EELS (Ethical, Economic, Legal & Social) Article

The Pharmacogenomics Journal (2005) 5, 2–5. doi:10.1038/sj.tpj.6500288

Stored tissue samples: through the confidentiality maze

Y Joly1, B M Knoppers1 and M T Nguyen1

1Centre de recherche en droit public, Universite de Montreal, Montreal (Quebec), Canada

Correspondence: Y Joly, Universite de Montreal, CP 6128, succ. Centre-ville Montreal (Quebec), Montreal, Canada H3C-3J7. Tel: +1 514 343 6111 (5020); Fax: +1 514 353 6233; E-mail:

Over the last 20 years, developments in the field of genetics have allowed the use of new knowledge concerning genes and DNA to improve public health in multiple ways. One of the most promising developments in genetics, pharmacogenomics (PGx) was made possible by the completion of the human genome sequence as well as the development of biomedical data processing (throughout this article 'PGx' will be used to indicate pharmacogenomics, including pharmacogenetics. The terms pharmacogenomics and pharmacogenetics will be used separately when a distinction is made between the two disciplines). The terms 'pharmacogenetics' and 'pharmacogenomics' are often used interchangeably to refer to the relationship between genes and drug responses. However, both terms refer to slightly different concepts1 and there remains a lack of consensus on appropriate definitions.2

'Pharmacogenetics' was first coined in 1959 when it was observed that individuals differ in the way they respond to medicines due to individual genetic profiles.3 According to the Food and Drug Administration (FDA), pharmacogenetic testing involves: 'an assay intended to study interindividual variations in DNA sequence related to drug absorption and disposition (pharmacokinetics) or drug action (pharmacodynamics).'4

In the 1980s, with the identification of the molecular bases of hereditary traits and then later the completion of the human genome draft, pharmacogenomics became possible.5 The FDA makes a distinction between pharmacogenetics and pharmacogenomics and specifies that the latter involves:

[A]n assay intended to study interindividual variations in whole-genome or candidate gene single-nucleotide polymorphism (SNP) maps, haplotype markers, and alterations in gene expressions or inactivation that may be correlated with pharmacological function and therapeutic response.4

Thus, compared to pharmacogenetics, pharmacogenomics goes even further by identifying genes or whole genomes responsible for modifying an organism's response to drugs.6 Pharmacogenomics also includes the use of genomics in the search for new therapeutic targets.7, 8

PGx promises to provide specific treatments to tailor doses to the genetic profile of an individual.9 Being able to reduce the adverse effects of drugs is of great interest since it has the potential to reduce high morbidity rates and medical expenses.

According to some reports,10, 11 the ethical, social, and legal challenges relating to PGx are less prevalent than those resulting from research tests for disease susceptibility genes. However, this position remains controversial. Several authors argue that PGx research presents its own important challenges, which need to be addressed.2, 12, 13

Since the nature of the genetic information obtained through PGx might present different ELSI privacy challenges and a different level of risk, a number of issues specific to the protection of privacy for research participants will need to be answered. Accordingly, after studying the different guidelines and recommendations from major international, regional, and national organizations, we propose points to consider for researchers and institutional review boards (IRB) to aid in determining the suitable level of protection of genetic data.


Privacy Challenges of Pgx

According to optimistic reviews:

The Renaissance period of the world history is analogous to the renewal of healthcare that will arise from pharmacogenomic discoveries. Just as geography, science, art and communication were reawakened by the works of Columbus, da Vinci, Michelangelo and Gutenberg; genetic science will revitalize the clinical role of pharmacists and [s]timulate a gradual rebirth in the way patient and healthcare providers think about disease prevention and treatment.14

Rhetoric aside, understanding the association between genetic variation and drug responses promises many benefits.2, 3 The scientific community has not yet reached a consensus on the global effect that pharmacogenomics will have. Most agree that, in the long term, it will have a positive impact on the management and care of many common, significant diseases such as cancer, AIDS, and various types of mental disorders.15, 16, 17

There are numerous potential benefits for PGx. Yet, it is necessary to articulate the specific privacy risks and challenges that may hinder the development of PGx research in order to separate myth from reality: 'Pharmacogenomics is not a panacea for all that ails individuals, corporations, and governments. [D]eflating the hype and 'unpacking' the rhetoric is essential for a social and ethical critique of the overly optimistic vision of pharmacogenomics'.18

It is important to recognize that testing for PGx research comes in a variety of forms. Some research will look at the genotypic variation, others at the gene expression (ie RNA or protein level)19 and others at the viral17 or tumoral genotype.15 Therefore, PGx research involves informational risks which, in some circumstances, could be lower than the risks associated with genetic research for disease genes.20, 21

While the level of physical risk associated with most PGx research is generally low, like other categories of genetic research, the informational risk relating to a breach of confidentiality is perceived to be an important issue in PGx.22 This informational risk is due to the personal, familial, and social nature of genetic information as well as its potential to discriminate and stigmatize.23 Like other genetic tests, information obtained through PGx has a familial component that goes beyond the individual being tested.

