Restoring Balance to Industry-Academia Relationships in an Era of Institutional Financial Conflicts of Interest: Title and subTitle BreakPromoting Research While Maintaining Trust

In medicine, academia is no longer as autonomous as it once was. Impelled by, among other things, the Bayh-Dole Act,¹ US academic centers have partnered with industry so that academic innovation can be rapidly and efficiently brought to market. Academic discoveries are often patented by universities and then licensed to industrial sponsors for commercialization. This translates into greater patient access to therapeutic advances, and ultimately serves the public good. Yet the nature and scope of this economic partnership have outpaced what was originally intended and have developed into a highly interwoven relationship extending to all levels of academia and the research enterprise. In addition to engaging in licensing agreements with private industry, academic institutions may own stock or options in the sponsors of research being conducted at the institution; incorporate start-up companies to develop faculty inventions in which they and their faculty members are major shareholders; accept cash compensation for granting preferred industry partners with first refusal rights on the discoveries of investigators or departments; and in some cases even develop their own brand-name products to be sold on the market.^{2 - 3} Most universities have established technology-licensing offices to manage growing research-related business operations.

Academia's relationship with industry extends beyond the university's corporate interests. Researchers, institutional review board (IRB) members, and institutional decision makers (eg, trustees, presidents, chancellors, provosts, deans, department chairpersons) also have developed extensive financial ties with industry. These individuals may own stock or options in drug or device manufacturers or other industry sponsors of research; be beneficiaries of actual or expected royalty payments from the sale of industry products tested or developed at the institution; receive private research support through grants or contracts; be consultants to or directors of private research corporations; receive fees for serving as expert witnesses on behalf of industry in legal proceedings or for supporting industry lobbying or marketing activities; receive honoraria for speaking on behalf of industry at scientific conferences; or receive research-related gifts, such as discretionary funds, biomaterials, or research equipment.^{4 - 8}

While entrepreneurship in academia has accelerated scientific innovation, on occasion it has also marred academia's reputation as independent truth-seeker, reduced public trust in the research enterprise, and resulted in a burgeoning literature on conflicts of interest.^{9 - 15} This literature is focused on the problems of investigator conflicts of interest and faculty conflicts of commitment, with little scholarship having been dedicated to institutional conflicts of interest.^{16 - 18} Several empirical studies and anecdotal reports demonstrate that financial conflicts of interest can affect the professional judgment of physicians and researchers,^{19 - 24} and there is growing concern within the research and regulatory communities that institutional financial conflicts of interest similarly may affect professional judgment. This concern has partly arisen from recent well-publicized research-related injuries or deaths in which the institutions hosting research, or noninvestigator officials within such institutions, reportedly had significant financial interests in the research.^{18 ,25 - 27}

Some who have studied the conflict-of-interest issues raised by the ever-deepening academia-industry partnership have lost hope for the possibility of a middle ground in which financial incentives spur innovation without corrupting or appearing to corrupt academic and ethical values,²⁸ including protection of human participant safety and welfare, academic freedom, objectivity, data integrity, the right to publish, and scientific collaboration.^{29 - 30} Doubtful of the prospect of a balanced alternative, physicians have been prompted to adopt positions at either end of the regulatory spectrum, some stressing the overreaching value of entrepreneurship in medical research while advocating a laissez-faire approach to financial conflicts of interest in research, and others highlighting the dangers to research and institutional integrity while emphasizing the need to reduce significantly or to eliminate industry-academia relationships.^{31 - 33}

We suggest that, even in an era of institutional conflicts of interest, it is still possible to promote (and even accelerate) the progress of research while maintaining (and even enhancing) public trust in the research enterprise by restoring balance in, but not eliminating, industry-academia relationships. To that end, we (1) discuss the nature and operation of institutional financial conflicts of interest; (2) propose a test for determining when financial interests should be eliminated and when they should be tolerated with oversight; and (3) set forth practical strategies for dealing with institutional financial conflicts of interest.

