Recommendation 1. Federal regulations governing human subjects research (45 CFR 46) that apply to research involving human biological materials should be interpreted by the Office for Protection from Research Risks (OPRR), other federal agencies that are signatories to the Common Rule, IRBs, investigators, and others in the following specific ways:
- Research conducted with unidentified samples is not human subjects research and is not regulated by the Common Rule.
- Research conducted with unlinked samples is research on human subjects and is regulated by the Common Rule, but is eligible for exemption from IRB review pursuant to 45 CFR 46.101(b)(4).
- Research conducted with coded or identified samples is research on human subjects and regulated by the Common Rule. It is not eligible for exemption unless the specimens or samples are publicly available as defined by 45 CFR 46.101(b)(4). Few collections of human biological materials are publicly available, although many are available to qualified researchers at reasonable cost. Therefore, OPRR should make clear in its guidance that in most cases this exemption does not apply to research using human biolResearch conducted
NBAC’s first recommendation offers guidance to “improve the interpretation and implementation of [the Regulations]3 with respect to human subjects research. NBAC proposes specific criteria, based upon the identifiability of samples, to determine whether a given research project involving human biological materials should be considered human subjects research. To place the discussion of this recommendation in context, the Working Group refers to the definition of “human subject” in 45 CFR 46.102(f): “a living individual about whom an investigator (whether professional or student) conducting research obtains (1) data through intervention or interaction with the individual, or (2) identifiable private information.”
The Working Group agrees with Recommendation 1(a), which is consonant with current interpretation of federal human subjects protection regulations.
Although the Working Group agrees that there are instances in which study of anonymized human tissue is human subjects research, the Working Group is concerned that NBAC’s proposal in Recommendation 1(b) goes beyond the regulatory definition of human subjects research and raises questions about whether this broader interpretation is necessary or enforceable under current regulations.
NBAC defines “unlinked” or anonymized samples as those that “lack identifiers or codes that can link a particular sample to an identified specimen or a particular human being.” The Working Group agrees with this definition and notes that an appropriately and effectively anonymized sample is functionally indistinguishable from an unidentified sample. Therefore, the Working Group believes that use of appropriately and effectively anonymized samples presents no more and no less potential for harm to an individual than use of anonymous samples. Whether the sample was made anonymous at the time of collection, at some point following the collection, or just prior to usage in research, once the sample has been anonymized, it can no longer be linked to an individual. Consequently, the researcher using such samples cannot be said to be performing human subjects research. The Working Group remains persuaded that, in all but one important case (see response to Recommendation 3), research conducted with effectively unlinked or anonymized samples that cannot be identified by the researcher is not human subjects research and is therefore not covered by the Regulations.
The Working Group notes that whether or not a proposal involves human subjects research is not always obvious. Current regulations, supported by OHRP guidance, hold institutions – not the researcher – responsible for making such a determination.4 Certification of whether a research project involves human subjects is only one example of the ways in which the system of human subjects protection relies upon institutional policies and practices, established under the Regulations and OHRP’s assurance process, to determine whether research is subject to the Regulations.
While the Working Group agrees that all research on identified samples and most research on coded samples is research on human subjects, the Working Group notes that some research on coded samples may fall outside the definition of human subjects research. In addition, the Working Group has serious concerns about NBAC’s interpretation of the term “publicly available”.
Coded samples can only be linked to individuals when the key needed to decipher the code is accessible. When the key exists but is inaccessible to the researchers and collaborators, those coded samples are effectively unlinked or anonymized for purposes of that research project. Since some important research is conducted using this model, OHRP has developed and published guidance for determining whether a researcher using coded samples is engaged in human subjects research.5 OHRP has determined that if a repository operating under the Regulations has a written agreement with a researcher that prohibits the researcher and collaborators from gaining access to the key, then the researcher has no means of identifying the source of the sample and therefore is not engaged in research on human subjects. OHRP has also, on a case-by-case basis, extended the application of this guidance to other situations where a similar written agreement exists. The Working Group endorses this guidance, which recognizes circumstances in which research using coded samples is not human subjects research.
The above example demonstrates the difficulty of using categories of samples (unidentified, unlinked, coded, identified) exclusively to determine whether or not research involves human subjects. The Working Group recognizes that, in the day-to-day research environment, involvement of human subjects and degree of research risk do not always have a one-to-one correspondence with the types of human biological samples used.
The recently issued HIPAA privacy rule also addresses the issue of research conducted with coded information. HIPAA refers to information as “de-identified” if specific data elements have been coded and permits use of “de-identified” information for research without IRB review or patient authorization. Therefore, some research using coded samples may be covered by the Common Rule but exempt from the HIPAA privacy rule. The need to analyze the implications of the interaction between the HIPAA privacy rule and NBAC’s recommendations is addressed in the Working Group’s response to Recommendation 14.
