Ethics and Challenges of Delivering Genomic Medicine
By: Michael Szego, PhD, MHSc
In 2000, one of the most significant milestones in genomics was achieved: the first draft of the human genome was completed (1). The genomic sequence was based on DNA samples pooled from several individuals and it established a reference genome for future sequencing projects. At the press conference announcing this achievement, Bill Clinton proclaimed that the completed genome sequence would “revolutionize the diagnosis, prevention and treatment of most, if not all, human disease” (2). Francis Collins, who led the effort to complete the human genome project, may have been trying to manage expectations when he suggested that a complete transformation of therapeutic medicine would take up to 15 to 20 years (2). One of the biggest hurdles that needed to be overcome before Clinton’s claims could be realized was the ability to sequence individual genomes. This feat required technological advances in sequencing and a drastic reduction in sequencing costs.
Since 2000, whole genome sequencing (WGS) of individual genomes has been developed, enabling the identification of new gene variants associated with disease that can subsequently be used for genetic testing in the clinical context (3). The WGS is still cost prohibitive with its use is limited to select research projects; however, its cost is decreasing rapidly and will soon be cheaper than currently employed genetic tests that assess one gene at a time. The development of affordable WGS may be the catalyst for the transformation that Dr. Collins predicted and the “revolution” that Clinton anticipated.
Since WGS has the potential to unlock every person’s unique disease risk profile (4), it may be one of the most significant technological breakthroughs in history. Therefore, WGS deserves special consideration from an ethics perspective. In this short article, I will focus on two key ethical topics with respect to WGS in that are found in both research and clinical settings, that is 1) informed consent and, 2) return of results.
The principle of informed consent is recognized as a main pillar in the practice of ethical research and medicine. For consent to be informed within the general research context, the research subject must be informed of the purpose for the research, its potential applications, the methods that will be employed, and any anticipated benefits and risks (5,6). However, many of these criteria are unrealistic when applied to genomic research specifically. At the time a DNA sample is taken from the research subject, all possible future research, its applications, and methods are usually not known. One main risk/benefit associated with most genomic research projects is the potential identification of incidental findings. For instance, the disclosure of a pathogenic variant that is clinically actionable which is identified over the course of research is an example of a potential benefit, while possible genetic discrimination is an example of a potential risk. However, the actual risks and benefits are not known beforehand because they depend on the research subject’s genomic sequence and the scientific knowledge at the time the analysis.
In order for genomic research to be performed, the traditional informed consent process has been modified to include broad consent. In the broad consent model, participants consent to a range of possible research activities (7). Under this paradigm, research subjects are educated about genomic research to ensure they understand the general risks and benefits. Consequently, they can make the most informed decision possible, even though the actual risks may not be completely known.
Informed consent is a different process within clinical medicine as compared to research. In the clinical genetic testing context, patients must be told the nature of the diagnostic test, the expected risks and benefits of the test, and any alternative tests for consent to be informed. When WGS is used as a clinical genetic test, the nature of the diagnostic test and any alternative tests are known and can be described to the patient. Current standards of care for clinical genetic testing include patient counseling about the risks and benefits and potential outcomes of the proposed genetic test. However, current genetic tests can examine one or a handful of genes, making counseling more straightforward than counseling on a WGS test, which examines over 20,000 genes and many of which have known pathogenic variants. That said, counseling is critically important within the WGS context; however, the information needs to be more general in nature, analogous to the research setting. Healthcare providers need to educate patients about genomic testing using theory and case-based examples, identify any likely clinical implications of WGS, and discuss the general risks and benefits of undergoing WGS in order for patients to make an informed decision as to whether they want their genome sequenced.
Return of results
Whether researchers have a duty to return individual research results in genetic studies has been the subject of much debate. Should research be purely for research sake? This debate is especially important within the WGS context since the likelihood of identifying pathogenic variants is high. Fortunately, a consensus is emerging. The World Health Organization has identified three conditions to be met before disclosure should occur: 1) the data should be clinically beneficial; 2) disclosure should avert or minimize significant harm; and 3) there is no indication that the individual in question would prefer not to know (8). Consistent with this approach, Canadian research ethics guidelines outline an obligation to disclose any material incidental findings or unanticipated discoveries made during the course of research that are interpreted as having a significant welfare implication for the participant (9).
Different strategies have been suggested to manage the return of clinically relevant research results in the WGS context. One approach would categorize human genes according to clinical parameters (10). In such a scheme, all gene mutations that are medically actionable are labeled as “Bin 1” genes and would trigger an automatic return of the result. “Bin 2” genes are defined as gene mutations associated with human disease that could not be acted upon medically. Finally, “Bin 3” contains all other genes whose association with human disease is unknown. It has been estimated that there are currently only 100 “Bin 1” genes in the human genome (10).
