March 2010: Volume 36, Number 3
NCI Launches National Biobank
Why High-Quality Specimens Have Researchers Excited
By Bill Malone
With all the bad news about the banking industry and the economy of late, the American public might not notice the emergence of a relatively new type of bank that is catching the interest of the lab community. Biobanks—also known as biorepositories—collect, store, process, and distribute biological materials and the data associated with those materials. Typically, these biological materials are human biospecimens—such as tissue or blood—and the clinical information pertaining to the donor of that biospecimen. Although individual researchers and institutions have long contended with storing their own frozen specimens, now a surge of cash from last year’s American Recovery and Reinvestment Act (ARRA) is speeding the launch of the first centralized, standardized biospecimen resource in this country.
Spearheaded by the National Cancer Institute (NCI), the new national biobank, called the cancer Human Biobank (caHUB), will at least initially have a more narrow objective than some of the independent biobanks already in the U.S., such as Mayo Clinic’s venture that began collecting specimens in April 2009. And as of yet, there are no plans to recruit and monitor massive numbers of healthy patients such as the half million-participant UK Biobank.
But while it will not replace other burgeoning efforts, caHUB will play an important role in meeting the growing demand for high-quality, well-documented biospecimens that are needed to keep translational research moving forward, according to Stephen Thibodeau, PhD, professor of laboratory medicine at Mayo Clinic College of Medicine, co-director of molecular genetics, and co-founder of Mayo Clinic Biobank. “Biobanking is really one of the cornerstones of a translational research program. It all comes down to the quality and the availability of biospecimens,” he said. “We in lab medicine and the Mayo research community are going to be so dependent on availability of biospecimens, that I think caHUB is really an important initiative that NCI needs to continue. But it won’t be sufficient on its own, and other biobanks will continue to be important.”
Perhaps the most well known biobank involves an entire country. deCODE Genetics in Iceland—where some 60% of the population has given a specimen—and other such projects in other countries are matched by a patchwork of more modest independent endeavors in the U.S., a disparity that has not gone unnoticed by the scientific community here. Other countries, including Japan, Estonia, and the U.K., grappled with the shortage of high-quality, well-documented biospecimens by setting up enormous national population-based banks of specimens and data.
Why is a national biorepository so important? Some researchers have complained that progress in pharmacogenomics and other cutting-edge lab medicine areas has been exceedingly slow. The frustration, they complain, is the lack of the blood specimens for biomarker discovery and validation. A chorus of reports from the Institute of Medicine, the Department of Health and Human Services, and the President’s Council of Advisors on Science and Technology (PCAST) have also lamented the slow progress of personalized medicine and identified the critical need for high-quality biospecimen banks.
“Despite tremendous excitement about the potential value of molecular biomarkers…this potential has largely gone unfulfilled,” noted the 2008 PCAST report, Priorities for Personalized Medicine, with bottlenecks in validation as “the most important constraint on progress.” The report’s recommendation to address this problem includes both disease-specific biobanks and population biobanks for longitudinal health and disease studies. NCI’s caHUB will be first national project that tackles the first with an opportunity to take on the second in the future.
One of the world’s largest single biobanks, UK Biobank belongs to this second category, with a collection of nearly 500,000 sets of blood and urine samples from people age 40–69. Central to the project is a plan to annotate precisely each participant via a thorough initial questionnaire and physical workup and then to follow their health records indefinitely. Another reason that this biobank will be such a treasure chest of information is that The U.K. National Health Service treats the single largest group of people anywhere in the world and keeps detailed records on all of them from birth to death. As a result, follow-up of UK Biobank participants through routine medical records will allow researchers to pinpoint those who develop a wide range of diseases and make studies that utilize the biobank’s samples that much more meaningful.
Key to making this large project successful was almost a decade of careful planning and pilot projects that have ensured a smooth collection process, high-quality specimens, and the necessary ethical-legal framework for an endeavor with huge implications for society at large, said UK Biobank CEO and principal investigator Rory Collins, FMed Sci, FRCP, British Heart Foundation professor of medicine & epidemiology at Oxford University. Before a single sample was collected, special working groups conducted wide consultation with the scientific community and the public. “These working groups looked at key questions to ask participants about exposure to a wide range of potential risk factors, what kinds of initial measurements to make, what kinds of samples to collect, and how to collect them in such a way that they would be usable for a range of assays, including assays that haven’t been invented yet,” Collins said. “And particularly in all of this, bearing in mind that we were trying to do a study that was a hundred times bigger than Framingham, efficiencies of scale have been critical.”
