Live Blogging: Post-Trial Responsibilities Conference, Session 4

By Rachel Sachs

Panel Discussion: Christine Grady, Mitchell Warren, Richard Saver, Luann van Campen; Moderator: Mark Barnes

Session four of the Post-Trial Responsibilities conference featured a panel discussion of a complex hypothetical scenario. The hypothetical used in this session (available here) was made complex by design, to avoid the easy answers and to force readers to make difficult choices regarding both people and institutions.

In this hypothetical, the Ministry of Health in Angola begins a collaboration with a Belgian biotech company, BelgiqueTec. BelgiqueTec is not an established multinational pharmaceutical company, but is instead more like a start-up that’s been funded by some investment capital but which lacks deep pockets. BelgiqueTec has the rights to a diabetes drug and is interested in testing it in clinical trials. The Angolan government contributes $25 million to finance trials of the drug in Angola, due to the population’s burgeoning diabetes problem and concomitant development of wait times for treatment in the government’s diabetes clinics. In return, the Angolan government will take some ownership stake in the company and therefore in the outcome. There is an in-kind commitment of resources from Angola in terms of the study sites, which will be hospitals or clinics owned by the Angolan government or ministry of health (MOH). The agreement between BelgiqueTec and the government establishes that Phase I and II trials will be done in Angola, and any Phase III trials would feature a site in Angola (as well as elsewhere).

In the hypothetical, the Phase I trial was completed and showed no adverse effects. Now, a Phase II study has enrolled 200 treatment-naïve Angolan citizens with moderate to severe diabetes. Half showed improvement in the control of their diabetes, but three patients showed decreased cardiac function associated in time with drug administration, suggesting a potential causal link between administration of the drug and the observed cardiac side effect.

The first set of key questions for discussion was as follows: if drug development is halted at Phase II based on the adverse events, what are the obligations of BelgiqueTec and the MOH to continue to deliver diabetes treatments of any kind to the 200 Phase II subjects? Since a number of the patients showed improvement on the experimental drug, should the treatment be the experimental drug or the standard of care? For how long must the treatment continue?

There was general agreement that if the company had decided based on adverse events not to proceed in developing the drug, then there is no obligation to provide the experimental therapy. Indeed, the therapy might not even be available for use. But there may well be an obligation to refer the participants in the trial to existing diabetes treatments. Additionally, the government’s involvement here by investing in BelgiqueTec might raise the bar for its obligations to its citizens who have participated in the trial. Yet as the hypothetical set out, the government clinic has an extensive waiting list, and it’s not apparent that the 200 participants should jump the queue in the clinic over those who did not participate in the trial, particularly if the patients on the waiting list were not eligible or invited to the trial. The idea of reciprocity has limits where the result is to burden others in the community.

There might also be an obligation to continue monitoring those three individuals who reported adverse events. If the view is that BelgiqueTec and Angola feel a fiduciary responsibility to the participants, they may have an obligation not to abandon the patients at the end of the study. The limits of this obligation, though, are unclear. For instance, the panel then puzzled over the question of what kind of treatment should be provided to the 200 and for how long. Should they receive not just the standard of care, but also laboratory testing and hospitalization? To put it starkly, even if there is general agreement that these 200 should receive the standard treatment (the same as the others in the clinic), should that treatment continue for the rest of their lives?

The panel raised two key points in response to this issue. First, we might look at what the informed consent documents say, and what commitments were made by the MOH and BelgiqueTec about what would happen at the end of the trial. You likely have an ethical responsibility to honor your commitments. Most likely, the documents contemplate a transition of care, eased by the sponsor. But it’s an open question whether (and if so, when) you can ethically get out of post-trial obligations by saying up front that there will be no post-trial access.

Second, the disconnect between the end of treatment relationships in the doctor-patient context and in the research context was noted. Specifically, in the United States, there are very few healthcare obligations when a provider decides to end a treatment relationship – if there’s no emergency, the provider can generally end the relationship and transfer the patient. Or with discharge planning, you refer the patient to the next appropriate intervention. So what’s so different about research? The risk taken by the patients with uncertainty of benefit is important, from a reciprocity standpoint, and so continued access for them might be required. One panelist pointed out that the research enterprise almost creates a social contract. But there must be limits.

