By Maxine Bookbinder

June 11, 2018 | A Harvard professor and Bio-IT World 2018 Best Practices Award winner is banking on old-fashioned collaboration and compassion, along with 21^st century technology, to help principal investigators collect data for clinical research and to facilitate the development of new therapies and cures.

NeuroBANK is a platform that helps researchers aggregate and link patient data from multiple observational clinical research and natural history studies, encourage cooperation among global researchers and clinicians, and increase the chances of new discoveries for rare diseases.

In the United States, 30 million people suffer from one of more than 7,000 rare diseases, of which only 5% have any type of medication, therapy, or cure. “We are trying to bring these diseases to the surface and make them ready for clinical trials,” says Alexander Sherman, Director of the Center for Innovation and Bioinformatics (CIB) at the Neurological Clinical Research Institute (NCRI) at Massachusetts General Hospital. “We need to aggregate as much information as possible, as part of a concerted effort, to discover and validate disease-specific biomarkers and outcome measures, so companies can prove that their products are effective.”

Traditional double-blind randomized clinical trials are difficult to design for rare diseases, due to the small number of patients; some diseases, especially those with rare genetic mutations, afflict fewer than 500 people globally. That makes collecting patients’ data and organizing research collaborations more critical.  “The more we are able to collect about certain patients and patient populations, the more useful the data will be, allowing everyone to benefit.” says Sherman. “Every data point, every vial, and every image are valuable in such rare populations.”

Launched in 2014, NeuroBANK collects patient data from multiple clinical studies. The platform captures and imports disease-specific clinical research data, harmonizing, anonymizing, and then aggregating it into a single resource. It also offers study design, common data elements creation and maintenance, data curation, data analyses, and customer support. Researchers can follow research participants throughout various projects, linking data to external resources such as biorepositories, WGS files, and image banks, and discover previously unknown disease connections. 

The platform also allows the enrollment of research participants in multiple studies and sharing of patient data, such as demographics and family history information, if investigators agree to do so. Approximately 8% of patients participate in more than one research project. “We want to involve and follow patients who participate in multiple studies and then merge this information and analyze it or bring in researchers and companies to play with data to discover new biomarkers, or to locate sub populations of interest, or to fine tune eligibility criteria for future trials,” says Sherman. 

After individual study data are captured, curated, analyzed, and the results are published, the dataset is anonymized and released into a central pool of disease-specific information available to international clinicians and researchers studying that disease. The NeuroBANK analytical module allows users to “play” with data and inventory of the corresponding resources, including biospecimen and image collections, and request information or tissues for the identified cohort. The NeuroBANK team does not control outside resources; it is up to the parties involved to decide what level of cooperation, such as a joint project, co-authorship in a paper, or free exchange of information, they will pursue.

In a more traditional, siloed research environment, researchers collect data, analyze, and publish, but rarely share the original information. Patients participating in multiple studies must give repeated histories and the same data are collected multiple times, costing money, time, and frustration. Additionally, data from the same research participants cannot be merged, as there is no proper and safe way to identify research subjects across multiple studies.

Systemic competition within pharma and academia for discoveries, funding, and publishing has stifled progress, says Sherman. NeuroBANK “strives to give researchers access to global information about a particular disease in real time. Everyone is contributing and using information for research. It’s the concept of a virtual sandbox, in which all stakeholders in the research continuum collaborate and share. It is just a bigger sandbox, the toys are more sophisticated, and the collaboration is performed on a global scale.” 

NeuroBANK’s collaborative mission encouraging information sharing between disease-specific networks opens new opportunities for investigators, including validation of biomarkers relevant to certain diseases, searches for common biomarkers in patients with dissimilar diseases, checks for fast or slow progressors, discovery of connections between unrelated diseases, and usage of data from one disease against another as controls.

For example, investigators might assume that multiple, unrelated diseases have nothing in common until they find a shared mitochondrial genetic mutation. If a company can find this, it can then figure out how to fix it, possibly simultaneously impacting more than one disease.  That is why data from NeuroBANK may be of interest to analytical companies, such as Microsoft, IBM, and Google, and institutions such as MIT, that have algorithms and can help in disease modeling and possibly new drug discoveries.

