21st century pandemic, prehistoric clinical trials

Guest commentary: it’s long past time to embrace randomized controlled trials as the fastest path to answers in a pandemic

BioCentury is providing this content for free given the urgent need for information about the coronavirus crisis. Further analysis can be found in our COVID-19 Resource Center. For more, sign up for our daily email.

The U.S. is trying to fight a 21st century pandemic with prehistoric clinical trials -- and achieving the kind of efficiency you might expect.

In the last six months, tens of thousand of patients have been enrolled in thousands of clinical trials all over the world, yet these studies have produced actionable answers on only three agents: remdesivir, dexamethasone and hydroxychloroquine.

While it is possible that only remdesivir and dexamethasone, of the hundreds of agents evaluated, are effective, it is more likely that the poor quality of the studies has prevented identification of other useful therapies.

Delays in finding treatment options for COVID-19, especially in view of the resources and number of patients who have received some type of experimental treatment, are unacceptable. These delays represent a collective failure of investigators, funders, regulators and IRBs.

It is not too late to change course and devote our collective resources to testing the most promising therapies in scientifically valid trials.

To get the answers that are desperately needed, we can’t continue to let misplaced hope stand in the way of good science.

It is time to embrace what we learned during the Ebola crisis -- that randomized controlled trials are the fastest path to answers during an epidemic -- and adopt the most promising study designs for getting to definitive answers quickly. Ideally, those designs should be carried out in the pragmatic setting, meaning embedded in routine care to include diverse patients reflective of the real world.

Rushing to repurpose

As is common in public health emergencies, many therapies already approved for other indications have been hastily repurposed as potential treatments for COVID-19 despite scant or conflicting evidence of benefit. Clinicians have again forgotten that even repurposed approved drugs require careful evaluation, and that efficacy in treating another disease isn’t evidence of effectiveness for patients with COVID-19.

In this instance, the most notable examples of the rush to repurpose have involved antiretroviral drugs for HIV, hydroxychloroquine, and convalescent plasma. Though the latter isn’t FDA approved, the therapeutic approach has been tried for many infectious diseases over the last two centuries and, for the most part, yielded disappointing results.

Physicians everywhere, including from well-respected academic medical centers with strong clinical research records, began administering these therapies off-label or through poorly designed clinical studies that had no prospect of generating interpretable data or ensuring patient safety.

In the case of hydroxychloroquine, thousands of patients were exposed to an ineffective and toxic drug for months before data from randomized controlled clinical trials demonstrated its futility.

The story of convalescent plasma is also disappointing. Since March, more than 20,000 COVID-19 patients in the U.S. alone have been treated with convalescent plasma through an FDA expanded access program, yet we still don’t know if it confers benefit, does nothing or is harmful.

Data on the first 5,000 patients treated showed a mortality of 14.9% at 7 days, but because there was no control arm, these data are not interpretable. For reference, the mortality in the placebo arm in the NIH-led study evaluating remdesivir was 11.9% at 14 days.

Despite the lack of data, demand for convalescent plasma has been so strong that it has created shortages of source material for the manufacture of hyperimmune globulin, a more potent and potentially more promising form of passive immunotherapy.

Repeating failure

Until recently, there was a collective belief that conducting clinical trials during emergencies was not feasible. As a result, the most promising investigational products were made available to anyone who needed them outside of a clinical trial and little was learned about benefits or harms before the emergency was over.

This was the case during the 2009 swine flu pandemic. At the time, peramivir, a promising intravenous therapy in clinical development with good activity against the flu virus, was made available to hospitalized patients. Clinical trials were conducted only after the emergency was over. It took years to find out that the drug was ineffective in the treatment of complicated flu.

The situation changed during the 2014 Ebola epidemic in West Africa, when NIH and industry partners led the effort to conduct randomized clinical trials for treatments and vaccines during one of the most complex public health crises of our time.

A new paradigm was established in Ebola - yet this paradigm seems to have been largely forgotten.

In a pandemic, the fastest, most efficient, and definitive way to know what works and what doesn’t is through properly designed clinical trials. The time it takes to show a drug is effective is predicated on how fast patients are enrolled -- which should not be difficult in a pandemic -- and how truly effective the drug is.

In other words, if we have a blockbuster, we will know in a matter of a few weeks.

If clinicians had heeded the advice of Thomas Chalmers, an American physician who pioneered randomized clinical trials, to randomize the first patient, we would certainly know by now whether convalescent plasma and other repurposed therapies have a place in the COVID-19 therapeutic armamentarium.

Six months into this pandemic, the current state of affairs should finally put to rest the idea that valuable data can be generated from uncontrolled studies or off-label use.

Getting into gear

Some trials outside the U.S. have been extremely innovative.

The Randomised Evaluation of COVid-19 thERapY (RECOVERY) trial, led by the University of Oxford, involves almost all acute hospitals across the whole of the U.K. RECOVERY has established the lack of benefit from hydroxychloroquine and the benefit from dexamethasone. It may be our best hope to establish whether there is a role for convalescent plasma in the treatment of COVID-19.

The World Health Organization is leading the SOLIDARITY trial, an adaptive randomized multinational trial involving more than 100 countries.

In the U.S., gradually but surely the clinical trial apparatus is being mobilized to conduct appropriately designed and powered trials. The Adaptive COVID-19 Treatment Trial (ACTT), led by NIH’s National Institutes of Allergy and Infectious Diseases (NIAID), already generated the data on remdesivir. Recently the NIH launched the Accelerating COVID-19 Therapeutic Interventions and Vaccines (ACTIV), a suite of platform trials that leverage existing federally funded clinical research networks.

There are also several well-designed, industry-led trials, including from Regeneron Pharmaceuticals Inc. (NASDAQ:REGN), BioCryst Pharmaceuticals Inc. (NASDAQ:BCRX), Eli Lilly and Co. (NYSE:LLY), and Alexion Pharmaceuticals Inc. (NASDAQ:ALXN), but these are testing one drug per trial, which is not the most efficient path during a public health emergency.

In contrast to the RECOVERY trial, which embeds a pragmatic design to select, enroll, and randomize patients within the routine delivery of care, most of the U.S.-based trials, whether led by NIH, academic or industry groups, rely on existing networks where clinical research runs in parallel with clinical care. This limits their size and speed, as well as the ability to generalize the data to the real-world setting.

One exception is the randomized, embedded, multi-factorial, adaptive platform (REMAP) trial, led by the University of Pittsburgh Medical Center, which is able to evaluate a number of treatments simultaneously and efficiently. The clinical research, from enrollment to treatment allocation, is conducted within the electronic medical record of patients.

Unlike the U.K., the U.S. lacks a common, coordinated health care system to facilitate the conduct of such a massive pragmatic randomized clinical trial. But NIH is in a position to establish similar capabilities in major medical systems across the U.S., and it should do so quickly.

The major providers of electronic health records, such as a Cerner and Epic, must be integral to these efforts for them to succeed.

Such capability would finally bring patients the cures they need for this and future pandemics.

Luciana Borio is Vice President at non-profit investment firm In-Q-Tel. She previously served as director for Medical and Biodefense Preparedness Policy on the National Security Council and was acting chief scientist at FDA.
Signed commentaries do not necessarily reflect the views of BioCentury.

Trial Subscription

Get a two-week free trial subscription to BioCentury


Article Purchase

This article may not be distributed to non-subscribers