Brisk trade in tissue for proteomics and genomics research

March 2003
Anne Paxton

The rapid evolution of clinical genomics has transformed tissue banking from a relatively informal resource for academic researchers to a commercial linchpin of the drug and diagnostics industries.

According to industry estimates, the current market for microarray technology and associated human tissue and RNA is $400 to $800 million—and on track to double in size soon. Alan D. Proia, MD, PhD, associate professor and vice chairman of pathology at Duke University, Durham, NC, noticed the trend a few years back. "I’d get a call a week from pharmaceutical companies, diagnostic companies, people at Duke and at Research Triangle Park," he says. "Everybody and his brother wants human tissue for genomic and proteomic research."

"This is a relatively new branch of tissue banking that’s come into being in the last five years," says Michael J. Becich, MD, PhD, chairman of pathology at the University of Pittsburgh Medical Center Shadyside Hospital and director of the university’s Center for Pathology Informatics. Dr. Becich, who is also founder of TissueInformatics, in Pittsburgh, and a former co-chair of the CAP Informatics Committee, says the number of high-volume formalized tissue banking efforts around the country has gone from "a handful" 10 years ago to thousands today.

In 1991, he helped launch what was then called the Western Pennsylvania Genito-Urinary Tissue Bank, which was housed at the University of Pittsburgh and drew tissues from a large cluster of hospitals. The university is now one of four sites funded by the National Cancer Institute (along with George Washington University in Washington, DC, the Medical College of Wisconsin in Milwaukee, and New York University) to maintain national tissue banks for prostate tumor samples as the Cooperative Prostate Cancer Tissue Resource.

These sites, says Dr. Becich, make more than 3,000 prostate samples available to researchers, along with serum, whole blood, lymphocytes, and a rich set of clinical data. And demand for the tissues is surging.

The reason, says Jim Wittliff, MD, PhD, professor of biochemistry and molecular biology and research professor of surgery at the University of Louisville (Ky.) James Graham Brown Cancer Center, is, "You now need human tissue to mine the data that have evolved from the Human Genome Project."

The term "tissue banking" often creates confusion because it is commonly used to refer both to tissues collected for research and those used for transplantation. Members of the American Association of Tissue Banks, for example, are the relatives of organ donation centers; they meet an equally growing demand for skin, veins, bone, fetal tissue, and other tissues used in burn treatment, other therapies, and plastic surgery.

Research-oriented tissue banks, however, do not have to meet standards for human transplantation and do not have a formal regulatory body. While collection and preservation standards are essential, these tissue banks are mainly concerned with the quality of their samples as scientific information.

Clinical genomics is the correlation of molecular changes—including differences in patterns of gene expression—with the characteristics of human disease to accelerate the discovery and validation of targets for new diagnostics and therapeutic agents. "For the first time it is moving beyond the thought phase," says Alan Buckler, PhD, chief science officer of Ardais Corp., Lexington, Mass. "We’re now marrying genomics and proteomics with actual human disease, and some very compelling patterns of disease are already emerging in the literature."

"Back in 1991," says Dr. Becich, "a lot of people would see the utility of frozen tissue, but there were a number of problems. To really create a lot of value for it, you need to provide a rich clinical annotation, with a lot of information about the patient, the disease, the organ site the tissue is taken from, and so on." That required "a lot of heavy lifting," he says, to work through patient consent issues amid changing regulations to ensure privacy and confidentiality.

The new demand for frozen tissue to perform genomic and proteomic profiling for diagnostic biomarkers and to develop therapeutic targets has injected the cash to fulfill that task. Relatively new companies like Ardais, TissueInformatics, and Asterand (in Detroit) cater to large pharmaceutical and biotech customers—the Lillys, Abbotts, Pfizers, and Millenium Pharmaceuticals, Dr. Becich says. And academic institutions are reaping many of the benefits.

