Portal de Periódicos da CAPES

Diagnostic Exome Sequencing: A New Paradigm in Neurology

2013; Cell Press; Volume: 80; Issue: 4 Linguagem: Inglês

10.1016/j.neuron.2013.09.011

1097-4199

Norman Delanty, David Goldstein,

Cerebrovascular and genetic disorders

Genomic tools have evolved with remarkable rapidity, but their clinical relevance and application have lagged behind. Now, consistent clinical applications have finally arrived and bring with them the promise of identifying the underlying causes of complex neurological disorders in a patient-specific manner. Genomic tools have evolved with remarkable rapidity, but their clinical relevance and application have lagged behind. Now, consistent clinical applications have finally arrived and bring with them the promise of identifying the underlying causes of complex neurological disorders in a patient-specific manner. Since the origin of large-scale genomics, two primary motivations have been to connect genetic variation to diseases and outcomes in order to identify validated drug targets (Manolio et al., 2008Manolio T.A. Brooks L.D. Collins F.S. J. Clin. Invest. 2008; 118: 1590-1605Crossref PubMed Scopus (740) Google Scholar) and to subclassify patients into groups relevant to treatment. These ambitions have been partly realized. Genetic studies have indeed led directly to drug development programs with the potential for wide therapeutic application. For example, mutations in PCSK9, encoding proprotein convertase subtilisin kexin-9 (PCSK9), were first identified in a family exhibiting hypercholesterolemia. Loss-of-function alleles were later shown to lead to reduced low-density lipoprotein (LDL) cholesterol and protect against cardiovascular disease without any adverse effects. Thus, genetic insights from patients told drug developers that PCSK9 inhibition may be an effective new tool in cholesterol management (Wierzbicki et al., 2012Wierzbicki A.S. Hardman T.C. Viljoen A. Expert Opin. Investig. Drugs. 2012; 21: 667-676Crossref PubMed Scopus (35) Google Scholar). Another example is SOST, encoding the protein sclerostin, a critical inhibitor of bone formation. Mutations in this gene are associated with a rare bone disorder, and modulation of normal sclerostin function (via specific monoclonal antibodies) may play a role in the treatment of common bone disorders such as osteoporosis (Paszty et al., 2010Paszty C. Turner C.H. Robinson M.K. J. Bone Miner. Res. 2010; 25: 1897-1904Crossref PubMed Scopus (104) Google Scholar). Despite these promising translational developments, the indispensable genomic application of the moment is arguably not in charging the pipelines of big pharma, but rather in diagnosing genetic conditions that are difficult to elucidate. Clinical genetics is a field that has traditionally focused on individual gene tests indicated by the specific clinical presentation. Recently, steps toward more comprehensive assessments have been made, including both disease-related gene panels and array-based technology for detecting genome-wide copy-number variation; these offer higher resolution than traditional karyotype analysis. The culmination of these steps toward more comprehensive assessment, however, is clearly next-generation sequencing (NGS). The era of comprehensive NGS in clinical genetics began with diagnostic reports appearing in late 2009 (Choi et al., 2009Choi M. Scholl U.I. Ji W. Liu T. Tikhonova I.R. Zumbo P. Nayir A. Bakkaloğlu A. Ozen S. Sanjad S. et al.Proc. Natl. Acad. Sci. USA. 2009; 106: 19096-19101Crossref PubMed Scopus (1004) Google Scholar) and early 2010 (Ng et al., 2010Ng S.B. Buckingham K.J. Lee C. Bigham A.W. Tabor H.K. Dent K.M. Huff C.D. Shannon P.T. Jabs E.W. Nickerson D.A. et al.Nat. Genet. 2010; 42: 30-35Crossref PubMed Scopus (1484) Google Scholar) in the form of whole-exome sequencing (WES). Some recent examples of WES in clinical diagnosis include an infant of consanguinous parents with failure to thrive and dehydration, who was diagnosed with congenital chloride diarrhea due to a homozygous missense mutation in the SLC26A3 gene (Choi et al., 2009Choi M. Scholl U.I. Ji W. Liu T. Tikhonova I.R. Zumbo P. Nayir A. Bakkaloğlu A. Ozen S. Sanjad S. et al.Proc. Natl. Acad. Sci. USA. 2009; 106: 19096-19101Crossref PubMed Scopus (1004) Google Scholar). Similarly, a compound heterozygote mutation in the DHODH gene was discovered in four affected individuals in three independent kindreds as a cause of a multiple-malformation disorder, Miller syndrome, a disorder that had previously been intractable to more traditional approaches of discovery (Ng et al., 2010Ng S.B. Buckingham K.J. Lee C. Bigham A.W. Tabor H.K. Dent K.M. Huff C.D. Shannon P.T. Jabs E.W. Nickerson D.A. et al.Nat. Genet. 