January 15, 2015 (Vol. 35, No. 2)
Bulging with molecular riches, next-generation biobanks plan to back personalized medicine.
There is a growing consensus that next-generation biobanks are central to the realization of personalized medicine. Next-generation biobanks are collecting, tracking, and maintaining a wide array of biological specimens—the very resources today’s researchers need if they are to bring about tomorrow’s clinical breakthroughs.
CHI’s recent Leaders in Biobanking Congress addressed both the business and science of biobanking. Key contemporary issues include enhancing cost-effective management, assuring quality of biorepositories, focusing on scientifically driven data management, and addressing unanticipated ethical considerations.
Establishing, sustaining, and optimizing the use of large biorepositories can be huge (and costly) undertakings, a point well appreciated by Claire Zhu, Ph.D., program director, Division of Cancer Prevention, National Cancer Institute (NCI). Dr. Zhu’s perspective on biorepositories is informed by her familiarity with the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial Biorepository. The PLCO, a large-scale research effort begun by the NCI more than 20 years ago, is now one of the largest and most actively utilized biorepository resources in the United States.
“The PLCO was designed not only as a randomized controlled trial of screening for the four cancers, but also a research enterprise with a well-characterized cohort with all cancer outcomes, and a biorepository,” said Dr. Zhu. “It now contains nearly 3 million specimens.”
According to Dr. Zhu, successful biorepositories focus on three basic issues: management, productivity, and costs.
With respect to management, Dr. Zhu offered the following advice: “Biospecimens and data generated must be actively managed. We have personnel dedicated to inventory and database management. We regularly query researchers who have obtained PLCO biospecimens for updates on their work.”
Productivity, Dr. Zhu suggested, may be broadly understood. For example, one way to gauge the success of a biobank is to keep track of resulting publications. “We follow up and regularly determine what has been published,” noted Dr. Zhu. “We also look at the impact factor of the publishing journals and the number of citations an article generates to assess productivity and scientific impact.”
Finally, addressing the issue of costs, Dr. Zhu cited lessons gleaned from her experience with the PLCO: “The costs associated with establishing a biorepository are relatively minor compared to the costs associated with conducting the trial. One cost-effective strategy is to piggyback biospecimen collection onto an existing trial or study. There are lots of different trials or cohorts that already exist that can be ‘repurposed’ to include a new biospecimens collection.”
Reporting Unanticipated Results
As the extraordinarily powerful and innovative technology of biobank genome sequencing is more frequently incorporated into mainstream research and medicine, the data generated has the potential for both benefit and harm. “There has been much discussion in the past decade as to whether investigators should return information to research participants about pathogenic mutations in actionable genes,” said Wylie Burke, M.D., Ph.D., a professor of bioethics and humanities at the University of Washington.
The key word here is “actionable.” According to Dr. Burke, the word itself is not clearly defined.
“People disagree on a crisp meaning of the term. Basically, we are talking about mutations in genes for which there is a medical treatment,” contended Dr. Burke. “For example, a study could conceivably identify a person with a gene variant associated with a high risk of developing colon cancer. This is actionable because there are therapeutic interventions that could be suggested such as earlier and more frequent screening.”
But what if a study were to find a risk variant for diabetes? “It’s conceivably actionable,” suggested Dr. Burke. “For example, some might be recommended to monitor blood sugar or make dietary changes. But is the risk sufficient to inform the participant? Now, we are dealing with threshold issues.”
Genetic studies may also reveal a participant has carrier status but is unaffected. This raises the question of how investigators or medical professionals might handle the possibility that a participant might partner with someone who is also a carrier. Might participants be advised to screen potential partners or seek prenatal diagnostic screening?
What’s needed, said Dr. Burke, is more of a consensus in the community as to when to report unanticipated, actionable findings: “Basically, to develop optimal policies, we need empiric data, ethical analysis, and multistakeholder deliberation. Professional organizations are beginning to weigh in on this issue. NIH research consortia are generating discussions. I believe we will continue to see efforts to develop consensus and more clearly resolve this issue.”
For now, regulatory oversight will be provided by each organization’s internal review board (IRB). According to Dr. Burke, investigators will need to propose how they will make decisions in order for their IRBs to perform a review and evaluation.
“It would help investigators to be able to indicate that the study will be following the evidence-based guidelines of a nationally recognized consortium or other such group,” explained Dr. Burke. “The community is definitely moving in this direction, but we are not there yet.”
At first glance, preserving and storing biological specimens seems straightforward: procure the sample and stick it in the freezer. However, it’s not quite that simple.
“The bar for quality is continually rising, and researchers need to better understand factors affecting the quality of their biorepositories, such as the most effective methods of processing, preserving, and storing biospecimens,” noted Allison Hubel, Ph.D., professor of mechanical engineering and director of the Biopreservation Core Resource (BioCoR) at the University of Minnesota.
