In recent years, several research subjects have told us that they had bought or intended to buy stock in the companies sponsoring the clinical trials in which they were enrolled. This situation has led us to ask what, if any, are physician-investigators' scientific, ethical, and legal responsibilities concerning research subjects who choose to buy stock in the companies sponsoring the clinical trials in which they are participating. Although the scope of this problem is unknown and is likely to be small, this commentary examines the scientific, ethical, and legal concerns raised by such activities on the part of research subjects enrolled in early phase clinical trials. In addition, this commentary also outlines the basis for our opinion that research subjects involved in an early phase clinical trial should avoid the financial conflicts of interest created by trading stock in the company sponsoring the clinical trial.
In August 2002, page 1 of The Wall Street Journal(1) announced the discovery that biotechnology stock analysts were attempting to “peek inside” clinical trials to gain advance information about drugs in development. In at least one case, an analyst posed as a patient and registered for a clinical trial so as to have access to inside information about a new drug.
Research subjects and others, whom we will refer to collectively as “noninvestigator insiders,” often gain access to unpublished clinical information about new therapies when they gather information about participation in or enroll in early phase clinical trials. Access to such inside information could serve as the basis for buying and trading securities. In recent years, several research subjects and some family members have told us (via personal communications) that they had bought or intended to buy stock in the company sponsoring the clinical trial in which the research subject was enrolled. Moreover, clinical investigators from other institutions have occasionally recounted that their own research subjects have expressed the same intentions. More recently, and perhaps more ominously, evidence has surfaced (2) that healthy individuals have insinuated themselves into clinical trials to obtain access to private information about drugs that are under clinical trial. Although the number of research subjects involved in such financial activities has probably been small to date, these admissions by subjects have led us to consider the following question: What, if any, are physician-investigators' scientific, ethical, and legal responsibilities concerning research subjects who are enrolled in early clinical trials and who choose to buy stock in the companies sponsoring the clinical trials in which they are involved?
Although much has been written about physicians' financial conflicts of interest while engaged in research (3–22), no one (to the best of our knowledge) has commented on potential financial conflicts of interest faced by research subjects and other noninvestigator insiders. Guidelines for disclosing and managing potential conflicts of interest for physician-investigators involved in research are already in place, but no guidelines address such potential conflicts of interest for research subjects.
Clearly, many noninvestigator insiders may have or gain access to unpublished data from ongoing early clinical trials. Such individuals include members of the research team (e.g., study monitors, data managers, research nurses, statisticians, and others) but may also include individuals who legitimately seek information about an ongoing early clinical trial, such as a referring physician seeking a clinical trial opportunity for a patient. It is our view that these individuals, who are in a sense coinvestigators, occupy a position analogous to that of physician-investigators and should be bound by the same obligations not to use inside information acquired as a result of participation in research investigations.
Other noninvestigator insiders, as the above-described example of the stock analyst posing as a patient suggests, may attempt to acquire inside information about a particular drug or treatment through nefarious means. This phenomenon may be more common than has been assumed and has in fact received attention from the legal community (2). Take, for example, the following excerpt from a Yahoo.com message board concerned with the biotechnology company Onyx Pharmaceuticals (Richmond, CA):
“I have found that you can get quite a bit of information from the trial administrators if you persist. They are not accustomed to [Securities and Exchange Commission]-type confidentiality, they are more patient oriented. In one trial, I was able to find out how many patients were still in the study, [thereby obtaining] good mortality results. The source only worried about names, etc. We have about five commercial trial centers in Phoenix, besides hospitals that participate. They all leak information, one way or the other.” (Seehttp://finance.messages.yahoo.com/bbs?. mm=FN&board=7077128&tid=onxx&sid=7077128&action=m&mid=18425; last accessed February 20, 2004.)
The contemptible behavior of those individuals who seek to illegitimately gain inside information about early phase clinical trials, such as stock analysts posing as patients or as referring physicians, should be dealt with by the legal system and by the Securities and Exchange Commission.
Although similar in some ways to other noninvestigator insiders, an actual research subject—who without forethought acquires inside information about a particular drug or treatment through enrollment in an early phase clinical trial—is substantively different because of the unique position research subjects have in clinical trials. Hence, it is research subjects and their potential conflicts of interest—not the potential conflicts of interest of other noninvestigator insiders—that are the subject of our inquiry.
