Not What Anyone Signed up for: Unnecessary and Insurmountable Barriers Encountered in Conducting Clinical Trials in COVID-19

WBE's perspective

My 2020 started with an email entitled, "Happy New Year! Have You Read about this Virus that's Coming to Kill Us All?" I remember the last day my life was "normal." It was the end of February, 2020 and I was on the phone with my father, and he was asking me my thoughts about what we were seeing in the news about the novel coronavirus—then still regarded as a distant enemy in a faraway land. By way of background, I am an infectious diseases specialist with training in epidemiology, infection control, and implementation science. In the past, I was a hospital epidemiologist, in charge of local outbreak control, and more recently, I work in clinical research, focusing on weighing risks and benefits of different infection prevention and antimicrobial use strategies and on expanding infection prevention services to relatively uncovered aspects of the healthcare system. My father was calling to ask me my thoughts about the short- and long-term prospects for an epidemic in the United States, and what my thinking was on how this might impact the financial markets. In the middle of the call, I received word of the first "suspect case" in our hospital. I abruptly hung up the phone and did not look back. In fairly short order, cases began to peak, schools closed, my elementary school-aged children were suddenly house-bound, and the state was put under near-lockdown. I did not have another day "off" until sometime after Memorial Day.

PM's perspective

I remember exactly where I was on the morning of September 11, 2001, and also when I heard that the Challenger had exploded in 1986. I had no "COVID moment," but over the course of March my obsession with the pandemic grew exponentially until it peaked at the point of controlling everything I did other than activities of daily living (ADLs). I knew I would be low on the list of people to be called in for "risky" inpatient work, having one of those "preexisting conditions," but I had to do something; with apologies to the patients I take care of at the VA who really did risk their lives for something in the past, I had to enlist. The opportunity arose when Westyn, whom I knew only because we served together on the local institutional review board (IRB), asked for my input on the use of rheumatologic drugs, specifically hydroxychloroquine and tocilizumab, in an institutional "treatment guideline" she was developing to advise inpatient teams on what labs to follow and what medications to use, based on very limited anecdotal evidence, to help the droves of patients we expected to descend on our hospital. As it turned out, my experience with clinical trials, including the plethora of bureaucratic processes and acronyms and numbered forms, would come in handy. Strangely, one of the few specific times where I remember where I was is when Westyn called to tell me she was being pushed to do a clinical trial.

WBE and PM: Designing and Implementing a Clinical Trial

In March, faced with a deadly disease descending upon our city and our hospital, there was an urge to be able to offer our patients "something" beyond the under-appreciated supportive care. Clinicians widely acknowledged that no available drug had sufficient evidence to support indiscriminate use in a purely clinical setting. The question was whether to use medications off-label based on limited anecdotes or to conduct a clinical trial—the first of many ethical issues we have confronted in the half-year since then. There was a desire by many, both among research leaders and some clinicians, for a "clinical trial" banner, so that patients would be appropriately informed about the potential for a lack of benefit—and potential for harm—associated with almost any COVID-19 intervention. With these realities in the background, the two of us, both clinical researchers and members of our facility's IRB, received our marching orders: designing and implementing a clinical trial in time for the "first wave." Why were we chosen? Partially because of our position on the IRB—one of the providers assigning us this Herculean task specifically said, "We picked you because we think you are the only ones who will be able to get the paperwork approved."

Under normal circumstances, the process of designing, refining, and conducting a clinical trial would take months, at least. However, with COVID breathing down our necks, time was short. We reviewed the evidence available, limited to advice from our more experienced overseas colleagues—and then found ourselves constrained by which medications were available for purchase. We were lucky to have pre-existing relationships with the important institutional stakeholders—including the leaders in Research and Development, the clinical trials Coordinating Center, the IRB, and the pharmacy—to move the trial from conception to implementation. We designed a pragmatic, adaptive randomized controlled trial comparing the addition of IL-6R inhibition to standard care for hospitalized patients with a confirmed diagnosis of COVID-19. With members of the team working literally around the clock—emails and approvals were flying during the hours between 2 and 4 AM—we were able to advance from a 2-page summary "pitch" to IRB approval in 6 days and enrollment of the first patient 4 days later.