Dr. Hubel said that it is all too easy for biobankers to “get caught in a silo” and imagine that no one else has the same problems. But biobankers are mistaken if they think they are the only ones who struggle with quality control and preservation protocols. The same issues challenge individuals who maintain cell therapy products, stabilize pharmaceuticals, or even preserve food.
“Lots of biobankers do temperature monitoring. However, what they need to do is employ the same best practices that pharma utilizes for documentation and monitoring,” asserted Dr. Hubel. “We focus on helping biobankers use common sense and scientific principles to avoid common pitfalls.”
There are many resources biobankers can utilize to better educate themselves about best practices. One such resource is the International Society for Biological and Environmental Repositories. Another resource, expressly recommended by Dr. Hubel, is the BioCoR library, where biobankers can access “scientific literature, protocols, standards, and best practices.” Dr. Huber added that the BioCoR website has FAQs on topics such as how to freeze whole cord blood, storing in a mechanical freezer versus liquid nitrogen, and pooling/freezing cells.
Large epidemiological studies from the United States and Europe have provided much of what is known about causes of chronic disorders. “These are often prospective cohorts or population studies,” pointed out Philip Awadalla, Ph.D., a professor of medicine at the University of Montreal.
Dr. Awadalla also directs Canada’s very own biobank, CARTaGENE (CaG). This biobank, which was initiated in 2008, now consists of more than 40,000 participants from Quebec.
“This is the largest ongoing prospective health study in Canada,” noted Dr. Awadalla. “CaG targeted the segment of the population most at risk for developing chronic diseases (that is, ages 40–69). These are participants, not patients.”
CaG, continued Dr. Awadalla, created a detailed snapshot of each individual: “In-depth health and sociodemographic information, physiological and biochemical measures, and biological samples (blood, serum, and urine) were captured for a total of 650 variables. In fact, CaG is one of the few population-based cohorts in the world that stores biologics not only for DNA and protein-based analyses, but also for gene expression analyses.
“We currently have a number of recontact programs to follow the health status of these participants. This facilitates mining of genetic and environmental factors associated with disease-related quantitative traits.
“Further, it was designed to maximize the ability to harmonize with other Canadian and international cohorts. Specifically, we are part of the Canadian-wide Partnership for Tomorrow Program, which has collected similar data now for over 300,000 participants and is supporting other in-depth recontact studies such as MRI phenotyping. This program will be open for researcher access in March 2015.”
Dr. Awadalla has exploited the detailed clinical and phenotypic data for his own research. He and colleagues performed a high-resolution genomic analysis of human mitochondrial RNA sequence variation, and they found that mutations in the genome of mitochondria are associated with a number of biological processes as well as some diseases.
“We performed next-generation sequencing of RNA from the blood of ~1,000 individuals from the CARTaGENE biobank,” explained Dr. Awadalla. “We found a previously underappreciated role of the gene MRPP3 in modulating the rate of mitochondrial RNA modification.
“The gene’s expressed product, mitochondrial ribonuclease P protein 3, is involved in mitochondrial tRNA maturation. Overall, the study demonstrated the potential of genome-wide association studies to identify genes involved in basic cellular processes and cellular energy production.”
Biospecimen sciences could be a key element for the ultimate success or failure of personalized medicine, advised Michael H.A. Roehrl, M.D., Ph.D., director of the University Health Network (UHN) Program in BioSpecimen Sciences at the University of Toronto: “Information technology is an integral part of biospecimen sciences. The biggest challenge and one of our major areas of interest is building the best data management software and hardware for next-generation specimen-driven personalized molecular healthcare.”
Some high-profile companies, Dr. Roehrl explained, are still juggling Excel spreadsheets to make decisions about their research specimens and data. “When trying to analyze multiple types of tests and different patient samples, a simple spreadsheet is not sufficient for a sophisticated analysis,” remarked Dr. Roehrl. “Entering the era of personalized medicine requires ramping up scientifically driven data management.”
According to Dr. Roehrl, if the field is to move forward, it must make progress toward three key goals:
- Make biobanking a science-driven enterprise. It is important to know the exact details of how and what is collected. The biobank must develop a suitable information technology infrastructure that can link various kinds of clinical, research and biospecimen data in real time.
- Ensure consistent specimen quality. Specimens must be collected by trained staff, and resected specimens must be immediately transported from the operating room to a skilled pathologist. Before beginning a research project, it is essential to know the nature of the biological materials being accessed and whether it is appropriate for the studies envisioned.
- Adopt an omics-driven approach. Comprehensive profiling, on the scale of the proteome, the metabolome, etc., will allow researchers to mine all that is possible out of each and every biospecimen.