2010; 42: 30-35Crossref PubMed Scopus (1484) Google Scholar). In addition to new disease gene discovery, WES may also be useful in refining clinical therapeutic decisions in individual patients, as exemplified by the beneficial addition of 5-hydroxytrptophan (a serotonin precursor) to L-dopa therapy in two twins with dopa-responsive dystonia (Bainbridge et al., 2011Bainbridge M.N. Wiszniewski W. Murdock D.R. Friedman J. Gonzaga-Jauregui C. Newsham I. Reid J.G. Fink J.K. Morgan M.B. Gingras M.C. et al.Sci. Transl. Med. 2011; 3: re3Crossref Scopus (228) Google Scholar). Another illustrative case is that of a young boy with a severe Crohn's disease phenotype who was found by exome sequencing to have a novel, hemizygous missense mutation in the X-linked inhibitor of apoptosis gene and who went on to do well following an allogeneic hematopoietic progenitor cell transplant (Worthey et al., 2011Worthey E.A. Mayer A.N. Syverson G.D. Helbling D. Bonacci B.B. Decker B. Serpe J.M. Dasu T. Tschannen M.R. Veith R.L. et al.Genet. Med. 2011; 13: 255-262Abstract Full Text Full Text PDF PubMed Scopus (574) Google Scholar). Furthermore, in a recent pilot program of WES in 12 patients with unexplained and apparently genetic conditions, a specific genetic diagnosis was made in half of the patients (Need et al., 2012Need A.C. Shashi V. Hitomi Y. Schoch K. Shianna K.V. McDonald M.T. Meisler M.H. Goldstein D.B. J. Med. Genet. 2012; 49: 353-361Crossref PubMed Scopus (333) Google Scholar). Despite some encouraging examples, however, successful diagnoses will not always, or even (at present) often, lead to improved treatments. The reality is that the majority of known Mendelian diseases cannot be effectively treated, at least as of yet. Nevertheless, the importance to affected families of receiving a specific, correct diagnosis after years of uncertainty and soul searching cannot be overstated. Individuals with intellectual disability and epilepsy often require full-time care from a young age, the burden of which falls on the parents and family. It is extraordinary to witness the dedication and love that families show in caring for individuals with complex neurological deficits and needs over many years. For them, simply knowing the real explanation for the underlying disorder can provide comfort, reassurance, and closure. The correct diagnosis can also facilitate the provision of appropriate state health and social services. Of course, the hope is that knowing the correct diagnosis will also allow a more targeted approach to future therapies as they are discovered. Early application of NGS can bring to a close an often previously tedious, expensive, and emotionally wrenching "diagnostic odyssey"; for all of the reasons listed above, the use of NGS is simply good medical practice. There are likely few therapeutic areas set to benefit more from this new paradigm in clinical genetics than neurological disorders, particularly those affecting children. There are several interconnected reasons for this: much of neurological illness has already been shown to have a genetic basis; it is often difficult to predict the genetic defect on clinical grounds; new causative variants are being described weekly; and it is expensive and burdensome to test on a gene-by-gene basis. In addition, the global burden of unexplained neurological disorders is immense. Epilepsy alone affects 6o million people worldwide, and the diagnosis of epilepsy encompasses a large group of brain disorders characterized by the occurrence of recurrent unprovoked seizures; one third of these individuals have medically refractory, poorly controlled seizures. Although there may be a recognized proximate cause in an individual patient (e.g., traumatic brain injury), in about 50% of those with epilepsy, no known etiology is apparent. It is likely that a large proportion of these individuals have an underlying genetic underpinning to their epilepsy. Many may be due to individual mutations affecting a variety of proteins and pathways necessary for normal brain development and function. Similarly, 1%–3% of the population has a lifelong intellectual disability (ID; from mild to profound) with associated significant long-term personal, family, social, and economic consequences. Again, the etiology of intellectual disability is unknown in about half of individuals. Recent evidence confirms that, as with epilepsy, the underlying causes of ID are molecularly diverse, with a significant proportion accounted for by functionally deleterious de novo mutations across a spectrum of genes (de Ligt et al., 2012de Ligt J. Willemsen M.H. van Bon B.W. Kleefstra T. Yntema H.G. Kroes T. Vulto-van Silfhout A.T. Koolen D.A. de Vries P. Gilissen C. et al.N. Engl. J. Med. 2012; 367: 1921-1929Crossref PubMed Scopus (1130) Google Scholar, Rauch et al., 2012Rauch A. Wieczorek D. Graf E. Wieland T. Endele S. Schwarzmayr T. Albrecht B. Bartholdi D. Beygo J. Di Donato N. et al.Lancet. 