In PGx research, information obtained will generally be limited to the association between a gene, a polymorphism (ie SNPs), and a reaction to a particular drug or a class of drugs. Consequently, several authors have affirmed that the level of privacy risk associated with PGx research results will usually be lower than other types of genetic or genomic research. Yet, these authors are generally only referring to genotyping for inclusion in pharmacogenetic research, so this assertion should not be applied to all types of PGx research (such as the research for new therapeutic targets).2, 21, 24 It should also be noted that, contrary to other types of genetic studies, participants who are recruited for profiling for PGx research are often already diagnosed with a disease and may have at their disposal a marketed drug as a treatment.25

However, given the complexity of the human genome and the interactions between genes, as well as between genes and the environment, it is possible that certain discoveries in PGx research could enable a genetic diagnosis of an individual's susceptibility to a certain disease.26 For example, by using the gene candidate approach, it is possible to isolate a particular gene that is associated both with an individual's drug response and to his or her risk of disease.3 The fact that an individual may be more susceptible to having adverse reactions, or, may not respond to certain classes of drugs, may have implications for insurance and employment in countries lacking a universal health care system.12

Although these and other findings make it difficult to adopt a stable classification into 'high-risk' test (eg genetic testing for disease, newborn screening) and 'low-risk' test (eg Pgx),27 it must still be acknowledged that pharmacogenomics involve informational risks and that in some research these risks will be lower than, for example, the risks associated with genetic research for disease genes or genetic screening of newborns.


Points to Consider for Confidentiality in Pgx Research

We have seen that the objectives and implications of PGx research may differ from other types of genetic research; the applicable level of confidentiality of the information will have to be adapted to the realities of this particular type of research. Aware of this need as well as the need for uniformity in the field, the Pharmacogenetics Working Group in the United States and the European Agency for the Evaluation of Medicinal Products in Europe have defined various levels of confidentiality within the framework of PGx research (presented in Table 1).21, 28 PGx researchers will have to decide the level of protection that will best suit their research protocols.

The objective in choosing the appropriate confidentiality level should be to maximize individual and societal benefit from PGx in accordance with recognized ethical principles. As concerns PGx information, recognized ethical principles demand that the data banks and DNA samples be protected against breaches of confidentiality. However, requiring the highest level possible of confidentiality, for example, having recourse to anonymized samples only, would not only limit the efforts of research but would ultimately result in preventing the participants themselves from receiving any results.24 It must also be recognized that it might be technically impossible to achieve perfect anonymization.29 This demonstrates that maximizing the protection of confidentiality to reduce the level of risk to zero will often be inappropriate, since this may result in sacrificing other important values.24 Unless the potential benefits of a specific PGx research are adequately evaluated, it is difficult to choose appropriately between the various options in reducing informational risks. It will therefore be necessary to determine the level of protection of confidentiality that is most appropriate and to evaluate the risks and benefits according to each particular case.6, 22 It could well be argued that in pharmacogenetics trials, the fact that drugs are being tested mandates a coded or double-coded approach for the possible immediate benefit of participants, since in this way these individual results can be returned. The same would not hold true in studies of haplotype or of genetic variation.

Points to consider found through a selective analysis of the literature and guidelines relevant to confidentiality in PGx research can be used as guides to determine the appropriate level of protection for particular research protocols. These same points to consider can also be used by the IRB when reviewing different research protocols.

It goes without saying that the general core elements of review for research involving human tissues still apply:

  • nature of the research;24
  • intended use of the samples;24
  • length of conservation;30
  • secondary uses;25
  • legal context;24
  • specific concerns raised by the researcher, the ethics committees and sponsors;31
  • selection of the greatest degree of privacy protection compatible with the objectives of the research.6

The points to consider are as follows:

