Compute for personalized medicine

It’s Changing Faster than Moore’s Law, but Is U.S. Policy Keeping Pace? In 1990, the U.S. launched an audacious scientific endeavor with the potential to changethe practice of medicine when the National Institutes of Health and the Department ofEnergy joined with the international community in a quest to sequence all 3 billion let-ters, or base pairs, in the human genome, which is the complete set of DNA in the humanbody. This concerted, public effort was the Human Genome Project (HGP). By 2000, sci-entists broke the code and paved the way for an explosion of investment in genetic andgenomic testing, generating 116,000 U.S. jobs and USD 16.5 billion in national economicoutput. These developments are being repeated in Oslo, Beijing, and around the world.
Stemming from the human genome sequencing is a new field referred to as personalizedmedicine, where providers and patients use diagnostic tools to identify specific molecularcharacteristics to help assess which medical treatments and procedures are best for thepatient. By combining an individual’s medical history and circumstances with this informa-tion, providers can develop customized treatment and prevention plans for patients whowill benefit, sparing side effects and expense for those who will not. For example, teststhat read the DNA structure of the most common form of leukemia in children havehelped boost the 10-year survival rate from 4 percent in the 1960s to more than 80percent today.1 Using the guidance from genetic tests, in the future physicians will more be increasingly able to prescribe the right drug, at the right time, in the right dosage.
methods is expected to make whole-genomesequencing both affordable and essential in giving a multi-faceted view of the patient’s USD 3 billion to complete in 2001, can now health, the biological basis of cancer, infec- be accomplished in about a day for less than tious diseases, inherited diseases, and drug USD 10,000 (Figure 1). Soon, that cost will response. Technology advances will make it likely drop below USD 1,000. Illumina, whose possible for the sequencing of individual HiSeq* DNA sequencing systems produce the genomes to become the standard and routine offers its sequencing services in bulk for as Changing Policy to Capture theOpportunity Now that technology has moved us towarda new environment where understandingthe genome is leading to unforeseen break-throughs in identifying new strains of com-plex diseases and offering treatments indi-vidualized for patients, we need to rethinkthe laws and regulations governing geneticinformation.
More than 90 percent of U.S. physicians areusing electronic health records (EHRs)according to a May 2013 Accenture report.
This is incredible progress and a criticalbasis for becoming the data engine to pro-vide a comprehensive data summary of apatient’s health and well-being. However,we only see the tip of the iceberg until we Figure 1. Sequencing Cost per Genome (Source: National Institutes for Health) have the genetic information built into thepatient information document and canensure that the records are available to the Making genomic data and tools interopera- patient’s care team Then, through analytics ble in a secure and trusted way will gener- medicine will replace today’s trial and error efforts. That is where policy can help deliv- er the comprehensive record—by supporting researchers, and healthcare professionals.
the integration of whole-genome-sequenceddata into health records (including clinical decision support tools) using the U.S.
"meaningful use" requirements to activate widespread adoption by providers. This will value of the data held in millions of patient not only allow physicians to have a full pic- face (API) supporting apps and services.
records—data that could help researchers ture of a patient’s medical history, but it may also serve as an invaluable platform for interoperability, scalability, stability, treat high blood pressure than a less costly generic, or what the rate of increase is in markers with clinical data. The EHRs give the diagnosis of Alzheimer’s. Today, the use of much-needed context to the genomic infor- this information is regulated by a series of mation. Clinical decision support tools inte- data for disease cures and treatments.
consent requirements constructed for a very grated with medical records are essential to different kind of research. In clinical trial allow physicians easy access to new patient- research, when comparing groups of patients appropriate diagnostic tests, as well as to use of millions of variants for each indi- taking new drugs to those who take placebos, automated resources for the interpretation discreet consent is necessary. But how do we keep them all? How do we access theinformation we need in real time? use today’s population data, now document- mant, and robust to effectively mitigate risk workarounds when security gets in the way.
groups? And what is appropriate for broad consent, for now and the future, when look- ing at de-identified data to use in research Accountability Act of 1996 (HIPAA) has many Today, it is not possible to predict which privacy provisions designed to safeguard medical changes in DNA sequence lead to clinical con- information and restrict access to it, yet it serves sequences. When held against a large reposi- as an important reminder of how policy can have tory of other such data, robust patterns and unintended consequences. Since becoming law, relationships can be identified which will HIPAA has had a chilling effect on scientific require millions of samples from real patients, investigation. For example, retrospective, chart- their treatments, and conditions. Researchers based research is difficult to conduct due to reg- are hungry for the ability to access data on a ulatory hurdles, and response rates in follow-up much wider basis than registries offer today.
investigations are very low given the challenges Yet, growing concerns regarding the ability to keep the data private need to be addressed.