At the first Clinical Genomics Symposium, in June 2002, Mark Boguski, MD, PhD, a computational biologist at Seattle’s Fred Hutchinson Cancer Research Center, said, "What we are witnessing today is the shifting upstream towards discovery research of the clinical trial discipline: namely, the power of large numbers of patient cases and the systematic application of relevant clinical data to correlate with molecular change."

Ardais announced in February that AstraZeneca, one of the five largest pharmaceutical companies in the world, has licensed access to Ardais’ library of tissue samples and related information for its drug discovery program. Ardais has commercial agreements with some 25 other pharmaceutical and biotech companies as well, including Bristol-Myers Squibb, Aventis, and CuraGen Corp.

It was Oxford Bioscience Partners, a venture capital fund with about $600 million invested in 80 biotech companies, that saw the need for a company like Asterand. "The senior partners, Dr. Alan Walton and Jonathan Fleming, realized with the mapping of the human genome that there eventually would be a massive demand for researchers to actually go beyond the animal models, using mice and so on, that drug companies used for years as the basis of early research work," says Asterand chief executive officer Randal Charlton. "With the knowledge of how humans are made up, it makes sense to look at gene and protein expression in the human cells, for which, of course, human tissue is required. However, scientists in the Oxford group of companies were not able to find the tissues they needed for this new approach, and they said, ’Look, we need a special company that specializes in this discrete task.’"

"Ten years ago," says Dr. Becich, "the number of new biomarkers in the pipeline as potential new therapeutic avenues probably would number in the low teens or 20s. Today there are probably 300 to 500 candidates in various stages, making their way through various frameworks. The majority really come from discoveries with genomic and proteomic methodologies on tumor tissues."

Adds Charlton: "With the mapping of the human genome, there was awareness that the growing genomics and proteomics industry would require well-characterized tissue. In other words, tissue collected in such a way that genetic information can be extracted. So it had to be collected under a very strict protocol which we developed in the first year of our existence. We probably went through 20 iterations of collection protocols." The result was a process in which the tissue is frozen quickly.

"If it’s diabetes, we’d collect a variety of tissues, including different fats and samples of other organs, like the pancreas, because diabetes is not just in one place—whereas for arthritis, we’d collect synovial fluid," says Charlton. In cancer, he points out, "the tissues tend to be from the tumor and the adjacent normal tissue, so there might be several tissues from the same patient."

Just as important as the tissue, he says, "are the clinical information and pathology information that go with it, and this is what I mean by ’well-characterized.’ We’ve got complete patient data, both the patient history and the family history, as well as the pathologist’s report on the tissue that has been collected."

There are two primary sources for research tissue, Charlton adds. "One is surgical collection, where someone is having an operation for colon cancer or lung cancer, for example, and we get surplus material, in excess of what is required for patient care, that would otherwise be destroyed. The other source is postmortem or so-called rapid recovery. What happens is, a patient has preconsented to donate their body to medical research and our pathology team goes in and collects material that is required." This collection must be done rapidly, within 10 hours of death, he notes, and ideally within six hours or less. Biopsies are another source of tissue, but they tend not to produce enough tissue to make each sample commercially valuable.

After a molecular biologist takes a tiny piece of the sample and checks its RNA, about eight percent of Asterand’s tissues are rejected because of degradation—but that percentage rises to 20 or even 40 percent with postmortem tissue. Surgical material "comes straight from the operating room to pathology and it’s frozen very quickly, within minutes," Charlton explains, in contrast to the several hours allowed for salvaging postmortem tissues.

Founded only three years ago, Asterand receives tissues from 20 sites all over the world and already has about 50 research clients, the majority of them large pharmaceutical and biotech companies. The company has 22,000 tissue samples, including all cancer material (breast, lung, colon, esophageal, prostate, and pancreatic, and rare cancers) and non-oncology disease targets such as cardiovascular disease, rheumatoid arthritis, diabetes, Alzheimer’s disease, Parkinson’s disease, bipolar disorder, schizophrenia, and amyotrophic lateral syndrome.