With its scale and depth of involvement with participants, UK Biobank has become a sort of proof of concept that with careful planning and technological creativity, large biobanking projects can be successful and financially feasible. Relying on a mix of public and private money has kept the project moving below budget and on schedule, with more than 80% of the 500,000 participants already recruited. In fact, Collins and his colleagues were recently invited to Washington to meet with NIH leaders and share their expertise.
Researchers Voice Their Opinions
The Critical Need for Quality Biospecimens
As part of the planning phase for the National Cancer Institute’s (NCI) new national biospecimen resource, called caHUB, the Office of Biorepositories and Biospecimen Research conducted a survey of researchers funded by NCI, as well as federal agencies, cancer centers, industry, foundations, and advocacy groups about the need for quality human biospecimens. The survey results indicated that acquiring quality biospecimens was a major barrier to progress in many research areas, including discovery and validation of new diagnostic assays.
Source: Office of Biorepositories and Biospecimen Research, 2009. Available online.
Focus on Quality
At the same time that other countries have pushed forward with national biobank schemes, researchers in the U.S. have been burdened by a surprisingly ineffective system composed of silos of samples that vary greatly in quality, according to Jim Vaught, PhD, Deputy Director of NCI’s Office of Biorepositories and Biospecimen Research (OBBR). “The rationale for caHUB is that we think the standards need to be raised across the board,” Vaught said. “Samples in U.S. biorepositories were not consistently collected, processed, and stored. So, after observing biospecimen-related quality issues in studies we’ve been a part of at NCI, we believe that the best plan is to prospectively collect samples under very carefully designed conditions, standard operating procedures, and careful quality control—all well documented and applied consistently.”
Before embarking on detailed planning for caHUB, OBBR started with an analysis to see what researchers were up against when it came to biospecimens, surveying primary investigators at NCI and other institutions. The survey showed that about 40% found it difficult or very difficult to acquire enough biospecimens, with half of the respondents saying it was equally challenging to get the quality of biospecimens they needed. Even more disturbing, 20% said they often or always question their data because of the poor quality of biospecimens. About 40% said they at least sometimes question data.
Even with the best intentions and protocols, individual investigators face innumerable challenges trying to collect and maintain their own biospecimens, said Allison Hubel, PhD, Director of the University of Minnesota Biopreservation Core Resource (BioCoR), a knowledge base, research, and education resource for the biospecimen community. Hubel is a professor of mechanical engineering and former member of the department of laboratory medicine and pathology at the University of Minnesota. “When individual investigators store their own samples, they just can’t match a biorepository’s monitoring, maintenance, and database management system, which allows you to have a sense of security about the quality of your samples,” she said. “In addition, a software system should be assigning identification to and tracking each sample. An individual investigator is going to have trouble justifying implementation of a laboratory information management system, while for the central repository it can be part of the infrastructure.”
After its survey and other study of the issue, OBBR decided that its previous efforts at developing best practices documents and an online network among independent biobanks—while successful—were not enough. “We decided that the best way to address the problem was to develop our own procedures and then have the various clinics, hospitals, laboratories, and repositories collect samples for us in carefully prescribed ways so there is a consistent set of samples and data,” explained Vaught. “This way when collaborators request specimens and data, they know what they’re getting.”
Commercial and other independent biobanks will continue to provide tissue and blood samples, and many do have high quality standards of their own or have adopted OBBR’s Best Practices for Biospecimen Resources, Vaught said. However, caHUB will be aiming still higher. “The difference between what they do and what caHUB will do is that we’ll have a much richer data set that will be collected in a longitudinal fashion, and the combination of the quality of the samples and the rich clinical, demographic, and specimen quality data set will be of a much broader and higher quality than the commercial providers.”
A better quality of specimens is the link from discovery to practice in lab medicine, and especially personalized medicine, emphasized Vaught. “We are promoting the connection between caHUB and personalized medicine because we think it’s really important. Sample integrity and quality are essential for the research that leads to new biomarkers and more personalized treatment. The Cancer Genome Atlas (TCGA) was a prime example that required very high quality samples and produced results.” TCGA is a project working to catalogue the genetic mutations associated with cancer.
