XAV-19 a Glyco-Humanized polyclonal antibody targeting SARS-CoV-2 accelerates the recovery of mild to moderate COVID-19 patients and keeps its neutralizing activity against Omicron and its subvariants

Study design

EUROXAV (NCT04928430; EudraCT: 2020-005979-12) was a multicenter, international phase II/III, double-blind, placebo-controlled, randomized clinical trial conducted in 14 hospitals from 5 countries (Bulgaria, Greece, Romania, Spain, Turkey) between March 2021 and October 2022. This trial followed the International Council for Harmonization E6 guideline for good clinical practice and the principles of the Declaration of Helsinki. An independent data safety monitoring board (DSMB) examined the data for safety on a regular basis during the trial.

Participants

Adult patients must have SARS-CoV-2-confirmed infection (RT-PCR, RT-qPCR or antigen test in the last 10 days) and present signs of respiratory disease with at least 2 clinical symptoms related to COVID-19 started less than 10 days prior to screening visit. Patients should have SpO2 > 90% (at ambient air) and require or not low-flow oxygen therapy (score 2, 3 or 4 on WHO clinical progression 7-point ordinal scale (15)). Exclusion criteria were a positive SARS-CoV-2 test > 10 days, multiorgan failure, immediate ICU hospitalization, critical respiratory illness, requirement of oxygenation at flow rate > 6L/min, or signs of severe systemic illness (respiratory rate ≥ 30 per min, heart rate ≥ 125 per min, PaO2/FiO2 < 300), participation of another trial, pregnancy or breastfeeding. Eligible patients were randomized in a 1:1 ratio to receive either XAV-19 (150mg) diluted in sterile NaCl 0.9% or placebo (NaCl 0.9% only) as a 1-hour intravenous perfusion. Randomization was stratified by center and by WHO score (WHO score = 2 or 3 versus WHO score = 4). Patients were monitored during the infusion and the following hour. On the investigator judgement, the patient was either hospitalized or discharged (ambulatory patients). Dexamethasone, antithrombotic prophylaxis, antibiotics and antiviral medication (except anti-SARS-CoV-2 mAb or convalescent plasma) were allowed according to the judgement of the investigator and to the national guidelines or standard of care.

Sample size calculation

Sample size calculation initially assumed 15% progression of COVID-19 in the placebo arm versus 8% in the XAV-19 arm and a two-sided 5% type I error rate. A sample of 650 patients would provide the trial with 80% power. Considering a drop-out rate of 25%, 870 patients were planned.

Primary endpoint

The primary endpoint was the proportion of patients with an aggravation of COVID-19 within 8 days. The aggravation was defined as a worsening of the score of at least 1 point on the WHO 7-point ordinal scale compared to the score at randomization.

Secondary endpoints

Secondary endpoints were proportion of patients with an aggravation of COVID-19 within 8 days after treatment (defined as a worsening of the score of at least 2 points on the WHO clinical progression 7-point ordinal scale compared to the score at randomization); proportion of patients with an aggravation/improvement of COVID-19 between day 8 and day 15 (defined as a worsening/improvement of the score of at least 1 point on the WHO 7-point ordinal scale); time to aggravation/improvement measured by the proportion of patients with an aggravation/improvement at each measurement days (3, 5, 8, 15) after randomization. Safety outcomes included the cumulative incidence of any Adverse Event (AE). A stratified log rank-test will be used to compare treatment effect between groups, after adjustment on the stratification variables at baseline.

Statistical analyses

The global alpha type I error with bilateral hypotheses for this study was fixed at 5%. Binary logistic regression model was performed for the primary efficacy analysis. Association between the randomization group and the progression of COVID-19 was adjusted on centre, WHO score, country, age, body mass index (BMI), gender and comorbidities (defined as BMI>30, diabetes DNID, diabetes DID, cardiac disorder, vascular disorder, hypercholesterolemia, renal failure, lung disease COPD and/or asthma). For efficacy analyses, the strategy “Missing=Failure” is performed for patients who withdraw before D8 or for patients with missing data. Under this strategy, patients without an available assessment at D8 will be considered as having an aggravation. Replacement of missing data is done in the ITT and TARGET populations. The comparisons between XAV-19 group and placebo group of two or more qualitative variables was made using the χ2 test, the continuity-corrected χ2 test or Fisher exact test, according to the expected values under the assumption of independence. Comparisons of quantitative variables between XAV-19 and placebo groups were made using a Student t-test (parametric test comparing means) or Mann-Whitney-Wilcoxon test (non-parametric test comparing ranks) depending on the distribution of the variable of interest. A log rank test was used to compare time to aggravation/improvement between randomization groups. Kaplan-Meier curves are presented by randomization groups and by WHO score at baseline.

