July 27, 2005 NCI position paper on PedCIRB conflicts of interest
July 27, 2005
Central IRB Membership: Enrolling Subjects, Conflict of Interest and the Central IRB Initiative
When 45 CFR 46 was promulgated 30 years ago, the conduct of biomedical research differed dramatically from the contemporary practice of multi-institutional clinical trials. Federally-funded clinical trials were most often conducted at a single institution or among a handful of institutions; therefore, when the federal government created the regulatory scheme for human subject review, an institutionally-based approach was logical. Federal regulations specified that IRBs be “sufficiently qualified through the experience and expertise of its members … to promote respect for its advice and counsel in safeguarding the rights and welfare of human subjects”. During that same era, conflict of interest of Board members was considered so self-evident that the IRB regulations did not define “conflict of interest” for IRB members, stating only “No IRB may have a member participate in the IRB’s initial or continuing review of any project in which the member has a conflicting interest, except to provide information requested by the IRB.”
Today, however, a typical phase 3 cancer clinical trial is conducted at 50-500 sites, and the focus on investigator conflict of interest has intensified. New models of human subject protection have been developed which require reassessment of how investigator-Board member conflict of interest is defined and managed. In this context, the discussion below focuses on conflict of interest, specifically that of physician members who sit on a nationwide central IRB and who also have the potential to enroll patients on those same trials at their local institutions.
In 2001, the NCI piloted a central IRB (CIRB) for multi-institutional phase 3 adult clinical trials conducted by the NCI-supported cancer cooperative groups. A facilitated protocol review model, one of the options for central IRBs acceptable to OHRP, was established in which the CIRB conducts full board review and generates a review packet including the minutes of the CIRB meeting, the Board’s final comments on the protocol and an IRB application form for the protocol. The institutional IRB chair or IRB subcommittee utilizes the protocol application and review packet to conduct a subsequent brief review to address local issues related to conducting the clinical trial at the individual institution. If the local IRB chair has no institutional concerns and accepts the CIRB review, the CIRB becomes the IRB of record and the CIRB assumes responsibility for amendments, adverse events (AEs) (with the exception of AEs occurring at the local site which are reported to the local IRB to keep them alerted to local safety issues) and continuing reviews for the life of the protocol. In late 2004, NCI established a second CIRB to review phase 2, 3 and pilot pediatric protocols that also used the facilitated review model.
As stated above, federal regulations and guidance do not explicitly define conflict of interest, thereby leaving IRBs, including the NCI CIRBs, to define conflict of interest for themselves. The CIRB conflict of interest policy states that CIRB Board members who are involved in the development, conduct or analysis of the trial under review or who have prominent roles in the disease committee submitting the trial are considered to have a conflict of interest and are recused from the protocol discussion and vote. With the advent of the Pediatric CIRB (PedCIRB), conflict of interest was revisited with OHRP and led to more specific direction. The most recent advice provided by OHRP requires Board members to recuse themselves from the review of trials on which they have a “reasonable expectation” to enroll patients during and after their PedCIRB tenure:
In general, when a member of one of the NCI-sponsored oncology groups is serving as a member of one of the NCI Central IRBs, OHRP would consider such a member to have a conflicting interest for an oncology group protocol undergoing initial or continuing review whenever that IRB member, given the scope of his or her clinical and research duties, has a reasonable expectation of enrolling and/or managing subjects on that oncology group protocol in the future, regardless of whether the IRB member serves on any oncology group committee related to the development or management of the protocol.
The current review procedures adopted by the PedCIRB to comply with these terms are provided at the end of this document.
Following extensive discussions with PedCIRB members and ongoing discussions within CTEP/NCI, we take the position that PedCIRB members whose participation in the clinical trials process is limited to enrolling patients in a manner consistent with pediatric oncology clinical practice have no conflict of interest in reviewing trials. No benefits accrue to these members that would bias either their review of the trial or any subsequent decision to enroll a patient. Taking the position that PedCIRB members who enroll patients onto clinical trials are conflicted relegates practicing pediatric oncologists to a consulting role on the PedCIRB. In the short-term, this approach may not compromise the quality of PedCIRB reviews. However, in the long-term, this approach will make the recruitment and enthusiastic participation of pediatric oncology Board members more difficult. These are the very individuals “possessing the professional competence necessary to review specific research activities” as specified in federal regulations, and their diminished participation will compromise both the quality of PedCIRB reviews and the perception of the quality of the review by local IRBs. Inevitably, this perception of lack of expertise will slow or halt the affiliation of local IRBs with the CIRB and jeopardize the entire Initiative.
Prior to discussing specific issues related to conflict of interest for practicing pediatric oncologists who serve as PedCIRB members, it is necessary to first review how the practice of pediatric oncology has evolved over the past four decades. Pediatric cancer specialists in North America have cooperatively designed and conducted sequential, randomized phase 3 clinical trials for over four decades in order to use evidence-based information to refine the treatment of children with cancer. The "state-of-the-art" treatment for a child with cancer derives from the best therapy identified in prior clinical trials and this best available treatment serves as the standard arm of subsequent phase 3 clinical trials.