Now is the time for laboratorians to provide feedback on the caHUB project as OBBR evaluates what kind of specimens the research community needs, Vaught said. Currently caHUB has several existing clients, including TCGA and various clinical trials under NIH, from which caHUB will collect samples, in addition to a separate “benchmark” or reference collection of tumor and normal tissues collected under contracts being solicited over the next few months. “Initially this will be a disease-based collection, but we have interest from population-based study investigators. We have to take their concerns into consideration and think about whether there are certain biomarker assays or validation studies that will require a different approach to specimen collection. We will be forming a committee to assess potential study designs,” he said.
Mayo’s Own Bank
The dearth of quality biospecimens has also lead to other new biobank projects in the U.S., such as the Mayo Clinic Biobank. Similar to the population-based UK Biobank model, Mayo Clinic aims to recruit 20,000 participants from among the institution’s own patients and then track them through their Mayo Clinic electronic medical records indefinitely. The focus of this collection will be on a generally healthy population as opposed to a disease population.
These biobanking projects have, and will continue to have, a large impact on new discoveries and their translation into clinical practice, Thibodeau stressed. “Access to well annotated, high quality material will be essential for a wide variety of studies. When I think about the use of biospecimens for these purposes, I think of lab medicine as being one of the many beneficiaries. These translational discoveries will ultimately lead to the development of a variety of novel assays—diagnostic markers, prognostic markers, predictive markers for drug response, and tests for genetic predisposition,” he said. “And with the implementation of new high-throughput, high-content technologies such as next generation sequencing, along with studies that require larger and larger number of participants, the need for a substantial number of quality biospecimens will be even more important. This will be true for samples obtained from both disease as well as healthy donors.”
Thibodeau also sees a trend where lab medicine and pathology are becoming more involved in the discovery of new biomarkers, especially in personalized medicine. However, this growing involvement of laboratorians will stall without the ability to properly validate novel tests before they are used clinically, he said.
Before one of these novel diagnostic assays can be implemented in a clinical laboratory, there’s an important next step. “There’s been a tremendous amount of work performed in both industry and research laboratories developing new markers, but what’s really critical before these things can be introduced into the clinical lab is our ability to clinically validate them: to demonstrate that they’re doing what they’re supposed to be doing,” Thibodeau explained. “And the only way that you can complete such studies is to have a large number of high-quality samples. The work in clinical labs is fundamentally changing, and with those changes, the access and need for biospecimens has become more acute.”
Pushing the Envelope with Technology
Just as breakthroughs in technology are driving trends in research and clinical labs, the biobanking community is grappling with the need to improve the technology behind preserving the high-quality specimens. Biobanks themselves are under pressure to keep up and are now drawing upon cutting-edge technology from other disciplines in science and engineering in order to collect and maintain the biospecimens that the tests of the future will demand. “Most of the technology that has been developed to actually facilitate the preservation of biospecimens is on the order of 30 to 40 years old,” said Hubel. “It was really developed for systems for which a very small number of samples are preserved, which translates into very labor-intensive and operator-intensive processes.” Hubel, an engineer herself, has been working on projects that will make biobanking as up-to-date as the assays researchers need them for.
One of these projects developed through Hubel’s BioCoR consortium at the University of Minnesota is an innovative microfluidic device to process biospecimens and reduce the amount of labor involved. When preserving a specimen, multiple steps have to occur. For example, components like red and white cells, plasma, and serum must be separated. Then a cryopreservative solution has to be introduced into the biospecimen being preserved. In order to be used, the specimen must be thawed and the protective agent removed. A microfluidic device could be used at each one of these steps in order to handle the volume of liquid and cells in a semi-automated and controlled fashion, thereby reducing dependence on an operator for some of the processing steps, as well as reducing loss of cells and samples to human variability.
BioCoR is also developing 3-D Raman confocal microspectroscopy techniques to understand the relationship between the state of the intracellular cryoprotective agent during cryoprocessing and the ensuing thermal and osmotic damage to cells. “The Raman study is allowing us to take the latest in molecular characterization techniques and apply that to the science of preservation for biobanking. Now that we have a tool that allows us to image down to the nanoscale, we can use it to expand our understanding of damage or stabilization, and possibly develop a new paradigm of preserving biospecimens,” said Hubel.
One of the discoveries that has come out of the BioCoR Raman study is helping researchers understand why cells within a block of tissue fare so poorly compared to cells isolated in a suspension. “The conventional wisdom has been that heat and mass transfer limitations frame the freezing of a tissue versus the freezing of an isolated cell suspension. Those issues are true, but we believe there are other issues involving cell-to-cell communication that are present in a tissue versus isolated cells which play a role in long-term freezing response,” Hubel explained. “Those are very exciting discoveries, in particular because we have a lot of challenges freezing tissue samples right now, whether they be normal or pathological samples.”