Spike/ACE-2 neutralization assay

Recombinant SARS-CoV-2 RBD (1µg/ml in carbonate/bicarbonate buffer pH 9.0) were immobilized overnight at 4°C on Maxisorp microtiter plates. After washes and saturation with PBS-Tween 0.05%-BSA2%, successive dilutions of XAV-19 or tixagevimab/cilgavimab were added for 30 min, followed by ACE-2-mFc Tag ligand (Sino Biological; final concentration 125 ng/ml). After 1h incubation and washing, the ACE-2 mFc Tag was revealed using a peroxidase-conjugated anti-mouse secondary antibody. Binding intensity was revealed by addition of TMB. Reaction was stopped with 50 μl of 1M H2S04 and optical density was taken at 450 nm.

Cytopathogenic Effect (CPE)

Vero cells (CCL-81) and Vero E6 cells (CRL-1586) were obtained from ATCC. SARS-CoV-2 clinical isolates were isolated from SARS-CoV-2 RT-PCR positive patients and viral stocks were generated by one passage on Vero cells. Titration of viral stock was performed on Vero E6 by limiting dilution assay allowing calculation of tissue culture infective dose 50% (TCID50). Serial dilutions of XAV-19 (50μl) were incubated with 50μl of virus (2 x 10E3 TCID50/ml) in a 96-well plate at 37°C for 60min in 8 replicates. Hundred μl of Vero E6 cell (3 x 10E5 cells/ml) were then added to the mixture and incubated at 37°C, 5% CO2 until microscopy examination on day 4 to assess CPE. For viral load (VL) quantification, RNA extraction of the 8 pooled replicates of each XAV-19 dilution was performed with NucliSENS EasyMag (BioMerieux) according to manufacturer’s protocol. The relative VL were assessed from cycle threshold values for ORF1ab gene obtained by the TaqPath™ COVID-19 RT-PCR (ThermoFisher, Waltham, USA) and by linear regression in log10 copies/ml with a standard curve realized from a SARS-CoV-2 positive nasopharyngeal sample quantified by Droplet-Digital PCR (Bio-Rad).

Demographic and patient characteristics

EUROXAV was initiated in 2021 and expected to enrol 870 patients. Yet, due to the very low enrolment rate, the study was unlikely to be completed in a reasonable timeline and was suspended on October 31, 2022. An interim blind analysis was performed in November 2022, and the results were shared with the Independent DSMB who recommended to stop the study. At that time, 293 patients with confirmed COVID-19 were screened. Eligible patients were randomized to receive XAV-19 (N=139) or placebo (N=140) (Figure 1). Final analyses were performed on target population defined as ITT patients having received at least one dose of treatment and fulfilling the main inclusion criteria, i.e. aged 18 or over, weighing between 40 and 120kg at the time of signing the informed consent, requiring or not low-flow oxygen therapy (≤6L/min) with WHO score at 2, 3 or 4, (129 and 130 patients in the XAV-19 and the placebo arms respectively). The demographic and clinical characteristic at baseline were similar between groups. The median age was 56.4 years and 54.5% were male patients. Percentage of patients with a WHO score of 2, 3 and 4 was 25.6, 45.7 and 28.7 in the XAV-19 arm and 25.4, 46.2, 28.5 in the placebo arm (Table 1). While comorbidities were slightly higher in the XAV-19 arm (62% in XAV-19 group vs 49 in placebo group), both groups were comparable for age, gender, BMI, and WHO score.

• Download figure
• Open in new tab

Figure 1. Clinical Trial Design

Primary efficacy endpoint

The primary endpoint was the proportion of patients with an aggravation of COVID-19 within 8 days after treatment. No differences on aggravation according to the treatment group was observed (12 vs 9 patients in the XAV19 and the placebo arm respectively, OR: 1.379, 95% CI 0.560 to 3.394, p=0.4831) (Table 2). Bivariate analyses showed that age at screening or comorbidities were associated with COVID-19 progression, while gender, BMI, or WHO score were not (Table 2). Logistic regression confirmed the association between age (over 70) and aggravation of COVID-19 (p=0.001, OR = 6.675 [2.158; 20.645]) (not shown).

Secondary endpoints

There was no difference in the proportion of patients with COVID-19 aggravation within 8 days of treatment in the XAV-19 arm vs the placebo arm. Proportion of patients with COVID aggravation at day 15 was similar between groups. Proportion of patients with COVID-19 aggravation/improvement at day 15 vs day 8 was also similar between XAV-19 and placebo arms (not shown). Although proportions of aggravation/improvement were not different between groups, improvement did occur earlier in the XAV-19 arm (p=0.0340) (Figure 2A). This benefit was even more significant for non-hospitalized patients (WHO score of 2 at day 1) or patients not requiring supplemental oxygen (WHO score of 2 or 3 at day 1) (p=0.0003 and p=0.0159) while not significant for hospitalized patients (WHO score of 3 or 4) or hospitalized patients requiring supplemental oxygen (WHO score of 4) (p=0.7511 and p=0.8059). Furthermore, median time to improvement was reduced in the XAV-19 arm compared to the placebo one (4 days vs 14 days and 7 days vs 14 days for patients with a WHO score at baseline of 2 and 2/3 respectively) (Figure 2B). By contrast, time to aggravation was not different between XAV-19 and placebo groups (not shown).

