The antibodies given to Trump. REGN10933 binds at the top of the RBD, extensively overlapping the binding site for ACE2, while the epitope for REGN10987 is located on the side of the RBD, away from the REGN10933 epitope, and has little to no overlap with the ACE2 binding site. Proof of principle was shown in in a cell model, using vesicular stomatitis virus pseudoparticles expressing the SARS-CoV-2 spike protein. Simultaneous treatment with REGN10933 and REGN10987 precluded the appearance of escape mutants (Baum 2020, Hansen 2020). Thus, this cocktail called REGN-COV2 did not rapidly select for mutants, presumably because escape would require the unlikely occurrence of simultaneous viral mutation at two distinct genetic sites, so as to ablate binding and neutralization by both antibodies in the cocktail.

  • The first clinical data on REGN-COV2 (REGN10933 + REGN10987) were published online on September 29 (not peer reviewed). Regeneron called it “a descriptive analysis on the first ~275 patients”, derived from a broad ongoing clinical development program. Adult, non-hospitalized COVID-19 patients with symptom onset ≤ 7 days from randomization were randomized to receive single doses of REGN-COV2 at 2.4 g or 8 g IV or placebo. Before treatment, serology was used to divide patients into positive (n = 123) versus negative (n = 113). As expected, “viral load” in nasopharyngeal (NP) swabs was higher in seronegative patients (7.18 versus 3.49 log10 copies/mL). Main results showed a modest viral load reduction mainly in seronegative patients and a lack of a numerical dose-response relationship: REGN-COV2 appeared to reduce viral load through day 7 mainly in seronegative patients: the mean NP viral load reduction was -1.98 (high dose) and -1.89 log10 copies/mL (low dose), compared to -1.38 with placebo (difference versus placebo -0.56 for both dosage groups, p = 0.02). If all patients were included (including seropositives), the reduction was -1.92 and -1.64 log10 copies/mL, compared to 1.41 with placebo (significance only seen with high dose). Patients with higher baseline viral levels had correspondingly greater reductions in viral load. Median time to symptom alleviation for the overall population (median) was 8, 6 and 9 days for high, low dose and placebo, respectively (seronegative only: 8, 6 and 13). As for medical visits, there was a numerical reduction versus placebo, but with just 12 visits in total there was no way of discerning the relevance. Most non-hospitalized patients recovered well at home. Both doses were well-tolerated. Infusion reactions and severe adverse events were balanced across all groups, no deaths occurred.

Did this save Trump’s life? There is no doubt that larger data are needed in patients with more severe disease. Let’s see what happens. Half a log viral load reduction is not impressive although it may be clinically relevant. If approved, Regeneron will distribute REGN-COV2 in the US and Roche will be responsible for distribution outside the US.

Find the entire treatment chapter at



AcalabrutinibAnticomplement therapiesAzithromycinCamostatChloroquineColchicineConvalescent plasmaCorticosteroidsCytokine blockersFamotidineFavipiravirG-CSFHuman recombinant soluble ACE2HydroxychloroquineIbrutinibIloprostInterferonsJAK inhibitorsLeflunomideLopinavirMonoclonal antibodiesN-acetylcysteineOseltamivir(other) Protease inhibitors(other) RdRp inhibitorsREGN-COV2Umifenovir

Outlook | References

By Christian Hoffmann