Trends in clinical research require updated policies on tissue banking

Experts offer guidance and examples

Researchers and clinical trials managers may have noticed that the best practices standards in tissue banking have been evolving in recent years due to scientific advances, greater public scrutiny on human subject protection, and new privacy rules under the Health Insurance Portability and Accountability Act (HIPAA).

For example, investigators in years past may not have asked subjects to provide informed consent for something unknown, such as a future use of a tissue that is not specified at the time their consent was sought. But now science has produced so many striking examples of how tissue samples stored for one purpose later produce valuable information for an entirely different purpose that it seems wise to anticipate unforeseen future uses of a subject’s tissue and data, experts note.

"It used to be often held in IRB circles that if the subject didn’t know what the research proposed was, he or she couldn’t give informed consent," says Elizabeth Hohmann, MD, chair and director of the Partners Human Research Committees at Massachusetts General Hospital and Brigham and Women’s Hospital in Boston.

"Over the past five to eight years, I believe we have moved away from that with the widespread creation of tissue banks and DNA collection," she reports.

The key is for researchers and clinical trials managers who desire to keep open future options for a particular database of information and tissue samples to find ways to help subjects understand what they are doing and why, Hohmann says.

"We need to help people understand who is overseeing this collection and that they are being used in scientifically and ethically appropriate ways," she adds.

At the same time, there’s a greater scientific need for consent to cover future uses, the privacy regulations under HIPAA make this more difficult to accomplish.

There are research challenges, particularly with regard to large multicenter databases and the application of the common rule and HIPAA regulations, notes Marianna Bledsoe, MA, program director for resources development at the National Cancer Institute in Rockville, MD.

"There’s a wide range of interpretations by IRBs and institutions, and these are creating some considerable challenges," she says.

For example, a study’s informed consent document is different from the study’s privacy authorization statement.

HIPAA uses the term authorization to refer to permission for the use of that person’s medical information for the purposes of research. The requirement for an authorization may be waived if certain conditions are met. IRBs are responsible for determining when and how the waiver is sought and used, explains Roger Aamodt, PhD, chief of the Resources Development Branch of the National Cancer Institute.

Under HIPAA, greater scrutiny is being given to privacy and confidentiality protections that are built into the research than in the past, he notes.

"The changes in the law have certainly changed how people have to do business, and the biggest example of that is that, under HIPAA, the information that’s being protected is any identified information, and the regulations have specific terms for what is identifiable in a covered entity, which includes anyone who bills for medical services," Aamodt says.

Cooperative research projects that involve hospitals, medical centers, doctors’ offices, and others may be subject to privacy regulations.

"So data that’s released appropriately from one center can find its way into another one and become protected health information even when it wasn’t protected originally," Aamodt says. "That means potentially that for multi-institutional studies, research groups at the very least have to have multiple authorizations or deal with multiple privacy boards or IRBs to get approval to do the study."

Also, the HIPAA authorization form for research is required to be specific to a research project, comments Julie Kaneshiro, MA, policy team leader of the Office for Human Research Protections (OHRP) of the U.S. Department of Health and Human Services (HHS).

"In other words, an authorization is only valid if it is authorizing a use for a specific research project, and it can’t be for a class or unspecified future research projects," she says.

"This is somewhat different than HHS regulations regarding informed consent, where we have said that informed consent can be broader in the event of tissue banks and repositories," Kaneshiro says.

So while HHS rules would allow a database connected to tissue samples to be used in an identifiable form for future studies of the specimens, the privacy rule does not without specific authorization for each of these future uses, she explains.

For an informed consent document to be used appropriately, it would require some sort of language indicating that an investigator doesn’t know what specific research project he or she might want to conduct down the road because the knowledge and data are not yet available at the time of a first study, Kaneshiro says.

"Assuming you’re using identifiable information about individuals, it would be human subject research and need IRB review and informed consent under the IRB or waiver of informed consent," she adds.

Another difference between the HHS informed consent regulations and the privacy rule is that human subjects research under HHS is limited to people who are living, whereas the privacy rule also extends to the deceased, Kaneshiro notes.

