ONCOLOGY Editor in Chief Howard S. Hochster, MD stresses the need for properly designed, controlled and randomized trials during the time of COVID-19
Never in the history of this country has the mainstream media spent more time speaking about clinical trials than in the past few months. A search for “clinical trials” in the New York Times cites 857 articles since the beginning of this year alone.
With the onset of the coronavirus disease 2019 (COVID-19) outbreak in the United States, President Donald J. Trump announced on March 19 that he “hears good things about hydroxychloroquine” and it could be a “game changer,” rather than concentrating on a federal strategy to deal with millions of infected individuals. And on April 4, he gave his considered recommendation of “What do you have to lose?” The FDA gave emergency approval in days, and the federal government purchased millions of doses, rather than needed personal protective equipment, based on biased anecdotal reports of a 100% cure rate at the Instituts Hospitalo-Universitaires in France (New York Times, May 5). At the same time, Anthony Fauci, MD, climbed out on a limb to suggest that the efficacy had not been proved and that appropriately controlled clinical trials were lacking. Placebo-controlled trials, such as one being conducted at Rutgers, are still pending results, looking at viral clearance on day 5 as the primary end point.
It was therefore with great disappointment that I read about the underwhelming results of the remdesivir SIMPLE trial, which sent Gilead stock soaring and the Dow Jones average bounding upward. In a randomized trial of 640 patients treated with 5 versus 10 days of the drug, the time to discharge was about the same (median, 10-11 days), although patients treated within 10 days of symptom onset seem to have been discharged sooner, and more than 60% of patients were discharged by day 14. However, without a control group, we really do not know the meaning of any of this. Fortunately, the National Cancer Institute followed up shortly with the interim analysis of the ACTT trial, which randomized 1063 patients to remdesivir or placebo. The results demonstrated a moderate effect of 31% shorter time to recovery (median, 11 vs 15 days, respectively) and lower mortality (8.0% vs 11.6%). These results have been issued by press release and have not been published with actual data for review.
At the time of this writing, we now have more than 2 million cases of proven COVID-19 infection in the United States and possibly 10 times that many individuals who have actually been infected. And we are rapidly approaching 115,000 deaths.
Despite these crushing numbers, we do not have a single appropriately conducted randomized clinical trial result in the public domain. Just 1 trial of approximately 3000 patients, randomized 2:1, to either recieve remdesivir or any appropriate drug versus placebo would give us hard information on the hazard ratios for viral clearance, duration of hospitalization, time to recovery, and which subgroups benefit the most, whether they are distinguished by age, ethnicity, or severity of illness. We can never develop these data without appropriate control groups, and therefore we are left to proceed in ignorance when tending to millions of those afflicted for the lack of having the moral courage to conduct a randomized controlled trial.
As physicians, we must bring the message to our patients that the control group in such a trial is not receiving inadequate or inferior treatment because they did not receive a drug having uncertain toxicity and efficacy. These several thousand volunteers are necessary to define the use of any treatment before it is unleashed on the public. It is only through this process that we will be able to treat the millions to come in the next year with this disease, and to render care with appropriate medical knowledge and certainty. Now, more than ever, we need properly controlled, randomized clinical trials to show us the way. We must educate and treat the clinical trial subjects as the heroes they are for their voluntary assistance in proving the benefit of these treatments.