for 1 or more genes that are located near the SNP in the strand of DNA. Previously, investigators thought that the occurrence of some common, complex diseases could be predicted using just a few so-called candidate genes encoding for proteins that were related biologically to disease pathogenesis. However, investigators in large agnostic GWASs have validated only approximately 1.2% of previously identified candidate genes. Moreover, GWAS investigators generally report small magnitudes of association (odds ratios) for most genetic variants (SNPs) associated with a given phenotype. GWASs having large
sample sizes (thousands of participants) increase the likelihood of finding associations of SNPs with common, complex diseases, but the strength of the associations found in such studies is often small. The expectation that a common genetic variant can affect one’s phenotype directly may be overly optimistic given the myriad regulatory or modifying processes in the cascade of events that occur from the start of gene transcription until final protein synthesis. In complex polygenic diseases, multiple genes act in combinations with one another and with the environment. The environment includes many endogenous and exogenous influences, such as antibiotics, drugs, bacteria, and viruses that could influence gene expression. Therefore, it is unrealistic to expect that GWASs necessarily will result in genetic tests that can be used to predict disease in a manner identical to that of well-known mendelian single gene diseases. Nevertheless, GWAS investigators have identified many SNPs that may be associated with complex diseases. In the 10 years since this technology became available, investigators in more than 2,300 articles have reported more than 17,000 statistically significant associations.8 Many of these SNPS are located in genes previously unsuspected of being related to the associated disease phenotypes, and nearly one-half are in regions of the genome containing no protein-encoding nucleotides. A catalog of statistically significant GWAS findings accessible online is provided jointly by the National
JADA
( )
- -
http://jada.ada.org
-
2016 3
SPECIAL REPORT
Figure 1. DNA double helix. Reproduced from the US National Library of Medicine, National Institutes of Health, Department of Health and Human Services.5
Human Genome Research Institute and the European Bioinformatics Institute.8 Many of the SNPs found in GWASs appear to be associated with multiple diseases, and their identification is shedding new light on the pathophysiology of complex diseases, as well as defining disease phenotypes better. GWAS investigators also are identifying promising targets for drug development. WHOLE GENOME SEQUENCING
In contrast to GWASs, which investigators use to assess common SNPs that occur once in approximately every 100 to 300 nucleotide base pairs in the genome, whole genome sequencing assesses the sequence of nucleotides in a person’s entire genome. Whole genome sequencing studies might be more effective than GWASs in helping to identify genetic markers of disease risk because investigators can use them to assess both common and rare genetic variants. In general, results of such studies have been variable. Occasionally, however, single markers or combinations of markers have been clinically useful, such as the BRCA gene mutations for predicting breast cancer risk. However, for most genetic variants detected, particularly as related to oral diseases, there has been considerable disagreement as to their clinical relevance and usefulness in routine practice. EPIGENETICS
Epigenetics refers to mechanisms that alter how genes are expressed without altering the underlying DNA nucleotide sequence.9 One can think of the DNA that codes for
4 JADA
( )
- -
http://jada.ada.org
-
2016
genetic traits as genomic hardware, and the epigenome is the software that controls the function of the hardware influencing the how, what, and when of the DNA genetic code expression. Moreover, environmental factors can modulate epigenomic chemicals and influence the phenotypic expression of DNA. An excellent primer of epigenetic mechanisms was published in The Scientist,10 illustrating how DNA is modified by methylation, in which a methyl group is added to DNA, or by histones, which are proteins that organize DNA into more complex structures called nucleosomes. Epigenetic modification does not change the DNA nucleotide sequence but rather modifies or regulates gene expression—for example, differentiation of stem cells into various cell types such as liver, brain, or gingival cells. Epigenetic modification also may regulate how the environment influences gene expression, such as activation of immune cells during an infection.11 Understanding how these “software instructions” or epigenetic mechanisms regulate our DNA “hardware” in interactions with ribonucleic acid, protein synthesis, cellular function, and the environment is being pursued by huge consortia of researchers, such as those involved in the International Human Genome Consortium.12 Epigenetic research also has focused on specific cells of oral tissues that are involved in inflammation and how these cells interact with environmental influences such as smoking, bacteria, nutrition, and oral hygiene. This focus is likely to be an exciting and productive future research area. TRANSLATION OF GENOMICS INTO CLINICAL PRACTICE
Translating genomics into clinical practice for assessing disease risk involves finding a genomic association that indicates increased disease risk, determining whether the finding is useful for separating patients who will benefit from an intervention from those who will not, and assessing whether patient outcomes improve as a result of using the genomic information. There are instances in which the addition of genomic information to disease risk assessment can be deceptive and of little value. For example, adding a genetic variant to a risk assessment that includes early clinical signs of a disease will do little to improve the ability to predict whether later stages of the disease will develop. Similarly, if the common clinical or behavioral risk factors for a disease are known and included in a risk assessment model, the addition of genetic information will do little to enhance predictive ability. A challenge in translating genomic findings to patient care is the ability to communicate accurate and reliable information to health care professionals and patients. Traditional teaching methods in education for health care professionals are not necessarily the most effective aid to learning in today’s environment. A concept in
