ABSTRACT
Background Hospitalized patients with Covid-19 experience high mortality rates, ranging from 10-30%. Casirivimab and imdevimab (REGEN-COV®) is authorized in various jurisdictions for use in outpatients with Covid-19 and in post-exposure prophylaxis. The UK-based platform RECOVERY study reported improved survival in hospitalized seronegative patients treated with REGEN-COV, but in most of the world, anti-spike monoclonal antibody therapy is currently not approved for use in hospitalized patients.
Methods In this phase 1/2/3 double-blind placebo-controlled trial, patients on low-flow or no supplemental oxygen hospitalized with Covid-19 were randomized (1:1:1) to 2.4 g or 8.0 g REGEN-COV or placebo and characterized at baseline for viral load and SARS-CoV-2 endogenous immune response.
Results 1336 patients on low-flow or no supplemental oxygen were treated. The primary endpoint was met: in seronegative patients, the LS mean difference (REGEN-COV vs. placebo) for TWA change from baseline viral load was -0.28 log10 copies/mL (95% CI: -0.51, -0.05; P=0.0172). The primary clinical analysis of death or mechanical ventilation from day 6-29 in patients with high-viral load had a strong positive trend but did not reach significance. REGEN-COV reduced all-cause mortality in seronegative patients through day 29 (RRR, 55.6%; 95% CI: 24.2%, 74%). No safety concerns were noted overall nor in seropositive patients.
Conclusions In hospitalized patients with Covid-19 on low-flow or no oxygen, REGEN-COV treatment reduced viral load and the risk of death or mechanical ventilation as well as all-cause mortality in the overall population, with the benefit driven by seronegative patients and no harm observed in seropositive patients.
(ClinicalTrials.gov number, NCT04426695.)
INTRODUCTION
The clinical progression of coronavirus disease 2019 (Covid-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is highly variable; while many cases manifest with relatively mild symptoms, others progress to severe respiratory failure requiring supplemental oxygen and/or mechanical ventilation.1-4 Casirivimab and imdevimab (REGEN-COV®) is a monoclonal antibody combination approved or authorized for emergency use for the treatment of outpatients with mild to moderate Covid-19 and for post-exposure prophylaxis in the United States and other jurisdictions.5-7 In outpatients with Covid-19, REGEN-COV reduced hospitalization or all-cause death, reduced viral load, and shortened symptom duration.8-10 Data show REGEN-COV is also highly effective in preventing asymptomatic as well as symptomatic Covid-19, evidenced by a single-dose subcutaneous administration in showing ∼80% lower risk of developing Covid-19 for household contacts living with an infected individual.11 The totality of evidence suggests that benefit is greatest when treated early.12
Based on the potent anti-viral activity of REGEN-COV it was prospectively hypothesized that reducing viral burden as early as possible would also decrease morbidity and mortality associated with SARS-CoV-2 infection in hospitalized patients. In a recent open label platform trial of hospitalized patients with Covid-19 in the United Kingdom (RECOVERY), REGEN-COV met its primary endpoint in improving overall survival in patients who had not mounted their own immune response at baseline (seronegative) by 20%, and also improved duration of hospitalization.13 Here, we describe the final efficacy and safety results from the first phase 1/2/3 double-blind placebo-controlled trial of REGEN-COV in hospitalized patients with Covid-19 on low-flow or no supplemental oxygen.
METHODS
TRIAL DESIGN
This was an adaptive, phase 1/2/3, double-blinded, placebo-controlled trial to evaluate the efficacy, safety, and tolerability of REGEN-COV in hospitalized adult patients with Covid-19, conducted at 103 sites in the United States, Brazil, Chile, Mexico, Moldova, and Romania (NCT04426695).
Patients were randomized 1:1:1 to 2.4 g REGEN-COV (1.2 g casirivimab and 1.2 g imdevimab), 8.0 g REGEN-COV (4.0 g casirivimab and 4.0 g imdevimab), or placebo as a single intravenous dose. Randomization was stratified by standard-of-care treatment (antiviral therapies, non-antiviral therapies; phase 1/2/3) and country (phase 2/3). The trial included a screening/baseline period (days –1 to 1), a hospitalization/post-discharge period, a monthly follow-up period, and an end-of-study visit (phase 1 day 169, phase 2/3 day 57) (Fig. S1). Trial adaptations, including pausing of enrollment and early termination, are described in the Supplementary Appendix.
PATIENTS
Patients were 18 years or older and hospitalized with confirmed SARS-CoV-2 ≤72 hours with symptom onset ≤10 days from randomization. Standard-of-care treatments for Covid-19, per the investigator, were permitted. Full inclusion and exclusion criteria are presented in the Supplementary Appendix.