Unlike investigator financial conflicts of interest that are addressed by the Public Health Service, the National Science Foundation, and the US Food and Drug Administration,³⁴ no laws or regulations directly govern the financial conflicts of interest of institutions or institutional decision makers. A few regulatory agencies and professional associations have offered guidance on institutional conflicts of interest. Among the first were the Draft Interim Guidance issued by the federal Office of Human Research Protections in late 2000,³⁵ and commentary on that draft by the National Human Research Protections Advisory Committee in August 2001.³⁶ Additional guidance has been issued by the Association of American Universities,³⁷ the Pharmaceutical Research and Manufacturers of America,³⁸ and most recently by the Association of American Medical Colleges (AAMC).³⁹

Institutional financial conflicts of interest may be understood as circumstances in which professional judgment regarding a primary interest (eg, patient welfare or research integrity) may be compromised by the actual or expected pecuniary corporate interests of the institution or its departments, the actual or expected individual economic interests of noninvestigator institutional decision makers, or the actual or expected individual economic interests of IRB members or the members of other institutional bodies responsible for research oversight. The research-related financial interests of institutional decision makers, IRB members, and members of other research oversight bodies are properly characterized as leading to institutional conflicts of interest when they threaten to compromise a primary interest because they arise from the individuals' authority and influence over research at the institution. Thus, institutional conflicts of interest can arise either from corporate or from individual financial holdings or relationships in research, and should be distinguished from investigator conflicts of interest.

Developing strategies for managing institutional conflicts of interest requires understanding the mechanisms through which such conflicts may operate and be expressed. Institutional conflicts typically involve institutional decision makers or IRB members. The institutional conflicts may inappropriately influence decisions of institutional decision makers or IRB members, or the conflicts may be transferred onto and then expressed through others at the institution, such as staff or investigators. Yet even when an institutional conflict is transferred onto and expressed through, for example, an investigator, the conflict remains institutional in nature since the financial interests that produced the conflict belong not to the investigator, but rather to the institution, an institutional decision maker, or IRB member. Therefore, conflict-of-interest policies should require investigators to disclose not only their personal financial interests in research, but also any information they may have regarding the financial interests of the institution, of an institutional decision maker, or of an IRB member in that same research. A conflict-of-interest oversight system thus may assess the need to manage any institutional interests that might inappropriately influence the investigator. Institutional conflicts may operate before, during, or after the review and performance of research.

First, institutional conflicts may result in inappropriate decision-making by institutional decision makers or IRB members. For example, an IRB member engaged in initial or continuing review of a research study may be improperly influenced by the fact that the IRB member, the institutional department with which he or she is affiliated, or the institution as a whole stands to profit significantly from US Food and Drug Administration marketing approval of the product under investigation. This may lead the IRB member to be more lenient or forgiving during initial review (eg, inadequate disclosure of study risks, insufficient description of eligibility and exclusion criteria, exaggeration of potential study benefits) or continuing review (eg, enrollment of subjects not meeting eligibility criteria, failure to exclude subjects meeting exclusion criteria, failure to report adverse events). The conflicted IRB member could even hesitate to suspend or terminate a study, based on awareness of institutional interests implicated.

Second, institutional conflicts may be transferred onto, and result in inappropriate decision-making by, others at the institution. For instance, an institutional decision maker may pressure support staff, IRB members, or investigators to achieve a research end point that is favorable to the pecuniary corporate interests of the institution or the personal economic interests of the institutional decision maker. This pressure may vary in the level of its directness and vigor. For example, while investigators who are informed of their department's significant economic interests in the outcome of their study may not be unduly influenced by this information alone, the conflict of interest that is created through such knowledge might be exacerbated to the point of affecting professional judgment if these investigators are also notified of their department's financial shortfalls or are notified of important upgrades that could be implemented within the department should additional funds become available. Moreover, depending on the manner and content of the information conveyed, investigators might assume that they are being implicitly or explicitly directed to exercise their discretion so as to favor the department's pecuniary interests. The risk, then, is that institutional support staff, IRB members, and/or researchers may act on the bias-generating information that is transmitted to them, directly or indirectly, by institutional decision makers.