  1. The use of identified samples in a pharmacogenomic research protocol will be difficult to accept and must always be justified by the researcher.5 In fact, according to the Pharmacogenetics Consortium, this protection level places the full responsibility (and possibly legal liability) on the researcher to ensure that sensitive information is not conveyed to other parties.24
  2. The use of coded samples is acceptable on the condition that it is justified by the researcher or sponsor. This level of protection will make it possible for the researcher to report clinically significant personal results to individual subjects or their treating physicians. However, the researcher will have to ensure that all information and samples will be provided maximum protection (to ensure the reliability and safety of the system protecting the information). This level of protection will be especially useful within the framework of research having immediate implications for the participant.26, 28
  3. Double-coded samples offer an interesting alternative in cases where the research is not likely to provide information of immediate clinical interest for participants. Even though it remains possible to give out individual results, the process is more complex than in the case of coded samples. The researcher having recourse to double coding will have to specify in the research protocol the particular cases where it will be necessary to ask the key holder to break the codes (ie in case of emergency, possible medical significance, if the participant wishes to withdraw, etc).26, 28
  4. Anonymized samples radically reduce the value of the patient's participation in PGx research. Subsequent to anonymization, it will no longer be possible to obtain additional information from the patient's medical file. Once the trial is completed, for example, this constraint prevents the researcher from verifying if the observed responses to medication during research combined with other research results are due to genomic or genetic variations, or due to other factors (TBA).24, 28 Anonymized samples will be useful for research aimed at generating assumptions, but much less interesting in the case of pharmacogenetics clinical trials whose goal is to support 'label claims'.27 It will also imply an important role for the IRB during subsequent research studies since the participant will no longer be able to express personal wishes, receive individual results, or withdraw from the study. Participants will need to be informed of the limitations, which this type of protection imposes on their rights. As well, participants will have to be informed precisely concerning how samples will be used subsequent to anonymization. On this, the Nuffield Council on Bioethics considers that it is possible to request broad consent with regard to the uses of samples that are already anonymous or to be anonymized.6 However, the anonymization of the data and samples should not be used as an excuse for researchers to simplify ethical or legal hurdles or to neglect the participant's right to self-determination.
  5. Completely anonymous samples are extremely rare and of marginal interest. Furthermore, regulatory authorizations as well as the advancement of science could be impeded since certain validation criteria cannot be met.24, 26, 28



For a long time, it has been suspected that heredity may have an impact on individual response to drugs. Developments in PGx research promise multiple applications that could transform our health systems, and some even speak of a new paradigm.32

However, if the public is to fully profit from the benefits of pharmacogenomics, policy makers must recognize the specificity of this type of genetic research and develop requirements for the confidentiality levels that correspond to the realities and needs of PGx research.

Nevertheless, we are only beginning to understand the degree to which research subjects value and insist on coding and other methods of ensuring anonymity and confidentiality.33 PGx researchers will have to avoid uniformally applying the same level of protection of confidentiality to all research protocols. More precisely, researchers will have to proceed with each project on an individual basis, taking guidance from the reference points suggested in this article. The researcher will also have to inform the participants of the degree of protection applicable to the research protocol in order to obtain a truly informed consent.



None declared.