Patients were enrolled in one of four cohorts based on disease severity (Fig. S1) as described in the Patient Cohorts section in the Supplementary Appendix. The trial proceeded through phase 2 for patients requiring no supplemental oxygen (cohort 1A) and phase 3 for patients requiring low-flow oxygen (cohort 1); together, these patients are the subject of this manuscript. As phase 1/2 data from patients on low-flow oxygen was previously unblinded in an interim analysis on December 22, 2020, it was not included in the phase 3 efficacy analyses. For patients requiring high-intensity oxygen (cohort 2) or mechanical ventilation (cohort 3), enrollment was paused and data were not included due to low sample size; this is further described in the Trial Adaptations section in the Supplementary Appendix.
OUTCOME MEASURES
The primary virologic efficacy endpoint was the time-weighted average (TWA) daily change from baseline (day 1) viral load in nasopharyngeal samples through day 7.10 The primary clinical efficacy endpoint was the proportion of patients who died or required mechanical ventilation from day 6-29 and day 1-29. Secondary efficacy endpoints examined all-cause mortality and discharge from/readmission to hospital. Safety endpoints included the proportion of patients with treatment-emergent serious adverse events (SAEs) and adverse events of special interest (AESIs): infusion-related reactions (IRRs) through day 4, and grade ≥2 hypersensitivity reactions through day 29.
STATISTICAL ANALYSIS
The statistical analysis plan was finalized prior to database lock and unblinding. The full analysis set (FAS) was used for safety analyses and includes all randomized patients who received any amount of study drug. The modified FAS (mFAS) was used for efficacy analyses and excludes patients who had negative central lab SARS-CoV-2 RT-qPCR at baseline.
Primary endpoint analyses are described in more detail in the Supplementary Appendix. As the trial was stopped earlier than planned (due to low enrollment prior to the surge associated with the Delta variant), the sample size was smaller than anticipated and it was elected to combine the REGEN-COV dose groups and pool patients on no supplemental oxygen (phase 2) and low-flow oxygen (phase 3) for efficacy measures. To control alpha at an 0.05 level, the primary endpoints were tested hierarchically (Table S1). After virologic efficacy, the first two clinical endpoints were the proportion of patients who died or went on mechanical ventilation for REGEN-COV vs. placebo from 1) day 6-29 in high-viral load patients and 2) day 6-29 in seronegative patients. High viral load was selected based on previous experience with treatment in the outpatient setting8,10; it was expected that the high-viral load population would be highly correlated with the seronegative population, and assessing viral load could be easier than assessing serostatus in a clinical setting.
Safety was assessed in separate analyses for patients receiving no supplemental oxygen (phase 2) and low-flow oxygen (phase 1/2/3). Prespecified subgroup analyses using baseline serostatus and viral load were selected based on previous results.10 Sample size calculations and missing data handling are described in the Supplementary Appendix.
TRIAL OVERSIGHT
Details are provided in the Supplementary Appendix.
RESULTS
DEMOGRAPHICS AND BASELINE CHARACTERISTICS
As of April 9, 2021, 1364 patients on low-flow or no supplemental oxygen were randomized into the study, of whom 1336 were treated. Of those, 1197 (89.6%) tested positive centrally for SARS-CoV-2 (constituting the mFAS) with 406, 398, and 393 in the REGEN-COV 2.4 g, 8.0 g, and placebo groups, respectively (Fig. S2).
Baseline demographics were well-balanced. The median age was 62 years, 54.1% were male, mean body mass index was 31.1 kg/m2, 12.1% identified as Black/African American, and 30.1% identified as Hispanic/Latino (Table 1). Covid-19 characteristics were similar except for a higher proportion of seropositive patients in the placebo group (51.1%) compared to the combined REGEN-COV group (45.9%) (Table 1). The number of days of Covid-19 symptoms, viral load, and C-reactive protein concentrations at baseline were balanced, and similar proportions of patients received remdesivir and/or systemic corticosteroids (Table 1). Demographics and baseline characteristics by serostatus are presented in Table S2.
VIROLOGIC EFFICACY
REGEN-COV significantly reduced viral load in seronegative patients on low-flow or no supplemental oxygen; the least-squares (LS) mean (95% confidence interval [CI]) TWA daily change in viral load from baseline through day 7 was -1.03 log10 copies/mL (CI: -1.22, -0.84) in the placebo group compared with –1.31 log10 copies/mL (CI: -1.43, -1.18) in the REGEN-COV combined dose group, with an LS mean difference vs. placebo of -0.28 log10 copies/mL (CI: -0.51, -0.05; P=0.0172) (Table 2).
