In the early days of biobanking, biological sample collections were usually associated with a specific research area and stored near the collection point. Today, biobank collections are made up of a range of sample types that are used to create large research cohorts. Biobank collections may be tied to specific diseases or phenotypes, or they can be gathered as part of a multigenerational population-based collection. While the findings of studies conducted with these samples make news, research data is only as good as the quality of the sample, and managing and storing hundreds of thousands, even millions, of samples is a sophisticated process involving state-of-the-art equipment and workflows.
The idea of storing, managing and tracking millions of anything is daunting. Yet without careful sample management and accurate, auditable information-logging, results may be questioned and rejected in peer review. Researchers seeking to reproduce studies attribute low levels of success to poor sample quality. For example, one article reported on a project to confirm the published findings of 53 papers, which were deemed “landmark” studies. Scientific findings were confirmed in only six (11%) of the cases.1
Samples have traditionally been stored manually in conventional freezers. The drawbacks of this method, such as potential sample degradation caused by temperature changes and lack of sample-tracking data, are now widely recognized, and the move to sophisticated biobanking systems is well underway. Newer biobanking systems offer uniform environments, including a temperature-controlled hatch for sample delivery and deposit; and automated sample handling that extends into the controlled environment, tracks every sample and interfaces with LIMS to provide audit trails on sample history. As the field matures, new trends develop.
Centralizing collections
Many hospital systems are moving toward centralized sample storage and management. Instead of storing small sample collections in each clinic for department-specific research needs, samples from multiple sites are sent to a centralized biobank facility for processing, storage and tracking. With centralization, organizations can establish and implement unified standard operating procedures (SOPs) to ensure quality and maintain the sample data that is critical to research studies. The central facility then reviews and fills requests for access to the samples from groups within or outside the organization.
Inselspital, Bern University Hospital (Bern, Switzerland) is piloting such an institutional biobank, recruiting from every clinic in the hospital system to collect as many patient samples as possible, along with complete records. Dr. Carlo Largiader, a population geneticist and academic head of the institutional biobank, said, “Safety, automation and sample quality were our top concerns when we were planning this biobank. The goal is to get samples from 15,000 patients with a turnover of every 15 years or so.” There are a total of nine aliquots of serum, plasma and buffy coat for DNA per patient. “In addition, people from outside our institution can store their samples with us,” said Largiader. “Because we have such a good infrastructure, we can meet a wide range of demands.” Bern’s samples will be used in a variety of research applications, including validation of biomarkers (diagnostic and predictive), genotyping for rare conditions and pharmacogenetic studies.
Biobank Graz in Austria is another centralized facility. Established in 2007, this biobank took over its medical center’s pathology archives and quickly became one of the largest medical biobanks in Europe, with 7 million samples. Graz collects samples from clinical sites and ultimately distributes them to academic and industrial research partners.
Growing standardization
“There is an increasing distinction between a professional biobank and a sample storage facility,” explained Marcel Bruinenberg, project manager at the LifeLines LifeStore in Groningen, The Netherlands. “You can no longer just put a sample in a –80 °C freezer and keep track of it manually. Collection needs to be standardized and the history of the sample must be registered. The research community now understands the implications that poor sample quality can have on findings, and the industry demands accurate data on what has been happening to that sample at all times.”
LifeLines’ biobank was established in 2006 to support a general population study and collect samples over time from the same cohort of 167,000 individuals. Approximately 6.5 million blood, urine, fecal, hair and buccal swab samples have been collected to date, at 10 collection centers throughout The Netherlands. All samples are processed within 10 hours of collection and are then transported in special cassettes in ultralow-temperature freezers to the central storage facility. LifeLines turned to automation to help standardize its collection and maintain sample quality. Samples are stored in a Hamilton BiOS system (Hamilton Storage GmbH, Bonaduz, Switzerland) (Figure 1), which has the capacity to store up to 10 million sample tubes at –80 °C (Figure 2).
Figure 1 – The Hamilton BiOS stores 692,000 samples at Inselspital, Bern University Hospital.
Figure 2 – The Hamilton BiOS tube picker processes samples at –80 °C to provide temperature stability for sample integrity.Berthold Huppertz, director and CEO of Biobank Graz at the Medical University of Graz, noted, “The complex issue is how to define quality. Experts differ on how long blood samples can be at room temperature before being frozen. This really depends on what analysis you plan to do. So a general measurement of quality is not always possible for a biobank.”
When it comes to data, standardization is also a large task. Clinical sample data may include multiple codes, and data may be stored in a variety of text formats. Most biobanks institute data harmonization protocol for samples, with the same data being collected from clinic to clinic, and standard wording being used to define diseases and conditions.
Huppertz is certain that the International Organization for Standardization (ISO) will develop standards for biobanks soon and ISO certification will become an essential quality designation for biobanks. These standards should point to the importance of the automation of storage, pipetting and retrieval as well as automated data management, particularly in linking data to the sample. “If you are doing manual pipetting of blood samples, information such as patient ID number must also be recorded manually, and errors will occur,” he noted. “Without automation, a simple typing error can result in a lost sample. The 2-D barcode reduces these errors.” (See Figure 3.)
Figure 3 – Automated storage systems use 2-D barcode readers to accurately track samples throughout the system.Dr. Largiader added, “We hope to participate in a network of biobanks throughout Europe for sharing samples. This will be especially valuable for the study of rare diseases. Harmonization of protocols will be very important if samples from multiple biobanks are shared for a single study.”
Biobank funding
“People think that because biobank contains the word ‘bank,’ we must have a lot of money,” said Huppertz. “But getting funding is one of the most important things that biobanks need to do today.” When a biobank project is created, the vision is for the long term, but future funding is not guaranteed.
LifeLines’ goal is to collect and store samples for 30 years, to facilitate research on healthy aging. With such a long-term goal, ensuring the necessary funding on an ongoing basis may be challenging. Last year the LifeLines biobank became an independent company and is now acting as a profit center.
Some biobanks offer storage to outside research groups that do not have the resources to maintain an automated biobank. “Many pharmaceutical companies, for example, do not have biobanks in-house and need to move beyond their manual freezer systems,” explained Bruinenberg. “Many are looking to outsource storage, so the industry-wide effort to standardize collection protocols will be helpful here.”
Some biobanks also allow outside researchers to use their samples, often for a fee. “As a general-population biobank (as opposed to a disease-specific collection), it’s difficult to make a profit,” said Bruinenberg. “Only a small percentage of the samples will be relevant for a particular study. Researchers are looking for a particular condition or disease state or a particular measurement. However, in the future the income can offset as much as a third of our operating cost.” Alternatively, Biobank Graz generates revenue by conducting contracted analytical services on its samples, sending the results rather than the actual samples to its research customers.
One challenge of this source of funding is that not all researchers are aware of the millions of samples stored all over the world in these biobanks or know how to request them. All of the regulations involved in obtaining and using the samples need to be understood as well. Several good portals list availability, such as the repository locator maintained by the International Society for Biological and Environmental Repositories (www.irlocator.isber.org).
Conclusion
Like any industry in the initial growth phase, biobanking is evolving and becoming more sophisticated, as are the guidelines for using stored samples. Automation will play an important role as the push for centralization and standardization continues.
Reference
- Begley, C.G. and Ellis, L.M. Drug development: raise standards for preclinical cancer research. Nature Mar 2012, 483, 531–3.
Martin Frey, Ph.D., is vice president of Hamilton Storage GmbH, Via Crusch 8, 7402 Bonaduz, Switzerland; tel.: +41 58 610 10 30; e-mail: [email protected]; www.hamiltoncompany.com