The impact of on the future of and health

John S. Mattick1, Marie Dziadek1, Bronwyn Terrill1, Warren Kaplan1, Allan D. Spigelman2, Francis Bowling1, and Marcel E. Dinger1 1. Centre for Clinical Genomics, Garvan Institute of Medical Research and 2. The Kinghorn Cancer Centre; St. Vincent’s Hospital & UNSW St Vincent’s Clinical School, Sydney, NSW 2010

Over the past decade there have been extraordinary advances in DNA technology, with hyper-exponential decreases in cost, such that it is now practicable to readily acquire and analyse individual sequences. The pace of data acquisition has accelerated the consequent understanding of the molecular basis of disease and contributed to new strategies and to prevent and treat disease.

While the molecular stratification of cancer is the immediate opportunity and a growing medical imperative, genomic medicine is being increasingly used to accurately diagnose monogenic diseases, and to identify and assess health risks of acquired genetic variants. It seems likely that within a decade or two, individual genome sequences, most likely done at birth and at indicated times throughout life, will become an integral part of patient medical records and management.

A key challenge is to build a sophisticated infrastructure that links comprehensive genome knowledge databases to extensive e-health record networks. This infrastructure would be accessible to clinicians for effective decision-making support based on patient’s genomic sequence information, as well as to researchers and health system managers to derive informative conclusions from aggregated population data.

Health systems worldwide are being challenged by aging populations, the growing burden of chronic disease and rapidly increasing medical costs. New genomic approaches to healthcare delivery and management will fuel a drive towards personal health optimization through early identification and targeted treatment of individuals at risk of disease. This will have a transformational impact on personal health and wellbeing, health economics and national productivity. ______

In their report to British parliament on Genomic Medicine in 2009 [1], the House of Lords and Technology Committee concluded that:

"Every so often, a scientific advance offers new opportunities for making real advances in medical care. From the evidence given to this inquiry, we believe that the sequencing of the , and the knowledge and technological advances that accompanied this landmark achievement, represent such an advance."

In recent years, there has been an unprecedented leap in information and insight into the human genome and its role in health and disease. A decade ago, researchers were tentatively exploring the first ‘reference’ human genome sequences [2,3], which cost over $1 billion to produce. Now, thousands of peoples’ , from a cross-section of cultures, have been sequenced [4-8]. This genomic explosion has been enabled by extraordinary advances in technologies that can sequence a person’s genome – more than 6000 million bases – in hours or days, rather than years. A human genome currently costs around $5,000 to sequence (Figure 1), but costs are likely to fall to less than $1,000, and possibly only a few hundred dollars, in the next few years.

1

Figure 1: The hyper-exponential decline in the cost of human genome sequencing (graph courtesy of the US National Human Genome Research Institute) [9]

Making sense of genomic data requires sophisticated computational technologies. These are evolving rapidly. Advances in genomic and informatics technologies enable an ever increasing capacity for diagnosis of existing disease and development of effective clinical treatment strategies. They also offer opportunities to assess predisposition to disease, potentially prompting more focused clinical monitoring and lifestyle changes.

The accelerating pace of data acquisition is similarly accelerating our understanding of the molecular basis of health and disease. Although our knowledge of the human genome is far from complete, life- saving examples continue to demonstrate that even limited genomic understanding can be powerful in the clinic. Currently genome sequencing is having the most impact in stratifying patients for appropriate cancer treatment, characterising genetic disease, and providing information about an individual’s likely response to treatment. However, it is also clear that contributes to most complex diseases, including infectious diseases, by influencing individual risk in a context-dependent manner. Much of this variation has not been effectively characterized, due in part to the inadequacy of genome- wide association studies in quantitatively identifying most of the relevant genetic variants contributing to complex disease. There seems little doubt that genome sequencing analyses will make inroads into our understanding of relevant genotype-phenotype correlations in a range of diseases such as diabetes, osteoporosis and cardiovascular disease.

Cancer: stratifying tumours for treatment Genomic medicine has shown clinical benefit in refining diagnoses and guiding therapeutic approaches for cancer [10]. Since the late 1990s, the clinician’s cancer ‘toolkit’ of , radiation and chemotherapy has been increasingly supplemented by a range of molecularly targeted therapies, such as trastuzumab (Herceptin®) and imatinib (Gleevec®), which target specific molecular pathways in cancer growth and development [11]. Genomic information can now help clinicians to decide their therapeutic approach by stratifying a patient’s tumour according to its and corresponding drug sensitivities [see Cases 1 and 2].