Harvard University Professor Latanya Sweeney Another important privacy and safety provision published findings in April 2013 showing that is the Common Rule, which protects patients in federally funded clinical trials. It requires researchers to obtain informed consent and participants in a recent DNA study, highlight-ing the risk that many see to sharing their puts in place other provisions to protect patients in those trials, explicitly safeguardingvulnerable populations such as pregnant To enable greater sharing of patient data for women, children, and prisoners. The principles research, privacy and security risks must be of the Common Rule are also transferred to kept manageable. De-identification is a key non-federally-funded research through other safeguard, but not a panacea. As the Harvard regulatory guidelines. Thus, the overall system Since DNA cannot be effectively de-identified, example shows, there is residual risk with de- of clinical trials is coordinated to ensure privacy how do we design informed consent while bal- identified data that can enable re-identifica- and safety, and these rules extend to studies ancing with considerations of patient privacy in tion and breach, especially with new types of data, such as DNA, that we are just starting to patients the tools to decide who will see and How does the U.S. balance the need for patient research require some fields, such as the data that will accelerate research and provide research/public health or give patient discounts patient's age or zip code, that would normally individualized treatments with common and on health coverage and early access to clinical be eliminated during de-identification. To keep rare diseases against the interest in keeping privacy and security risk manageable, enable much broader sharing of data, and support As data sharing increases in scope, research research that requires more than fully de- participants will no longer be asked to consent access by researchers to vast collection of data, identified data, the best practice of a multi- to a single study, but rather to make their but must adhere to the research participants' layered approach to security should be used.
data available to a large number of researchers, conditions of access. When and how will partici- De-identification is combined with other safe- likely from different countries. Public trust in pants be allowed to access their own data? We guards, including encryption, tokenization, and the procedures used to store and access data need policies that allow the return of individual access controls, which must be usable, perfor- Today’s privacy and consent requirements Regardless of the stakeholder position on need a twenty-first-century review of the products. Because reimbursement for high- the agency of record, it is time for policy- data that will allow efficient, secure access for value products must be driven by true clinical makers to clarify agency jurisdiction, reduc- benefit for the covered population, criteria for appropriately protecting the confidentiality of demonstrating both clinical utility and validity growth of this critical industry to the detri- individual patients. Patients should be empow- must be developed and standardized. These cri- ered to have a right to share their data as well teria can then be used to guide both product as to protect the data they choose. This will development and reimbursement decisions.
require new protocols where individuals can The U.S. has one of the world’s most far- dictate access to their information for medical reaching protections for genomic informa- increase CMS funding unless a comprehensive or research purposes. Privacy is ensured by tion. GINA, the Genomics Information Non- cost analysis is considered, including misallo- limiting data access to authorized users and cated treatments, missed diagnosis of early stage disease, and the total costs of trial and The penalties for lack of controls need to be health. GINA creates federal rules to protect specified, much like the breach notification regulations in the U.S. HITECH Act, where employers and insurers. It is a milestone and institutions are required to alert patients of Is a new framework needed for regulation of a best practice for the rest of the world. clinical laboratory tests that generate genetic-based data? Regulation of in vitro diagnostics in To use patient-specific health information available through genomic mapping, we must Insurers, including Medicare, are funding remove the stigma associated with genomic genomic testing for specific chronic diseases— testing and integrate it ubiquitously as yet operating under its authority to regulate like cancer, HIV, and heart disease—in patients another set of lab test results. If we want to and their families. This has significantly increased both predictive and preventative CMS, operating under the authority of the options. However, the limits on reimbursement place both the patient and the healthcare sys- lower-cost treatments as the new standard tem at risk. As Congress considers cost-saving of care. As patients, we can take control and policies that will transform healthcare delivery, Are these regulatory systems equipped to deal manage our diets, behavior, and exercise, but genomic mapping will need to be one of the with complex, high-value tests that draw on not be penalized for what we cannot control.