"We are very sharply focused as a service company," says Charlton. "We do not do research ourselves, other than research our customers ask us to do. We’re different from your normal academic tissue bank, which basically takes a bunch of material in and hordes it, to put it crudely. So, for example, we’re not involved in collecting diabetes tissue and saying we’re not going to sell it because we’re going to identify the genes that cause diabetes. Some of our competitors do; we don’t.

"We simply accrue the tissue," Charlton adds, "and it can be frozen or fixed, or we’ll make it into tissue microarrays, or cut slides for you, or if you want, we’ll extract RNA and deliver that. We’re increasingly doing some of that front-end work." A client wanting to do a study of renal cell carcinoma, for example, may want only the results from extracting RNA. The company then provides the customer with digital images of the tissue.

Finding sources for tissue remains a challenge, but companies frequently rely on established relationships with hospitals to obtain tissues. "It’s extremely difficult," Charlton says. "Very often the hospitals themselves want to do their own research, and what we do is make arrangements whereby in return for their supplying tissue and becoming part of our donor network, they can have access to our tissues for their own research. For example, one hospital we deal with is studying colon cancer, but that’s only 10 percent of what they’re collecting, so they give us all the other stuff they do; then they have the right to the rest of our colon cancers."

"We also have a few academic customers who say, ’We can’t afford to pay the value of the tissue, but if you’ll supply them at a serious discount, we’ll give you the option on any intellectual property that results,’" says Charlton.

Ardais Corp., which has more than140,000 samples in its repository, chose a different strategy. It formed "strategic alliances" with several established medical institutions by launching the National Clinical Genomics Initiative, which now includes the medical centers at the University of Chicago, Beth Israel Deaconess in Boston, Maine Medical Center in Portland (a teaching hospital of the University of Vermont College of Medicine), and Duke University.

In a presentation to other potential partners, Ardais said the advantages of participating in the initiative include access to Ardais’ research-quality samples and clinical data for academic research as well as access to a network for commercial genomic research collaboration opportunities.


"For most tissue banking efforts," Ardais’ Dr. Buckler says, "normal tissue is typically remnant tissue from surgical procedures that is considered characteristically normal and non-neoplastic. Our focus has actually been to look for a true representation through postmortems. We are just in the process of developing a postmortem sample-collection process, which is the only true approach to getting normal tissue."

Affiliating with Ardais allowed Duke to have a quality resource for its own investigators without footing the entire bill, Duke’s Dr. Proia says. "It’s pretty common to have multiple tissue banks at an institution—maybe a lung person who collects lung tissue, and so on," he says. "It’s part of the problem at most academic places that there are all these little repositories, and the deans hoped to set up a one-stop shop, so if you had IRB approval to do research you’d be able to call up and say, ’I need 10 prostates, 20 livers, and are they in stock?’ That was the goal—to have a comprehensive tissue bank with informed consent, with appropriate clinical data, with a great diversity of tissues, and the only way to fund it was to do it in collaboration because it’s horrendously expensive."

In the old system, Dr. Proia says, the investigators got tissue without much data. "Now, because of Duke’s affiliation with Ardais, I have two research nurses who meet with all patients scheduled for surgery, and if the surgery is likely to yield bankable tissue, they ask for their consent, explain that we’re affiliated with Ardais, that it’s a for-profit company, and there are all these provisions to ensure their care won’t be compromised."

More than 99 percent of the subjects will consent, Dr. Proia says, "because medical research has everything to gain and they have nothing to lose." In 2001, about 900 patients gave their consent for tissue to be collected and Duke was able to bank approximately half of it; in 2002, the number of consenting patients rose to over 1,400 and Duke was able to bank about 800.

Even now, however, too many people are probably "hobby collecting," Dr. Wittliff says. "It’s dangerous because people don’t store tissue and sera in a standardized format. There’s no standard way to evaluate tissues for research."