• Download figure
• Open in new tab

Figure 2. Time to clinical improvement

Kaplan-Meier curves by randomization group on whole TARGET population (A) or on TARGET population with a score of 2, 2 to 3, 3 to 4 or 4 at baseline (B). Improvement was defined as a diminution of at least 1 point on the WHO score compared to the WHO score at day 1. Dotted lines represent the median time to improvement.

Safety

A total of 83 adverse event (AE) (45 in the XAV-19 arm and 38 in the placebo arm) were observed from 29 patients in each group (Table 3). The percentage of mild/moderate AE in the XAV-19 arm was 91% (vs 76.3% in the placebo arm). Twenty-two severe AE (SAE) were reported in the XAV-19 group compared to 13 in the placebo group. None of them were considered as related by the investigators. Interestingly, SAE >3 account for 86% of SAE in the XAV-19 arm vs 92% in the placebo arms. Number of fatal SAE was equivalent between the two groups with 5 deaths in each group, from respiratory failure, multi-organ failure or disease progression. No anaphylaxis or hypersensitivity reactions and no infusion-related events were reported in any patients receiving XAV-19.

To further explore XAV-19 clinical potential, the neutralizing activity of XAV-19 against Omicron and its subvariants was determined.

XAV-19 antiviral efficacy against Omicron and its subvariants

Binding of XAV-19 to Omicron, BA.2, BA2.12.1, BA.4/5 and BQ1.1 RBD appeared slightly lower than binding to the original strain RBD. Yet, similar plateau was reached whatever the variant tested (Figure 3A, top). In comparison, binding of the association tixagevimab/cilgavimab (Evusheld®) to Omicron or BA.4/5 RBD was dramatically reduced, and binding to BQ1.1 RBD was completely abolished. Although binding of BA2 and BA2.12.1 subvariants was less affected, a lower plateau was obtained (Figure 3A, bottom).

• Download figure
• Open in new tab

Figure 3: XAV-19 antiviral activity against Omicron and its subvariants

Binding of XAV-19 or Evusheld to Wuhan or Omicron and its subvariant RBD (A). XAV-19 or Evusheld were added to RBD coated plates at the indicated concentration and revealed with an HRP-conjugated secondary anti-pig or anti-human antibody. Neutralizing activity of XAV-19 or Evusheld to Wuhan or Omicron and its subvariant RBD (B). Recombinant His-Tag-RBD pre-incubated with XAV-19 or Evusheld were added to human ACE2 coated plates. Bound RBD were then detected with HRP-conjugated anti-His-Tag antibody. Neutralizing activity of XAV-19 on whole replicating viruses (C). Vero E6 cells were infected with Wuhan or Omicron SARS-CoV-2 strains. CPE was assessed by microscopy examination and viral load percentage was determined by quantitative RT-PCR.

XAV-19 neutralization potential was then tested in an RBD/ACE-2 binding competition assay. Comparable dose response profiles with full neutralization by XAV-19 were obtained with all the variants tested (Figure 3B, left); IC50 was moderately increased (5.9 ± 0.2 and 9.1 ±-0.6µg/mL for Wuhan and Omicron RBD respectively). By contrast, neutralization efficacy of tixagevimab/cilgavimab was limited to 60% for Omicron versus 100% for Wuhan RBD (Figure 3B, right). XAV-19 neutralization of viral infectivity was confirmed by CPE assay using SARS-CoV-2 clinical isolates. XAV-19 neutralized 100% of CPE for all the variants tested and reached 100% viral load reduction (Figure 3C) with slightly varying NT50 concentrations. Interestingly, NT50 against Omicron was in the low range in CPE and lower than for all other variants in VL assessment.

XAV-19 targets epitopes outside the Omicron mutations

Four main target epitopes of XAV-19 (347-fasvyawnr-417, 409-qiapgqtgn-417, 445-vsgnynylyrlfrksnlkpferdisteiy-473 and 530-stnlvk-535) were identified on Omicron RBD by proteolytic epitope mapping (Figure 4, bold), including 6 amino acids out of the 17 (in blue) directly involved in ACE-2 receptor binding. Among the 15 Omicron mutations in RBD (in yellow), only 2 lie inside the major XAV-19 target epitopes (N417 and S446) compared to 5 for the association tixagevimab/cilgavimab (underlined; K440, S446, N477, K478, A484)(16).

• Download figure
• Open in new tab

Figure 4: XAV-19 target epitopes lie outside the Omicron mutation sites

Amino acid sequence of the SARS-CoV-2 spike RBD variant Omicron and XAV-19 target epitopes (amino acid sequence numbered according to DBSOURCE sequence reference NC_045512.2). Bold: XAV-19 target epitopes confirmed by proteolytic epitope mapping; blue: amino acids in contact with ACE-2; yellow: mutations found in Omicron, differentiating from the original Wuhan RBD, underlined: Tixagevimab/Cilgavimab target epitopes.