1. Health Council of the Netherlands. Advisory Report—Pharmacogenetics. Health Council of the Netherlands: The Hague 2000;.
2. Nuffield Council on Bioethics. Pharmacogenetics Ethical Issues Nuffield Council on Bioethics: Brimsdown 2003; Website: date accessed: May 27 2004.
3. Lerer B & Understanding pharmacogenetics. Psychiatric Times 2003; 20:5 Website: date accessed: June 1 2004.
4. Food and Drug Administration. Guidance for Industry: Pharmacogenomic Data Submissions Food and Drug Administration: Rockville 2003; Website: date accessed: May 27 2004.
5. Binzak BA. How pharmacogenomics will impact the federal regulation of clinical trials and the new drug approval process. Food Drug Law J 2003; 58: 103−127. | PubMed |
6. Vazar LT, Rosen GD & Raffin TA. Pharmacogenomics and the challenge to privacy. Pharmacogenomic J 2002; 2: 144−147. | Article | ChemPort |
7. Paul NW & Roses AD. Pharmacogenetics and pharmacogenomics: recent developments, their clinical relevance and some ethical, social, and legal implications. J Mol Med 2003; 81: 135−140. | PubMed |
8. Meyer UA. Introduction to pharmacogenomics: promises, opportunities, and limitations. In: Licinio J, Wong ML (eds).Pharmacogenomics—The Search for Individualized Therapies Wiley-VCH: Weinheim 2002; pp 1−7. | ChemPort |
9. Poste G. Extract from the nature biotechnology round-table entitled: pharmacogenomics: revolution or reaction? Genetic variation in drug development. Nat Biotechnol 1998; 16: 16−19. | PubMed |
10. Roses AD. Pharmacogenetics and the practice of medicine. Nature 2000; 15: 857−865 Anderson DC, Gomez-Mancilla B, Spear BB, Barnes DM, Cheeseman K, Shaw PM et al. Elements of informed consent for pharmacogenetic research. J Pharmacogenomics 2002; 2: 284−292. | Article | ChemPort |
11. Sass H-M. Bioethical issues in genetic screening and patient information. T Klin J Med Ethics, Law History 2004; 12: 1−9.
12. Rothstein MA & Epps PG. Ethical and legal implications of pharmacogenomics. Nat Rev Genet 2001; 2: 228−231. | Article | PubMed | ISI | ChemPort |
13. Bolt LLE, Van Delden JJM, Kalis A, Derijks HJ & Leutfkens HGM. Tailor-made pharmacotherapy: future developments and ethical challenges in the field of pharmacogenomics CBG: Utrecht 2002;.
14. Brock TP, Valgus JM, Smith SR & Summers KM. Pharmacogenomics: implications and considerations for pharmacists. Pharmacogenomics 2003; 4: 321−330. | Article | PubMed |
15. Cobleigh MA, Vogel CL, Tripathy D, Robert NJ, Scholl S & Fehrenbacher L et al. Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2-overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease. J Clin Oncol 1999; 9: 2639−2642.
16. Masellis M, Basile VS, Gubanov A & Kennedy JL. Psychiatric pharmacogenetics: prediction of treatment outcome in schizophrenia. In: Licinio J, Wong ML (eds).Pharmacogenomics: The Search for Individualized Therapies Wiley-VCH: Weinheim 2002; pp 369−375. | ChemPort |
17. Hirsch MS, Brun-Vezinet F, D'Aquila RT, Hammer SM, Johnson VA & Kuritzkes DR et al. Antiretroviral drug resistance testing in adult HIV-1 infection. JAMA 2000; 283: 2417−2426. | Article | PubMed | ISI | ChemPort |
18. William-Jones B & Corrigan OP. Rhetoric and hype where's the 'ethics' in pharmacogenomics? Am J Pharmacogenomics 2003; 3: 375−383. | PubMed |
19. Freund CF & Clayton EW. Pharmacogenomics and children. Am J Pharmacogenomics 2003; 3: 399−404. | PubMed |
20. Haga SB, Khoury MJ & Burke W. Genomic profiling to promote a healthy lifestyle: not ready for prime time. Nat Genet 2003; 34: 347−350. | Article | PubMed | ChemPort |
21. Epstein CJE. Genetic testing: hope or hype? Genet Med 2004; 6: 165−172. | Article | PubMed |
22. Pharmacogenetics Working Group (Anderson DC et al). Elements of informed consent for pharmacogenetic research—perspective of the Pharmacogenetics Working Group. Pharmacogenomics J 2002; 2: 284−292. | Article | ChemPort |
23. Kosseim P, Letendre M & Knoppers BM. Protecting genetic information: a comparison of normative approaches. GenEdit 2003; 2: 1−8 Website:
24. Consortium on Pharmacogenetics (, Buchanan A, McPherson E, Brody BA, Califano A, Kahn J, McCullough N & Robertson JA. Consortium on Pharmacogenetics (, Buchanan A, McPherson E, Brody BA, Califano A, Kahn J, McCullough N & Robertson JA. ). Pharmacogenetics: ethical and regulatory issues in research and clinical practice. Consortium on Pharmacogenetics 2002;.
25. McPherson EC, Buchanan AE & Monasco PK. Pharmacogenetics: consideration of ethical, legal and social implications. Presented at the 2000 ASHG Annual Meeting, Philadelphia, 2000;.
26. Italian Society of Pharmacists et al. Italian Proposed Guideline for the Evaluation of Pharmacogenetic Research Italian Society of Pharmacists et al: Verona. 2001; Website: date accessed: June 1 2004.
27. Netzer C & Biller-Andorno N. Pharmacogenetic testing, informed consent and the problem of secondary information. Bioethics 2004; 18: 344−360. | Article | PubMed |
28. European Agency for the Evaluation of Medicinal Products. Position Paper on Terminology in Pharmacogenetics European Agency for the Evaluation of Medicinal Products: London 2002; Website: date accessed: June 1 2004.
29. Lin Z, Owen AB & Altman RB. Genomic and human subject privacy. Science 2004; 305: 183. | Article | PubMed | ISI | ChemPort |
30. National Bioethics Advisory Commission. Research Involving Human Biological Materials: Ethical Issues and Policy Guidance National Bioethics Advisory Commission: Rockville 1999;.
31. GlaxoSmithKline. The Science of Genetics and Modern Medicine GlaxoSmithKline: Bruxelles 2001; Website: date accessed: June 1 2004.
32. Finkelstein SN, Sinskey AJ & Cooper SM. The coming paradigm shift in pharmaceuticals. PharmaGenomics 2002; 26−28.
33. Rothstein MA & Hornung CA. Public attitudes about pharmacogenomics. In: Rothstein MA (ed).Pharmacogenomics—Social Ethical and Clinical Dimensions John Wiley & Sons, Inc.: Hoboken 2003; pp 3−27.


We acknowledge the invaluable assistance of Ms Ruth Essebag, Genetics and Society Project ( Research funding for this article came from Valorisation Recherche Quebec (VRQ), Genome Quebec and Genome Canada.