Industry Watch Biobanks: Will They Help Promote the Revolution? Ethical concerns notwithstanding, databanks are beginning to spur scientific debate about their utility in deciphering disease and furthering the cause of .

Stan N. Finkelstein, Anthony J. Sinskey and Scott M. Cooper

recent Google™ search on the word “biobank” yielded the Genomic in the African Diaspora (GRAD) biobank. close to 40,000 hits. It’s hard to say when the word — On the other side of the globe, the BioBank Japan Project Awhich combines “biology” and “databank” — first was began in fiscal 2003 with the support of the nation’s Ministry of used, but there’s little doubt that it is fast becoming a staple in Education, Culture, Sports, Science and Technology (Tokyo, the vocabulary of life science researchers. Japan). It aims to provide researchers with a means to clarify the What is a biobank? Essentially, it’s any collection of biologi- causes of disease and side effects on a wider scale than ever be- cal samples and associated clinical data. There are biobanks for fore, focusing on different genetic codes and with the ultimate diagnostic purposes, such as pathology; for therapeutic treat- goal of developing new drugs and diagnostics. ment, similar to banks (which means that biobanks have Biobanks offer an approach to overcoming something we been around for quite a while, even if that’s not what they’ve wrote about two years ago in this column: the “tyranny of data” been called traditionally) and — increasingly — for pure re- faced by researchers who have made such tremendous genomics- search into specific populations or specific diseases. It’s this third related advances in but who, for completely le- type of biobank that is on the lips of the genomics world. And gitimate reasons, are finding it to be a daunting challenge to fig- that’s where we see the link to a topic about which we’ve writ- ure out everything that’s necessary for drug development. A ten many times: personalized medicine. biobank can supply the scientific community with biological Of course, you don’t need to take our word about the link. samples and associated clinical data that can be used specifically Researchers everywhere are spelling out the connection. Accord- to research a disease’s genetic basis. From that, biomarkers can ing to Science magazine, “Hoping to jump-start an era of per- be identified for developing diagnostic tools to predict and mon- sonalized medicine for black Americans, researchers at Howard itor the disease optimally. All of this gets us closer and closer to University in Washington, DC, USA, want to build the first large the ultimate day of personalized medicine, where an individual DNA and health database on people of African descent. The patient’s health needs can be addressed with medicines that are project…aims to collect samples from 25,000 volunteers over right not for the patient’s broader “class” but specifically for her five years and use the data to probe how and lifestyle or him. factors contribute to common diseases” (1). It will be known as It all sounds great. But not everyone is convinced. Consider this headline from the Institute of Science in Society (ISIS, Lon- Stan N. Finkelstein and Anthony J. Sinskey are co-direc- don, England, United Kingdom): “Human DNA ‘BioBank’ tors of the Program on the Pharmaceutical Industry (POPI) in Cambridge, Worthless” (2). That’s a specific reference to the UK Biobank Massachusetts, USA and members of the Pharmaceutical Discovery (Manchester, England, United Kingdom), one of the larger and editorial advisory board. Finkelstein is a senior research scientist at MIT more public biobank projects underway. But the sentiment could Sloan School of Management in Cambridge. Sinskey is a professor of apply to any biobank. They are generating a lot of controversy microbiology at MIT in Cambridge. Scott M. Cooper, an affiliate and frequent collaborator of POPI researchers, is a visiting scholar in the MIT — about ethical and scientific considerations. Department of Biology. Ethical Controversy Abounds A survey article in Science (published by the American Associ- ation for the Advancement of Science, Washington, DC, USA) two years ago catalogs nearly every controversial aspect of biobanks (3). It’s worth reviewing these aspects. In the realm of ethics, concerns revolve primarily around how donors of biological material or data will be assured that their privacy and interests will be protected. This is especially prob-