In some key cases, patients have been spared costly and complex procedures, such as transplants, based on a molecular diagnosis [12,13]. Yet other studies have stabilized a patient’s cancer development – for a time at least – by targeting specific molecules or pathways in the tumour cells

2

[14,15]. More controversially, clinicians are starting to use genomic information to refine a cancer diagnosis and alter an individual’s quality of life, even when therapies are not currently available. For example, a patient’s five-year survival outcome after developing melanoma of the eye can be accurately predicted by the pattern of genomic activity in their tumour [16, 17].

Genome-wide sequencing is now also being applied to the analysis of circulating DNA in the plasma of cancer patients as well as in individuals with other pathological diseases [18]. This technology presents non-invasive approaches for tumour detection and monitoring responses to that will significantly improve patient management.

Case 1: Precision treatment of esophageal cancer Case 2: Identifying therapies for a rare cancer of the tongue A 60-year-old man diagnosed with metastatic A patient presented with a rare tongue adenocarcinoma and no esophageal cancer was initially underwent cytotoxic standard treatment options. The tumour showed elevated EGFR (a chemotherapy, which is the usual standard of care for growth factor receptor), but treatment with an EGFR inhibitor erlotinib this type of cancer. However, genomic analysis of the did not slow down tumor growth. Genome sequencing identified tumour identified a in the bcr-ablkinase mutations in other that overrode the effect of erlotinib, explaining , which suggested that the cancer might respond its inefficiency. They also identified that the oncogene RET was to Gleevec (Imatinib). “You would never think of duplicated. “After much agonizing and hand-wringing” the clinicial team giving Gleevec for an esophageal tumor,” Dr. Boguski prescribed a RET-inhibiting drug, sunitinib, usually prescribe for renal cell said, “but the target that the drug is normally . The first treatment stabilised tumour growth for four months. effective for was present in [the man’s] particular When the cancer spread again, they used another RET inhibitor and anti- tumor. It was an incidental finding.” “I think there may inflammatory agent sorafenib, which stabilised the disease for four be countless additional incidental findings when you additional months. Analyses of the recurring tumours showed that other start sequencing tumors,” he says [15]. cancer pathways had been activated, but by then it was too late [13].

Drug prescription and development Cancer treatment is also set to benefit from using an individual’s genomic information to predict how he or she will respond to drugs (known as ) or surgery. For example, testing for the presence of an increasing number of genetic variants can be used to predict whether an individual will suffer severe toxicity from chemotherapy [19]. In another example, a lack of responsiveness to imatinib treatment was found to be due to a genetic variation in a required gene (called Bim), which is common in East Asian populations but not in African or European populations. This discovery led to an improved genomic test to identify non-responders and the development of new drugs to treat such individuals [20].

The use of genomic information to prescribe the appropriate drug or dosage for a patient extends far beyond cancer into daily clinical practice. Currently, only a handful of DNA tests for variants implicated in drug metabolism have been recommended for use in Australia [21]. The first of these approved in Australia was for the drug Abacavir®, which is used to treat HIV infection but causes a potentially life- threatening allergic reaction in 5-8% of patients [22]. However, there are many more applications for pharmacogenomic tests for prescribing drugs such as antidepressants, painkillers, and anticoagulants [23]. Better targeting of existing drug use avoids wasteful and risky therapy. To move more examples into common use, researchers are calling for widespread clinical evaluation of genomic variants that could assist in selecting the appropriate drug or prescribing the appropriate dosage for a patient [24].

In the longer term, genomics is expected to dramatically change the testing and use of pharmaceuticals [25]. New research into molecular pathways underlying health and disease will continue to inform rational drug development and design. Alongside this effort, researchers are using genomic data to suggest new therapeutic applications for existing drugs (“repositioning”) – with significant cost savings – and to better target individuals for clinical trials to find uses for drugs that failed an earlier trial (“rescue and repurposing”) [26]. An example from earlier genetic analyses is the recognition that mutations associated with Tuberous Sclerosis cluster in the mTor pathway, which can be successfully treated with

3 the immunosuppressant rapamycin [27]. Such examples illustrate the enormous and transformational potential of genomically-informed drug development, selection, re-purposing and dosages, providing treatments for individuals together with savings for the healthcare system and benefits for the economy at large.

