New YorkBloodCenterPosition on the
National Cord Blood Stem Cell Bank Network
The Institute of Medicinestudy was requested because of theapparently irreconcilable controversy between the vision of the initiators of the legislation and the objections of the National Marrow Donor Program (NMDP) over the operation and structure of the National Cord Blood Bank Network.
A. The Controversy.
The legislation calls for an independent Network of high quality Cord Blood Banks with its own Board of Directors. The NMDP proposes, instead,that the Network Banks should operate under NMDP. NMDP asserts it is ready now to take on this responsibility because of its broad array of resources and its experiencein facilitating donation by volunteer donors (it is the world’s largest bone marrow donor Registry, partially funded by federal resources). NMDP proclaims it has developed the necessary mechanismsto govern the Network (choose Banks, serve as conduit for funding from HRSA, organize Banksthrough a standing Cord Blood Subcommittee, evaluate Bank performance, provide information technology support for Banks, perform searches for matching cord blood units through a “one-stop shopping” resource, help educate donors and patients, provide patient advocacy services, conduct recipient follow up and promote and fund research). The NMDP argues that an independent Cord Blood Bank Network would need to duplicate everything it has created, which it considers a wasteful and unnecessary effort.
The issues at stake are essential, not just trivial differences between alternative approaches. We show below that the Cord Blood Bank Network must operate as an independent entity and, further, that the Banks must control the search mechanism, whether performed by the Network itself or out-sourced to another organization under contract. Analysis indicates that independence is criticalto thesuccess of the Network andof cord blood bankingin general and,consequently, is vital for the patients who need this resource, because of the absolute requirement for Bank responsibility and accountability,effective inventory management, collegial competition, full transparency and the expectation of eventual financial self-sufficiency. These issues are explored more fully below, followed by specific recommendations to the IOM panel.
B. Critical Requirements forthe National Cord Blood Bank Network.
1. Cord Blood Bank Responsibilityand Accountability.
The Cord Blood Bank hasthe technical, regulatory and legal responsibility for a stem cell product used in clinical transplantation and will be held accountable for it. It is expected that a participating Cord Blood Bank will be a recognized, experienced, high quality entity operatingwithin a reputable sponsoring institution. A Bank must accomplisha variety oftasks, including collection of cord blood units, assessment ofdonor eligibility, testing, processing, freezing, storage, unit selection, distribution and shipping,as well as tracking transplant outcomesas quality control in a comprehensive cGMP-GTP quality management environment. A Bank must have a well-functioning data and inventorymanagement system. Banksmust be accredited by an appropriate, independent agency and ultimately licensedby the U.S. Food and Drug Administration. Banks may contract with a separate entity for one or more specific tasks such as infectious disease testing, blood grouping or HLA typing. Ultimately, however, the Bankbears all responsibilityformaking a clinically appropriate hematopoietic stem cell product that meets all clinical and regulatory requirements. The Bank’s responsibility cannot be diluted by or assigned to a third party. Thus, the proposed legislation designates the Network Board of Directors as responsibleonly for monitoring Bank compliance with accreditation and regulatory requirements, not for meeting them.
2. Cord Blood Bank Accountability to the Federal Government through HRSA.
Each participating Cord Blood Bankwill be accountable for meeting its obligations as a recipient of federal funds. The legislation expects that Banks participating in the Network will be geographically dispersed and will obtain high quality cord blood units from an ethnically diverse donor population so that all ethnic groups in the U.S. population may be served effectively. Each Bank, however,will be individually accountable to the government for meeting numerical and allother obligations it has agreed to undertake. The legislation assigns the Network Board of Directors responsibility for reporting to HRSA on Bank performance.
3. The Mechanisms to Search for, Reserve and Distribute Matching Cord Blood Units Need to be Optimized to FitBank Requirements for Inventory Management.
Cord Blood Banksmust manage theirown inventories of available cord blood units. Cord blood units are currently distributed through multiple mechanisms, including marrow donor registries, NetCord, Bone Marrow Donors Worldwide (BMDW) and even intramurallyfor Banks that operate in institutions also having Transplant Centers. Multiple distribution options foster potential unit reservation conflicts. That cord blood grafts do not have to be perfect HLA matches improves patients’ chances to find suitable cord blood units, but also makes it much more likely that one unit will match multiple patients and increases the likelihood of encountering reservation conflicts. Acceptability of mismatched transplants provides patients with another advantage that depends on access to the most current inventory. Unlike the case with bone marrow, a Center need not commit to a specific unit until the actual time of shipment. Thus, a last minute repeat of the search may identify a superior unit. Centers often tentatively reserve several units for one patient, while confirmatory typing and other pre-release testing is being completed. Thus, cord blood inventories change dynamically reflecting unit reservations by multiple “clients” as well as changes in unit status (out and back into the list of available units), unit shipmentsanddaily acquisition of new units. An active Cord Blood Bank, therefore,must have its own effective, computer-based system to manage its own inventory at the local level.
