The first oral antiviral for treating COVID-19, Merck & Co.’s molnupiravir, received approval from the U.K. Medicines and Healthcare products Regulatory Agency on 4 November. But the approval, for people at high risk of severe disease, comes as a prominent virologist has suggested using molnupiravir could do far more harm than good, potentially unleashing new, deadlier variants of SARS-CoV-2. Other virologists say the concern is worth tracking but is largely hypothetical, for now. “I don’t think we are in the position of withholding a lifesaving drug for a risk that may or may not happen,” says Aris Katzourakis, a viral evolution expert at the University of Oxford. Molnupiravir, which Merck and Ridgeback Biotherapeutics developed from an earlier, experimental antiviral, works by interfering with viral replication, littering the viral genome with mutations until the virus can no longer reproduce. Last month, Merck and Ridgeback officials announced results of a clinical trial that found giving the drug to COVID-19 patients early in the disease reduced their risk of hospitalization and death by 50%. The drug’s ability to mutate RNA has raised persistent fears that it could induce mutations in a patient’s own genetic material, possibly causing cancer or birth defects; studies so far have not borne out those fears. Now, William Haseltine, a virologist formerly at Harvard University known for his work on HIV and the human genome project, suggests that by inducing viral mutations, molnupiravir could spur the rise of new viral variants more dangerous than today’s. “You are putting a drug into circulation that is a potent mutagen at a time when we are deeply concerned about new variants,” says Haseltine, who outlined his concern Monday in a Forbes blog post. “I can’t imagine doing anything more dangerous.” He notes that patients who are prescribed antibiotics and other drugs often don’t complete a prescribed medication course, a practice that can allow resistant germs to survive and spread. If COVID-19 patients feel better after a couple of days and stop taking molnupiravir, Haseltine worries viral mutants will survive and possibly spread to others. “If I were trying to create a new and more dangerous virus in humans, I would feed a subclinical dose [of molnupiravir] to people infected,” Haseltine says. “The possibility [of generating variants] is there,” agrees Raymond Schinazi, an infectious disease expert at Emory University. But neither he nor anyone else contacted by ScienceInsider voiced as much concern as Haseltine. Katzourakis says, “I don’t share the alarm in this. If you force an organism to mutate more, it’s more likely to be bad for the virus.” Underpinning Haseltine’s worry are studies that show coronaviruses can survive with molnupiravir-induced mutations. Two years ago, for example, Mark Denison, a virologist at Vanderbilt University, and colleagues repeatedly exposed coronaviruses to sublethal doses of a form of the drug called EIDD-1931 to test whether drug-resistant viruses would emerge. They reported that in populations of two coronaviruses—murine hepatitis virus and the virus that causes Middle East respiratory syndrome—30 rounds of such drug treatment caused up to 162 different mutations that did not kill the viruses. But Denison notes that his study didn’t catalog mutations in individual viruses; rather, up to 162 mutations arose in populations of cells infected with one of the two coronaviruses. Most of the mutations harmed the virus, slowing growth. “If I take away anything from our work, it is that if the virus tries to adapt, say through resistance [to molnupiravir], it continuously develops deleterious mutations,” Denison says. However, Ravindra Gupta, a microbiologist at the University of Cambridge, cautions that mutated viruses may have better odds of flourishing in the people most likely to take molnupiravir: patients with compromised immune systems. Because vaccines are less effective at protecting those patients, he says, “These are precisely the people who are most likely to receive [molnupiravir].” Daria Hazuda, who heads infectious disease discovery for Merck, notes that the company hasn’t seen any evidence of people who take molnupiravir generating viruses with new and dangerous mutations. In patients who completed the 5-day course of the drug, Hazuda says, “we don’t see any infectious virus”—let alone mutated variants. The mutations that arise along the way have been random, she says—not concentrated in a particular gene that would make the virus more likely to survive. “There is no evidence for any selective bias,” she says. What’s more, Hazuda and others note, SARS-CoV-2 is plenty good at churning out variants naturally as it replicates in infected people. “There is no shortage of viral variation out there,” Katzourakis says. The more important question is whether molnupiravir provides selective pressure that drives the virus toward transmissibility or virulence. “I find it difficult to imagine,” he says. “But I can’t rule that out.” More likely, Denison and others say, is that use of molnupiravir will drive the emergence of virus that is no more deadly or transmissible but is resistant to the drug, a common outcome for anti-infectious agents. But the 5 November news that another antiviral drug, from Pfizer, is highly effective against SARS-CoV-2 suggests a way to forestall resistance: using both pills in combination, the same multiprong strategy used to treat HIV and other infections. On 30 November, a Food and Drug Administration advisory committee will review possible emergency use authorization for molnupiravir in the United States.