Understanding genetic cause of and predisposition to disease With the exception of a few key pharmacogenomics tests and cancer-based treatments, clinical in Australia is currently limited to diagnosing hereditary and rare genetic diseases. Every baby born in Australia is offered screening for approximately 30 genetic conditions (the Guthrie test) and more than 300 tests for genetic disorders are available through the healthcare system [28]. However, clinicians worldwide are currently embracing genome sequencing to search for variants implicated in undiagnosed genetic diseases [29,30] and using this information to guide treatment [31,32] [see Case 3]. In one of the most dramatic cases, a young boy underwent a risky, but seemingly successful, bone marrow transplant that was proposed in response to molecular data [33,34]. As the field matures and we have an improved understanding of the affected pathways there will be many more genomic regions implicated in rare disease, with improved treatment prospects. The Human Variome Project, an international consortium modeled on the Human , seeks to further define the significance of variants, with its exemplar being the InSiGHT (International Society for Gastrointestinal Hereditary Tumours) mismatch repair database relating to Lynch Syndrome [35].

Case 3: Accurate genetic diagnosis of disease A 14 year-old pair of twins were diagnosed with dopa-(3,4-dihydroxyphenylalanine)- responsive dystonia (DRD) based on their symptoms. Treatment with l-dopa alleviated many, but not all, of their symptoms. Whole genome analysis showed that they did not have mutations in the genes normally causing DRD, but rather a mutation in the SPR gene, coding for an enzyme important in serotonin action. When treatment was supplemented with a serotonin precursor, 5-hydroxytryptophan, both twins showed marked improvement in coordinated movements and attention in school within 1-2 weeks, such that they were able to undertake sporting activities [32].

Rather than sequencing the entire genome, rapid sequencing of only the protein-coding portion (about 1.5% of the total), called the , can offer comparatively cost-effective analysis of many genetic disorders. Forms of inherited heart disease such as cardiomyopathies result from mutations in one of many different genes, and currently each one would need to be assessed individually to achieve a precise diagnosis [36]. Similarly, severe eye diseases can result from a single mutation or a combination of many mutations [37]. identifies any and all of the mutations in a single analysis, serving both as a diagnostic tool and a method to discover new genes and mutations that underlie these classes of disease. However, recent experience suggests that whole genome sequencing provides not only a more comprehensive picture of genomic rearrangements, particularly in cancer, but is also faster and more accurate in sequencing of exonic regions. It is expected that whole genome sequencing will quickly replace exome sequencing as sequencing costs decrease.

4

Case 4: Accurate diagnosis of diabetes Case 5: Predisposition to type 2 diabetes Diabetes can be misdiagnosed in childhood or Michael Snyder heads up a laboratory at adulthood. People who are diagnosed with Type 2 Stanford University and became his lab’s diabetes but are not overweight may have “Type research subject. Over the course of two years, 1.5”: either a slow-developing autoimmune Snyder and his lab members studied his body in diabetes or a set of genetically-linked forms of great detail, from his genomic sequence to the diabetes known as MODY or “Maturity Onset workings of his immune system. Snyder’s tests Diabetes of the Young”. MODY forms of diabetes revealed a predisposition to type-2 diabetes. He are caused by at least nine different genes; you only then went on to develop the condition during have to inherit a mutated version of one of these the course of the study. “"When I went to see genes to have the disorder. More precise diagnosis my doctor, she said there was no way," recalls for MODY could have a major impact on an Snyder.” His genomic information, integrated individual: some MODY subtypes can be managed with more routine tests, helped to identify by diet and exercise alone, while others can benefit diabetes early and take steps to lessen its enormously from sulfonylurea therapy [38]. impact [39,40].

Increasingly, genome sequencing is being used for accurate assessment of complex disease where multiple gene variants may be involved in disease development and progression. It is also used to determine individual predisposition to disease [see Cases 4 and 5]. Moreover, with genome sequencing becoming more accessible it is becoming more common for individuals to seek this information themselves. A number of ‘direct-to-consumer’ genetic testing services, such as those offered by 23andMe, have come to market in recent years [41] and are rapidly gaining in popularity. Inevitably, personal genome sequences, likely obtained at birth, will become an integral part of patient medical records and incorporated into their Electronic Health Record (EHR), where it can be integrated with other clinical and environmental data and interrogated throughout the individual’s lifetime [42]. Clinicians and patients may then query the sequence to accurately prescribe treatments, determine disease susceptibilities and identify drug sensitivities, and determine a course of action to monitor, manage, ameliorate risk or prevent the disease.