chief considerations to enable our healthcare cutting-edge genomics technology to directly An informed partnership among the patient, practices to target and accelerate care. Genetic inform high-stakes clinical decisions? Of the mapping will provide the research that will lead two agencies with jurisdiction, CMS has respon- patients for test results but reward proac- to savings from eliminating unnecessary tests sibility for laboratory-developed tests (LDTs) through its regulation of the quality and safety standards for labs. The FDA regulates in vitro Today’s static system of codes prevents the diagnostic products (IVDs) as medical devices Centers for Medicare and Medicaid Services and has stated its authority to regulate LDTs, How can the U.S. close the loopholes that (CMS) from differentiating single tests for dis- but the agency has exercised its enforcement were left in the Act, including nondiscrimina- ease rather than creating a dynamic testing discretion and refrained from regulation.
tion based on genetic information for mort- system based on the value of the test. CMS However, the FDA has stated its intention to gages, long-term care insurance, the disabili- payment needs to be realigned with the com- apply risk-based oversight of LDTs as medical ty and life insurance marketplaces, and cov- plexity of the test and the risk associated devices. Final guidance providing illustrative with the disease. Additionally, CMS should set examples that distinguish products that will be standards for the evidence that the agency subject to pre-market approval from those that and other payers will require to validate the benefits of the tests on a value-based cover- The future is here, the research is ongoing,and the challenge is to integrate the new datainto advanced clinical decision support soft-ware connected to EHRs with the clinicaltraining required to use the patient data andbring this science into daily clinical operations.
Centers like Cleveland Clinic are holding sum-mits for clinicians to earn CME credits whilebuilding upon the genetic data that is famil-iar—such as family history—and migrating topharacogenomics and more as clinicians beginto weave these additional sources of data intotheir workflows. At Mayo Clinic, the goal is toget every physician to use personalized medi-cine, starting with an alert system issued bythe EMR. For example, when a physician pre-scribes abacavir, an AIDS drug that causessevere reactions in some people who havecertain genetic variants, the EMR alert popsup and gives the name and pager of an expertin Mayo whom the prescriber can call foradvice. The Clinical Decision Support tool,combined with genomics data, is changingmedicine today.
How to change workflows and educationrequirements to convince doctors to adopt institutions, technology corporations, and poli- Five years later, many of the issues remain to personalized medicine within their practices.
be addressed and resolved. The roadmap is In 2008, The President’s Council on Science and In the absence of an open and interopera- Technology issued “Priorities for Personalized next-generation sequencing, we need to get Medicine,” calling for “the Federal government, the policy right to ensure the benefits from through the leadership of HHS, to join with the this miraculous new science, providing indi- private sector to create a public/ private sector vidualized treatments for patients at a speed ‘Personalized Medicine R&D Roadmap’ for coordi- nating discovery and translational research in personalized medicine.” The report focused on Integrating personal health into clinical prac- policy recommendations for technology, regula- tice requires policy and practice changes which are today being led by academic medical 1New England Journal of Medicine, 2006, 2001, Personalized Medicine Coalition, 20062Forbes, Harvard Professor Re-Identifies Anonymous Volunteers In DNA Study, April 25, 20133Creating a Global Alliance to Enable Responsible Sharing of Genomic and Clinical Data,http://www.ucsf.edu/sites/default/files/fields/field_insert_file/news/White%20Paper%20May%2029%20DAPG.12.54pm.pdf, May 27, 2013, pp 6-7.
4Personalized Medicine Coalition, Personalized Medicine Regulation Pathways for Oversight of Diagnostics, 2012http://www.personalizedmedicinecoalition.org/sites/default/files/files/Personalized_Medicine_Regulation-Pathways_for_Oversight_of_Dx.pdf Copyright 2013 Intel Corporation. All rights reserved.
Intel and the Intel logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States andother countries. *Other names and brands may be claimed as the property of others.
INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPELOR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL'S TERMS ANDCONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIEDWARRANTY, RELATING TO SALE AND/OR USE OF INTEL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR APARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT.

Source: http://www.intel.es/content/dam/www/public/us/en/documents/white-papers/compute-for-personalized-medicine-paper.pdf