It is our opinion that research subjects involved in clinical trials, especially early phase clinical trials, should avoid the financial conflicts of interest created by trading stock of the companies sponsoring the clinical trials on the basis of private, unpublished information gained while enrolled therein. This commentary will examine the scientific, ethical, and legal concerns and their implications that support this opinion.
Scientific Concerns About Potential Conflicts of Interest for Research Subjects
Although financial conflicts of interest may arise for research subjects enrolled in any clinical trial, the dangers are especially great in early phase clinical trials that are not performed in a blinded or randomized manner, such as phase I trials, which are designed to determine the maximum tolerated dose and to describe the toxic effects of new agents. In an early phase clinical trial, the candid reporting of side effects is vital to the scientific goals of the trial, and when assessing toxicity, physician-investigators rely on research subjects' forthright reports of symptoms.
When research subjects participating in a clinical trial have a financial conflict of interest, they have an incentive to understate the toxicity of the agent being tested and to overstate the subjective benefits of investigational treatments. This situation raises two scientific concerns. First, research subjects may actually underreport subjective symptoms; second, physician-investigators may not trust research subjects' reports of symptoms if the investigators know of a financial conflict of interest. Real or perceived uncertainty surrounding the reporting of effects and side effects of new agents or treatments in early phase clinical trials may compromise the scientific goals of such trials.
The notion that research subjects may have the potential to interfere with scientific inquiry in this manner raises the question of whether research subjects are in some way responsible for the accuracy of the research investigations in which they participate. Clearly, it is well accepted that physician-investigators are responsible for the integrity of the scientific research of which they are a part, and it is therefore important for them to disclose or avoid any potential conflicts of interest that may compromise the scientific accuracy and integrity of the research. Nevertheless, unlike physician-investigators, research subjects do not share direct responsibility for the integrity of the scientific research. Kass (23) has interestingly described the position of research subjects as that of coinquirers distinct from the investigator, a term that more appropriately describes those individuals who are directly involved in the conduct of the research. Kass’s formulation conceptualizes research subjects as joining in the efforts to inquire into a scientific question, even though they are not invested in either the conduct or the outcome of the scientific research itself. Through their coinquiry, research subjects become an integral part of the research. In the context of clinical trials, the fundamental moral responsibilities that derive from their position as coinquirers require that research subjects should be held to basic moral responsibilities such as keeping promises and telling the truth.
In addition to bearing the responsibility to report symptoms truthfully, research subjects also have an interest in the accuracy of the research because false information has the potential to harm them and others. For example, research subjects could be harmed if substantial drug-related toxicity is underreported either by them or by other research subjects. Thus, financial conflicts of interest that compromise the accuracy of research or that create uncertainty regarding its accuracy should be avoided.
Ethical Concerns About Potential Conflicts of Interest for Research Subjects
The financial conflicts of interest for research subjects of early phase clinical trials raise two potential ethical concerns. First, such conflicts of interest may interfere with the process of informed consent; second, they may interfere with the patient–physician relationship.
Prior to enrolling research subjects in a clinical trial, physician-investigators have a duty to ensure that the subjects have given their informed consent. As part of the informed consent process in early phase clinical trials, physician-investigators often disclose to research subjects (and to their families and other individuals accompanying them) relevant data, such as toxic effects and responses to new agents or treatments, that were collected during the course of the trial to date (24). Such free and open disclosures are particularly important in early phase clinical trials, in which much information about a particular new treatment may have been learned during the course of the trial. If physician-investigators suspect that a research subject may be tempted to use disclosed information for financial gain, the investigators may be reluctant to disclose updated toxicity and response data, thereby compromising the quality of the informed consent. This hesitancy on the part of physician-investigators may be exacerbated if they suspect that such information might be shared with a larger circle of individuals who could then use this unfiltered and unpublished information to make financial decisions.