Many individuals do not realize they carry disease mutations until their symptoms manifest in late-stage disease. An early genomic sequence analysis could enable interventions that prevent disease onset and potentially severe medical consequences. For example hemochromatosis, an inherited and potentially fatal iron overload disorder can be easily treated if diagnosed early. It is one of the most common genetic disorders in Europeans, with one in nine individuals carrying a genetic mutation and one in 200 to 400 people affected. Early symptoms include chronic fatigue and joint pain, and if undiagnosed an individual can succumb to extensive organ damage due to excessive iron build-up [43]. Factor V Leiden deficiency is another common inherited disorder, which causes excessive blood clotting in affected individuals that may lead to deep vein thrombosis, pulmonary embolism and transient ischemic attack and stroke [44]. Mutations are present in approximately 5% of Caucasian populations but rarely found in Asian groups. This common blood clotting disorder can be readily prevented by low dose aspirin or other anti-coagulant therapy, thus reducing the incidence of life-threatening disease in at-risk individuals. These examples, exemplify the potential benefits of broad spectrum genomic testing to individuals and to the healthcare system. Forewarned is forearmed.

Capitalising on the integration of genomic and clinical information Ultimately, the era of personalised medicine/genomic healthcare will be realised when genomic information can be used routinely to improve the health, diagnosis and treatment of all individuals. The clinical utility of genomic information depends on the state of knowledge in associating genetic variations with phenotypic/disease characteristics and drug responses. The pathway from patient DNA sequencing to a clinical treatment plan relies on the integration of the individual’s genomic information with knowledge databases that contain known genotype/phenotype correlations and genomic and

5 clinical associations from large populations of individuals (Figure 2).

Clinical decision-making currently relies on the knowledge of individual practitioners and genetic counselors. These practitioners work as a multidisciplinary team (MDT) to assess the genomic variants identified through genomic sequencing and arrive at a treatment plan. There is at present insufficient readily-available data upon which to make an evidence-based assessment of the clinical validity or utility of an individual’s genomic analysis, and clinicians are not generally well-trained in how to interpret genomic data. Purpose-built evidence-based databases of human genotype-phenotype associations are urgently needed, with computational tools to access and interrogate the ever-increasing information in an automated way.

patient

clinical referral

Genotype Phenotype treatment Genome sequencing/ Consultation informatics E-health record outcome Patient data

Interrogation of genotype/phenotype knowledge Current: Multidisciplinary team consultation (MDT) Future: State/national clinical decision support database + MDT

Value add Value Ideal: Global clinical decision support database + MDT

Clinical report Diagnosis/treatment plan/clinical trial

Figure 2: The patient pathway in personalised genomic medicine

The volume of information contained in a genome sequence is enormous. While some of it may never be useful in medical practice, there is a continual and increasing rate of acquisition of knowledge about genetic variants that play a role in disease; particularly those whose contribution to disease predisposition can be modified by lifestyle, environmental factors and/or pharmaceutical intervention. Consequently genome knowledge and genotype-phenotype databases will need to be constantly updated with well curated evidence-based information and have appropriate data query tools to make this information accessible and clinically useful. Information on how variation in the genotype alters the natural history of both genetic and non-genetic diseases is also needed.

An ideal clinician decision support database would consist of two important elements (Figure 3):

1. A comprehensive genome knowledge database that contains evidence-based data on genotype-phenotype correlations from multiple sources, including basic and clinical research, clinical trials and case reports from across the world.

2. An EHR database, where genome sequence information and longitudinal medical records from large populations provide progressively expanding information to link genomic variants with health information.

6

Europe

Clinical decision support database Genome knowledge database – data from experimental studies, clinical research, clinical trials, case reports): Asia • Function of normal/variant sequences • Pathogenic mutations • Drug resistance genes/sequences • Potential drug targets Australian • Drug responses EHR database Local linked EHRs: • Genomic sequence • Phenotype • Medical history USA • Drug responses • Familial disorders EHR networks

Figure 3: Components of a clinical decision support database for genomic medicine

The inherent computational challenges in integrating genomic and clinical data necessitate a significant investment in capability. Given the considerable overheads of both storage and computational power, Australia will most effectively be served by a single or very small number of centralized genome knowledge repositories.