Using more advanced technology and cutting-edge methods is, in fact, the only way that the larger projects have been successful. This strategy fits right in with the advice that the UK Biobank’s Collins has for caHUB. “If you’re going to do a large study, you have to be efficient. You can’t just take what you would normally do with a few thousand people and multiply it up,” he said. “It’s really worthwhile having an industrial approach to the way in which the study is done and the way in which the quality is maintained. You can be a bit ad hoc if you’re just studying a few thousand people. But you can’t if you’re studying a few hundred thousand people. I think that’s the main thing. Design the study for its scale, and in particular invest in systems that manage a very large-scale study in a very efficient way in order to get much higher quality data.”
The Ethical Side
Simultaneous with the unique needs of biobanks that are driving new technology and preservation science, caHUB will also have to grapple with an increasingly complex burden of ethical questions surrounding collecting samples and using them for research. Most of these questions focus on obtaining informed consent to use a person’s blood or tissue in the first place. Plus, the ensuing information associated with that person’s sample creates more ethical issues down the road.
“The expectation and the classical requirement in traditional research is that research participants have to be fully informed, which includes understanding the nature of the research, and knowing exactly what it is they are consenting to,” explained Karen Maschke, PhD, a research scholar at The Hastings Center, an independent nonprofit bioethics research institute. Maschke also sits on caHUB’s ethics subcommittee. “The problem that emerged with research involving biospecimens was that the classical approach to informed consent was not always going to be possible, especially when biospecimens are collected and stored for future use that researchers may not be able to specify at the time they’re collected. So you can’t always fully inform a person about all the potential types of research they’re consenting to with their biospecimens.”
Due to the often large leaps that science makes, such as the sequencing of the human genome, researchers and biobanks learned that to make biospecimens useful, it was important to make consent for their use wider, accommodating future advancements in science and unknown directions that researchers might take. However, having a completely open-ended consent in which a researcher or a biobank asked to be able to perform any kind of valid research, even if it wasn’t yet conceived, left a lot to be desired for many observers concerned with ethics, Maschke said.
This has led to two other approaches. Closer on the spectrum to specific consent, so-called tiered or layered consent evolved from studies at the NCI where participants were given a sort of multiple choice on how their specimen might be used, with check offs for each type of research or disease like cancer, diabetes, or Alzheimer’s. “The challenge with tiered consent is that the biospecimen resource that collects the samples must then maintain a database to track each combination of consent for each specimen,” Maschke explained. “Some researchers have argued this is too complicated and expensive, while others don’t like it because it limits the use of the biospecimens and makes it hard for people doing less common kinds of studies to get the number and quality of specimens they need.”
While tiered consent opened up the use of biospecimens beyond a single, specific type of consent, the compromise just doesn’t work for the scope of a new generation of biobanking initiatives that must include as-yet unknown research possibilities. The desire of UK Biobank, Mayo Clinic, and others to be a library for future research has lead to a third, hybrid approach: essentially open-ended consent, coupled with independent oversight that keeps the participants’ interests in mind.
This is why going forward, caHUB and other large projects will have to put so much stress on governance by data access committees and ethics committees, Maschke emphasized. Though not without controversy, UK Biobank has been a leader in this new model, depending on an independent committee called the UK Biobank Ethics and Governance Council to act as an autonomous guardian of the overall governance and ethical framework. The council monitors and reports publicly on the biobank’s conformity to the framework and advises on the interests of research participants and the general public. This way, the biobank can use a very open-ended consent from participants and still maintain the public’s trust as new uses for biospecimens develop beyond what any original participant might have imagined.
As caHUB begins collecting its first samples later this year, it will have the advantage of observing larger and more complex initiatives in other countries, while Mayo Clinic and others continue with independent projects that have their unique features. Due to the distribution of care in the U.S., it’s unlikely that caHUB will ever match the depth and breadth of a project like UK Biobank, Thibadeu predicted. Yet the Mayo Clinic initiative and others can produce good results and all deserve the support of laboratorians.
“Quality biospecimens will lead to new diagnostic, prognostic, and therapeutic markers, and that’s why NIH is devoting so many resources to their own program. Downstream, it’s going to mean a lot to patients,” he said. “Our end goal here is to have these assays in a clinical environment, but how do we get there? These biorepositories are one of the tools, so we need to continue to invest in them.”