SPECIAL REPORT
8
P = 4.8 × 10–7
–LOG10(P)
6
4
2
0 0
50,000
100,000
SINGLE NUCLEOTIDE POLYMORPHISMS Figure 2. Manhattan plot of statistical probability (P values) of a genomewide association scan for genes that affect the risk of developing age-related macular degeneration. Vertical dotted lines show chromosomal boundaries. The dotted horizontal line shows the cutoff for P ¼ .05 after correction for multiple comparisons. The arrow indicates the peak for SNP rs380390, the most significant association with age-related macular degeneration. log: Logarithm. Adapted with permission of the publisher from Klein and colleagues.7
education is that instructional methods are based best on less teaching and more on problem-based participatory learning. Because information gleaned during formal learning sessions about a topic such as genomics or genetic testing may not be retained for long periods, it is useful to have a means of obtaining information conveniently at the time it is actually needed. An example of a useful resource is the Genetic Home Reference website13 that can be accessed using a mobile device and provides a way of rapidly obtaining information during, for example, a clinic session when time is limited. GENETIC TESTING—AN OVERVIEW
The genetic basis of a given trait or disorder (phenotype) may be due to mendelian transmission of specific gene mutations, chromosomal anomalies, or complex interactions between genes and environmental factors. Many health care providers do not distinguish between single gene (mendelian) inheritance of rare diseases and
complex patterns of inheritance that involve many genes, variable gene expression, and interaction between the genes and the environment. Identification of specific gene and chromosomal variations has been used for many years to diagnose mendelian disorders such as amelogenesis imperfecta and chromosomal disorders such as Down syndrome. However, mendelian or chromosomal inheritance and methods of diagnosing these diseases or disorders must not be confused with use of genomics in the prediction, prevention, or treatment of diseases that have complex etiologies involving both genetic and environmental influences. The genetic component of polygenic diseases is not causative but rather is indicative of a small increase in the risk of developing disease. Whether a person will develop a disease is a function of the interaction between the genome and environmental factors. Diseases having mendelian patterns of inheritance involving 1 or a few genes affect a small fraction
JADA
( )
- -
http://jada.ada.org
-
2016 5
SPECIAL REPORT
of the population, have high heritability, and can be predicted with a high degree of accuracy. In contrast, diseases having low heritability, genetic complexity involving many genes, and strong environmental influences have poor predictability. Genetic tests are in use for diagnosis, disease prediction, determination of carrier status, prenatal and preimplantation testing, newborn testing, prediction of therapeutic response, and forensic analyses. Factors that health care professionals should consider when evaluating any genetic test include the following: - What is the type of genetic variation being tested: single gene mutation, multiple genes, a single genetic variant (SNP), or multiple SNPs? - Is the test for a rare disease caused by a single gene mutation and inherited in a mendelian pattern, or is the disease common and likely to have a complex etiology? - Does the test look for mendelian patterns of inheritance, or does it use GWASs to look for associations of genetic variants and common, complex diseases? - Is the test intended for research or clinical use? - Does the test have analytic validity—that is, are the laboratory assays precise, accurate, and reproducible? Can the test accurately detect whether a specific genetic variant is present or absent?13 - Does the test have clinical validity—that is, is the test clinically meaningful? How well does the genetic variant being tested relate to the presence, absence, or risk of developing a specific disease?13 - Does the test have clinical usefulness—that is, will having the genetic information change patient treatment or the clinical outcome? Does the test provide information about the diagnosis, treatment, management, or prevention of disease that will be helpful to the consumer?13 In summary, the purpose for which a genetic test is used must be defined clearly. Tests for mendelian disorders are used for the diagnosis or prediction of disease occurrence. These disorders are based on a single gene or a small subset of genes, and the test results are generally definitive and predictive. People with mutations in certain genes either will develop the disease (for example, Huntington’s disease) or will have a high probability of developing the disease (for example, breast cancer, colorectal cancer). Tests for common complex diseases involving multiple genes (polygenic) are not diagnostic but may be only indicative of small increases in the risk of developing a disease. To date, most genetic tests to assess the risk of developing complex common diseases, including periodontal disease and dental caries, have not been clinically useful. Genetic testing of rare mendelian dental disorders. Inherited dental disorders can be isolated and affect only the teeth (nonsyndromic) or can occur as a part of a syndrome with other diseases or tissue malformations.