Infrastructure support will also be critical for the development of e-systems and software interfaces. These must enable clinicians and healthcare providers to interrogate a patient’s genomic sequence against the clinical decision support databases and access useful information as an automated clinical report.

Storing and sharing population data The speed with which personalised genomic medicine becomes a reality for most of the population will be greatly influenced by the sharing of clinical information – in this case an individual’s genomic sequence together with their medical record – as part of a large population dataset that is accessible to clinicians and translational researchers. Typically only a small subset of the genetic information yielded in a genomic test is used for a clinical diagnosis, so that a by-product of genomic analysis is a potential treasure-trove of information in an EHR database for translational researchers to further understand both disease and normal genetic variation in the human population.

By aggregating and analyzing large datasets it will be possible to uncover patterns and relationships that would not otherwise be revealed. This will be best achieved through the addition of genomic sequence information to EHRs and linking local and state-wide patient EHRs, on a coded and de-identified basis, into a national EHR database. As a national resource it is vital that such a database is access-controlled and appropriate consent processes are in place (Figure 3).

The Australian Federal Government has committed significant funds to building a Personally Controlled Electronic Health Record (PCEHR) system [45]. Adding genomic information and linking large population datasets will add significant value to the healthcare system through the acceleration of personalised genomic medicine. “Data-mining” of a national EHR database will allow evidence-based forward planning of healthcare needs and allocation of resources by government health departments. However, the implementation of comprehensive EHRs requires substantial investment, and has not been successful in many countries, including the USA [46].

Establishment of a large population EHR database requires all EHRs to have a common language to

7 permit appropriate retention, integration, processing and exchange of unambiguous genomic and medical data. Clinical documentation is frequently imprecise and highly variable in use of descriptors and incorporation of high-quality clinical data into EHRs will require training of the health workforce. Genomic and clinical data also needs to be de-identified and protected to an extent whereby it safeguards patient privacy and cannot be harmful to them or their families.

The enormous value of data sharing in the acceleration of progress in genomic medicine is now well recognized by international leaders active in establishing capability in this area. A consortium of nearly 70 institutions in 13 countries has recently been established to develop a framework for data sharing, including the technical standards required for storage and sharing of genomic and clinical data and the management of privacy, informed consent and security [46,47]. Core principles and policies will be established to enable participating institutions to share data while operating within their individual country’s regulatory frameworks. The consortium plans to create a network of computing platforms and analysis tools to provide access to shared data, and will encourage the development of tools to allow patients to maintain control over their own medical and genomic information.

In Australia, all medical testing, including genetic testing, must be undertaken in a facility accredited by NATA, the National Association of Testing Authorities. The medical testing accreditation program is administered by NATA in conjunction with the Royal College of Pathologists of Australasia (RCPA). The introduction of genomic testing in a clinical setting in Australia will require operation within the regulations set by NATA and RCPA, with standardized processes for DNA sequencing, annotation and informatics. The computational aspects in particular are sophisticated, diverse and rapidly evolving, presenting particular challenges for the development of standardised procedures. A combination of centralization and close cooperation between genomic testing facilities will be critical: to ensure consistent testing outcomes; to facilitate the efficient introduction of improved analytical approaches and; to allow the promise of the genomic era to be fulfilled.

Economic benefits of genomic medicine Genomic medicine will transform healthcare and the national economy, especially in a population whose average lifespan is increasing. The personal economic benefits accrue from genomically informed restoration of health and consequent earning capacity. Higher precision in risk identification reduces health costs for an individual and the entire healthcare system through avoidance of adverse reactions and unnecessary treatments.

Genomics has the potential to make genetic diagnosis of disease a more efficient and cost-effective process, by reducing genetic testing to a single analysis, which then informs individuals throughout life. If just 10% of those with monogenic diseases, which collectively comprise 1-2% of the population, can have their conditions characterized by genomic testing, the financial effect on their families and communities will be significant over their lifetime. The same applies to the identification of individuals at heightened risk for cancers and other complex diseases such as Type 2 diabetes or stroke. Indeed, while not all diseases can (yet) be avoided, we expect that personal identification of risk will result in much more effective monitoring and preventative strategies by individuals, as already occurs in individuals with BRCA1 and BRCA2 mutations. Broad-based warnings pitched at the entire population have often been ineffective because many people assume or hope that they themselves are not at risk. Knowledge of their actual personal risk is expected to focus an individual’s attention on disease prevention.

