Following mixed results in the clinic against COVID-19, antibody engineers are further optimizing SARS-CoV-2 antibodies with the aim of improving outcomes.
The US government has bought almost a million doses of Eli Lilly’s SARS-CoV-2 neutralizing antibody, with the latest purchase announced 2 December. Lilly’s monoclonal antibody (mAb) drug, bamlanivimab, was given an Emergency Use Authorization (EUA) on 9 November by the US Food and Drug Administration for use in high-risk COVID-19 patients with mild or moderate disease. Regeneron’s two-antibody cocktail (casirivimab and imdevimab) also received an EUA for the treatment of mild-to-moderate COVID-19. Though both drugs represent new options for treating the coronavirus infection, neither is ideal. On the basis of a phase 2 trial report, the overall reduction in viral load between the bamlanivimab and placebo arms was small. The phase 2/3 data on Regeneron’s cocktail have yet to be reported in a peer-reviewed venue, but the company claims a similar small reduction in viral load compared with that in placebo-treated patients, although patients with high viral load did better and there were fewer medical visits than in the control arm. In the meantime, several other companies are applying antibody-engineering strategies to enhance antiviral potency, improve mAb safety profile, provide greater convenience of administration and reduce cost.
The use of antibodies as antivirals is gaining momentum. In the 20 years after Synagis (palivizumab) was approved for use against respiratory syncytial virus (RSV), no other mAbs targeting viral pathogens were registered. In 2018, Roche/Genentech received a green light for Trogarzo (ibalizumab) for the treatment of multidrug-resistant HIV; and early last year, the Food and Drug Administration approved Regeneron’s Inmazeb antibody cocktail treatment for Ebola. But the pandemic has now put at least a dozen anti-SARS-CoV-2 mAbs into the clinic (Table 1).
For patients with early or mild disease, mAbs for COVID-19 face several limitations. They need to be given intravenously, and outpatient infusions are costly and difficult. Also, the limited supply forces institutions to make decisions about which patients are likely to benefit. “There’s always a concern about healthcare equity,” says Adam Lauring, an infectious disease doctor at the University of Michigan. mAbs drugs are expensive. Lilly, for example, is charging the government $1,250 a dose, not including the infusion cost.
So far, the clinical results from these mAbs have been less than stellar in moderate COVID-19 and even worse in severe disease. The US National Institutes of Health stopped treating hospitalized patients with Lilly’s bamlanivimab in October, on the basis of phase 3 data showing no improvement in clinical outcomes. The same month, Regeneron paused enrollment of sicker patients to one of its hospital trials “based on a potential safety signal and an unfavorable risk/benefit profile,” according to the trial’s independent data monitoring committee.
But even with vaccines rolling out, finding effective antibody therapies for passive immunization continues to matter. “They’re extremely complementary approaches — we need them both,” says Erica Ollmann Saphire, a biologist at the La Jolla Institute for Immunology. Mass vaccination will take time and vaccines won’t work in everyone, she points out, and many people can’t be vaccinated with certain vaccines because they’re elderly and immunocompromised.
The Lilly and Regeneron mAbs are standard engineered formats: they are basically wild type, unmodified IgG1 mAbs that target the SARS-CoV-2 spike (S) protein. They were cloned from B cells taken from patients convalescing from COVID-19 or, in the case of one Regeneron antibody, generated in humanized mice. “Everyone just had to make a choice and launch their therapeutic, to get something where it needed to be, rather than wait for the perfect thing,” says Saphire, who directs the Coronavirus Immunotherapy Consortium, which is evaluating over 200 SARS-CoV-2 antibodies.
Many next-generation mAbs to the new coronavirus are heavily engineered, specifically in the antibody Fc region. The Fc binds to receptors on immune cells and elicits a broad array of Fc effector functions during viral infections, including macrophage phagocytosis and natural killer (NK) cell–mediated cytotoxicity. Certain Fc mutations extend antibody half-life and increase lung bioavailability, something clearly desirable in respiratory infections. But mutating the Fc can also alter antibody binding to NK cells or macrophages. And, because no one yet knows the clinical benefits and risks of doing so in COVID-19, there remains a lack of consensus as to the best approach. Different companies are placing different bets.
Some are choosing to knock out the antibody’s effector cell functions, by introducing the so-called LALA double mutation into the Fc. By interfering with antibody binding to Fc receptors on NK cells and macrophages, such mutations can mitigate activation of these effector cells, which mediate inflammation. “They cause cytokine release, they cause direct attack on the infected tissue, they cause inflammation,” says Jake Glanville, CEO of Centivax, which has an anti-SARS-CoV-2 mAb in preclinical development. For a newly infected patient, that may be okay, he says. “You can put up with a little inflammation in exchange for better viral clearance. But that’s a terrible idea to give to someone who has their lungs totally colonized.” Centivax therefore knocked out effector function in its lead antibody. Lilly’s second antibody, etesevimab, does this too, as do Sorrento Therapeutics’ antibodies.