After securing a medication amidst supply chain barriers, which necessitated completely revising the study within the 6-day period of design, the regulatory and ethical challenge we noted immediately when moving the study to the real world was: how do we weigh the research requirement for documentation of informed consent against the need to keep staff safe, while also limiting use of personal protective equipment (PPE)? Problems with PPE shortages in the clinical setting are well-known, but the impact on the research service line is not. Although questions about whether a wet signature was required were under review, all agreed that the patient should receive a paper copy of the consent form. A member of the staff doing this strictly for research purposes would risk exposure and have to use PPE to simply hand the sick and potentially morbidly ill patient a 7-page stack of paper—at a time when providers were instructed to wear their masks "for as long as possible." We solved this first dilemma through collaboration with our clinical colleagues, who were working in the COVID units. They agreed to bring the consent forms to the patients' bedsides during morning rounds, so that we could avoid redundant exposures and use of PPE.

After the dust settled on the problem of delivering a paper copy of the informed consent form, we experienced another: what were the requirements for the informed consent processes? As a VA facility, we are required to abide by FDA regulations, but early in the pandemic, when we were at our peak caseload, these "regulations" (technically Guidances) were muddy at best. During an early online presentation, we were heartened to learn that the FDA would allow a remote consent process, with an impartial phone witness signing the consent form to affirm the patient's desire to participate—a process not dissimilar to what has led to the highly successful RECOVERY Trial in England. This appeared to solve our PPE and staff risk problem: we could conduct the entire process over the phone, not worry about "poison paper" (that was more of a concern then than it is now), and limit exposures. Unfortunately, only a few weeks later, the FDA issued an update clarifying that the witnessed consent process could only be used for clinical consents, not research consents (particularly in interventional trials), and we were forced to develop more complicated and cumbersome processes, all in the name of collecting proof of a wet signature on a page from the patient or legally authorized representative (LAR).

Which brings us to the third, and most troubling, major ethical issue we faced: therapeutic misconception. One of the tenets of conducting clinical trials is that there should be "clinical equipoise," and in line with that, patients should be deterred from expecting benefit from participation. Our experiences conducting a trial in COVID patients laid bare why these concepts are unrealistic, particularly for fatal diseases with no known effective treatments, such as COVID-19 in elderly patients. High mortality coupled with lack of evidence creates a perfect storm: patients and providers desperate to receive or prescribe "something," whether in the context of a clinical trial or not. No one was willing to accept the idea that they would just let patients die without trying to do something more than the "supportive care" recommended in early societal guidelines.

Research consent processes require that investigators emphasize to patients that they may not benefit from participation in a clinical trial. However, this concept is divorced from reality: except for comparative-effectiveness studies, the reality is that patients participate because they hope for benefit, and providers refer their patients because they hope that the trial will help their patients more than will standard care. We encountered this challenge first-hand during our own trial. Because our study included a standard care arm—meaning no active therapy—we witnessed psychological distress not only among patients and their relatives, but also among treating physicians as evident in immediate abandonment of the scientific principles that they had supported during design of the trial. When patients were randomized to standard care, physicians quickly revolted, and requested open-label use of the unproven study drug within 24 hours—even in stable patients. Honestly, we felt the same thing ourselves: as a clinician, and even as a researcher, it is very difficult to tell a patient that you have nothing to offer them, particularly when you feel responsible for that limited choice. Plus, we have been on the other side.

WBE

At age 33, I was diagnosed with early-stage, HER-2 positive breast cancer. Because of my "extremely" young age, and the aggressiveness of the tumor, there was a long discussion about how I should be treated: Standard chemotherapy regimens? Clinical trials? Something else? My oncologist and I considered enrolling in a clinical trial of chemotherapy plus Herceptin versus TDM-1, a newer, more expensive option that linked the chemotherapy agent to the Herceptin. The upshot of the trial for patients was that TDM-1 did not cause hair loss, and based on trials in other populations, there was hope that the medication would be more effective than the current standard of care. The prospect of avoiding hair loss is a big draw for most cancer patients, and especially young women: no one wants to look "sick." In this trial, patients, with high hopes of not losing their hair, and for the opportunity to receive a potentially more effective therapy, were randomized 3:1 to the novel treatment arm. Not surprisingly, despite being fully informed about the possibility of randomization to the standard of care arm, patients randomized to the taxol-herceptin arm were disappointed—they were hoping for a better drug, fewer infusions, and a tangible benefit (no hair loss) that they didn't receive.

