On biobanking: Deposits and withdrawals
It is encouraging to see the challenge of well-documented samples being met using electronic medical records to make queries more feasible and less costly.
Register for free to listen to this article
Listen with Speechify
0:00
5:00
In the last several months, ddn has featured several articles on the topic of biobanking.Academic bioanalytical colleagues have been complaining for years about thedifficulty of obtaining well-characterized samples in sufficient numbers tosearch for biomarkers. Others wish to validate a few that looked promising inpreliminary, underpowered animal or clinical studies.
Many putative biomarkers have been reported for smallnumbers of subjects, often in late stages of disease. The sizes reported aretypically large, because without that, there would be no publication. In thosecases where larger follow-on studies have been done, the results have tended tobe disappointing.
What at first seems superficially simple is, in fact,necessarily complicated and costly. After all, who among us realizes thepotential economic value of urination, an activity I perform more frequentlywith age? To increase the value, I'd have to sign an informed consent, record thetime of day, note any prescriptions I've dosed and make some notes on meals,weight, the last time I had sex and any diseases encountered by my parents andsiblings. Then I'd have to freeze it. A mundane task now starts to sound likework. Can I sell an aliquot to recover my costs? I don't have time.
There clearly is a need for archives of disease-associatedhuman biological materials as distinct from more traditional collections suchas the American Type Culture Collection with its broader mission. There aremany pre-omics collections such as the Armed Forces Institute of Pathology(AFIP), which has archived 55 million traditional glass slides as well as hugenumbers of paraffin blocks and wet-tissue samples. The AFIP is scheduled toclose and the disposition of those samples is unclear. The value of samplearchives has long been recognized, even before butterflies were first mountedon pins.
The terms biobank and biorepository today are usedinterchangeably. An industry has developed to fill the growing need forfreezers, cold rooms, software, barcode labels, sample transport technology androbotic means to pick specific samples from the deep freeze. Other firms serveas librarians or sample warehouses and have put in place backup electricalsystems and elaborate means for sample tracking. They keep records of what youdeposited and charge you rent for as long as you wish. Pharmaceutical firmsoften archive samples from specific clinical trials in the event they'd want toreexamine them at a later date. Most, however, are destroyed without furtherscrutiny.
In some recent cases, there is evidence of sharing amongpharma in the "open innovation" sense, where all could benefit from trackingmarkers in a larger pool of test subjects. There also are a number of public ornonprofit biobanks, such as the National Cancer Institute, which provide auseful service to help investigators find relevant samples via their Office ofBiorepositories and Biospecimen Research.
Another is the Rutgers UniversityCell and DNA Repository, a very impressive archive that defines its mission asplaying "a key role in research aimed at understanding the genetic causes ofcommon, complex diseases." BioServe defines part of its mission as providing"academic and industry researchers with access to over 600,000 human DNA,tissue and serum samples linked to detailed clinical and demographic data from120,000 consented and anonymized patients on four continents." The VeteransAdministration (VA) this year announced the Million Veteran Program, proposingto link its large database of health information with blood samples collectedfrom volunteers over the next five to seven years.
The focus here appears to be purely on genomics, and noother disease associations are to be monitored in those archived samples. Thisdramatically simplifies things by removing concerns about diet/fasting or worryabout diurnal variations or unstable analytes. Given the size of the VAhospital system and the huge number of subjects, there should be greater powerto see relatively rare, but important, genetic predispositions for phenotypeswhere organized sample collections can be challenging to obtain. The samplesand medical records for this project will be anonymized, a crime againsthumanity, but not against samples.
There also are smaller and far more focused organizations.One of these off to a good start is the INbank here in Indiana, an initiativeof the Fairbanks Institute for Healthy Communities. INbank is especiallyfocused on longitudinal studies, following the same participating individualsover time as their disease progresses with age. It currently has two studies inprogress, focused on coronary artery disease and type 2 diabetes. For eachcohort, there are carefully matched controls. Electronic medical records foreach participant enable queries on more than 8,000 variables that can bematched to the collected specimens. Given that longitudinal changes ofbiomarkers within a subject are far more likely to be interesting forpersonalized medicine than averages among subjects, this approach affords newopportunities. No doubt, its value will increase with time over many years.
It is encouraging to see the challenge of well-documentedsamples being met using electronic medical records to make queries morefeasible and less costly. Nevertheless, covering the cost of all this is nottrivial, and it changes the value proposition for the aliquot of body fluidsamples we will be feeding to our DNA arrays, mass spectrometers and high-fieldNMRs. Will actionable markers be found in the earliest stages of degenerativediseases such as cancer and Alzheimer's? Not necessarily, but let the chemistrytalk. Much more will be known over the next decade. Happy holidays.
Peter T. Kissinger isprofessor of chemistry at Purdue University, chairman emeritus of BASi and adirector of Chembio Diagnostics, Phlebotics and Prosolia.