Third, while institutional financial conflicts of interest that operate during the review or performance of research most commonly would be expressed through IRB members and investigators, those conflicts operating before or after the research process may be directly expressed through institutional decision makers and their support staff. Institutional decision makers may decide in advance, for example, preferentially to allocate institutional resources, including funds, equipment, or laboratory space toward industry-sponsored or clinically patentable work. After research has concluded, institutional pressure may, for example, delay publication or restrict oral communications of research results beyond what is reasonably necessary for the institution's office for technology licensing to secure patent rights to academic discoveries. However, the most worrisome institutional financial conflicts of interest are those that operate on IRB members, research administrators, and investigators during the review or conduct of research because these may directly jeopardize the health and safety of human research subjects or lead to inappropriate data manipulation.

One means of restoring balance to industry-academia relationships that would reduce both the appearance of bias and the potential for actual bias, but would not eliminate the financial incentives that genuinely promote innovation in research, would be to require individual and corporate possessors of significant industry-related financial interests and relationships to have a legitimate justification for such interests and relationships. That is, possessing such interests and relationships would be a privilege and responsibility, rather than a right. Absent a legitimate justification, divestiture of significant industry-related financial interests and relationships or recusal from research oversight responsibilities would be expected.

The treatment of investigator conflicts of interest is the historical precedent for managing institutional conflicts. Most policies, including those suggested by the AAMC and the National Human Research Protections Advisory Committee (NHRPAC), would allow some exceptions to the general presumption that an investigator should hold no significant financial interests in research he or she is conducting. According to the AAMC and the NHRPAC, investigators should be allowed to maintain such interests in cases in which the researcher is the inventor of the device or drug under study and may be the best positioned to conduct research safely and competently. Such exceptions would also serve the social purpose of encouraging investigator-entrepreneurs to continue their interest and involvement in their own inventions, thus providing incentives for new inventions and ideas. In these contexts, the social purposes of encouraging entrepreneurship are greatest, even though dangers to data integrity also may be highest. Therefore, when inventors are allowed to conduct human participants research on their own ideas or inventions, independent oversight and tough management of the personal conflicts are indicated.

When institutional conflicts of interest arise from IRB members' or institutional officials' personal research-related holdings, the social purpose of tolerating conflicts to encourage entrepreneurship vanishes, and only a palpable risk to principled research oversight is left. There are no legitimate justifications for allowing IRB members to have significant financial interests related to studies they review. Moreover, because these members have primary responsibility for protecting the safety and welfare of human research participants in trials at the institution, there are compelling reasons for requiring divestiture. The importance of distinguishing between persons responsible for overseeing biomedical research and those carrying out such research is gaining acceptance in both the legal and the biomedical communities. In general, a zero-tolerance policy regarding financial conflicts of interest typically is applied to the former category of persons. For example, the law has already instituted a zero-tolerance policy with respect to IRB members who may not hold any financial interests in the research they review.^{40 - 41} Moreover, according to the Uniform Requirements for Manuscripts Submitted to Biomedical Journals,⁴² editors who make decisions about manuscripts must have no personal, professional, or financial involvement in any of the issues they might judge. Peer reviewers either should disqualify themselves from reviewing specific manuscripts or disclose any conflicts of interest that could bias their opinions of a manuscript.

