Institutions play an integral role in translating “bench” research into practical applications that advance human health. To do this, researchers need strong research infrastructure to function as an “engine” that can accelerate discoveries. What vulnerabilities do clinical data repositories pose? Will researchers be able to communicate and fulfill responsibilities with their academic partners? Can they assure that they never misplace a sample? And, how do they appropriately track data and sample sharing when needed?
In this article, Sonia Abrol, VP of Software Development and Operations at WCG Velos, shares her insights into these questions. Ms. Abrol has worked with several prominent institutions on data quality systems that support complex parent-child biobanks, integrated laboratory associations, varying workflows with study questionnaires, longitudinal data, samples and consent tracking, and the associations between single plasma donations with multiple recipients.
What is a data or biorepository? Could you give some examples of the different types?
Sonia Abrol: Biorepositories are a central location set up to manage, collect, store and disperse bio-specimens of living things such as whole blood, urine, tissues, etc. and subsequent derivatives such as plasma, serum, DNA, infectious agents, etc. Biorepositories can vary in design and may include biobanks, which are human based repositories, as opposed to animal or plant-based repositories. Internal biobanks serve institutional research needs for a single or multiple studies with samples and/or data. External repositories are not attached to a single research institution, such as those at the National Cancer Institute (for NIH funded studies) or bio-specimen consortiums in which multiple researchers, or academic partners collaborate across institutions or countries. Non-profit, disease-oriented biobanks or longitudinal, population-based biobanks are common within academic institutions. The terms “biorepositories”, “biobanks” and “biospecimen repositories” are often used interchangeably.
Biobanks are used to support many different types of research. Samples collected during treatment can include malignant tissues, saliva, or blood, and preserved (subject to patient consent) for future research. Samples can also be collected in a clinical trial for a sub study, or as donations from patients at the point of care for future use, running a new trial or program for a treatment. For example, the Stanford Blood Center (and many other institutions) is working with Stanford Medicine to collect plasma donations from people who recovered from Covid-19. They plan to transfuse this plasma into patients who are critically ill with the disease, with the aim of providing disease antibodies to the recipients.
What is the importance of linking clinical data with samples in plasma transfer studies?
Sonia Abrol: Currently, there are several plasma transfer studies designed to find treatment and prevention options for Covid-19, for example. Such studies may involve samples that need to be collected, managed and dispersed accurately as per the protocol. The plasma treatment process approved by FDA for this study involves collecting and tracking plasma donations that are received from individuals who have recovered from Covid-19, as well as track the use of such plasma transfusions in Covid-19 patients. Without a scalable, accurate and secure process to track this, participants may be exposed to safety risks. A repository linking the plasma (bio-specimen) with studies and patients (research and clinical information) can help institutions efficiently track the progress of their studies, efficacy of treatment, and analyze multi-dimensional data through one system.
While plasma transfer in Covid-19 trials is a notable recent use case, biorepositories have been an integral part of clinical and translational research for many years. In fact, WCG Velos eSample has been in use for more than ten years as a natively integrated module within WCG Velos eResearch clinical research management system (CRMS) to support translational researcher needs. Researchers can track clinical, research and biospecimen data, lab results, notes and images all in one system. For example, more than three years ago a cancer center simplified their specimen management process and created a dedicated single platform for entry and retrieval of participant and specimen data after migrating from an outdated NIH system to WCG Velos with eSample . More recently, a customer created a harmonized, statewide approach to collaboration across three Arizona hospitals with the primary goal of centralizing human biorepositories to raise visibility and improve usage of collected biosamples for basic and translational research.
What are the most frequently requested infrastructure needs for sample and data repositories management? What has been your experience in transferring data or samples from other biorepositories?
Sonia Abrol: To facilitate clinical and translational research, the most requested infrastructure needs include designing a bio-specimen repository or bank, virtual storage design and management, sample location, sample processing and status history, and managing specimen details pertaining to different types of samples, observations and notes.
Another key need is security or authorization control on biobanks and patients, non-association, de-identification, or association of samples with protocol and patient data obtained from participants who donated specimens and have consented to their use for research purposes.
Flexibility and extensibility are very important as well. No two biorepositories capture the same information. Different sample types, and their respective derivatives, may need different data sets. The ability to define additional fields and forms to capture unstructured and structured information is also commonly requested.
We frequently hear from clients that they also need the ability to transfer information from existing systems. WCG Velos has experience migrating data from a variety of sources; eSample implementations involve bringing in data from old systems, including caTissue, Freezerworks, disparate spreadsheets, etc. Transferring old data requires data review, clean-up, and defining corresponding field mappings in the new system. It is also helpful during a migration to have the ability to create new forms to capture biospecimen information, define additional fields, configure pull-down menus, and host data sets coming from any system.
Finally, interoperability is catching-up in the biospecimen management-research world as well. A huge win for CRMS implementation is the ability to interface with an institution’s electronic health/medical records EHR/EMR to get lab results associated with their biospecimens which feed directly into Velos eSample. Having these data available, along with the specimen and patient tracking details, provides a more holistic view of the data without increasing user workload.
Considering the global nature of clinical trials, and now the remote nature of clinical trials, how are researchers gathering data, samples and consent centrally using existing capabilities within WCG Velos?
Sonia Abrol: Velos clients are taking advantage of Velos eSample in a variety of ways, including these workflows:
- Establishing independent internal repositories: data from samples are collected at the point of care, consent, authorization for use, medical history (or study specific forms), and location of specimen is recorded in Velos eSample. These repositories are used to search for specimens for research, to track specimen allocation, manage specimen lineage, etc.
