Science’s COVID-19 reporting is supported by the Pulitzer Center and the Heising-Simons Foundation.
A second company has now produced strong hints that monoclonal antibodies, synthetically produced versions of proteins made by the immune system, can work as treatments in people who are infected with the pandemic coronavirus but are not yet seriously ill.
The biotech Regeneron Pharmaceuticals has developed a cocktail of two monoclonal antibodies that attach to the surface protein of that coronavirus, SARS-CoV-2, and attempt to block it from infecting cells. Yesterday at an investor and media webcast, the firm revealed early results.
The company showed slides with detailed data from 275 infected people in a placebo-controlled trial that ultimately plans to enroll 2100 individuals who are asymptomatic or, at worst, moderately ill. The analysis divides patients into two groups: those who had detectable antibodies against SARS-CoV-2 at the trial’s start and those who did not, a so-called seronegative group. The monoclonal cocktail showed little effect on people who already had antibodies against the virus. But it appeared to help the seronegative patients, powerfully reducing the amount of virus found in nasopharyngeal swabs and alleviating symptoms more quickly. “These are provocative results,” says Myron Cohen of the University of North Carolina, Chapel Hill, who was not involved with the study but is helping Regeneron test its monoclonal cocktail as a preventive.
Cohen notes that Regeneron’s data look similar to those in a press release from Eli Lilly 2 weeks ago about early results from a trial of its single monoclonal antibody against SARS-CoV-2. “Both of these reports go in the same direction,” Cohen says. But he cautions that neither has been published, both trials are ongoing, and more data are needed to understand how—or whether—these experimental medicines can best help patients. Lilly, oddly, did not see an impact at the highest dose of antibody tested, and Regeneron saw no difference between its low- and high-dose preparations used in the study.
James Crowe, a viroimmunologist at Vanderbilt University who is working with AstraZeneca to develop COVID-19 monoclonal antibodies, welcomed Regeneron’s detailed preliminary results. “I applaud Regeneron for releasing so much information,” Crowe says. “They’re contributing to public health by releasing this as soon as possible.” But he notes even people who did well on the monoclonal cocktail still had low levels of virus detectable after their treatment, which in theory could cause problems. “I was surprised that there was any virus at all given that these are such potent antibodies,” he says, adding that the residual virus detected in the swab tests may not be capable of copying itself.
The monoclonal antibodies from the two companies are clones of potent SARS-CoV-2 antibodies that can “neutralize” the virus in test tube studies. Researchers plucked the genes for these antibodies from humans who recovered from COVID-19 or from mice artificially infected with the virus. The companies then put the genes in Chinese hamster ovary cells to bulk manufacture the antibodies, which were given to the COVID-19 patients as infusions.
At the webcast that announced Regeneron’s results, George Yancopoulos, president and co-founder of the company, emphasized how the “target population” for the monoclonal cocktail are SARS-CoV-2 infected people who have “not yet mounted their own immune response” and have exceedingly high levels of the virus. “What we really want to do is turn them into patients who have already started to effectively fight the virus,” Yancopoulos said.
In the Regeneron data, the most dramatic drops in SARS-CoV-2 were seen in seronegative patients who had the highest levels of virus at the trial’s start. In comparison with patients who received the placebo, the results were clearly statistically significant.
Daniel Skovronsky, Lilly’s chief scientific officer, says the Regeneron data are “quite confirmatory” of their own. “I don’t expect there to be large differences between good neutralizing antibodies,” Skovronsky says. “Antibodies will work best in people who can’t clear the virus on their own.” One key difference between the two studies, he says, is that Lilly enrolled fewer seronegative people and still found an impact—although the company, in contrast to Regeneron, is withholding details until it publishes results. Lilly also stressed that people receiving its antibody were shown to have fewer hospitalizations or emergency room visits: five out of 302 (1.7%) treated patients versus nine out of 150 (6%) in the placebo group. “Yes, these are small numbers by some measures,” Skovronsky says, “but by other measures, there are significant differences in hospitalization.”
Regeneron hasn’t yet accumulated enough data to show the same protection. Its trial had only 12 patients who had “COVID-19–related medically attended visits.” Although there was a trend toward more of these in the placebo group than treatment arms, only one was hospitalized.
Regeneron’s data raise difficult questions about when to use its cocktail. People who test positive for SARS-CoV-2 aren’t routinely screened for antibodies to it or for levels of the virus. “If the decision is going to be made to deploy such a therapeutic solution in the patients who might benefit the most and need it most, we’re going to have to solve the problem of using the right point-of-care diagnostic tools, either for serology or high viral load,” Yancopoulos said, noting that their partners—including Roche—are developing these types of assays.
Skovronsky says Lilly has a simpler plan: Offer monoclonals to people who test positive for the virus if they are in high-risk groups for developing severe disease, which include the elderly and people with underlying diseases such as diabetes or who are overweight. Running extra tests before treating people, as Regeneron suggests, “is just not going to meet the needs of the population,” he says
Both Lilly and Regeneron say they are discussing their data with regulators to see whether their monoclonal antibodies might warrant moving to widespread use more quickly through mechanisms like the U.S. Food and Drug Administration’s emergency use authorization process. Additional studies of their monoclonal treatments are underway in hospitalized COVID-19 patients and, separately, as preventives in uninfected people.
Monoclonal antibodies are more difficult to make than many drugs and often are extremely expensive, which means that demand could outstrip supply and many countries might not be able to afford them. The U.S. government’s Operation Warp Speed has invested $450 million in Regeneron to produce up to 300,000 “doses” of its cocktail by the end of the year, which would be distributed to Americans free of charge. “A substantial fraction of those are already available,” Yancopoulos said—although it’s not yet clear what constitutes a single dose of the company’s cocktail. Nonetheless, Regeneron, which is partnering with Roche to increase production capability, says it hopes to ramp up to produce 250,000 doses per month.
Skovronsky says if the lowest dose Lilly is testing works, it could have up to 1 million doses by the end of the year. Lilly is partnering with Amgen to scale up production to “several million doses” next year. “We’re rooting for Regeneron’s success, just as Regeneron is rooting for Lilly’s success,” he says. “None of us can make enough antibodies to meet the need.”