This policy of zero tolerance regarding financial conflicts of interest should also be applied to institutional decision makers. At minimum, institutional decision makers should not be permitted to have any significant financial interests implicated in research being conducted at the institution. As with IRB members, there seem to be no legitimate justifications. Institutional decision makers are not usually the progenitors of academic discoveries, and technology transfer is not furthered when institutional decision makers have significant research-related financial interests. In those rare instances in which an individual is both inventor and institutional decision maker at the institution where the invention is being tested or developed—a recent example being the clinical drug trials of cetuximab at the M. D. Anderson Cancer Center at the University of Texas, where the conceiver of the drug serves as president⁴³ —the individual should be entitled to maintain his or her financial interests in the invention. However, removal of the study to an impartial institution should be considered. When decision makers are not also inventors, divestiture by institutional officials of personal significant industry-related financial interests vindicates their duty to uphold institutional integrity by ensuring compliance with laws, codes of ethics, and institutional policies.

The elimination of institutional financial conflicts of interest arising from the individual economic interests of institutional decision makers, IRB members, and any other persons at the institution who oversee clinical trials or safeguard the safety and welfare of human research participants will promote regulatory consistency and administrative simplification, enhance public confidence in the research enterprise by reducing the appearance of bias, and promote institutional integrity by reducing the likelihood that actual institutional bias affects research. Moreover, although such persons would be prohibited from maintaining any relevant research-related financial interests, they would remain free to invest in matters unrelated to research.⁴⁴

Legitimate justifications exist for permitting institutions to derive income through licensing agreements with industry or to own equity in start-up companies aimed at developing faculty discoveries. Both types of interests serve the intent and purposes of the Bayh-Dole Act by promoting the commercialization of academic inventions. Moreover, the licensing and equity proceeds that are eventually received by the institution may be used to fund additional research at the institution.⁴⁵ Technology transfer is promoted through licensing because commercial entities have the resources to bring laboratory discoveries to market. Prior to the Bayh-Dole Act, the federal government retained ownership of technologies derived through federal research funding, which resulted in significant delays or total impasses in getting these technologies to market because "few companies were willing to take licenses on government-held patents."^{45 - 47} By 1978 (the year Bayh-Dole was introduced), only 4% of the 28 000 patents owned by the government had been licensed to the private sector for commercialization.⁴⁸ Congress enacted the Bayh-Dole Act to increase the speed with which innovations are brought to market, thus enhancing public access to these innovations and increasing the United States' world market competitiveness.⁴⁹ Commercialization of research and development has significantly accelerated after the Bayh-Dole Act. For example, between 1991 and 1995, licensing activity increased by 68%.⁵⁰ Between 1991 and 1999, licensing increased by 129%.⁵¹

In exchange for equity interests, academic institutions provide start-up companies with shareholder capital that then is used to finance the testing and development of one or more faculty discoveries. The nation's biotechnology industry and its continued world dominance have in fact been credited to the Bayh-Dole Act.³ While few small or newly formed start-up companies (which are the bedrock of the biotechnology industry) had the resources to surmount the bureaucratic red tape associated with obtaining a license from the federal government, 66% of licenses issued by universities in the year 2000 were to small or newly formed corporations.⁵² Nevertheless, the financial conflicts of interest created as a result of the pecuniary corporate interests of the institution should be subject to the oversight and management jurisdiction of a specially constituted conflict-of-interest committee.

Because institutional conflicts of interest may arise from the research-related financial holdings of IRB members, institutional decision makers, or the hosting institution, a comprehensive policy on institutional conflicts of interest will address each of these sources of conflicts. There are no justifications for, and there are compelling reasons against, allowing IRB members and institutional decision makers to maintain their research-related financial interests. Consequently, policies on institutional conflicts of interest should require IRB members to divest themselves completely of any financial interests they may have in any research they review or to recuse themselves from reviewing research in which they maintain an interest, and should require institutional decision makers to completely divest themselves, or to divest themselves beyond a threshold of significance, of any financial interests they may have in any research taking place at the institution. This could be effected by requiring IRB members and institutional decision makers to disclose annually their research-related financial interests to the institution's conflict-of-interest committee, and to update that committee when those interests materially change. This compliance strategy would build on that already existing for investigator conflicts. The conflict-of-interest committee could be charged with the responsibility for ensuring compliance with the policy on institutional conflicts of interest by IRB members and institutional decision makers. That committee then could be given the necessary powers to audit for the purpose of verifying the accuracy of financial disclosures and the power to impose sanctions for noncompliance. This single step would eliminate 2 of the 3 sources of institutional conflicts of interest.