6 JADA
( )
- -
http://jada.ada.org
-
2016
Syndromic conditions may occur simultaneously with or subsequent to the clinical appearance of dental malformations. Genetic diagnosis of these disorders is essential to determine whether the cause is a gene mutation that affects only the teeth or whether it is a mutation associated with a syndrome that involves other tissues as well. Determination of the genetic basis allows health care professionals and patients to minimize or deal with the potentially serious syndromic manifestations that may accompany mendelian tooth disorders. For example, with tooth agenesis (hypodontia), dentinal and enamel defects can be isolated inherited defects or also can be part of a syndrome. A mutation of the AXIN2 gene causes severe oligodontia, but mutations of this gene also can be associated with colorectal cancer or precancerous lesions of various types.14 Identification of this gene mutation in people who have many congenitally missing teeth could provide an early warning of predilection for colorectal cancer and the need for frequent colorectal cancer screening. Moreover, autosomal forms of DSPP and COLIA1 or COLIA2 gene mutations are present in most patients having dentinal disorders. Testing for the presence of these gene mutations is available at some academic institutions. People having a mutation in the DSPP gene have clinical manifestations that are limited to defects in dentin, but those having mutations in COLIA1 or COLIA2 have defects in other collagen tissues that cause bone fragility, as in osteogenesis imperfecta. It is impossible to tell the differences between these 2 conditions by examining the teeth; genetic testing is important to establish whether the dentin defect is caused by mutations in the DSPP gene or by genes that cause collagen tissue disorders in other organs. Genetic testing for common complex oral diseases. Rapid advances in genetic science have led to unrealistic expectations among health care professionals, patients, and the public, fostered by companies that offer genetic testing under the pretext of precision medicine and personalized care. Because a person’s DNA can be sequenced, it does not necessarily mean that it can be used to predict, prevent, diagnose, and guide treatment for common, complex diseases. Genetic tests offered to the public that claim to predict the risk of developing complex diseases often have limitations. For example, results from commercially available genetic tests involving saliva may indicate an elevated risk of developing disease on the basis of small increases in odds ratios considerably lower than the 2.0 often considered to be the threshold for clinically significant increased disease risk. Moreover, some genetic tests marketed to the public for predicting disease risk have been based on findings obtained from populations having a specific ethnic ancestry. Use of such a test would be of questionable value to people from other ethnic populations.
SPECIAL REPORT
Some companies offer genetic tests for common, complex oral diseases. However, there is no genetic test yet that can be used reliably to predict or assess risk of developing any complex oral disease, such as dental caries or periodontitis. Moreover, given the complexity of the human genome and its interaction with the environment, it is unlikely that genomic studies of association will lead to a meaningful predictive genetic test for the occurrence or progression of these diseases. CONCLUDING REMARKS
At the conclusion of the conference, co-chair Dr. Michael Glick shared key points that are important for practicing clinicians: - In common, complex diseases such as dental caries and periodontal disease, multiple genes interact with one another and with the environment in such a way that the genetic component of these diseases is not causative but may be indicative of a small increase in the risk of developing the disease. - The clinical usefulness or purpose of any genetic test must be considered, not only whether it is a valid and reproducible laboratory test. - The usefulness of genetic and genomic tests for diagnosing disease is different than the usefulness of such tests for predicting disease risk or disease progression. - Health care providers who use genetic and genomic tests need to be able to understand the meaning of the tests and discuss the findings, clinical implications, and test limitations with patients. - There are few genetic markers for common, complex diseases that have been validated by means of GWASs or replicated studies. - Finding SNPs that are associated with a disease or a disease biomarker does not necessarily imply causation or progression of the phenotypic expression of