In the 1970s, Dr. Wittliff managed a proficiency testing program for early tamoxifen trials and an international quality assurance program for clinical trials on estrogen and progestin receptors. In the 1980s, he started the biorepositories at the Brown Cancer Center.

The CAP established a formal Survey program on estrogen and progestin receptors around that time, Dr. Wittliff says, and it continued until about 2000. "Then it stopped, because most of the proficiency Survey assays were beginning to be performed by immunohistochemistry, although even then the biorepository helped with standardization. Now this biorepository is primarily a frozen tissue bank, which has turned out to be most valuable since the human genome has been uncovered."

One of the first tumor markers Dr. Wittliff, along with other investigators, used the tissue bank to look at was the protein product of the oncogene HER-2/neu, which indicates an aggressive breast cancer. "We conducted early studies of HER-2/neu proteins in our patients’ biopsies, and our studies and others helped define the ranges of expression," he says. "Then Genentech produced a new drug—which binds to this protein—to treat breast cancer, and the drug is called Herceptin."

"We also used the biorepository to define ranges for epidermal growth factor receptors, and a new drug has just been produced called Iressa, for breast, lung, and other cancers," he says. Another protein he studied was a new tumor marker called uPA, urokinase type plasminogen activator. "Our studies and studies of groups in Europe have indicated this is a prognostic factor in breast and endometrial cancer, perhaps in other cancers too, and it’s being integrated into new clinical trials. This is the current approach to cancer therapy, to synthesize targeted drugs, and this is where the biorepository is incredibly important."

Dr. Wittliff is even more enthusiastic about the possibilities for tissue banks since the commercial development of laser capture microdissection by Arcturus Applied Genomics, Mountain View, Calif. "What this instrument does is allow one to dissect a frozen tissue section and—without any destruction—to capture individual cancer cells and put them in one tube and the normal cells in another," he explains.

Working with Arcturus, the Brown Cancer Center has been developing gene expression profiles of pure breast carcinoma cells isolated from sections of the tissues from the biorepositories, using laser capture microdissection, and correlating these profiles with significant clinical data on patients’ nodal status, race, therapeutic outcome, disease-free survival, and overall survival.

In effect, Dr. Wittliff says, "I’m still mining data and using samples I stored 20 years ago. What’s extraordinary is that the protocols set up two decades ago have preserved the activity of the labile RNA. The proteins are still viable—and we can now perform gene expression studies from messenger RNA we’re extracting from biopsies that have remained viable for more than 20 years."

The cost of research tissues, however, is increasingly controversial. Depending on the type of tissue and amount of clinical information needed, the expense of a single tissue sample might reach four figures—out of range for many academic researchers. Asterand announced in August 2002 that it was revamping its price list for common cancer tissue samples in response to the different needs of academic and commercial researchers.

"We recognize that medical research is affected by budgets and commercial realities just like most other areas of human endeavor," says Asterand business development specialist Victoria Blanc, PhD. Noting that some experiments have more rigorous requirements than others, she says, "Rather than having one monolithic price for each kind of tissue, our new prices will reflect the level of quality assurance required." Common cancer tissue samples such as lung, breast, and stomach now range from around $100 to several hundred dollars if detailed pathology reports are required.

A myriad of complications surrounds informed consent as well. Margery Moogk, MS, says that researchers who request tissue from Seattle’s nonprofit Northwest Tissue Bank, which she directs, must sign a statement agreeing that the tissue will not be commercialized. "So if a biotech company were trying to develop a diagnostic tool that recognizes liver cancer, they may need to have normal liver tissue to compare how that diagnostic tool distinguishes between normal and abnormal cells, but the product may not have liver cells," she says.

The Cooperative Human Tissue Network funded by the National Cancer Institute, which distributes about 80,000 specimens per year in North America through six participating institutions, has adopted a similar restriction. "We basically make tissue available to people who call and ask, and pharmaceutical and clinical diagnostics companies have access as long as they are doing research and not incorporating the tissues into product development," says Roger Aamodt, PhD, chief of the NCI’s Resources Development Branch and president of the International Society for Biological and Environmental Repositories.