Given the legitimate justifications for allowing institutions to maintain their significant financial interests in research, these interests should be managed rather than eliminated. The primary methods for controlling institutional conflicts of interest should include adequately separating research operations from institutional investment activities, and instituting oversight by an independent review panel (IRP).¹⁷

The separation method, which encompasses both physical separation and certain information-sharing restrictions regarding the institution's corporate holdings and relationships, can be used to control institutional conflicts of interest arising from the institution's research-related pecuniary corporate interests. Physical separation entails housing the institution's office for technology transfer in quarters that are set apart from faculties and departments whose members conduct research. Thus, the office might be attached to the provost's office or to other senior-level offices in the university, rather than to the faculties of medicine or science under the control and supervision of the deans of medicine or science.

Adequate separation also requires certain limitations on communication between those responsible for the institution's financial investment activities and those engaged in the performance or oversight of research at the institution. In general, information regarding the pecuniary assets and relationships of the institution should be considered the confidential information of the office for technology licensing, and should only be disseminated on a need-to-know basis in accordance with the formal policy of that office. The policy should describe the persons or categories of persons to whom disclosures may be made, the types of information that may be disclosed, and the purposes for which disclosures may be made. Additionally, technology licensing offices should ensure that bias-generating information, such as descriptions of research projects or start-up companies from which the institution stands to profit significantly, are not distributed within the university and among research faculty. Although this would not prevent all information regarding institutional financial interests from leaking to faculty, it at least would signify appropriate modesty and restraint about possible conflicts of interest arising from institutional interests.

Institutions increasingly satisfy their legal obligation of regulating investigator financial conflicts of interest through conflict-of-interest committees. This same method seems equally fitting to the oversight and management of conflicts originating from the financial interests of the institution. A significant difference, however, is that in the former case the institutional body oversees investigators (who are typically employees of the institution), whereas in the latter case the institutional body oversees the institution itself. This is not unlike the judiciary's role in supervising the actions of government. Judicial independence is fundamental to its ability to serve as watchdog, and the characteristics that secure such independence (eg, security of tenure for IRP members, removal for good cause only, documentation of removal and cause of removal for audit purposes, and immunity from retaliation) should be drawn on when structuring the IRP that will oversee conflicts emanating from the institution's financial interests in research.¹⁷

The IRP should have expertise in financial investments, the handling of intellectual property, bioethics, and the process of research involving human participants.¹⁷ It also should be empowered to review and monitor research in which the institution has one or more significant financial interests, and to recommend strategies for managing institutional financial conflicts of interest to the institution's board or to its IRB.¹⁷ The IRP could be a committee of the institution's board of directors, in recognition of the board's fiduciary duty to ensure integrity in all institutional operations, or could report to a board committee (eg, audit committee) while being composed of persons from the community who are independent from the board, but who have some moral affinity to the institution itself. The ideal member of such an IRP would be a community leader, not on the board of trustees, and not dependent, financially or otherwise, on the institution, but would have some financial and research expertise as well as sufficient loyalty to the institution to lead him or her to volunteer for this unique oversight role.

Our recommendations are designed to address institutional conflicts of interest in a real and meaningful way, without damaging the incentive structures that have fostered so many scientific advances. Financial and nonfinancial incentives spur innovation. Industry-academia relationships are permitted only when there is a legitimate justification for them. In particular, no legitimate justification exists when a relationship with industry serves only the pecuniary interests of the holder, without directly and materially furthering scientific advancement. In these circumstances, elimination of the financial relationship seems appropriate. When industry-academia relationships promise to advance science, management of any potential conflicts of interest via independent and expert committees is necessary.