This means that even decade-old tissue samples of a cohort of patients who all have died from a particular disease would need to meet all documentation requirements under HIPAA. An exception to this is in the case of research that precedes the privacy rule’s implementation date of April 14, 2003. HIPAA permits a grandfathered informed consent, an IRB waiver of such informed consent, or authorization to disclose the information for research, according to recent guidance issued by OHRP. The document can be found at

The key is that the informed consent must have been on file prior to HIPAA’s implementation, Bledsoe says.

No blanket authorization

The gray area is in anticipating future needs.

"HIPAA doesn’t allow blanket authorization," Aamodt says. "It’s still possible to get consent for future research that you don’t specify at the time you’re giving consent, and HIPAA allows you to create a database of identified data with general authorization just to create the database," he adds. "The problem comes if you want to use that material; because the way the law is written, you have to have authorization for each specific use."

This means clinical staff will need to go back to the IRB or privacy board and obtain a waiver of authorization, or they will need to go back to each individual patient and obtain authorization for each specific use, Aamodt explains.

"So you can imagine a situation in which someone has created a tissue micro-array, a paraffin block into which they’ve inserted cores of samples from various tissue samples," he notes. "You may have several hundred patients represented on one slide, and if you had to get authorization from each patient for each study, and you cut 200-400 slides from each block, then you’re in a situation in which you would have to make a tremendous amount of effort to contact all patients to get information or to convince an IRB that it doesn’t need authorization."

While most clinical trials managers and investigators are handling this by de-identifying the information, this isn’t always in the best interest of science.

"The key thing that people are struggling with is how to do this and how to do this well," Hohmann says. "It’s important for investigators to think up front about what their research goals are."

For instance, in some cases, the goals might not include widely sharing samples for other uses, whereas in other cases this goal of sharing might be a priority, she notes.

"If you do anticipate that you’ll need a larger tissue bank and will make tissues more widely available for others, then you need to focus on that," Hohmann adds.

Tissue banking issues

Massachusetts General, for example, has begun to develop guidance for investigators and research staff about tissue banking and has been investigating the possibility of starting a centralized tissue bank, she notes.

Just as there are commercial companies that have created tissue banks and contracted with hospitals and other research institutions, it’s also possible for the research institution to create its own bank. The idea is that the institution’s physicians and research staff would need to give patients clear education and choices at the time of a surgery in which a tissue might be available for banking, Hohmann says.

The physical process of setting up a tissue bank is the easy part; it’s the education, informed consent, and storage of data/confidentiality issues that pose the biggest challenges, she adds.

The two main ways to store tissue samples and data are to store these after a pathological diagnosis in a way that doesn’t retain a link to the individual from whom it was obtained or to maintain that link with the sample, Hohmann explains.

"It’s much more valuable of course if you retain a link to the individual, so you then can find out whether they had chemotherapy A and how well they did and that might correlate with some research finding," she says. "So the richness of the information and its value scientifically is so much greater if you retain the link to the individual."

The flip side is that when samples are stored with identifiable information then they pose greater risk of an inadvertent release of information and, thus, psychological and financial harms, Hohmann adds.

Another challenge is informing patients about these possible uses of their tissue samples, and the process would need to be more involved than the casual mention of such a possibility.

"A surgeon can say, You’re having a breast biopsy; and after we’re done, we might use the leftover tissue, and here’s who you can contact if you want to talk about that,’" Hohmann explains.

Part of the informed consent process also might include reviewing with the patient a consent document, showing the patient a video about tissue banking, and asking the patient to decide whether they are interested in participating in this program, she says.

Lastly, clinical trials managers and investigators will need to have agreement letters between themselves and the recipients of any data and tissue samples that are shared with other researchers, Hohmann remarks.

These agreements should clarify the uses for what’s shared and specify that tissues are not to be subsequently transferred or sold and also that the recipient will not try to identify the person from whom the sample came, Hohmann says.

OHRP officials are planning to update guidance on tissue repositories and databanks and may release this additional guidance some time within the next year, Kaneshiro says.

"We’re not proposing any policy changes to existing guidance," she adds. "But because of all the questions arising about this — and not just because of HIPAA — there seems to be quite a bit of interest and confusion about how the regulations pertain to tissue banks and databases."