One of the most well-known rare diseases is Amyotrophic lateral sclerosis (ALS), or Lou Gehrig’s disease, with about 25,000 Americans currently afflicted. ALS was the first disease network to use the NeuroBANK platform. Currently NeuroBANK’s ALS ecosystem comprises 23 observational studies with 4,000+ participants from 14 countries, millions of data points, tens of thousands of vials with biofluids (CSF, plasma, serum, DNA, RNA, and PBMCs for stem cell lines), thousands of WGS files, and clinical, phenotypical, and cognitive outcomes.  The ALS Association and ALS Finding a Cure Foundation are offering the NeuroBANK platform free of charge to any ALS-related study that cannot afford such services.

To maintain patient confidentiality, Sherman’s team developed a Neurological Global Unique Identifier, or NeuroGUID platform, that assigns a random 11-character string identifier to every research participant.  Once generated, the NeuroGUID uniquely identifies that person in a clinical research continuum; it is impossible to re-identify a person from it. NeuroGUID usage ensures that clinical research information from multiple sources can be linked to the same person, avoiding inaccurate conclusions about a disease.

Research volunteers must first consent to be part of the NeuroBANK research platform, to receive the NeuroGUID, and to allow data sharing for any medical research purpose. Users wishing to access data available for distribution must register with the CIB and agree to not attempt to re-identify research participants, to not sell or re-distribute patient data, and to acknowledge in publications those investigators whose studies contributed to the data pool.

Industry users can search for specific patient cohorts or phenotypes, to match with trial eligibility criteria. “If someone is looking for a specific disease phenotype, say females between ages 55-65 with C9ORF72 genetic mutation, vital capacity above 80%, and available WGS files and plasma samples, NeuroBANK may return twenty patients, allow to download clinical data, and point to the institutions that have biofluids and WGS files. We want to share these resources with academia, foundations, patient advocacy groups, and certainly pharma, literally with anyone who eventually may lead us to cures,” says Sherman.   

“The more and longer we collect, the more information is available for research and analyses. We can follow patients in multiple research projects and link information.”  

Sherman proposed the concept in 2005, but says it took him almost 10 years to convince people that data sharing and collaboration “is not a bad thing,” especially in rare diseases, which require collaborative efforts to aggregate meaningful disease-specific information. The idea, he says, originated out of desperation. “All good ideas come out of pain or lack of resources. If there is insufficient information, we can’t afford not knowing what data others have captured. Consider rare pediatric cases, in which desperate parents rely on each other for funding, research, and hope.” 

The NeuroBANK team does not seek intellectual property rights, but rather collaboration among researchers. “We are changing the culture of research, how companies, foundations, patients, and academics are sharing data,” says Sherman. “People have discovered that there are more benefits in collaborating than competing. We are proud of this. The Bio-IT World award validates our funders’ vision and our efforts to create the single largest ALS clinical data and information pool, with eventually every bit of clinical information to be used and shared in research.”  

Sherman is passionate about the need for cooperation and its impact on hastening the pace of discovery.  “There are no means to make people share data. It is even more challenging to make them share tangible and finite items such as biofluids. Therefore, it is critical just to know where relevant resources are and what is available for sharing. And this is what we are advocating for, to register what investigators and institutions have and show their willingness to consider sharing.”

So far, NeuroBANK works with nine disease networks; in the future, Sherman foresees his team dealing with hundreds of illnesses and stakeholders forming research consortia with which to increase collaboration and to improve clinical trial readiness. “A similar approach is applicable for a single pharmaceutical or biotech company that runs development programs in multiple diseases and would benefit from the expanded knowledge, or is willing to collaborate with other industry partners to jointly develop a disease area for clinical trials,” says Sherman.

With more studies using NeuroBANK and more captured data, the team also is eager to enter the era of Artificial Intelligence, including machine learning.

In addition, Sherman believes that patients will soon be able to instruct investigators to share their data regardless of the researchers’ wishes; the General Data Protection Regulation (GDPR), a legal framework for collecting and processing personal information of European Union (EU) residents launched in May 2018, already encourages this. As a global network, NeuroBANK is technologically compatible and compliant with GDPR guidelines as well as U.S. federal guidelines and regulations. It also follows Clinical Data Interchange Standards Consortium (CDISC) format required by the FDA for data submission.

The 25 CIB NeuroBANK team members are excited and honored by the Bio-IT World award, Sherman’s second in five years. “NeuroBANK is all about collaboration and sharing. The NeuroBANK platform and the affiliated services ensure that accurate, clean, and attributable data are ready for discoveries, and poised for regulatory submission when the time is right.”