Again on the informed consent point, the panel noted that there has been a lot of work in assessing how people understand the informed consent process – but very little assessing their expectations about post-trial access. It might be that we should be trying to better understand people’s expectations of access as part of consent, and even that we have an obligation to be realistic about what expectations are. In one study, very few people thought only trial participants should have access to efficacious therapies – they thought everyone who needed it should get it.

Another concern raised by the hypothetical was whether the 200 subjects, assuming they received some sort of diabetes treatment post-trial, would be permitted to substitute in a sicker family member. The panelists had previously observed this phenomenon in the context of early antiretroviral studies in various resource-poor settings. In some cases, patients even impersonated others in an effort to gain treatment. Reactions on this subject were mixed, with the panelists noting the difficulty of continuing to monitor safety and efficacy under a protocol if patients were permitted to trade slots, but also recognizing that the realities on the ground and difficulty of preventing all attempts at substitution shouldn’t be allowed to significantly impede the progress of research.

Next, the panel was asked about BelgiqueTec’s status as an early-stage company. If BelgiqueTec does have an obligation to provide continuing care to the 200 patients and is required to pay for that treatment, then it will not be able to test two other compounds it has in development for diabetes. It will simply take its remaining capital and buy an annuity for the benefit of the 200. There’s a lost opportunity cost here, which might not be as true in the case of a large company, which troubled several panelists.

Although several panelists were sympathetic to this concern and did not want to chill future research, they were concerned about the difficulties of crafting a rule which carves out such institutions. Startups are required to comply with many legal requirements even though they might chill their ability to innovate. And trying to differentiate might create an incentive for large pharmaceutical companies to have startups conduct the trials and acquire them later. More fundamentally, we might care less about the institution’s financial capacity than its mission – to develop and manufacture drugs, or to practice medicine.

The panelists then were asked about the following scenario: what, if, aside from the three patients who suffered adverse effects, the other 197 demanded continued access to the experimental treatment. Do they have an ethical claim to it or should the company say, even though 100 of you are demonstrably better and 97 of you think you are, we’ll only give you the local standard of care?

Assuming that this issue wasn’t part of the initial informed consent discussion, the panel suggested that they have an ethical reason to demand that they get some kind of good diabetes care, but not the experimental therapy. There’s a reason we have a regulatory process to find when something is safe and effective. In Phase II, you can’t make that claim even though some patients think they did well. And when there are other options which are safe and effective, it isn’t ethical to keep them on the investigational drug.

The panelists then returned to the issue of informed consent and the importance of clear planning. There was consensus that an agreement between BelgiqueTec and the MOH and pre-trial consent by the patients about what would happen at the end of the Phase II trial would have been ideal. Although it’s difficult or even impossible to spell out all possible scenarios in advance, articulating a few common scenarios or even using phrases like “safe and effective” up front which can be debated later would be preferable. It might be most helpful not just to plan decisions, but also decisionmakers and legal processes, in advance, if there is no feasible way to define terms ex ante such that everyone will agree.

This brought back the issue of research literacy and the need to understand patients’ expectations and to help them manage them. There is in general a sense among patients that if they’re in a trial and the drug works, they’ll immediately get a product. But regulatory barriers, manufacturing challenges, and other delays might stand in their way. People should be made more aware of such delays. Too, many safety issues are discovered only after commercialization, and drugs may even be withdrawn. The panelists did express concerns about the potential need to add post-trial access information to the already-complex informed consent process, but they also noted that in some highly publicized United States court cases about post-trial access, courts have looked to informed consent documents to decide whether the patients had a reasonable expectation they would be able to access the drug.

The panel closed by echoing several of the themes that had been raised throughout the day and elucidated in the context of the case study: the difference between fiduciary relationships in the treatment relationship and in the research context, the importance and simultaneous difficulty of planning for post-trial access in advance (both substantively and procedurally), the challenge of finding the boundaries that limit the ethical obligation for post-trial access, the need to share responsibility across institutions, the need for better understanding of the ethical underpinnings of such responsibilities, and the difficulty of both understanding and complying with the Declaration of Helsinki.