20 Pharmaceutical Discovery November/December 2004 IndustrTitle y Watch lematic given the increasing number of data studied, should they be able to give ten in these pages, one of the key chal- private companies that are amassing large consent only for specific diseases? As the lenges for translating genomics informa- and entirely private biobanks of DNA and ISIS report states,“proper informed con- tion into something useful involves un- tissue samples. A perusal of Web sites re- sent must be obtained, and…it must be derstanding a disease by examining its veals that in the “world” of biobanks, there made clear to the participants as to what pathway and its phenotypical expression. are pharmaceutical firms amassing their they are consenting to. But what if the col- This is crucial if good choices are to be own data, as well as companies whose pri- lection is to be used for a different pur- made about which targets are involved in mary business appears to be amassing data pose in the future, to study other diseases, the establishment of disease. to provide to researchers, whether in com- or intelligence as was suggested for other Helen Wallace from GeneWatch con- mercial firms or in academia. behavior? Would participants be required cedes that a biobank in a place such as Ice- The Science article notes that,“compa- to give carte blanche consent” (2)? land, which has an “isolated” gene pool, nies say they provide secure data systems offers tremendous potential for research. for ensuring patient confidentiality, and Biobanks are generating But she questions whether similar bene- voluntarily follow federal rules for pro- fits can come in Britain or any multira- tecting human subjects. But…these pro- a lot of controversy — cial, multicultural society where genetic cedures often are shrouded in secrecy. information is likely to be more complex. Some private biobanks, for example, con- Referring to UK BioBank, Wallace says, sider key documents, such as consent about ethical and “There will be medical advances, but they forms, to be proprietary. And if compa- will be mainly in single-gene diseases, or nies go bankrupt, critics contend, tissue scientific considerations. inherited diseases…Once you start mov- and DNA samples might be sold off to ing on to multi-factorial conditions like practically anybody” (3). That’s exactly Another ethical issue is genetic discrim- heart disease and cancer, you might gain what happened in Japan in 2001, when a ination. GeneWatch UK (Buxton, Derby- some understanding of the mechanisms bankruptcy court auctioned off a human shire, United Kingdom), which has been of the diseases but it’s questionable cell collection that had been used by a sci- organizing opposition to the UK BioBank, whether studies on these half a million entific society as collateral for a loan (3). notes that Britain has no laws banning ge- people will tell us any more than we al- Serious contention over biobank pri- netic discrimination as a justification for ready know” (4). vacy has occupied Icelanders for years refusing insurance coverage or employ- This perspective is shared by Sir Alec now, ever since an company called ment.“Public trust in doctors could be se- Jeffreys, the scientist who invented DNA deCODE Genetics (Reykjavik, ) verely damaged if patients are harmed in- fingerprinting in 1984.“My concern with announced that it was contracting with stead of helped by this research,” the the project is with basic scientific do- the Icelandic government to put the health group’s deputy director, Helen Wallace, ability. Whether a study constructed in records of the entire 270,000 person pop- told The Scientist (Philadelphia, Pennsyl- that way will be effective at teasing out ulation into a single database that would vania, USA) (4). really quite subtle genetic factors influenc- be linked with detailed genealogy and ge- These are some of the ethical concerns ing disease, whether that is more cost- netic data collected from volunteers. most biobank projects face. Whether they effective to use focused studies on diabetes, deCODE’s contract allows it to provide can be allayed is a question we’ll leave to or specific cancers, which already has been drug companies with access to the data the experts. It is the scientific controversy done” (5). for a fee and to academic researchers pur- to which we’d like to turn. Proponents have answers, as you might suing “noncommercial” projects at no have expected. At BIO, a biotechnology charge. The data is being encrypted, but Does the Science Make Sense? conference held this past June in San Fran- the main controversy has been over the The ISIS report makes a claim that really cisco (California, USA), Francis Collins issue of “presumed consent” — that is, goes to the heart of the scientific contro- reiterated what he had written in Nature government health records on every citi- versy surrounding biobanks. “The quest (London, England, United Kingdom) just zen are included in the database unless in- for ‘personalized’ medicine based on in- a week earlier — that the United States dividuals specifically opt out (sort of like dividual genetic makeup,” the authors should establish its own biobank. He ar- the “Do Not Call” list for telemarketers write, “is simply scientific nonsense” (4). gued for creating a longitudinal database, here in the United States). That’s the starkest among many state- publicly accessible, with the biological ma- In Iceland and elsewhere, safeguards ments from scientists and others suggest- terial of at least a half a million people. have been or are being created to ensure ing that the goal of personalized medicine Collins, the director of the National privacy, but some bioethicists wonder how is not simply elusive, but downright un- Human Genome Research Institute much really can be done and whether pri- attainable. We disagree — but let’s look at (Bethesda, Maryland, USA), wrote,“In the vacy really can be verified. But privacy isn’t what others are saying. United States, a gene-environment cohort the only problem. Consent looms large as The scientific argument for biobanks study could be assembled by building on, an ethical issue, as the “presumed consent” boils down to how genomics data can be at least in part, already existing large stud- approach in Iceland would indicate. For made useful. The idea of a biobank is to ies such as the Women’s Health Initiative, instance, when patients agree to have their put the data into context. As we’ve writ- the Framingham Study, the Harvard stud-