For example, such a situation arose with one of our patients who was being treated in a phase I clinical trial at the University of Chicago. This individual told us that he owned stock in the company sponsoring the clinical trial in which he was enrolled. During the clinical trial, we happened upon an Internet chat room dedicated to sharing opinions and information about stocks and investing, including the stock of the company sponsoring this particular phase I trial. We discovered that real, essentially accurate, and clearly nonpublic information about this research subject’s positive clinical response to an investigational agent was being described and discussed on this message board, often within days of the information becoming available to us. Because the clinical information about the research subject was presented anonymously, we initially suspected that the research subject himself was the source of the disclosed information and was attempting to bolster the price of the stock by presenting his own clinical response as evidence of the drug’s effectiveness. Hence, our suspicion raised concerns about the accuracy of the research subject’s previous reports of his improving symptoms and about the subject’s potential exposure to some undefined liability for so-called insider trading. Why we might have felt this way about such insider trading on the part of this research subject and whether we should have been concerned are two different questions that merit consideration.
The first question was why physician-investigators would be concerned about a research subject’s ownership of stock in a company that sponsors the clinical trial in which the subject is enrolled. Physician-investigators might justify such concerns on the grounds of protecting the research subject from potential harm, such as the legal consequences of insider trading. Their concerns might also be justified out of a desire to protect the integrity of the research or to guard against the premature dissemination of what might be considered nonpublic information.
The second question was whether physician-investigators should alter their communications with patients to forestall any misuse of nonpublic information. We believe that physician-investigators caring for research subjects should not alter their communications or withhold any information out of fear of a research subject’s financial motives. The informed consent process and the physician-investigator’s obligations to honesty are paramount. That said, however, physician-investigators conducting a clinical trial also have an obligation to take steps to preserve the confidentiality of nonpublic information. Determining how this objective should be achieved is a delicate matter. Requiring physician-investigators to monitor the conduct of their research subjects would add further burdens to the already complex role that physician-investigators play in conducting clinical trials. Such a requirement also raises the legitimate question of whether physician-investigators can be held responsible for the inappropriate dissemination of confidential information by a research subject. Nevertheless, physician-investigators are in a position to raise the issue of disclosure of conflicts of interest with the research subject, an act that at the very least would allow them an opportunity to express concerns about the flow and uses of information obtained during the clinical trial.
Regarding the research subject described above, we felt that the situation did in fact affect our communications and interactions with that subject until the source of this nonpublic information that was found on the Internet was identified as a contact of the research subject who learned of the research subject’s condition and progress from the research subject himself and who disseminated the information without the research subject’s knowledge. Earlier disclosure and discussion of the implications of using or revealing nonpublic information with the research subject might have prevented the tensions that arose from this incident.
One way to limit the disclosure of nonpublic information from early phase clinical trials is to omit the names of drugs and sponsoring companies from the informed consent process. We reject such a solution. Patients seeking experimental treatments for their disease usually receive most of the information about the proposed treatment from the treating team. However, information about most new drugs, even those in the earliest stages of testing, is now widely available on the Internet. Hence, to withhold the names of these new drugs and their manufacturers (or the sponsoring company) would be to improperly and unnecessarily limit research subjects' options of obtaining from other sources further information about the treatment they are considering.
The patient–Physician Relationship
In addition to the ethical concerns about the impact of a research subject’s financial conflicts of interest on informed consent, we also have reservations about what impact these financial conflicts of interest may have on the already complex relationship between the research subjects and the physician-investigator.
The foundation of the patient–physician relationship is the assumption that patients and physicians work together to achieve shared health goals. Thus, the pursuit of other goals, whether these be scientific knowledge or financial gain, may compromise that relationship. Much has been written (25–27) about the complexities of the relationship between physician-investigators and research subjects because physician-investigators may be perceived as acting as double agents—that is, physician-investigators simultaneously attempt to help the sick patient and pursue scientific knowledge. This obvious conflict of interest is the basis for policies requiring the mandatory disclosure of financial relationships for physician-investigators conducting clinical research. When physicians invest in companies sponsoring clinical trials in which they serve as investigators, their conflict of interest may compromise not only their relationship with their research subjects but also the integrity of the ongoing scientific research of which they are a part.