The pediatric oncology community of physicians, other health care providers, and patient advocates generally accepts that it is appropriate to offer families of children with cancer the opportunity to participate in appropriately designed clinical trials, with families being free to accept or decline participation without penalty. This practice is deemed necessary because most current therapies are inadequate because of insufficient efficacy and/or excessive toxicity, and because well-designed clinical trials are required to identify more effective treatments. Implementation of this approach in an ethically acceptable manner requires multiple safeguards to ensure that the risks and benefits to research subjects remain in balance throughout the course of the clinical trial. Various groups share responsibility for this task, including the research team, government agencies with review responsibilities (e.g., NCI and FDA), Data and Safety Monitoring Boards, and IRBs. The close linkage in pediatric oncology between clinical practice and research requires acknowledgement of the tension between the practitioner’s roles as a clinician and as an investigator. As an example, the Children’s Oncology Group (COG) is addressing this tension by studying the informed consent process to clarify and to improve both the decision-making process that families encounter and the information provided by the pediatric oncology team.
The pediatric oncology model described above has important implications for discussions about conflict of interest for physician PedCIRB members. Because pediatric oncologists offer families participation in clinical trials as part of the clinical care process, the vast majority of such physicians eligible to serve on the PedCIRB would likely enroll patients onto clinical trials. Hence, it is critical to determine whether the tension that exists in this investigator-physician role rises to the level of conflict of interest for physician PedCIRB members.
Conflicts of interest are most commonly thought to occur when the investigator may be biased by financial gain as the secondary interest. In addition, other secondary gains may include authorship of trial-related publications, academic advancement, enhancement of professional reputation or increased institutional patient referrals. It can be argued that oncologists who serve on the PedCIRB who have a reasonable expectation of enrolling or managing subjects on PedCIRB reviewed protocol in the future do not fall into any of these four groups.
For at least the last 20 years, the large majority of children with cancer in the United States have been treated on or according to COG protocols. Virtually all physicians and health professionals caring for children diagnosed with cancer have some association with the COG. However, the large majority of these physicians and health professionals caring for these children have no direct involvement in the development or management of specific trials. (The existing PedCIRB conflict of interest policy excludes members of the PedCIRB who sit on COG study committees or data monitoring boards from participating in, or voting on, protocols with which they might be involved).
PedCIRB members have no potential for financial gain related to the future patient enrollment of subjects on a clinical trial. Local institutions are reimbursed on a per study entry case basis by COG from an NCI grant. These funds are usually used to subsidize the increased institutional administrative salary burden for Clinical Research Associates required for COG study entry and data compliance. It is widely acknowledged that NCI’s reimbursement for research costs, currently $2,000 per patient, is far less than the actual costs incurred by institutions to participate in NCI-sponsored clinical trials. Institutions essentially contribute the balance of the research costs, making clinical trials participation a net loss rather than a financial benefit. The local disbursement of these funds is controlled by the COG institutional principal investigator (PI), not by the oncologist registering the patient on study. The existing PedCIRB conflict of interest policy excludes local institutional PIs from serving on the PedCIRB due to the distribution of per patient reimbursement for protocol accrual to the institution.
There is minimal likelihood for publication authorship, academic advancement, or enhanced professional reputation associated with Board member protocol review. Such academic benefits go to members of the COG national study committee or the COG disease committee. These persons are excluded from service on the PedCIRB by PedCIRB conflict of interest policies.
There is no likelihood for increased institutional patient referrals related to Board member review of a COG trial. The large majority of pediatric oncologists in the United States are COG members and already have access to COG clinical trials for their patients.
For the reasons cited above, we conclude that physician PedCIRB members whose participation in COG clinical trials is limited to offering protocol participation to families in a manner consistent with pediatric oncology clinical practice do not have a conflict of interest that would preclude them from reviewing COG clinical trials and, in fact, are performing a service both for child research subjects and for the cancer research community.
Proposed Policy for PedCIRB Members.
CTEP believes that PedCIRB members who are practicing pediatric oncologists and who meet the PedCIRB’s conflict of interest requirements can conduct the initial and ongoing review of COG clinical trials without conflict of interest. These Board members would not obtain financial, academic or institutional benefit from either participating in PedCIRB reviews or from offering enrollment onto a PedCIRB reviewed trial to potential research subjects as part of their clinical practice. Therefore, we conclude that physician PedCIRB members whose participation in COG clinical trials is limited to offering protocol participation to families in a manner consistent with pediatric oncology clinical practice do not have a conflict of interest that precludes them from fully participating in PedCIRB reviews of COG clinical trials.
The PedCIRB benefits from the full participation of practicing pediatric oncologists as Board members and from the pediatric oncology expertise and practical experience that they can share with the Board. Alternatives that relegate practicing pediatric oncologists to consulting roles on the PedCIRB, while not necessarily compromising quality of PedCIRB reviews in the short-term, will make the recruitment and enthusiastic participation of pediatric oncology Board members more difficult. The eventual result of these alternatives will be a perception of lack of expertise for PedCIRB reviews that will slow or halt the affiliation of local IRBs with the PedCIRB. As a primary purpose of the NCI CIRB Initiative is to improve the clinical, scientific and ethical expertise of protocol review, failure of the PedCIRB to achieve its goals will diminish research subjects protection and mean that children with cancer are less well served.
Background Information - Interim Operating Procedures for PedCIRB
The PedCIRB is currently operating in a manner consistent with the most recent advice provided by OHRP that requires Board members to recuse themselves from the review of trials on which they have a “reasonable expectation” to enroll patients during and after their PedCIRB tenure. A general outline of procedures relevant to implementation of this advice is provided below: