Bamlanivimab reduces nasopharyngeal SARS-CoV-2 RNA levels but not symptom duration in non-hospitalized adults with COVID-19

Abstract

Importance: The antiviral activity and efficacy of anti-SARS-CoV-2 monoclonal antibody (mAb) therapies to accelerate recovery from COVID-19 is important to define.

Objective: To determine safety and efficacy of the mAb bamlanivimab to reduce nasopharyngeal (NP) SARS-CoV-2 RNA levels and symptom duration.

Design: ACTIV-2/A5401 is a randomized, blinded, placebo-controlled platform trial. Two dose cohorts were enrolled between August 19 and November 17, 2020 for phase 2 evaluation: in the first, participants were randomized 1:1 to bamlanivimab 7000 mg versus placebo, and in the second to bamlanivimab 700 mg versus placebo. Randomization was stratified by time from symptom onset (≤ or >5 days) and risk of progression to severe COVID-19 ("higher" vs "lower").

Setting: Multicenter trial conducted at U.S. sites.

Participants: Non-hospitalized adults ≥18 years of age with positive SARS-CoV-2 antigen or nucleic acid test within 7 days, ≤10 days of COVID-19 symptoms, and with oxygen saturation ≥92% within 48 hours prior to study entry.

Intervention: Single infusion of bamlanivimab (7000 or 700 mg) or placebo.

Main outcomes and measures: Detection of NP SARS-CoV-2 RNA at days 3, 7, 14, 21, and 28, time to improvement of all of 13 targeted COVID-19 symptoms by daily self-assessment through day 28, and grade 3 or higher treatment emergent adverse events (TEAEs) through day 28. Secondary measures included quantitative NP SARS-CoV-2 RNA, all-cause hospitalizations and deaths (composite), area under the curve of symptom scores from day 0 through day 28, plasma bamlanivimab concentrations, plasma and serum inflammatory biomarkers, and safety through week 24.

Results: Ninety-four participants were enrolled to the 7000 mg cohort and 223 to the 700 mg cohort and initiated study intervention. The proportion meeting protocol criteria for "higher" risk for COVID-19 progression was 42% and 51% for the 7000 and 700 mg cohort, respectively. Median time from symptom onset at study entry for both cohorts was 6 days. There was no difference in the proportion with undetectable NP SARS-CoV-2 RNA at any post-treatment timepoints (risk ratio compared to placebo, 0.82-1.05 for 7000 mg dose [overall p=0.88] and 0.81-1.21 for 700 mg dose [overall p=0.49]), time to symptom improvement (median of 21 vs 18.5 days, p=0.97, for 7000 mg bamlanivimab vs placebo and 24 vs 20.5 days, p=0.08, for 700 mg bamlanivimab vs placebo), or grade 3+ TEAEs with either dose compared to placebo. Median NP SARS-CoV-2 RNA levels were lower at day 3 and C-reactive protein, ferritin, and fibrinogen levels significantly reduced at days 7 and 14 for bamlanivimab 700 mg compared to placebo, with similar trends observed for bamlanivimab 7000 mg. Viral decay modeling supported more rapid decay with bamlanivimab compared to placebo.

Conclusions and relevance: Treatment with bamlanivimab 7000 mg and 700 mg was safe and compared to placebo led to more rapid reductions in NP SARS-CoV-2 RNA and inflammatory biomarkers, but did not decrease time to symptom improvement. The clinical utility of mAbs for outcomes other than hospitalizations and deaths is uncertain.

Trial registration: ClinicalTrials.gov Identifier: NCT04518410.

Key points: Question: What is the safety and efficacy of bamlanivimab monoclonal antibody (mAb) treatment for mild to moderate COVID-19?Findings: In this randomized, placebo-controlled phase 2 trial of 317 non-hospitalized adults with COVID-19, there was no relationship between symptoms or disease progression risk and nasopharyngeal (NP) virus shedding. Bamlanivimab was safe and reduced NP SARS-CoV-2 RNA levels and inflammatory biomarker levels more than placebo, but did not shorten symptom duration.Meaning: Nasal virus shedding was not associated with symptoms or baseline risk factors for severe COVID-19. Bamlanivimab, which has been associated with reduced hospitalizations in high-risk individuals, demonstrated antiviral activity with early post-treatment NP sampling but did not accelerate symptom improvement. The clinical utility of bamlanivimab for outcomes other than hospitalizations and deaths, including longer-term outcomes, is uncertain.

Figures

Figure 1A.. CONSORT Flow Diagram, Bamlanivimab 7000 Dose Group

aIncludes 1 participant assigned bamlanivimab 700 mg

Figure 1B.. CONSORT Flow Diagram, Bamlanivimab 700 Dose Group

aIncludes 2 participants assigned bamlanivimab 7000 mg; bIncludes 2 participants assigned placebo 7000 mg

Figure 2.. Associations between baseline virology and symptom scores and comparisons by subgroups, combining bamlanivimab 700 and 7000 mg dose cohorts.

(A) Baseline nasopharyngeal (NP) and anterior nasal (AN) SARS-CoV-2 RNA levels (viral loads) were highly correlated. The diagonal line indicates line of equality. NP and AN viral loads did not differ by protocol-defined risk category (“higher” vs “lower”) for COVID-19 progression (B), but were significantly higher among participants entering the study with fewer days of symptoms (≤5 vs >5 days) (C) and among viremic vs aviremic participants (D). The proportion with SARS-CoV-2 viremia at study entry was the same for participants at higher vs lower risk for COVID-19 progression (E). Total symptom scores reported at study entry in the participant diary were higher among viremic participants (F). Symptoms scores did not correlate with NP or AN viral loads (G). In Figures 2B–2D and 2F, Tukey boxplots were used to demonstrate the distribution of viral loads or symptom score. Wilcoxon rank sum test was used to compare values from different groups.

Figure 3.. Nasopharyngeal (NP) SARS-CoV-2 RNA levels (viral loads) by dose cohort, treatment arm and visit.

NP viral loads declined in all participants, with median NP viral loads lower at day 3 for bamlanivimab 700 mg (n=111) vs placebo (n=112) (2.9 vs 3.9 log10 copies/mL, p=0.002) (A), without a difference in proportion undetectable at any time points (B). Similar findings were seen, though differences in median viral load at day 3 were not statistically significant for the smaller 7000 mg dose cohort (n=48 bamlanivimab vs n=46 placebo, 2.2 vs 3.4 log10 copies/mL, p=0.07) (C and D). The lower limit of detection was 1.4 log10 copies/mL.