Patients optimize their chances of finding the best unit for them if the unit search mechanism is based on the Banks’ most up-to-date, current inventories. A centralized database, such as proposed by NMDP, unnecessarily duplicates localBank databases. In order to be current,however, such duplication must be complete(reflecting what is actually available in the Bank) and instantaneous. Some months ago, for example, staff at the Cardinal Glennon Cord Blood Bank explained that they had several hundred units available that had not been entered into the NMDP database. Other Banks considering accepting NMDP funds for collecting units are anxious about adopting NMDP programs because they may not allow inclusion of data in a local database. A more modern, economical and labor-saving system, instead,would distribute the search requests to each Bank,thereby automatically accessing the Bank’s current inventoryand avoidingthe unnecessary,cumbersome and technologically antiquatedduplication in a centralized database. So-called “one-stop shopping” can be accomplished for the benefit of Transplant Centers and their patients by either mechanism, but the duplicative centralized database clearly limits the system’s possibilities if units listed centrally,in fact, do not include all units thatare actually available.
4. Competition and Transparency.
Cord Blood Banks, whose financial stability depends on “sales”, inevitably compete with each other for distribution of units. Banks must be assured that their units have a fair opportunity to be selected. Each Bank must know that its units are listed appropriately in any combined Match Report sent to Transplant Centers. Thus, each Bank must approve any algorithm that consolidates individual Bank Reports and must receive copies of all Match Reports. Further, the Bank must knowwhether unit selection was appropriate. Banksshould bear some responsibility for, and participate in unit selection when there is any contention. Thus, each Bank must receive a report on units that are selected for transplant. Finally, where there is any ambiguity, the TransplantCentershouldexplain its choice. Such transparency and accountability also willensure that patients truly get the most appropriatecord blood unit available.
Cord blood is no longer just an option for patients who have no other choice. As experience has increased (now approaching 6,000 transplants worldwide), cord blood is becoming competitivewith bone marrow as a source of hematopoietic stem cells for one and the same patient. Patients and their physicians now have two options to choose from, not just a first choice and a last resort. Several TransplantCenter physicians now preferentially choose cord blood grafts over bone marrow or peripheral blood, even for adults. As a consequence, Cord Blood Banks and Marrow Donor Registries understandably and unavoidably have become competing organizations. It is even plausible that cord blood could eventually replace bone marrow as the hematopoietic stem cell source of choice in unrelated transplants. If unrelated marrow donors are no longer needed or justifiable, the rationale for and ethics ofrecruiting donors and harvesting marrow or peripheral blood stem cellswill become difficult if not untenable. Moreover, the primary source of Registry funds (reimbursement fees of diverse kinds) will progressively decline. Thus, theMarrow Donor Registry conceptmay be threatened by the success of cord blood banking and vice versa. On the other hand, competition might actually work to improve the performance of both organizations, to the benefit of all patients. Each organization, therefore, must control its own operations. Neither should be allowed to control the other. Their futures should be determined, instead, by independent scientific and clinicalevaluation of their reciprocal contributions and advantages.
NMDP, nevertheless, is a marrow donor organization that recently also undertook to distribute cord blood units for existing Cord Blood Banks, sponsoring affiliated Banks under an “umbrella” FDAIND. Its cord blood unit distributionsthus far, however, have been meager. The table below summarizes, for comparison, data from NMDP, NYBC and the Japanese National Cord Blood Program (through June 30th 2004).
Program Current 2004 Search CB Units Distributed .
Inventory Requests Total 2004to June 30, 2004 . . # # Rate/1000 Searches
NMDP* 35,707 ~3,000 340 ~80<30
NYBC** 22,420 8191,734 100 122
Japan*** 19,294 ~3,000 1,769 371 124
* from D. Confer, IOM meeting 8/18/04. 1999 through June 30, 2004.
** NYBC data, February 1, 1993 through June 30, 2004.
*** from Tsuneo Takahashi, January 1, 1998 through June 30, 2004.
Thus, while NMDP has distributed some units from its affiliated Banks, its track record cannot be seen as reassuring regarding its effectiveness, despite having access to most U.S. patients who seek a transplant. In the absence of evidence to the contrary, the data could suggest that the involvement of NMDP and its mechanisms may even have been detrimental to the Banks and, quite possibly, to the many patients who did not get a graft and who might have been helped.