While NCI is in close contact with the big pharmaceutical companies, it has no formal relationship with them. "We feel we’re serving pretty much a different constituency," Dr. Aamodt says. But commercial and academic researchers can use NCI’s specimen resource locator or "tissue expediter," which functions as a single point of contact for researchers looking for tissues. In many cases, researchers don’t have any other means of learning about these tissues because they’re not commercially available.

A more elaborate effort underway at NCI is the Shared Pathology Informatics Network, or SPIN, which will use state-of-the-art informatics techniques to establish an Internet-based virtual database. "This will allow researchers to query a large number of institutions’ electronic clinical information systems and pull out a listing of what pathology specimens are there that will meet their research needs," Dr. Aamodt says. "It will be done in a way that totally protects patient privacy and confidentiality and will essentially be infinitely expandable."

SPIN, a five-year project now in its second year, is developing software to give researchers limited access to de-identified patient data. NCI says the need for such a system has been fueled by the growing use of tissues and diagnostic specimens and their related clinical data in biomedical research.

Since most pathology laboratories store at least 10 years of pathology reports electronically, there is a wealth of archived tissue and searchable databases with patient data. SPIN will not create a central database but will facilitate communications among disparate computer systems, even among those that use different architectures and search strategies.

In the meantime, debate persists over how much regulation the tissue banking industry needs. "We’re not governed by any overall FDA regulations," Asterand’s Charlton points out. "In the U.S., we operate under the rules of each institution we deal with, whether it’s an academic university or a private hospital or the federal rules for institutional review boards, and we have our own as well that meets from time to time. It’s a completely different area from organ transplantation. This material isn’t going into anybody—just into a test tube."

Does this mean a cursory informed consent is adequate? The courts have indicated, Moogk says, that "if a company knows in advance it will use tissue recovered from a patient’s body and that it will be a profit-generating thing, you should know that.

"But if an organization is investing a lot of money in research or new product development, and they don’t have a particular target in mind, they’re using the tissue without any real knowledge of or guarantee that it will make a contribution, then it’s not so clear that they have any kind of obligation to go back to an individual and say, ’We extracted DNA from your cells that we used to make the target RNA that is now an important component of our product’—because it’s not really theirs anymore."

"Law regarding ownership of specimens is very complicated and confusing at this point," Dr. Aamodt says. The California Supreme Court, in Moore v. Regents of the University of California, ruled in 1990 that a patient does not have a continuing ownership interest in his excised cells and tissue used in research. However, the court also held that the failure to inform a patient that his collected tissue would be used for research purposes was a breach of the duty to obtain informed consent. This is the only published case on the issue of ownership of excised cells, and there is no overarching federal rule regarding ownership. In some other states, diagnostic specimens are considered part of the clinical record.

For tissue that is "unlinked" (rendered anonymous), the CAP opposes hampering research by well-intended but intrusive regulations. In November 2001, the CAP’s Ad Hoc Committee on Tissue and Organ Procurement took the position that consent forms should use simple (general or unspecified) wording to include donation of excess human tissue for research, teaching, and quality control testing, and that previously collected specimens for which no consent form exists should be grandfathered so they can be available for research. However, because of the evolving nature of tissue procurement and informed consent laws, the ad hoc committee is revisiting this position.

Many tissues cannot be "anonymized" because they must be linked back to clinical information from the patient to allow outcomes research. In these cases, the CAP cites the federal Office for Human Research Protection’s requirements for specific types of patient consent, including education about the operation of the tissue repository, specific information about the type of research conducted, conditions under which data or specimens might be released, and procedures to protect privacy and confidentiality.