Both doses of REGEN-COV exhibited similar viral load reductions, showing improvement over placebo starting at day 3 and reaching significance at day 7, after which viral load in the REGEN-COV groups continued to fall relative to placebo (Fig. 1, Fig S3). The maximum LS mean differences versus placebo in seronegative patients were at day 7 (described above), day 9 (–0.47 log10 copies /mL, CI: -0.71, - 0.23), and day 11 (−0.59 log10 copies/mL CI: -0.85, -0.34) (Fig. 1A). The overall population LS mean fell below the lower limit of quantification (LLOQ; 2.85 log10 copies/mL) 2 days earlier with REGEN-COV (day 9 REGEN-COV vs. day 11 placebo) (Fig. S3). Reductions of viral load were observed in the seronegative (Fig. 1A) and the overall populations (Fig. S3), with greater reductions in seronegative patients.
Panel A graph shows LS mean viral load following administration of REGEN-COV (2.4 g, 8.0 g, or combined analysis of 2.4 and 8.0 g) or placebo for patients who tested negative for all SARS-CoV-2 antibodies at baseline (seronegative). Panel B shows the same but for patents who tested positive for any SARS-CoV-2 antibody at baseline (seropositive). For both panels, the lower limit of quantification is 2.85 log10 copies/mL.
CI denotes confidence-interval; IV, intravenous; mFAS, modified full analysis set; LS, least-squares; PBO, placebo; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; SE, standard error; and TWA, time-weighted average.
CLINICAL EFFICACY
DEATH OR MECHANICAL VENTILATION
Endpoints were examined both from day 1-29 and day 6-29, and were evaluated in the seronegative, high-viral load, and overall populations, as described in the Methods. The analyses presented herein examine the pooled REGEN-COV dose group and pooled cohorts for low-flow and no supplemental oxygen (Fig. 2). Individual dose groups of 2.4 g and 8.0 g of REGEN-COV (Fig. S4) and separate cohorts by respiratory status (Fig. S5) were also examined and showed trends of benefit across all clinical endpoints.
Panel A shows a Kaplan–Meier curve for the proportion of patients who died through study day 29, after administration of REGEN-COV (combined analysis of 2.4 g or 8.0 g) or placebo. Results are analyzed separately for patients who were seronegative or seropositive at baseline. + indicates censoring. Panel B shows a forest plot of relative risk and relative risk reduction with 95% CI for REGEN-COV combined dose analysis (2.4 g and 8.0 g) vs. placebo. Parameters examined include death within 28 days, discharge alive from hospital from day 1-29, and death or mechanical ventilation from day 1-29. For all populations, the mFAS was comprised of patients who tested positive for SARS-CoV-2 at baseline. Populations analyzed include patients who tested negative for all SARS-CoV-2 antibodies at baseline (seronegative mFAS) patients who tested positive for any SARS-CoV-2 antibody at baseline (seropositive mFAS), those with inconclusive or missing baseline serology (other), and the overall population regardless of serostatus (overall mFAS).
CI denotes confidence interval; mFAS, modified full analysis set; and SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.
In the statistical hierarchy (Table S1), the first test for clinical efficacy on the endpoint of death or mechanical ventilation in the high-viral load population from day 6-29 showed a numerically lower risk compared to placebo but did not reach statistical significance (relative risk reduction [RRR], 25.5%; 95% CI: -16.2%, 52.2%; P=0.2) (Table 2); accordingly, all subsequent clinical efficacy analyses are considered descriptive. The second test for clinical efficacy on the endpoint of death or mechanical ventilation in the seronegative population from day 6-29 showed an RRR of 47.1% (CI: 10.2%, 68.8%) (Table 2). Improvements were also observed from day 6-29 in the overall population (RRR, 24.2%; 95% CI: 10.9%, 48.2%) (Table 2).
Treatment with REGEN-COV led to a reduction in the proportions of patients who died or required mechanical ventilation, with improvement from day 1-29 in the high-viral load (RRR, 35.0%; 95% CI: 6.6%, 54.8%), seronegative (RRR, 47.0%; 95% CI: 17.7%, 65.8%), and overall (RRR, 30.9%; 95% CI, 5.4% to 49.5%) populations (Table 2). While seronegative patients exhibited the greatest benefit from REGEN-COV treatment, no meaningful benefit or harm was observed in seropositive patients (RRR, 19.5%; 95% CI: -32.8%, 51.2%) (Fig. 2A).