22 Pharmaceutical Discovery November/December 2004

TitleIndustry Watch

A Sampling of the World’s Biobanks, Existing or Proposed looking for them — and we hope to re- The following is a very small sampling of some of the biobanks currently port what we find in a future column. In proposed or operational, to show readers that these are becoming a part the meantime, we do know that the gen- of the public, private and academic sectors. eral issue of biobanks is attracting increas- ing attention from practitioners and deCODE Genetics thought leaders, who are meeting both in- Private-sector company banking genetic samples of 100,000 Icelandic formally and formally to discuss how best volunteers linked to Icelandic Health Sector Database and genealogical to move forward. records (www.decode.com). Those of us who advocate personalized medicine will be watching closely. There’s Estonian Genome Project a lot at stake, just as Francis Collins wrote. Government effort to establish a national genetic/medical database of We want to realize the promise of reduc- 1 million volunteers (www.geenivaramu.ee). ing disease held out by genomics, not let it slip from our grasp. Genomic Research in the African Diaspora Howard University project to collect DNA and health information from References 25,000 Americans of African descent (www.genomecenter.howard.edu). 1. J. Kaiser, Science 6 June, 1485 (2003). 2. Institute of Science in Society (ISIS). In- side Human Genetics and Genomics: Karolinska Institute (Stockholm, ) Human DNA “BioBank” Worthless. ISIS Swedish academic bank collecting human biological material for molecular Special Series, January 2002, part 2 of a and genetic research (http://www.ki.se/corefacility/biobank/biobank.htm). 4-part series. 3. J. Kaiser, Science 8 November, UK Biobank 1158–1160 (2002). 4. P. Hagan, UK Biobank reveals ethics Government plan to collect genetic samples from 500,000 volunteers framework, The Scientist Daily News between the ages of 45 and 69 (www.ukbiobank.ac.uk). (online), 24 Sept. 2003. Accessed at www.biomedcentral.com/news/200309 24/03. ies of health professionals and some of the reducing disease burden on a population 5. “£62m Biobank may not be worth it, says professor”. The Daily Telegraph, 6 Sep- many large cancer cohorts. The obvious basis will remain out of reach” (6). tember 2004. advantages are that many years of follow- For its part, the UK Biobank argues that 6. F. Collins, Nature 429, 475–477 (27 May up already have taken place in these co- it will be “the world’s biggest resource for 2004). horts and, for many of them, DNA already the study of the role of nature and nur- 7. UK Biobank. Briefing note. Accessed at has been collected. But serious consider- ture in health and disease,” and that with- www.ukbiobank.ac.uk.PD ation must be given to whether the dis- out the biobank the opportunity simply ease-specific focus of many of these stud- won’t be there for “researchers to carry out ies has limited the phenotyping and case-control studies in which environmen- exposure measures, whether the minority tal exposures, lifestyles and genetics can representation is adequate, whether the be compared in those with a disease or consent obtained is sufficient for broad condition (cases) and those without (con- access to data and biological materials and trols).” Further, “Our scientific under- whether the study design is appropriate standing cannot be considered complete for the ambitious goals of a national gene- until we know how and environ- environment study. If those limitations ment impact on these complex pathways, turn out to be significant, an entirely new and this understanding could produce im- cohort project may need to be contem- portant clues to possible new therapeutic plated. Although the challenges in under- interventions” (7). taking such a prospective population study in the United States would be consider- Awaiting an Answer able, a serious evaluation of its merits now Until the research is underway, the jury is in order.” will be — and should be — out on the sci- His scientific argument boils down to ence issues. yet, the ethical questions are this: “If the conclusion is that this resource substantial and real. It will take a lot of is needed, then we must collectively seek work — undertaken with complete trans- ways to organize and implement it quickly parency — to resolve the concerns about and efficiently — or face the real possibil- privacy and consent. ity that a decade from now the promise of We don’t know the answer to any of genetic and environmental research for these questions, but we’re going to keep

24 Pharmaceutical Discovery November/December 2004