"The federal rule that applies, 45CFR46, has a provision that if there’s no way to identify a subject, you can waive consent with the approval of the IRB," Dr. Proia says. "We do have a general consent form with a sort of nebulous provision for materials to be used for research. That’s where a lot of institutions get into conflict, because the government says that for federally funded research you need fully informed consent, but others say the consent form for surgery allows them to use tissue." So there’s tension, he says, between the CFR, federally sponsored research requirements, and the desire of investigators to get tissue when they may not have the resources to obtain informed consent.

Five years ago, before Duke’s affiliation with Ardais, Dr. Proia recalls, "some investigators who may have gotten a piece of tumor from us would call up and say, ’We have interesting results with this tissue and we need followup.’ And we’d say, ’Didn’t you notice when you got it there was nothing on it that allowed us to track the patient down?’ That was part of the stimulus to affiliate," he adds, "because the ideal is to have a full repertoire of clinical data, pathology data, and followup data, and to do that you have to have fully informed consent from patients undergoing surgery."

For that reason, Dr. Proia says, "we do not collect tissues from autopsies because of the concern about approaching families after someone has died." While there have been discussions about piggybacking onto organ donor programs to request tissue, "due to the whole medicolegal climate, we haven’t gone there." Duke is considering a rapid autopsy program, however, in which people would give consent before death. "You would sign up to be a tissue donor before the research; that way it would be clean," he says.

More rigorous rules will take effect April 14, Dr. Aamodt says. "The regulation of tissue banking is the same as rules applying to any other human subject research, and the Department of Health and Human Services is about to implement new rules under the Health Insurance Portability and Accountability Act. These rules will significantly strengthen privacy and confidentiality controls—probably with some major negative impact on research," he says.

Most tissue research is considered to pose "minimal risk" to patients, but the new privacy rules have extensive administrative requirements that apply whether or not there is a determination of increased risk. If it is health data and is identifiable, then it requires additional administrative procedures.

"Because the rules are very complicated, it will take the research community some time to fully understand what they mean and how they apply," Dr. Aamodt says. In the meantime, "research institutions are struggling to get policies to assure confidentiality while allowing their research to go forward."

The International Society for Biological and Environmental Repositories, to which many research tissue banks belong, is now organizing standards for quality control and best practices, perhaps leading to a formal accreditation process. The group has already created subcommittees to work on best practices documents in sample collection, processing and retrieval, sample storage, sample tracking, sample packaging and shipping, biological safety, training, QA/QC, and human subjects.

"For research tissue banks, there’s no accreditation now, but it’s probably going to happen," Ardais’ Dr. Buckler predicts. "As the use of these resources gets closer to the rationale for clinical development of new therapeutic strategies, it’s going to be a necessary thing." Ardais is prepared, he adds, because of systems it has already put in place to document the validity of samples and strict safety protocols.

Dr. Wittliff emphasizes that the pathology profession must prepare as well. "Biorepositories are really a growing resource. But we’ve got to have a change in the culture of the surgery theatre and in the pathology suite because the tissues have to be handled differently than we are accustomed to for routine pathology. Clinical followup has to be collected, and patient identity must be secure."

Pathologists must be aware of the responsibilities and opportunities that accompany this new role in tissue banking, Dr. Wittliff says. "In this post-human genome era, with breakthrough technologies for proteomics, genomics, and metabolomics, these are the resources that are necessary. This is where it’s going—and we in pathology and laboratory medicine must be guardians of the anonymity of our patients and the integrity of their tissues for research."

Resources

1.  National Cancer Institute Specimen Resource Locator: www.cancer.gov/specimens

2.  NCITissue Expediter: www.cancerdiagnosis.nci.nih.gov/specimens/finding.html#expediter

3.  NCI Cooperative Prostate Cancer Tissue Resource: www.prostatetissues.org

4.  International Society for Biological and Environmental Repositories: www.isber.org

5.  Shared Pathology Informatics Network (SPIN): www.cancerdiagnosis.nci.nih.gov/spin/

Anne Paxton is a writer in Seattle.