8

Genomics and its application to technical developments, medical research and healthcare will also have a major impact on the national economy, not only by the reduction in lost productivity and decreased costs of treating disease, but also by the creation of new medical information industries. Recent data from the USA indicates that its investment in human genome research and the associated industry activity has generated an economic output of $1 trillion between 1998 and 2012, as well as 4.3 million person years of employment [48]. Each dollar of investment has thus far contributed to the generation of $65 in the US economy in addition to effecting transformational changes in medicine.

The trajectory for genomic medicine is clear and, whatever the rate of progress, it will undoubtedly transform the course of medical science and practice. The sooner we start strategic planning to develop and implement the full potential of genomic medicine, the sooner will we realise its health and economic benefits. The UK government has recognized this imperative, just announcing its investment in a project, to be run by , to sequence the genome of 100,000 patients over five years and introduce genomic technology into its mainstream health system [49].

Key challenges As genome sequencing moves into medical practice, the healthcare system will become the primary site for collecting the phenotypic and genomic data needed to establish the knowledge systems for clinical care and translational research. No hospital or individual practitioner will have the knowledge to interpret and translate genome sequence data into clinically relevant and actionable information. This advice will need to be provided online from central knowledge databases of evidence-based, well curated genotype-phenotype correlations.

For clinical genomics to provide full benefit for the Australian public as a whole, major challenges are to: • Build the infrastructure needed to establish EHR databases that integrate patient’s genomic and medical information for clinical and research applications, with appropriate patient consent, protection of privacy and data security in place. • Establish national/international knowledge sharing platforms using a standardised approach for recording, sharing and interrogating fully integrated clinical and genomic databases to provide clinically useful reports. • Engage and educate clinicians, the health workforce and the community to fully realise the medical potential of genomics [50]. Well-designed and integrated public and professional education efforts will be required nationwide to address this fast-moving field.

Opportunities The rapid aging of populations, rapidly increasing costs of healthcare and the growing burden of chronic disease present major challenges to health systems worldwide. Personalised/precision genomic medicine provides opportunities for new approaches to healthcare delivery and comprehensive population health management. For example, it will identify individuals at risk for many diseases and significantly reduce the incidence of these diseases. This will fuel a drive towards personal health optimization, with enormous benefits for individuals, the healthcare system and the national economy.

The medical and scientific communities around the world are just starting to seize the opportunities that personalised genomic medicine offer. With further investment in the infrastructure required to acquire and share clinical and genomic data, Australia will be positioned as one of the key leaders and beneficiaries in implementation of personalized genomic medicine.