Another safety concern is antibody-dependent enhancement (ADE), whereby the antibody potentiates virus uptake by a macrophage, enabling entry and replication, increasing viral load and worsening disease. But evidence is mounting that the ADE risk may be small. “SARS-CoV-2 doesn’t naturally target these cells [macrophages],” says University of North Carolina molecular virologist Tim Sheahan. “It’s not known technically if these cells are even able to support the complete life cycle of virus replication.” Hundreds of thousands of COVID-19 patients have received convalescent plasma therapy — containing a wide variety of antibodies — without ADE. Enhanced disease has also not been reported in human COVID-19 vaccine trials.
Nevertheless, an antibody with intact effector function could lead to an exaggerated immune response. This is a real threat to patients with advanced COVID-19. “In the very severe patients hospitalized on respirators, it probably is a concern,” says Vir Biotechnology CEO George Scangos. “In newly hospitalized patients, I think the jury is still out.”
To treat non-intubated patients more effectively, Vir selects its antibodies to recruit immune effector cells and mobilize them against the infection. “In all our preclinical models, effector function matters,” Scangos says. Neutralizing antibodies fight infection by blocking viral entry into cells. But if these antibodies also mobilize effector cells — triggering antibody-dependent cellular cytotoxicity (by NK cells) and antibody-dependent cellular phagocytosis (by macrophages) — they also indirectly kill cells already infected by the virus. “Two ways to eliminate the infection,” says Scangos. Vir expects to report phase 3 data for its lead fully human mAb, VIR-7831, by the end of January.
Vir is engineering additional effector function into its next antibody, making three Fc-region amino acid changes first described by Rockefeller University immunologist Jeff Ravetch. These mutations tighten binding to stimulatory Fc receptors on immune cells while reducing binding to inhibitory receptors. “The result is a dramatic increase in potency, in the short term,” says Scangos. In animal models, Vir found that the engineered antibody boosts not just NK cell and macrophage cell killing, but also the T-cell response, because the antibody Fc engages Fc receptors on antigen-presenting dendritic cells, says Scangos. A clinical trial is planned for early 2021.
But there’s limited published evidence that anti-COVID-19 antibodies require effector function to be fully protective. A mAb therapy might not need to kill infected cells with effectors: just blocking viral entry might be enough. On the other hand, mopping up infected cells could prove necessary because not all antibodies that strongly neutralize viruses in culture are protective in vivo. Sheahan’s group recently reported that, in a mouse model, immune effector function does, for some antibodies, help protect against the virus. “There is a difference in efficacy, yes,” says Davide Robbiani, director of the Institute for Research in Biomedicine in Bellinzona, Switzerland, and a coauthor on the paper. But “regarding efficacy and safety of antibody effector functions, it’s early to be conclusive.”
Besides effector function, another big divide in the field is cocktails. In addition to Regeneron, Lilly is also testing an antibody cocktail, as is AstraZeneca, which is developing two antibodies discovered at Vanderbilt University. An antibody pair can bind the virus spike protein at two distinct epitopes to overcome drug resistance if one mutates. “The virus may evolve; the virus may find ways to escape the treatment with monoclonal antibodies, especially if a single monoclonal antibody is being used,” says Robbiani.
But Vir is testing single-antibody therapy, in part because its antibody binds to a viral epitope that rarely mutates, says Scangos. It will be “more difficult for the virus to escape this antibody than a cocktail of two antibodies that Regeneron or Lilly or others are bringing forward,” he says. “Two isn’t better than one, necessarily. Quality matters.”
Centivax is also going with a single antibody. Glanville points out that resistance mutations to the anti-RSV mAb, Synagis, are rare. For SARS-CoV-2, as based on an analysis of viral sequences in public databases, “I think the amount of escape variants is going to be pretty low, and in the low cases that happens, there are these other antibodies people can take,” says Glanville. “My goal is to make a medicine that’s less expensive, [so] more people can take it.”
Furthermore, at lower doses, intravenous infusions could be replaced by intramuscular (IM) or subcutaneous injections. Regeneron is also teaming up with gene therapy pioneer Jim Wilson of the Perelman School of Medicine at the University of Pennsylvania to develop an AAV-based intranasal therapy that will express its COVID-19 antibodies in nasal epithelial cells. “Infusions are just not practical for mass release,” says Glanville, who is working on injectables. Scangos agrees. “The key is going to be an IM or subcu injection,” he says. “You need a lower dose to do that; you can’t possibly do an IM injection with the 2-mg dose.” Regeneron’s infusion dose is 2.4 grams — 1.2 gram per antibody — while Vir is dosing at only 500 milligrams. “We’re on our way to getting an IM formulation,” says Scangos. Single antibodies would also cost less. “The right choice is to make an inexpensive monoclonal that will be mass used,” says Glanville. Whether the newer antibodies can accomplish that goal is a question for 2021.
About this article
Cite this article
Garber, K. Hunt for improved monoclonals against coronavirus gathers pace. Nat Biotechnol 39, 9–12 (2021). https://doi.org/10.1038/s41587-020-00791-6