In contrast, almost nothing has been written about situations in which research subjects have dual goals (double agency) in their relationships with physicians. Such situations would include the example discussed above and those in which research subjects themselves have financial conflicts of interest relating to an investigational drug. As we have argued above, these types of conflicts of interest may affect both the research subject’s reports of side effects and, if the physician-investigator is aware of the conflict of interest, the physician-investigator’s communication with the research subject. Although “double agency” is a more important issue for physicians than for research subjects (given its higher occurrence in physicians and the power imbalance between physicians and patients), the pursuit of other goals by the physician-investigator may compromise fundamental parts of the patient–physician relationship. Moreover, the use of clinical trials as a means to other ends (e.g., financial gain) by research subjects also runs the risk of damaging the patient–physician relationship. For these reasons, and if financial conflicts of interest for patients turn out to be a substantial and widespread problem, we believe that it would be reasonable to ask research subjects to subscribe to the same conflict of interest rules that apply to physician-investigators—that is, to avoid financial conflicts of interest and to disclose any conflicts of interest they might already have. Thus, such measures would help protect the already complex relationship between physician-investigators and research subjects participating in early phase clinical trials.
As noted above, there are both scientific and ethical reasons that research subjects in early phase clinical trials should not invest financially in the companies sponsoring those trials. In addition, there is a close relationship between the ethical concerns outlined above and potential sources of legal liability. On the basis of current law, it seems possible, although not certain, that individuals participating in a clinical trial who gain nonpublic information and then make financial decisions based on that information would face liabilities under insider trading laws.
Disclosure of information about financial conflicts of interest is already required for physicians who participate in clinical trials as investigators because of the concern that their financial interest might influence their scientific judgment about the drug under study. Indeed, physician-investigators have been prosecuted for insider trading using information gained while conducting clinical trials (28). We are not aware of any analogous cases involving research subjects involved in financial conflicts of interest while enrolled in early phase clinical trials. Clearly, a research subject is likely to have much less depth and certainty of knowledge of the clinical developments within the clinical trial in which he or she is participating than a physician-investigator, so the risk of prosecution for insider trading for research subjects is likely to be smaller. However, given that the general prohibitions on insider trading cover all individuals who trade on the basis of material nonpublic information, the risk of prosecution is not negligible. This prohibition against insider trading could also apply both to research subjects of clinical trials who invest in the trial’s sponsoring companies and to their immediate families, friends, and referring physicians (i.e., all so-called tippees) who acquire this information from inside sources (29).
The recent revelations about insider trading in the ImClone case seem relevant to this discussion. In the ImClone case, several individuals with advance knowledge of the fate of the company’s anticancer drug sold their stock in the company. The nature of the information that research subjects acquire through participation in an early phase clinical trial may appear less certain than the certain knowledge acquired by senior executives at ImClone (who knew that their drug would not receive Food and Drug Administration approval). Even so, it is hard to see how research subjects, acting on the belief that the results from the clinical trial in which they gained nonpublic information as participants dimmed the prospects that a drug would receive Food and Drug Administration approval, could regard themselves immune from prosecution (30).
For many years, the SEC has followed rule 10b5-1 of the Securities and Exchange Act of 1934 (31–32), which holds that an individual who makes a sale or purchase of stock on the basis of material nonpublic information is subject to insider trading liability. Thus, the SEC rule requires awareness of nonpublic information on the part of the investor, in addition to materiality. The definition of materiality is complicated (33–36); in broad terms, it covers any information that would move a “prudent investor” to purchase or sell the traded stock. It seems clear that research subjects would, at times, have awareness of information that could rightly be considered nonpublic. The central question then becomes whether such information is considered material. An examination of the concept of the prudent investor just mentioned helps clarify the concept of materiality.
To establish whether a research subject involved in an early phase clinical trial has access to material information, we must first examine the nature of the information gained from within the clinical trial and whether this information is sufficient to move a prudent investor to buy or sell the stock of the sponsoring company.
What is the nature of the information a research subject could gain as a participant in an early phase clinical trial? For example, the research subject might learn about his or her own experience of the toxic effects of a particular drug or about its subjective or objective benefits. Indeed, having information on a major objective response (one’s own or that of another research subject) for an investigative drug would be important information about a new drug’s potential efficacy. Nevertheless, much of the information that could be acquired by a research subject during participation in a clinical trial would probably amount to little more than rumor. That is, such information would often consist of the reports of other patients (because patients commonly meet each other and discuss their experiences), the investigators' subjective reports of both responses and toxic effects, and various other bits and pieces of information gleaned from research nurses, pharmacists, and other members of the treating team.
