February 1, 2005 (Vol. 25, No. 3)

Present Challenges and Future Benefits Addressed

Biobanks are increasingly seen as an essential tool in translating biomedical research into real improvements in healthcare. Broadly speaking, a biobank (or biorespository) is defined as a collection of tissue samples and is, potentially at least, a rich source of both genetic and clinical data.

IBM Healthcare and Life Sciences (White Plains, NY) recently brought together many of the key players in this area at the “Worldwide Biobank Summit II” in Tarrytown, NY, to address the various ethical, legal, IP, scientific, and IT challenges posed by biobanking.

“Healthcare is the single biggest market for IT in the next decade,” said Caroline Kovac, Ph.D., general manager of IBM Healthcare and Life Sciences. IBM’s vision is one of information-based medicinea transformation of current practice into a more targeted and individualized approach by integrating data from clinical and biomedical sciences.

This can drive innovation in IT, says Dr. Kovac, and IBM is already developing many new systems and tools for the healthcare sector, with biobanking being one of the major applications.

Anna Barker, Ph.D., deputy director for advanced technologies and strategic partnerships at the National Cancer Institute (NCI), asserts, “We must do something about the biobanking issue or personalized medicine will be delayed by 30 or 40 years.”

NCI’s goal is to eliminate suffering and death due to cancer by 2015. This, claims Dr. Barker, is achievable, but only through targeted diagnosis and treatment.

Biobanking Roadblocks

There are many challenges that need to be addressed on the biobanking issue, according to Dr Barker. Although there are in excess of 300 million clinical specimens in the U.S. alone, the absence of high-quality clinically annotated tissue and a common information platform is seen as a major bottleneck in cancer research.

Many biorepositories are more appropriate to classical pathology rather than post-genomic research. One disturbing implication is that at least some research results may merely be a function of sample-to-sample variability.

Another critical challenge in biobanking is access. “How do we decide who gets access to precious tissue, and to do what?” asked Dr. Barker. Additionally, many are concerned about patient confidentiality, although this issue is solvable.

As to the future, Dr. Barker believes the biobanking challenge can only be solved through appropriate public-private partnerships.

NCI has a number of initiatives under way to help it “connect up” its own biobanking efforts; the Cancer Bioinformatics Grid is NCI’s largest ever undertaking and will, eventually, link everyone in the cancer community. Proteomics, cancer genome, and nanotechnology initiatives also existall new projects needing human specimens.

Current Biobank Models

Dr. Barker is one of the driving forces behind the establishment of a National Biospecimen Network, which is looking to biobank projects elsewhere for ideas and inspiration.

The U.K. Biobank is the largest civilian IT project in the world and will use the unique environment of the National Health Service to look at the genetic and environmental factors affecting the health of the nation. It will be the world’s largest resource of its kind.

Funded by government and the Wellcome Trust, pilot schemes are under way to recruit half a million volunteers, who will be followed for 20 to 30 years, with environmental and lifestyle factors being linked into medical records and biological samples. Full recruitment will be completed by Sept. 2005.

IBM’s Business Consulting Services has been advising the Biobank project on how to collect, store, retrieve, and analyze the samples and has produced a report that could help other biobanks with their development. “It is a challenge to build an IT architecture that will be robust for 30 years. This is an opportunity to get it right the first time,” says Steve Walker, CIO, U.K. Biobank.

Many decisions still have to be made on the U.K. Biobank project, but there will be no compromise in data privacy. While many patients don’t much care about the confidentiality issue so long as the research gets done, for others privacy is paramount.

Arthur Caplan, Ph.D., professor of bioethics at the University of Pennsylvania, pointed out how the privacy issue, if not taken care of, can hold back research, citing the example of Framingham Genomics.

Here was a company that could have extracted valuable knowledge on the interaction of genes and environment using the biobank linked to the long-running Framingham Heart Study. But the company foundered because it did not look for retrospective informed consent, a necessity in this kind of work. “There must be a chain of trust around tissue. Patients want this to protect their privacy, although some are not concerned about it,” said Dr. Caplan.

