Pharmaceutical research experts are raising concerns that attempts to offer desperate coronavirus patients some form of treatment may hobble crucial studies of which drugs actually work and which ones don’t.
More than 100 off-the-shelf and experimental therapies are being tested either formally or informally for the coronavirus disease, including hydroxychloroquine, the antimalarial touted as a treatment by U.S. President Donald Trump, and remdesivir, an experimental drug from biotechnology company Gilead Sciences Inc. There are so many even coronavirus experts can barely keep track.
The huge field of potential treatments represents a mobilization of medical science that has few historical parallels. But with tens of thousands of patients in the U.S. hospitalized and no proven drugs to treat them, the crisis has exacerbated an always-present tension in the parallel worlds of caring for patients and looking for new therapies: that what might seem best for an individual patient can undercut a research effort that will benefit the greater public good.
“We need to be true to the science and make sure that the proper trials are done so that we can really understand what works and what doesn’t,” Rochelle Walensky, chief of the infectious-diseases division at Massachusetts General Hospital, said on a call with reporters Friday. She said randomized clinical trials, the scientific gold standard for studying drugs, are essential to make sure doctors understand proper treatments.
Over the long history of medicine, the majority of experimental drugs tried in people either didn’t work or caused side effects that outweighed their benefits. Researchers try to identify the rare few that can help in a sea of useless or harmful chemicals. Human trials provide a clearer picture of how safe and effective a medicine is, but they take time.
“If we just said throw everything at everybody, it would be a catastrophe of uninterpretable results,” said Arthur Caplan, director of medical ethics at New York University Grossman School of Medicine.
In 1747, one of the first successful medical studies, conducted by James Lind, compared several regimens to show eating oranges and lemons could help cure scurvy, a deficiency of vitamin C that can aggravate wounds and lead to bleeding gums, in British sailors.It took two more centuries for the current system of controlled clinical trials to catch on. In the early 1900s, doctors tested anti-infective medications by giving every other patient an experimental therapy. But some doctors found subtle ways to cheat on the patient allocation, undermining the results, according to a 2016 essay by Harvard Medical School researchers in the New England Journal of Medicine.
That gave rise to the current system of blindly and randomly allocating some patients to a dummy pill and some to an experimental treatment. Neither doctor nor patient know who got what, making it impossible to game the results.
Rigorous methods developed over a long time can seem cumbersome in a public-health emergency. Previous outbreaks have tested the willingness of physicians and patients to wait for a clinical process. But past epidemics also show why it pays to be cautious about early signs of efficacy for drugs that haven’t run the full scope of a clinical trial.
When the world’s worst Ebola outbreak hit West Africa in 2014, doctors tried a range of experimental therapies. Numerous unproven remedies were given to patients outside of any formal study, in part because many researchers felt it was unethical to give patients a placebo in a deadly epidemic.
It took five years and a new outbreak for scientists to get a clearer handle on what worked. The outcome was a surprise: Results of a major trial showed last year that ZMapp, once considered the most promising anti-Ebola agent, fell short.
“Everyone thought that ZMapp was the definitive therapy for Ebola,” said Aneesh Mehta, an infectious-disease doctor at Emory University in Atlanta, Georgia. “It really wasn’t until we did a comparative clinical trial that we saw there were other agents out there that might have more benefit.”
As desperate doctors rush to find something to help patients with COVID-19, the disease caused by the coronavirus, some researchers fear that mistake is recurring. They warn uncontrolled testing could delay the hunt for a cure.
“I am worried that we are repeating what happened with Ebola,” says Andre Kalil, an infectious-disease doctor at the University of Nebraska Medical Center. “A lot of people are being given unproven and unsafe drugs without any controls.”
If doctors keep using drugs without carefully testing, “the COVID-19 outbreak is going to finish and we are going to know absolutely nothing about which drugs harmed and which drugs helped,” he said.
Since most patients recover from coronavirus on their own, without organized studies doctors could be fooled into thinking that a bad drug works. Similarly, they could miss a drug that is effective because they didn’t try the right dose in the right patients.
“The more panic we get the more it is important to hold on to the handrails and try to stay somewhat organized,” said Caplan, the NYU ethicist. “Otherwise we get nothing.”
The rush to use hydroxychloroquine before the evidence is in illustrates the problem. Much of the public’s excitement over it stems from a tiny French study that was shared on Twitter by Trump and others. U.S. researchers say it is full of methodological flaws. Among other weaknesses, it didn’t look at whether the drug can keep patients out of the ICU.
Hydroxychloroquine, which is also used to treat rheumatoid arthritis and lupus, has been shown to inhibit coronaviruses in a test tube. But a close chemical relative, chloroquine, has been tested in human trials against viral diseases including influenza, dengue fever and chikungunya. It has repeatedly failed.
Against viruses, “the track record of this drug is really poor” said Nebraska’s Kalil. Hydroxychloroquine may trigger heart arrhythmias and even liver problems, which would be dangerous for the elderly patients with cardiovascular problems who tend to be most sickened by COVID-19.
That’s why careful clinical trials are needed. It may seem cruel to give patients a placebo in a raging pandemic, said Mark Denison, a coronavirus researcher at the Vanderbilt University School of Medicine, but controlled trials the best way to find out what works while making sure experimental drugs aren’t causing side effects that make the disease worse.
“What is cruel is if you use an untested modality that theoretically gives people hope but either may not work or causes harm,” says Denison.
The Trump administration’s championing of hydroxychloroquine has put pressure on doctors to prescribe it. The drug is being used in some New York City hospitals to treat COVID-19 patients, and some health care workers are taking it to try to ward off infection. NYU told health-care staff they are eligible for a preventive prescription if they had been exposed to the new coronavirus, according to documents reviewed by Bloomberg News.
Miriam Merad, an immunologist and director of the Precision Immunology Institute at the Icahn School of Medicine at Mount Sinai, is studying COVID-19 patients’ inflammatory response to virus. She has seen informal, off-label use of drugs including hydroxychloroquine to treat the disease in the Upper Manhattan hospital.
Merad said she doesn’t blame the desperate physicians trying to treat sick patients flooding resource-strapped emergency rooms across the city.
“It’s not clear it’s going to work. We must remain scientifically driven, following guidelines without harming patients,” she said. “But we don’t need people, the president, to interfere, causing anxiety among those who feel they’re not able to get the drug.”
Several U.S. trials of hydroxychloroquine have recently begun. Some of the largest are focused on whether the drug helps prevent the disease in people who have been exposed to Covid-19, while others are exploring whether it will alleviate symptoms of the disease. Preliminary data could be available as soon as early May.
Antiviral researchers say a clear answer on whether Gilead’s remdesivir helps patients with coronavirus is more likely in the near term.
In February, weeks before the coronavirus pandemic exploded in the U.S., the National Institutes of Health began a trial comparing remdesivir with a placebo in 440 hospitalized patients. Half get remdesivir and half get a placebo. The trial has enrolled patients rapidly as the outbreak widened, and results could come by the end of this month. (Among the researchers involved in that study are Emory’s Mehta and Nebraska’s Kalil.)
The NIH trial employs a clever design that will allow other drugs to be added to the trial over time. If remdesivir works, the placebo could be dropped and newer drugs compared with remdesivir, limiting the number of people who have receive a placebo.
There is a long way to go; before this year, remdesivir has never been tested in humans for a coronavirus disease. In the 2019 Ebola study that included ZMapp, there were four drugs, including the two that beat ZMapp that showed sharply lower death rates.
The other drug in the trial that failed to lower the death rate was remdesivir.