Does the sum of such information acquired during the trial add up to enough material information to move a prudent investor to buy or sell the relevant stock? A prudent investor would understand the preliminary, uncertain, and incomplete nature of much of the information he or she possessed. However, in other cases, a research subject’s own objective major response to a novel cancer drug or the knowledge of others' confirmed responses to the same cancer drug might be enough to convince a prudent investor to rationally make trading decisions on the basis of this information. It is notable that most insider information gives the trader an edge rather than a certainty. Measured against this standard of information, and depending on the nature of the information, there is a serious risk that a research subject would be held liable for insider trading, with the consequent mix of civil and criminal penalties.
Who should be responsible for informing research subjects of early phase clinical trials of such potential liabilities? To give physician-investigators the task of warning research subjects about the risks they face in having financial conflicts of interest would be likely to further complicate their role in the research subject’s care. This new regulatory role might substantially change the nature of the physician-patient relationship. Adding such warning statements as part of the informed consent procedure might also change the nature and the goals of this process. In fact, we believe that any regulatory issues should be dealt with after enrollment so that the informed consent process is free from the negative influence that discussions of financial conflicts of interest might have. There may come a time when early disclosure of financial conflicts of interest is required; however, at present, when essentially nothing is known about the number of research subjects who might be involved in such conflicts of interest, it seems better to defer decisions regarding regulation of financial conflicts of interest among research subjects in early clinical trials. Instead, it may be best to keep a watchful eye on the scope of the problem to determine whether the number of research subjects involved in such conflicts of interest is sufficient to warrant the development of policies intended to manage these conflicts of interest. If these situations are found to be prevalent enough to warrant attention, documents that outline the scope of the risk and the rationale for the prohibitions on further trading of stocks of sponsoring companies could be presented to research subjects as part of the enrollment process for early phase clinical trials (distinct from the process of informed consent).
Research subjects participating in early phase clinical trials have potential conflicts of interest when they invest in the companies sponsoring those trials. In terms of science, such conflicts of interest may interfere with the honest and straightforward reporting of a treatment’s toxic effects and participant responses in ways that could compromise the primary scientific goals of early phase clinical trials. In terms of ethics, the process of informed consent and the patient–physician relationship may be negatively affected by concerns regarding the disclosure of what might be considered material nonpublic information or by the ways in which such information might be used. In terms of legality, research subjects who invest in the companies sponsoring the clinical trials in which they are enrolled conceivably face the legal consequences of insider trading.
On the basis of our analysis of this issue, we offer several suggestions about how to proceed in the management of the issue of financial conflicts of interest for research subjects enrolled in early phase clinical trials. First, the issue of financial conflicts of interest should be studied, and data about the extent of the problem should be collected. This could be accomplished by surveying a cross-section of subjects currently participating in phase I and II cancer trials at major centers across the nation about their current ownership of and intentions to buy stock in sponsoring companies. At the same time, such a survey could be sent to physician-investigators involved in early phase cancer clinical trials so as to assess their knowledge of subjects who might have such financial conflicts of interest. Although conceptually interesting, such financial conflicts of interest for research subjects may be so rare that no corrective action need be taken. If it is discovered that financial conflicts of interest for research subjects of early phase clinical trials are a widespread problem, then we believe that it would be helpful to have research subjects screened (by the sponsoring companies) for financial conflicts of interest after enrollment in a trial and to require subjects to disclose their holdings and to cease trading the stock of the sponsoring company until material information from the study becomes public. The exact forms of these agreements, as well as the full range of procedures and policies appropriate for dealing with financial conflicts of interest among noninvestigator insiders, are unclear and should not be pursued until the scope of the problem is better established. If procedures for dealing with these financial conflicts of interest should turn out to be necessary, these might include asking research subjects if they intend to invest in the company that is sponsoring the trial in which they are enrolled, informing them about the potential legal risks of insider trading, and encouraging research subjects to recognize that such financial conflicts of interest may compromise the accuracy and quality of scientific information.