Biobanking Solutions

In another initiative, IBM is working with the Karolinska Institutet in Stockholm to create Sweden’s first IT-enabled biobank. Researchers are to examine many thousands of human tissue samples along with genetic and environmental data to integrate existing biobanks across the country.

They will be supported by the new IBM Clinical Genomics Solution, which will help provide an IT infrastructure to integrate genotypic and phenotypic data, while assuring privacy and security to the patient.

“The IT infrastructure and expertise provided by IBM is absolutely critical to achieving our goal to impact international efforts in epidemiology, improve public health, and guide planning for new studies,” said Jan-Eric Litton, Ph.D., director of informatics at the Karolinska Institutet Biobank.

The clinical genomics solution is now finding many other applications worldwide, according to Brett David, manager of information-based medicine at IBM Health Care and Life Sciences.

“There is a lot of data in hospital departments. Clinical genomics comes in to integrate and consolidate it so researchers and clinicians can access it and ask questions they otherwise couldn’t, explained David. “We need this to realize the potential of the human genome project.”

He promises that the information-based approach is already having an impact at multiple levels of biobanking in medical centers and pharma and biotech companies. Translational researchdriving research from the bench to the bedsideis beginning to work. “There are drugs already in clinical trials thanks to the information-based approach.”

Meanwhile, Kathy Giusti, president and founder of the Multiple Myeloma Research Consortium (MMRC), employed one approach to building a biobank, using four centers specializing in this rare cancer.

They are all involved in genomics, target validation, and clinical trials, with compounds already in Phase I and II. The aim is to create a critical mass of patients who have multiple myeloma. MMRC is now focusing on the pharmaceutical industry. “We found that they were looking for access to tissue and cell lines,” said Giusti.

MMRC has been very successful in fund raising. “We want this to be a new research model for drug development and offer hope for new treatments for a fatal orphan cancer,” she added, pointing out that the model could also be applied to other diseases.

Data Mining and Integration

IBM Research is developing a number of other healthcare solutions. HealthMiner is a data-gathering system with clinical data mining, predictive analysis (an IBM business and market application used in medicine for the first time) and pattern-discovery elements.

HealthMiner is being applied in a project with the Universities of Virginia involving a collection of 667,000 patient records and has already revealed 30 new discoveries relating to chronic disease that are worthy of further analysis.

BioTeks System

Another system, the Biological Text Knowledge Server (BioTeKS), can analyze text within the biomedical literature, enabling the discovery of correlations that couldn’t be found in any other way. The system is being applied in a collaboration with the Mayo Clinic to the text analysis of clinical notes.

On a more individual level, IBM has been working with doctors at Maria Ferreri Children’s Hospital at Westchester Medical Center, NY, on developing an E-chart for healthcare. This addresses the physician’s need for on-demand mobile access to patient data with the ability to make updates while away from the hospital.

A Blackberry device acts as an interface for such clinical communications, allowing access to key medical information and giving the ability to record any patient interventions, such as a new prescription. The interaction can then be sent to the care team so records are updated. The system is easy to use, mobile, and fast.

Finally, at the other end of the healthcare spectrum, IBM’s new Blue Gene Supercomputer project will help researchers investigate basic biological problems such as protein folding.

The Blue Gene family is being optimized for the ability to handle huge amounts of data with only a fraction of the power and footprint required by today’s fastest systems.

One of the Blue Gene machines was recently declared the world’s most powerful supercomputertypically, about two million times more powerful than the average PC.

IBM is commercializing the machines but will dedicate oneset to be the largest privately owned supercomputer in the worldto its own research projects. One of the first applications will be in protein folding with a focus on studies of membrane-bound proteins, such as G-protein coupled receptors, a major class of drug targets.

Previous articleBuilding Momentum in Scandinavian Cluster
Next articleMerops: The Peptidase Database