9

REFERENCES

1. House of Lords Science and Technology Committee. Genomic Medicine 2009 Volume I: Report, pp. 1-127. Available from http://www.publications.parliament.uk/pa/ld200809/ldselect/ldsctech/107/10702.htm 2. International Human Genome Sequencing Consortium. Initial sequencing and analysis of the human genome. Nature 2001;409(6822):860-921. 3. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, et al. The Sequence of the Human Genome. Science 2001;291(5507):1304-1351. 4. Levy S, Sutton G, Ng PC, Feuk L, Halpern AL, Walenz BP, et al. The Diploid Genome Sequence of an Individual Human. PLoS Biol. 2007;5(10):e254. doi: 10.1371/journal.pbio.0050254 5. Wheeler DA, Srinivasan M, Egholm M, Shen Y, Chen L, McGuire A, et al. The complete genome of an individual by massively parallel DNA sequencing. Nature 2008;452(7189):872-876. 6. Wang J, Wang W, Li R, Li Y, Tian G, Goodman L, et al. The diploid genome sequence of an Asian individual. Nature 2008;456(7218):60-65. 7. Schuster SC, Miller W, Ratan A, Tomsho LP, Giardine B, Kasson LR, et al. Complete Khoisan and Bantu genomes from southern Africa. Nature 2010;463(7283):943-947. 8. The Consortium. An integrated map of genetic variation from 1,092 human genomes. Nature 2012;491(7422):56-65. 9. Wetterstrand KA. DNA Sequencing Costs: Data from the NHGRI Genome Sequencing Program (GSP) Available from: http://www.genome.gov/sequencingcosts/. Accessed 17 June, 2013. 10. Wagle N, Berger MF, Davis MJ, Blumenstiel B, DeFelice M, Pochanard P, et al. High-Throughput Detection of Actionable Genomic Alterations in Clinical Tumor Samples by Targeted, Massively Parallel Sequencing. Cancer Discovery 2012;2(1):82-93. 11. National Cancer Institute. Targeted Cancer Therapies. Available from http://www.cancer.gov/cancertopics/factsheet/Therapy/targeted Retrieved 17 June 2013, 12. Welch JS, Westervelt P, Ding L, Larson DE, Klco JM, Kulkarni S, et al. Use of whole-genome sequencing to diagnose a cryptic fusion oncogene. JAMA 2011;305(15):1577-1584. 13. Maher B. Human : Genomes on prescription. [News Feature]. Nature 2011;478:22-24. 14. Jones SJ, Laskin J, Li YY, Griffith OL, An J, Bilenky M, et al. of an adenocarcinoma in response to selection by targeted kinase inhibitors. Genome Biol. 2010;11(8):R82. doi: 10.1186/gb-2010-11-8-r82. 15. Lusky K. Cancer genomics: the exception now, but semi-routine soon? CAP Today: College of American Pathologists. June 2011 Available from http://www.cap.org/apps/portlets/contentViewer/show.do?printFriendly=true&contentReference=cap_t oday%2F0611%2F0611g_cancer_genomics.html 16. Onken MD, Worley LA, Char DH, Augsburger JJ, Correa ZM, Nudleman E, et al. Collaborative Ocular Group Report Number 1: Prospective Validation of a Multi-Gene Prognostic Assay in Uveal Melanoma. 2012;119(8):1596-1603. 17. Kolata G. A Life-Death Predictor Adds to a Cancer's Strain. New York Times 10 July 2012. 18. Lo YM, Chiu RW. Plasma nucleic acid analysis by massively parallel sequencing: pathological insights and diagnostic implications. J. Pathol. 2011;225:318-23. 19. Curtin C, Dublin M, Rizk C, Vence T. The Cancer Fight. Genome Web: Genome Technology, (April 2012). Available from http://www.genomeweb.com/mdx/cancer-fight. 20. Ng KP, Hillmer AM, Chuah CTH, Juan WC, Ko TK, Teo ASM, et al. A common BIM deletion polymorphism