- Ad hoc sample collection: Researchers can design sample collection kits and associate them with study calendars or treatment arms. This tightly weaves sample collection with the research and, as patients are seen, the coordinators are aware that one or more samples that will be collected during certain visits. Coordinators, research or lab staff are then able to prepare sample kits, sample records, labels, etc. with greater automation, as they complete study visits (or collection events). A designated ‘prepare samples’ area can be used for samples that will be collected with printed labels and sample ‘check-in’ processes when patients come in. This workflow reduces user workload and provides full visibility which samples are collected in specific studies and which can be expected to be collected in the future.
- Consortium: There are different use cases for researchers within consortia. For example, is the biobank for a rare disease? Will the team need to collect longitudinal clinical data, samples, ad hoc study questions and consent across generations? Will it be available statewide, nationally or globally? Will it be population based or disease based? The workflow for these is specifically designed to meet the objectives of the consortium and the associated researchers who aim to improve human health. Some consortiums also establish “storefronts” for participating sites/researchers to look for and request samples for their research.
Specific to compliance, what best practices have you seen implemented which help ensure that appropriate uses of data and samples are followed (including sharing)?
Sonia Abrol: The first thing that comes to mind with respect to compliance best practices for biobanks is automation. If a protocol involves sample collection, you can automate tracking of the sample type, number of samples, quantity, disposition, and processing type -- such as aliquoting, sectioning, printing labels and auto-generating such planned samples to decrease errors and burden. It is important to ensuring that a workflow associates the donated sample with the relevant studies. Forms can be created to track further information such as shipping details, who it is shared with (investigator, institution), etc. Clients also take advantage of tracking sample lineage and independent handling of children-grandchildren samples that ensure accurate and effective management of their inventory.
When using biosamples, it is important to honor the approved uses of samples as per the donor’s consent. If the consent for the sample has options, such as permission “for future research” or specific uses, it is best practice to build processes and workflows that support the distinct protocols. Once you track consent from choices using forms, then research teams can run queries, manage, share, and search through the biobank for available research samples.
Implementing such best practices also reduces risk by finding the right sample – researchers must be able to readily find samples for specific usage, along with gender, biopsy type, lab values, ethnicity, race, and other details which is difficult to do using spreadsheets or multiple siloed systems. The ability to run reports to find the right samples for studies helps drive protocol compliance. Without eSample, many researchers don’t have the full inventory of samples readily at hand in a reportable format.
Data and sample repositories can span years or even generations. How do you minimize vulnerabilities to data and sample loss? Also, how can researchers assure that they never misplace a sample?
Sonia Abrol: A couple of key things can help minimize vulnerabilities:
- Have proper authorizations – Access control can be based on the size of an organization, biobank or study level, for example. Research teams can assign samples to a study and properly track the transfer of sample custody, processing, and freezer storage along with associated access rights to those specimens in a mapped, virtual repository that mimics the real freezer.
- Have technology that supports documentation of sample transfer - documenting the location and transfer from one freezer to another should prevent sample loss. Sites have control over who they share with, that is a human responsibility; but the appropriate technical infrastructure makes it easier.
- Safeguard sample lineage – As samples are collected, the system you use should be able to document what you collect and what you do with it. You can create 50 serum samples from a single participant, allocate it to a study, associate it with generations of samples over many years, such as with grandparents, parents, and children. As samples are processed or reprocessed, you have a centralized system for tracking these activities. And as plasma transfer samples are used, you can track where samples came from.
What advice would you offer to an institution/researcher prior to implementing a biorepository?
Sonia Abrol: My advice:
- If you have old data, review what you will bring into the new biorepository. This involves electronic data clean-up, along with paper-based data to ensure that the new repository is accurate. Full migration means exactly that – what the research team tracks is what they actually have on hand.
- Explore and determine how to organize the biorepository – will it be a standalone biobank or biobank for a specific study? Will there be more than one biobanks?
- Designate authorizations for the data and sample usage within the system
- Welcome automation as much as possible (if appropriate) – this will help increase efficiencies in tracking and data entry.
- Create a simple hierarchy of centralized data fields and forms with enough flexibility and extensibility – tissue may require additional fields if you are recording DNA values/results, for example. Technology providers have tools, but researchers have a choice in how to implement them. I would advise keeping the registry simple. I have seen clients try to implement overly complicated relationships that create unintended reporting challenges. Do not overcomplicate it.
- By using the bulk upload capability for files, lab samples, and research related data, clients significantly reduce the more labor-intensive manual entry.
- Always consider integrations with other systems, for example EMR/EHR, etc.
What are things to look out for in 2020 and beyond?
- I believe biorepositories will be even more critical in research given the current COVID-19 pandemic. Treatment discovery, experimental procedures, plasma transfusion - easy access to search, export, use, and collaboration will be even more important.
- Integrations or interoperability with other systems such as EHR/EMR, device and/or freezers will also be important.
- Collaboration will become more common from institutional, to local, to regional to cross regional or global. Some of our clients have created custom storefronts. The storefronts facilitate a comprehensive and coordinated effort across the state to effectively generate enough biospecimens for clinical research. This allows researchers to browse, query and order from the specimen universe suitable for their needs and provide a platform for the researchers to communicate/manage their requests. No two consortiums will have the same set or workflow; and I expect growth and customization in this space to continue. I’m excited to do my part to advance scientific discoveries.