2012; 380: 1674-1682Abstract Full Text Full Text PDF PubMed Scopus (764) Google Scholar). Moreover, there is overlap between epilepsy and ID, whereby one third of individuals with ID have epilepsy as a manifestation of their underlying brain disorder, and approximately 20% of patients attending a tertiary referral epilepsy clinic have an associated intellectual disability. A recent study has shown that de novo mutations are important as a cause of previously unexplained childhood epileptic encephalopathies, conditions generally associated with severe epilepsy and intellectual disability (Allen et al., 2013Allen A.S. Berkovic S.F. Cossette P. Delanty N. Dlugos D. Eichler E.E. Epstein M.P. Glauser T. Goldstein D.B. Han Y. et al.Epi4K ConsortiumEpilepsy Phenome/Genome ProjectNature. 2013; 501: 217-221Crossref PubMed Scopus (1086) Google Scholar). It is becoming apparent that a significant proportion of common, but unexplained, diseases (e.g., the epilepsies, as well as other neurological and non-neurological conditions) may be a collection of rare and often private genomic disorders due to mutations in genetically intolerant genes (Petrovski et al., 2013Petrovski S. Wang Q. Heinzen E.L. Allen A.S. Goldstein D.B. PLoS Genet. 2013; 9 (Published online August 22, 2013): e1003709https://doi.org/10.1371/journal.pgen.1003709Crossref PubMed Scopus (640) Google Scholar). The International League Against Epilepsy classification of epilepsy includes information about seizure type, age of onset, response to antiepileptic drugs, electroencephalogram (EEG) and structural brain imaging information, and prognostic considerations. From a molecular and physiological perspective, however, it is clear that this scheme often bears little relationship with underlying biology. Copy-number variants are associated with a range of epilepsy subtypes (Heinzen et al., 2010Heinzen E.L. Radtke R.A. Urban T.J. Cavalleri G.L. Depondt C. Need A.C. Walley N.M. Nicoletti P. Ge D. Catarino C.B. et al.Am. J. Hum. Genet. 2010; 86: 707-718Abstract Full Text Full Text PDF PubMed Scopus (207) Google Scholar), including focal epilepsy, which responds to surgery (Catarino et al., 2011Catarino C.B. Kasperavičiūtė D. Thom M. Cavalleri G.L. Martinian L. Heinzen E.L. Dorn T. Grunwald T. Chaila E. Depondt C. et al.Epilepsia. 2011; 52: 1388-1392Crossref PubMed Scopus (14) Google Scholar); causal mutations in SCN1A show very complex genotype-phenotype relationships (Zuberi et al., 2011Zuberi S.M. Brunklaus A. Birch R. Reavey E. Duncan J. Forbes G.H. Neurology. 2011; 76: 594-600Crossref PubMed Scopus (185) Google Scholar); and mutations in the gene encoding DEPDC5 are responsible for a significant proportion of cases of familial nonlesional focal epilepsy (Dibbens et al., 2013Dibbens L.M. de Vries B. Donatello S. Heron S.E. Hodgson B.L. Chintawar S. Crompton D.E. Hughes J.N. Bellows S.T. Klein K.M. et al.Nat. Genet. 2013; 45: 546-551Crossref PubMed Scopus (253) Google Scholar). The National Academies has recently recognized the need for "a new taxonomy of human disease based on molecular biology" in its publication Toward Precision Medicine (National Research Council (US) Committee on A Framework for Developing a New Taxonomy of Disease, 2011National Research Council (US) Committee on A Framework for Developing a New Taxonomy of DiseaseToward precision medicine: building a knowledge network for biomedical research and a new taxonomy of disease. The National Academies Press, Washington, DC2011Google Scholar). NGS can facilitate individualized molecular diagnoses in patients and families with hitherto undiagnosed and unexplained disorders. The traditional diagnostic model in the evaluation of an individual with a putative genetic disorder includes formulation of a diagnostic hypothesis that may include a diverse range of possibilities. These possible diagnoses are then tested by a variety of biochemical (blood, urine, cerebrospinal fluid [CSF]), structural (MRI), functional (EEG), and specific gene analyses. A recent study examined the economic implications of WES-based diagnosis in the context of 500 patients evaluated using traditional genetic tests (Shashi et al., 2013Shashi V. McConkie-Rosell A. Rosell B. Schoch K. Vellore K. McDonald M. Jiang Y.