Treatment-Emergent Cilgavimab Resistance Was Uncommon in Vaccinated Omicron BA.4/5 Outpatients

Abstract

Mutations in the SARS-CoV-2 Spike glycoprotein can affect monoclonal antibody efficacy. Recent findings report the occurrence of resistant mutations in immunocompromised patients after tixagevimab/cilgavimab treatment. More recently, the Food and Drug Agency revoked the authorization for tixagevimab/cilgavimab, while this monoclonal antibody cocktail is currently recommended by the European Medical Agency. We retrospectively reviewed 22 immunocompetent patients at high risk for disease progression who received intramuscular tixagevimab/cilgavimab as early COVID-19 treatment and presented a prolonged high viral load. Complete SARS-CoV-2 genome sequences were obtained for a deep investigation of mutation frequencies in Spike protein before and during treatment. At seven days, only one patient showed evidence of treatment-emergent cilgavimab resistance. Quasispecies analysis revealed two different deletions on the Spike protein (S:del138-144 or S:del141-145) in combination with the resistance S:K444N mutation. The structural and dynamic impact of the two quasispecies was characterized by using molecular dynamics simulations, showing the conservation of the principal functional movements in the mutated systems and their capabilities to alter the structure and dynamics of the RBD, responsible for the interaction with the ACE2 human receptor. Our study underlines the importance of prompting an early virological investigation to prevent drug resistance or clinical failures in immunocompetent patients.

Keywords: SARS-CoV-2; Spike; cilgavimab; deletion; mAbs; mutations; quasispecies; tixagevimab.

Figure 1

Cycle threshold values (a) and Spike-specific immune response (b) for all patients from day 0 to day 7 of tixagevimab/cilgavimab treatment. **** p ≤ 0.0001.

Figure 2

(A) Per-residue RMSF of SARS-CoV-2 WT_T0 (BA.4) S trimer. Spike monomers 1, 2, and 3 are colored black, red, and green, respectively. Monomer 2 is in the up conformation (see Materials and Methods), ready for interaction with the ACE2 human receptor. Dotted lines indicate the NTD and RBD regions. The dashed line at residue 681 highlights the S1/S2 boundary. Note that, to allow the comparison among different S variants, the residue numbers are reported as WT (Wuhan) equivalents, not considering the deletion and insertion of residues in the mutated systems. The region of residues 138–145, where the two deletions in the quasispecies systems are located, is highlighted with a horizontal line; (B) as in panel (A) for the NTD domain.

Figure 3

Cartoon representation of the representative frame (341.5 ns of MD simulation) for the WT_T0 (BA.4) S trimer. Spike monomers 1, 2, and 3 are colored black, red, and green, respectively. Residues 138–145, where the two deletions in the quasispecies systems are located, are highlighted in blue in the insert for monomer 1, which showed the highest RMSF values.

Figure 4

Projections of the essential dynamic movement along eigenvector 1 for the five simulated systems. Ten frames are represented, colored from red to blue. The WT_T0 system is reported twice to facilitate the comparison.

Figure 5

RMSF of the S-filtered trajectory along the first eigenvector. (A) WT_T0/QS1/QS3 system; (B) the differences between the filtered RMSF along the first eigenvector of WT_T0 and QS1/QS3. Residues above the threshold of |± 0.5 nm| (dashed lines) are indicated. Panels (C,D) as for (A,B) for WT_T0/QS2/QS4 systems.