The ClinSeq project at the National Institutes of Health (NIH) has taken a different approach that is more research-subject-centered and engages research subjects to determine the type and extent of information they would find useful (11). In the NIH project, all pathogenic variants can be returned to research subjects, provided consent is obtained for disclosure. Additionally, they have set up an independent panel to periodically review any new evidence linking variants to human disease and to determine if the evidence is sufficient to warrant disclosure.
With all the effort spent on identifying variants associated with human disease, new clinically relevant variants will no doubt be identified in the future. This reality necessitates a long-term plan to deal with stored sequences. Researchers returning individual research results could implement software that can reanalyze past genomes and flag new clinically relevant data. Alternatively, if such a measure was not possible, research subjects should be informed during the informed consent process that reanalysis will not occur and that any results returned would only reflect current knowledge.
It is generally understood that test results are returned to patients, however, it is unclear what should be returned to patients when WGS is used as a genetic test. While any results that inform the original differential diagnosis seem appropriate to disclose, many other variants of known and unknown significance may also be detected that have nothing to do with the original query. A “binning” mechanism or a patient-guided approach may be borrowed from the research context described above. Lastly, the issue of clinical reanalysis was addressed in a recent article by the head of the NIH clinical sequencing project mentioned previously, Leslie Biesecker describes WGS as a resource not a genetic test (12). As such, a WGS dataset can be “interrogated by the patient and clinician in situations where it could be of potential use to the patient, when both agree to this use” (12). This type of integration would also solve the issue of reanalysis. If treated as a one off test, the physician who ordered it would not have a legal or ethical obligation to periodically reanalyze each of her patient’s genome for any new medical information. If comprehensively integrated into primary care, a patient’s whole genome could be part of the transformation of healthcare Francis Collins and Bill Clinton predicted over a decade ago.
While it is still unclear exactly how informed consent and return of results are going to look in the future context of WGS, many of the research projects employing WGS have an integrated ethics component, which includes research subject engagement. As such, I am confident we will establish ethical best practices when it comes to WGS in clinical and research settings. Furthermore, as WGS is exploited clinically, opportunities and challenges will be created for researchers and clinicians. Clinical WGS datasets could provide a rich source of data for researchers, provided that appropriate informed consent and privacy safeguards are put in place. This environment would represent a paradigm shift in which clinical medicine and research could occur using the same platform, facilitating knowledge exchange.
Michael Szego, PhD, MHSc
Centre for Clinical Ethics (A joint venture of Providence Healthcare, St. Joseph’s Health Centre, and St. Michael’s Hospital),
The University of Toronto Joint Centre for Bioethics
Research Ethics Consultant,
The Centre for Applied Genomics,
The Hospital for Sick Children,
The University of Toronto McLaughlin Centre for Molecular Medicine
1. Lander ES, et al. Initial sequencing and analysis of the human genome. Nature. 2001; 409(6822): 860-921.
2. Wade NA. Decade Later, Genetic Map Yields Few New Cures, in New York TImes.2010: New York.
3. Tucker TM, et al. Massively parallel sequencing: the next big thing in genetic medicine. Am J Hum Genet. 2009; 85(2): 142-54.
4. Tabor HK et al. Genomics really gets personal: how exome and whole genome sequencing challenge the ethical framework of human genetics research. Am J Med Genet A. 2011; 155A(12): 2916-24.
5. Human T.-C.P.S.E.C.f.R.I. 1998 Aril 9, 2009]; Available from: http://pre.ethics.gc.ca/eng/index/.
6. Beauchamp TL and Childress JF. Principles of biomedical ethics. 6th ed. 2009, New York: Oxford University Press. xiii, 417.
7. Singer PA and Viens AM. The Cambridge textbook of bioethics. 2008, Cambridge ; New York: Cambridge University Press. xvi, 538 p.
8. Organization, W.H. Genetic databases: assessing the benefits and impact on human and patient rights. 2003 [cited 2009 December 15]; Available from: http://www.law.ed.ac.uk/ahrb/publications/online/whofinalreport.doc.
9. Canadian Institutes of Health Research, N.S.a.E.R.C.o.C.a.S.S.a.H.R.C.o.C., Tri-Council Policy Statement: Ethical Consult for Research Involving Humans, 2010.
10. Evans JP and Rothschild BB. Return of results: not that complicated? Genet Med. 2012; 14(4): 358-60.
11. Biesecker LG et al. The ClinSeq Project: piloting large-scale genome sequencing for research in genomic medicine. Genome Res. 2009; 19(9): 1665-74.
12. Biesecker LG. Opportunities and challenges for the integration of massively parallel genomic sequencing into clinical practice: lessons from the ClinSeq project. Genet Med. 2012; 14(4): 393-8.