10

mediates intrinsic resistance and inferior responses to tyrosine kinase inhibitors in cancer. Nature Med. 2012:18(4):521-528. 21. Deloitte Economics. Improving the Quality Use of in Australia: Realising the Potential of Pharmacogenomics: Australian Centre for Health Research. 2008. Available from http://www.deloitte.com/view/en_AU/au/industries/Lifesciencesandhealth/c669bf2733101210VgnVCM 100000ba42f00aRCRD.htm 22. National Health and Medical Research Council. Clinical Utility of Personalised Medicine. (PS0001). Australian Government 2011. Available from http://www.nhmrc.gov.au/guidelines/publications/ps0001. 23. Sheffield LJ, Phillimore HE. Clinical use of pharmacogenomic tests in 2009. Clin Biochem Rev. 2009;30(2):55-65. 24. Harper AR, Topol E J. Pharmacogenomics in clinical practice and drug development. Nature Biotech. 2012;30(11):1117-1124. 25. Swen JJ, Nijenhuis M, de Boer A, Grandia L, Maitland-van der Zee AH, Mulder H, et al. Pharmacogenetics: from bach to byte – and update of guidelines. Clin Pharmacol Therap. 2011;89(5):662-673. 26. Collins FS. Mining for therapeutic gold. Nat Rev Drug Discov. 2011;10(6):397-397. 27. Franz DN, Leonard J, Tudor C, Chuck G, Care M, Sethuraman G, et al. Rapamycin causes regression of astrocytomas in tuberous sclerosis complex. Annals of 2006;59(3):490-498. 28. Royal College of Pathologists of Australia. RCPA Catalogue of Genetic Tests and Laboratories 2009. Available from http://genetictesting.rcpa.edu.au/ 29. Hahn SH. Targeted next-generation sequencing expands the spectrum of mitochondrial disorders. Genome Med. 2012;4(3): 22. 30. Lupski JR, Reid JG, Gonzaga-Jauregui C, Rio Deiros D, Chen DCY, Nazareth L, et al. Whole-Genome Sequencing in a Patient with Charcot–Marie–Tooth Neuropathy. New Engl J Med. 2010;362(13):1181- 1191. 31. Bainbridge MN, Wiszniewski W, Murdock DR, Friedman J, Gonzaga-Jauregui C, Newsham I, et al. Whole- Genome Sequencing for Optimized Patient Management. Science Translational Medicine 2011;3(87): 87re83. doi: 10.1126/scitranslmed.3002243. 32. Check Hayden E. Genome study solves twins' mystery condition. Nature 15 June 2011; doi:10.1038/news.2011.368. 33. Worthey EA, Mayer AN, Syverson GD, Helbling D, Bonacci BB, Decker B,et al. Making a definitive diagnosis: Successful clinical application of whole exome sequencing in a child with intractable inflammatory bowel disease. Genet Med. 2011:13(3):255-262. 34. Johnson M, Gallagher K. One in a Billion: A Boy's Life, A Medical Mystery. JSOnline: Milkwaukee Wisconsin Journal Sentinel. December 25, 2010. Available from: http://www.jsonline.com/features/health/111224104.html 35. The Human Variome website. Available from http://www.humanvariomeproject.org. 36. Williams SCP. in the Clinic. HHMI Bulletin 2012;25(2). Available from http://www.hhmi.org/bulletin/may2012/features/exome_sequencing_pop.html 37. Mihelec M, St Heaps L, Flaherty M, Billson F, Rudduck C, Tam PP, et al. Chromosomal rearrangements and novel genes in disorders of eye development, cataract and glaucoma. Twin Res Hum Genet. 2008;11(4):412-421. 38. Thanabalasingham G, Owen KR. Diagnosis and management of maturity onset diabetes of the young (MODY). BMJ;2011:343. doi: 10.1136/bmj.d6044.

11

39. Chen R, Mias GI, Li-Pook-Than J, Jiang L, Lam HYK, Chen R, et al. Personal Omics Profiling Reveals Dynamic Molecular and Medical Phenotypes. Cell 2012;148(6):1293-1307. 40. Conger K. Revolution in : First-ever integrative 'omics' profile lets scientist discover, track his diabetes onset. Inside Stanford Medicine (March 15, 2012). Available from: 41. National Health and Medical Research Council. Direct-to-Consumer DNA Genetic Testing: An information resource for consumers. (PS0004). Australian Government 2012. Available from http://www.nhmrc.gov.au/guidelines/publications/ps0004. 42. National Health and Medical Research Council. Medical Genetic Testing: information for health professionals. (E99). Australian Government 2010. Available from http://www.nhmrc.gov.au/guidelines/publications/e99. 43. Franchini M. Hereditary iron overload: update on pathophysiology, diagnosis and treatment. Am J Hematol. 2006;81:202-206. 44. Kujovich JL. Factor V Leiden thrombophilia. Genetics Med. 2011;13(1):1-16. 45. Department of Health and Ageing. The Personally Controlled eHealth Record System 2012. Available from http://www.ehealth.gov.au 46. Goodhand P, Orfeo M, Altshuler D (Eds.). Creating a Global Alliance to Enable Responsible Sharing of Genomic and Clinical Data. 2013. Available from: http://www.broadinstitute.org/files/news/pdfs/GAWhitePaperJune3.pdf 47. Goodhand P, Orfeo M, Altshuler D(Eds.). Fact Sheet: A Global Alliance to Enable Responsible Sharing of Genomic and Clinical Data. 2013. Available from: http://www.ucsf.edu/sites/default/files/fields/field_insert_file/news/AllianceFactSheet.pdf 48. United for Medical Research (UMR) report by Battelle Technology Partnership. The impact of genomics on the US economy. 2013. Available from https://www.dropbox.com/s/5fp88ytialrk3qz/The-Impact-of- Genomics-on-the-US-Economy.pdf 49. The 100K Genome Project. Available from http://www.genomicsengland.co.uk/ 50. Green ED, Guyer MS. Charting a course for genomic medicine from base pairs to bedside. Nature 2011; 470(7333):204-213.

12