-H. Xie P. Need A. Goldstein D.G. Genet. Med. 2013; (Published online August 8, 2013)https://doi.org/10.1038/gim.2013.99Abstract Full Text Full Text PDF PubMed Scopus (193) Google Scholar). This work showed that if the diagnosis is not clinically apparent at the first visit, then the cost on average per successful genetic diagnosis using traditional tests is approximately $25,000. The cost of WES, on the other hand, is now well under $1,000 per sample. Thus, when used in an appropriate setting, WES has the potential to provide significant cost benefit to the healthcare budget and to society. Diagnostic sequencing should, and probably will, find wide, immediate application in the care of patients with neurological disease. The realization of its full potential will require addressing a number of key bottlenecks. Of particular importance is the challenge of data integration. Clearly, to maximize the benefit of WES-based diagnostics, it is critical to be able to compare the sequences of patients evaluated in different academic medical centers. In this regard, it is increasingly clear that communication among affected families via social media is hastening the identification of novel causative variants, and this is a movement that should be fully embraced by the academic and clinical communities. For example, a recessive mutation in NGLY1, encoding N-glycanase, was recently discovered in a single family as a cause of a new disorder of deglycosylation (Need et al., 2012Need A.C. Shashi V. Hitomi Y. Schoch K. Shianna K.V. McDonald M.T. Meisler M.H. Goldstein D.B. J. Med. Genet. 2012; 49: 353-361Crossref PubMed Scopus (333) Google Scholar). Subsequent to this initial work, the efforts of that family were instrumental in the identification of further cases (http://matt.might.net/articles/my-sons-killer/) to confirm the putative diagnosis. There are also current plans to initiate and establish secure sequence data repositories to allow more dynamic evaluation of patient genomes than is afforded by the current diagnostic models. There are other hurdles and challenges along the way, but these are surmountable (Cavalleri and Delanty, 2012Cavalleri G.L. Delanty N. Adv Protein Chem Struct Biol. 2012; 89: 65-83Crossref PubMed Scopus (9) Google Scholar). For example, recent bioinformatic approaches that integrate gene-level and variant-level prioritization schemes (Petrovski et al., 2013Petrovski S. Wang Q. Heinzen E.L. Allen A.S. Goldstein D.B. PLoS Genet. 2013; 9 (Published online August 22, 2013): e1003709https://doi.org/10.1371/journal.pgen.1003709Crossref PubMed Scopus (640) Google Scholar) open the possibility of identifying candidate mutations in a genome-wide context, even without prior information implicating specific genes. Another issue is that relevant healthcare professionals often lack the necessary genomics expertise to counsel patients; however, this could and should be addressed through the integration of genomic medicine into relevant curricula at the level of theoretical instruction and also including practical clinical exposure in medical instruction and allied educational programs. A greater challenge will be to persuade contemporary clinicians of the power of clinical genomics. Other challenges include the use and appropriate release of incidental data, secure storing of genomic and updated phenotypic information on an electronic patient record, appropriate reimbursement, and—as genetic discoveries continue to be made—a system for regular reanalysis of genetic variants after the initial analysis of the patient's genome. The latter will become particularly relevant, as the secure interpretation of disease-causing rare variants will improve with the availability of increasing cohorts of control samples from different populations. In summary, despite the challenges, it is now likely that most patients with serious neurological diseases will soon have their genomes sequenced, certainly in the context of pediatric presentations. In some therapeutic areas, this will mean that many, and eventually perhaps most, patients seen will have an identified genetic cause of their condition. Ongoing efforts to sequence and understand large cohorts of well-phenotyped individuals, such as the Epi4K project in epilepsy, will help lead us to this goal (Epi4K Consortium, 2012Epi4K ConsortiumEpilepsia. 2012; 53: 1457-1467Crossref PubMed Scopus (66) Google Scholar). The clinical implications of these advances are hard to overstate. First, many more families would have a diagnosis, which is simply better medicine than what is currently offered. Moreover, evaluating appropriate treatments may in some cases be predicted by functional assays in laboratory models of particular genetic defects instead of through trial and error in the patient. Finally, NGS may bridge the divide between evidence-based medicine and patient-oriented care and help rehumanize clinical medicine. In an era in which physicians are being encouraged to see ever more patients using a formulaic, protocol-driven approach within a predetermined timescale, NGS reminds us of the unique biology of our patients and the need to treat each of them as an individual.