The corollary is unavoidable. Cord Blood Banks must participate in the establishment of criteria and control the mechanism for cord blood unit selection and distribution. The mechanism must be fully acceptable not only to the Banks,but also to the other interested parties including TransplantCenter physicians and patients. The legislation assigns the independent Network Board of Directors (which includes Bank, TransplantCenter and patient representation) the responsibility for establishing the specifics of the search mechanism. An outside agencymight compete for and be selected to perform searches for the Network. Any organization that performs searches for Network Banks, however, must answer to the Banks. If the organization does not perform effectively, the Banksmust have the authority to replace the organization. To assure accountability, Banks should pay, under contract, anyoutside agency selected to perform the search mechanism and the Banks must own the search programs the organization develops in order to guarantee transferability. The cost for establishing and operating the search mechanism, thus, should be part of each Bank’s budget.
5. Financial Self-Sufficiency.
Cord Blood Banks differ from other organizations involved in tissue or organ donation because they produce a tangible asset that, when sold, allows forcost recovery, and makes self-sufficiency possible. Legislators that sponsored the FY 2004 appropriation and pending legislation expect Cord Blood Banks to become self-sufficient after the five years of federal funding. Self-sufficiency is only feasible, however, if the Banks own the cord blood units they collect and are allowed to charge for those they distribute for transplantation. Ownership of the units must also take into account the fact that existing Cord Blood Banks already have made considerable upfront and continuing investment in establishing operations and building inventory (so far, more than $50 million in our case). Moreover, Banks that succeed in becoming part of the Network will likely continue to invest some of their own resources since federal funding probably will not cover all costs. At the first IOM panel meeting, we provided several scenarios regarding Banks’ self-sufficiency based on their ability to charge for units distributed,demand for cord blood unitsand the transplant community’s level of trust in the reliability of cord blood, including the level ofsafety and effectiveness.
C. What the Legislation Does Not Do.
1. Fund Research.
The legislation will provide that a portion of the cord blood units collected be made available for research. Funding for research is already the responsibility of other agencies such as the NIH, CDC, NSF, etc. This legislation does not create another funding agency.
2. Create an Infrastructure thatDuplicates NMDP.
The legislation is intended to fund the rapid building of a high quality cord blood unit inventory on a national scale; the critical resource that patients need. Much of the infrastructure that is essential for a bone marrow registry is irrelevant to cord blood banking and transplantation. The logistics required tocoordinate donation, harvest and transplantation of bone marrow are not needed in cord blood. Cord blood is not faced with same issues of protecting donor confidentiality and privacy. Thus, the TransplantCenter can go to the Bank directly with no need for a protective intermediary. Many of the patient advocacy issues that plague unrelated marrow donor registries do not apply to cord blood. Most importantly, there need be no upfront charges to patients. NYBC, for example, does not charge for the search, confirmatory HLA typing of the unit or patient, nor for other pre-release testing;samples are provided free to the Transplant Center as are shipping and post-transplant HLA typing, all elements that we consider integral to the transplant. We recently began to charge Centers for class I HLA typing by DNA sequencing, however, since we currently consider such testing not necessary for unit selection.
D. Recommendations.
1. Cord Blood Bank Selection.
Cord Blood Banks should be selected to participate in the Network along the following principles:
(1) Selection should be by open, fair competition with defined selection criteria published in advance of the actual competition.
(2) Selection should be based on applications to HRSA from individual Banks that demonstrate the Bank’s experience, quality of performance, capability to contribute to the Network goals for the national inventory and costs.
(3) Quality of candidate Bank performance should be evaluated based on pre-defined criteria, including validation of procedures and transplant outcome (engraftment and survival), appropriate accreditation and FDA licensure.
(4) Banks should be selected by an independent panel of experts with no real or apparent conflict of interest. Individuals employed by any organization which benefits financially fromcollection or distribution of cord blood units must be disqualified automatically from serving on the selection committee. Thus, employees of a candidate Cord Blood Bank or of marrow donor registries that distribute cord blood units (such as NMDP) should be disqualified.
2. Selection of an Organization to Perform Searches of Matching Cord Blood Units.
If outsourced to a separate organization, selection of the organization to perform the searches must follow the same principles outlined above for selection of Cord Blood Banks. The legislation assigns the Network BOD the responsibility of defining the specifics of the search mechanism that applicants must meet.
3. HRSA Should Distribute Federal Funds Directly to the Banks as Grants.
Funding Banks should be through grants rather than the contract mechanism to assure there will be no ambiguity about who owns the cord blood units, anessentialcondition for self-sufficiency. Our own experience and that of COBLT illustrate the desirability for this approach. The legislation mandates, however, that the Banks’ sponsoring institutions must guarantee continued access to Network Bank cord blood units. Thus, the sponsoring institution must agree to maintain the units if collections cease or the Bank closes for any reason or, alternatively, agree to transfer the units to another sponsor, with appropriate financial protections.