Ultimately, I was conservative and opted for a standard (and more aggressive, although not the most aggressive) chemotherapy option, and lost all of my hair. Having been through that trauma and having to walk around that way for far longer than I ever would have imagined, many years later, I followed up on the trajectory of the trial, and learned that it did not meet its primary endpoint of improved safety in the TDM-1 arm, although interestingly, one of the important "clinically relevant toxicity" endpoints—hair loss—was not included as one of the measured adverse events.

PM

I spent the first two months of COVID in Boston with a WBC count averaging 3000 cells/uL, a lymphocyte count averaging 900, and transaminases high enough that I would have been kicked out of a clinical trial of any drug on the basis of hepatoxicity. The offending drug was probably lomustine/ CCNU, although procarbazine was also in play. The reason I had been taking them since October was that the low-grade glioma in my right motor cortex, initially dormant after treatment with temozolamide in 2013–14, was growing again. These treatments, along with the proton radiation therapy I received in 2019, are the only standards of care, and on average they are effective—temporarily. The median survival after diagnosis of a low-grade glioma with a favorable cytogenetic profile is 15 years, with a steady ongoing risk of transformation rather than some period of risk after which one is safe. At some point I am going to need a different treatment. It is of more than academic interest to me for new treatments to be studied, so that another proven approach is available when I need it. I would prefer that companies not be deterred from studying such treatments by the regulatory burden. If I need a treatment that is still in the investigational phase, I would prefer to not have a high risk of randomization to placebo, or to be excluded from the study based on having already received standard-of-care treatment—design elements often included because that is what is most likely to meet FDA requirements with the smallest number of patients.

WBE and PM: Even the Curse of the Billy Goat Will Be Broken

When it comes to serious diseases, patients enroll in clinical trials with the hope of benefit. Pretending otherwise and adding barriers to try to ensure this does not happen is a fool's errand. Use of placebo controls can avoid immediate psychological distress in addition to bolstering scientific validity for subjective outcomes, but the use of placebos in diseases in which standard care is inadequate is

Table 1.

Ethical issues that have arisen in the planning and conduct of our clinical trial.

also ethically problematic. Rather, we as a medical research community should recognize motivations behind treatment decisions and adjust how we conduct interventional research accordingly.

We will also add: cumbersome and changing recommendations, put in place by people who never have to interact with a patient or treating physician, has made the conduct of our trial nearly impossible. Everyone who could influence the trial from a distance slowed it down. Everyone whom we knew personally or was within one degree of separation was incredibly supportive and helpful. We are extremely grateful to our friends and collaborators on the IRB, R&D leadership, the coordinating center, the pharmacy, and at other VA sites in New England, who worked tirelessly and off-hours in order to bring an option—any option—to our patients. Without them—and without the underlying trust we had all established as members of a group—the study would never have opened.

In the end, in addition to being grateful, we are exhausted, over-saturated with information about SARS-CoV-2, and more than a little bit angry. But we are also motivated to effect change. Advocates of pragmatic trials and learning healthcare systems have been arguing that the need for change, on ethical as well as scientific grounds, is urgent—since at least 2015. Since that time, how many patients have we literally "protected to death"?

2020 started with an email about a new virus that was coming to kill us all, and then has featured thousands of emails related to our efforts to keep that from happening, which brings us to our final question: Can we start 2021 with a different email: "Happy New Year! We have new treatments and a vaccine!"? For decades, the best year in Cubs' history was "next year." And then they won.

Author's Note: This story was submitted in October 2020. Much has happened since then, including that the trial re-opened and is finished, and the US has widespread availability of effective vaccines.