the disease. - GWASs having larger sample sizes may not be clinically informative because they likely will find additional associations of disease and genetic markers that have small effect sizes. - Samples of oral tissue may be better than blood for the genetic and genomic study of common and complex oral diseases. - The expression of our DNA code involves a cascade of many regulatory and modifying events during the process of gene expression as the phenotype. - There may be a need for a policy or professional guidelines for the clinical use of genetic and genomic information. - There is a need to explore the ethical issues of using genomic information for disease prediction, prevention, diagnosis, and therapy. n
Dr. Pihlstrom is a professor emeritus, School of Dentistry, University of Minnesota, Minneapolis, MN, and the associate editor for research, The Journal of the American Dental Association, Chicago, IL. Address correspondence to Dr. Pihlstrom at 4801 Fairmont Ave., Suite 902, Bethesda, MD 20814, e-mail [email protected]. Dr. Barnett is a clinical professor, School of Dental Medicine, University at Buffalo, State University of New York, Buffalo, NY. Disclosure. Dr. Pihlstrom and Dr. Barnett are members of the Task Force on Design and Analysis in Oral Health Research, which provided financial support for the conference. The conference was co-chaired by Dr. Sebastian Ciancio, executive director of the Task Force on Design and Analysis in Oral Health Research; Dr. Daniel Meyer, who was the chief science officer, American Dental Association, when the conference was held; and Dr. Michael Glick, editor of The Journal of the American Dental Association. Any errors of fact or interpretation of conference proceedings are the full responsibility of the authors. However, this summary would not have been possible without the encouragement, constructive criticism, and enthusiastic support of the following people who attended the conference and spent many hours reviewing drafts of this summary: Al Best, PhD; Sebastian Ciancio, DDS; Deborah Dawson, PhD; Michael Glick, DMD; Peter Imrey, PhD; Bryan Michalowicz, DDS, MS; and Malcom Snead, DDS, PhD. 1. Venter JC, Adams MD, Myers EW, et al. The sequence of the human genome. Science. 2001;291(5507):1304-1351. 2. Kingman A, Imrey PB, Pihlstrom BL, Zimmerman SO. Chilton, Fertig, Fleiss, and the Task Force on Design and Analysis in Dental and Oral Research. J Dent Res. 1997;76(6):1239-1243. 3. The Task Force on Design and Analysis in Oral Health Research. Task Force website. Available at: http://taskforceondesign.org/. Accessed November 12, 2015. 4. US National Library of Medicine, National Institutes of Health, Department of Health and Human Services. Genetics home reference. Available at: http://ghr.nlm.nih.gov/glossary=humanhenomeproject. Accessed January 15, 2016. 5. US National Library of Medicine, National Institutes of Health, Department of Health and Human Services. GeneEd: Genetics, Education, Discovery. Structure of the double helix. Available at: http://geneed.nlm. nih.gov/topic_subtopic.php?tid=15&sid=16. Accessed January 13, 2016. 6. Evangelou E, Ioannidis JP. Meta-analysis methods for genomewide association studies and beyond. Nat Rev Genet. 2013;14(6): 379-389. 7. Klein RJ, Zeiss C, Chew EY, et al. Complement factor H polymorphism in age-related macular degeneration. Science. 2005;308(5720): 385-389. 8. Burdett T (EBI), Hall PN (NHGRI), Hasting E (EBI), Hindorff LA (NHGRI), Junkins HA (NHGRI), Klemm AK (NHGRI), MacArthur J (EBI), Manolio TA (NHGRI), Morales J (EBI), Parkinson H (EBI) and Welter D (EBI). The NHGRI-EBI Catalog of published genome-wide association studies. Available at: www.ebi.ac.uk/gwas. Accessed January 15, 2016. 9. National Human Genome Research Institute. Epigenomics. Available at: https://www.genome.gov/27532724. Accessed November 6, 2015. 10. Kubicek S. Epigenetics: a primer. Available at: http://images.thescientist.com/content/images/articles/58007/epigenetics_primer.jpg. Accessed January 3, 2016. 11. Obata Y, Furusawa Y, Hase K. Epigenetic modifications of the immune system in health and disease. Immunol Cell Biol. 2015;93(3):226-232. 12. National Institutes of Health. International Human Epigenome Consortium. Available at: http://ihec-epigenomes.org. Accessed November 7, 2015. 13. US National Library of Medicine, National Institutes of Health, Department of Health and Human Services. Genetics home reference. Available at: http://ghr.nlm.nih.gov/. Accessed December 15, 2015. 14. Lammi L, Arte S, Somer M, et al. Mutations in AXIN2 cause familial tooth agenesis and predispose to colorectal cancer. Am J Hum Genet. 2004; 74(5):1043-1050.
JADA
( )
- -
http://jada.ada.org
-
2016 7