The Working Group agrees with NBAC’s statement that if coded or identified samples are publicly available, then research using these samples is exempt from the Regulations. However, the Working Group does not concur with NBAC’s conclusion that “few collections of human biological materials are publicly available, though many are available to qualified researchers at reasonable cost.6 The Working Group believes that NBAC’s definition and examples of human biological materials that are “publicly available” are overly restrictive and would, if implemented, impede certain research that is otherwise exempt under 45 CFR 46.101(b)(4).
Although OHRP has no formal guidance on the definition of “publicly available”, on a case-by- case basis OHRP has provided informal advice that interprets “publicly available” as commercially available or available in a reasonably unrestricted fashion. At times, OHRP has also interpreted “publicly available” as widely available to responsible parties (emphasis added), especially if there is a statute or other legal requirement mandating such availability. OHRP’s interpretation therefore recognizes that a research resource, such as a human biological materials repository, may meet the definition of “publicly available” even if access to the samples requires a reasonable commercial fee or handling charge and/or demonstration of the ability to manage the materials responsibly.
The Working Group calls attention to the American Type Culture Collection (ATCC), a source of publicly available samples, to illustrate that requiring the payment of commercial fees and demonstration of responsibility are compatible with the principle of public availability. The ATCC holds approximately 1500 cell lines that are derived from human cells. Catalogs of repositories such as the ATCC are freely available to the public. Access to its web sites, on which catalogs are posted, is also not restricted. Therefore, access to information about these cell lines is not constrained in any significant way. To obtain materials from ATCC, investigators pay a fee and sign an agreement stating that they will abide by ATCC rules regarding use of the materials, that they are qualified to work with the materials, and that they agree in writing to assume all risk and responsibility in connection with the receipt, handling, storage and use of the materials. Most of ATCC’s human-derived cell lines have been distributed to thousands of laboratories over the last few decades.
For other types of human biological materials and repositories, arguments that samples are not publicly available are valid. Samples should not be considered publicly available if information about the existence of samples is available to only a select group. In addition, if only a small and select group can meet conditions required for release of materials, samples are not deemed to be publicly available. There is often a correlation between the type of sample (for example, a renewable cell line or a finite amount of the originally collected tissue) and the conditions that a repository imposes before release of samples. However, the correlation is not sufficiently strong to permit an unequivocal determination a priori about whether or not a given sample is publicly available.
The Working Group concludes that interpretation of the term “publicly available” must be sufficiently broad to permit application to a wide variety of repositories and to the human biological materials in their custody. The Working Group expects there will be instances in which it will not be absolutely clear whether samples are publicly available and therefore recommends that OHRP continue to provide advice on a case-by-case basis.
The Working Group proposes that, with respect to existing guidance applicable to use of human biological materials in research, OHRP be charged with the following:
- Conduct outreach and educational efforts to underscore existing OHRP guidance, which recognizes that research with “unlinked” or anonymized samples is not subject to the Regulations. This education should highlight the fact that it is the responsibility of the institution and not the prerogative of the individual researcher to determine whether or not a proposed research project is covered by the Regulations.
- Require that all relevant institutions have policies and procedures for reviewing the validity of a researcher’s assertion that a proposed research project involving human biological materials is not regulated by or is exempt from the Regulations. Such assertions should be reviewed by an IRB, or by a designated institutional official with appropriate training in human subjects protections, in accordance with current OHRP guidance.
- With respect to determining whether or not research involving coded human biological materials meets the definition of human subjects research, broadly disseminate current guidance on reviewing written agreements between researchers and repositories to ascertain whether or not the key to the code will be made inaccessible to the research team.
Recommendation 2. OPRR should revise its guidance to make clear that all minimal risk research involving human biological materials–regardless of how they were collected–should be eligible for expedited IRB review.
The Working Group concurs with Recommendation 2 and agrees with NBAC that, for the purposes of determining eligibility for expedited IRB review, it is not necessary to draw a distinction between samples originally collected for clinical purposes and those obtained for research purposes. The Working Group also agrees with NBAC’s observation that current guidance regarding the types of research that IRBs may review through expedited procedures (63 FR 60364-60367 [HHS] and 60353-60356 [FDA], November 9, 1998) appears to exclude research utilizing existing specimens previously collected for research purposes. It is the understanding of the Working Group that this apparent exclusion is not intentional but rather resulted from a copy editing oversight and that OHRP and FDA intend to publish a Federal Register Notice to correct this error.
The Working Group proposes that OHRP and FDA (a) publish a Federal Register Notice as soon as possible to explicitly state that research involving human biological materials originally collected for research purposes is eligible for expedited IRB review if the newly proposed research poses no more than minimal risk to the study participants and then (b) revise its guidance accordingly.