ALL-CAUSE MORTALITY
Treatment with REGEN-COV led to improvement in all-cause mortality through day 29 in the seronegative, high-viral load, and overall populations. The greatest reduction in the relative risk of death occurred in seronegative patients; 24/360 (6.7%) died within 28 days in the REGEN-COV group, compared to 24/160 (15.0%) in the placebo group (RRR, 55.6%; 95% CI: 24.2%, 74%) (Fig. 2B). No harm or meaningful benefit was observed in the seropositive population (Fig. 2A). For the overall population, driven by the seronegative group, a substantial reduction in death was observed in which 59/804 patients (7.3%) died within 28 days in the REGEN-COV combined dose group, compared to 45/393 patients (11.5%) in the placebo group (RRR, 35.9%; 95% CI: 7.3%, 55.7%) (Fig. 2B). The improvement in all-cause mortality with REGEN-COV persisted through study day 57 (Fig. S6). Similar results were also observed in the secondary endpoints of hospital discharge and readmission, shown in Table S3 and Table S4 and described in the Supplementary Appendix.
SAFETY
SAEs were experienced by more patients in the placebo group compared to the REGEN-COV group for patients on low-flow oxygen (131/469 [27.9%] placebo vs. 224/941 [23.8%] REGEN-COV) and no supplemental oxygen (43/198 [21.7%] placebo vs. 61/399 [15.3%] REGEN-COV) (Table 3). More patients experienced treatment-emergent adverse events that resulted in death in the placebo group compared with REGEN-COV for both patients on low-flow oxygen (72/469 [15.4%] placebo vs. 108/941 [11.5%] REGEN-COV) (Table S5) and no supplemental oxygen (15/198 [7.6%] placebo vs. 15/399 [3.8%] REGEN-COV) (Table S6), consistent with treatment benefit of REGEN-COV. These events were generally considered by the sponsor as associated with Covid-19 and its complications.
Grade ≥2 IRRs occurred in few patients on low-flow (5/469 [1.1%] placebo vs.
18/941 [1.9%] REGEN-COV) and no supplemental oxygen (1/198 [0.5%] placebo vs. 8/399 [2.0%] REGEN-COV) (Table 3). Grade ≥2 hypersensitivity reactions also occurred in few patients on low-flow (1/469 [0.2%] placebo vs. 7/941 [0.7%] REGEN-COV) and no supplemental oxygen (1/198 [0.5%] placebo vs. 2/399 [0.5%] REGEN-COV) (Table 3). AESIs are further detailed in Table S7 and Table S8.
DISCUSSION
REGEN-COV is a potent antiviral monoclonal antibody combination shown to rapidly reduce SARS-CoV-2 viral load and modulate a patient’s disease course across various populations.8-11,14 REGEN-COV, which contains two distinct neutralizing antibodies,15,16 retains neutralizing potency against viral variants of concern, including the Beta, Gamma, and Delta variants.17
Hospitalized patients with Covid-19 experience high mortality rates, ranging from 10-30%.4,18-20 REGEN-COV has potent anti-viral activity, so it was prospectively hypothesized that reducing viral burden as early as possible would decrease morbidity and mortality associated with SARS-CoV-2 hospitalization. Until the recently released results from the RECOVERY platform trial,21 it was unknown whether reducing viral burden in patients who were already hospitalized would be early enough to meaningfully impact clinical outcomes; RECOVERY reported a 20% reduction in 28-day mortality with REGEN-COV treatment compared to standard of care in seronegative patients. The current placebo-controlled randomized trial (with approximately half of patients receiving concomitant remdesivir) demonstrated and extended the benefit preliminarily reported in the RECOVERY trial among seronegative patients, and also documented no harm signals among seropositive patients. When added to standard of care, REGEN-COV significantly reduced all-cause mortality for patients on low-flow or no supplemental oxygen. While the primary clinical endpoint, which had a strong positive trend, did not reach significance, all clinical endpoints demonstrated substantial numeric improvements, predominantly driven by the robust and nominally significant results in the seronegative population. In addition, REGEN-COV improved the rates of hospital discharge and death or readmission to hospital at day 29, which persisted through day 57, showing direct benefit to patients as well as the overburdened healthcare system, particularly in the context of the Delta variant which doubles the risk of hospitalization.22
The clinical benefits observed in the overall population of this study, including improvements in death or mechanical ventilation, all-cause mortality, and discharge from hospital, were driven by seronegative patients, but no harm was demonstrated in seropositive patients, for whom there were numerically fewer deaths through day 29 with REGEN-COV compared to placebo. In the current variant-rich world with widespread Covid-19 vaccination, the utility of serostatus is unclear; numerous publications cite that even vaccinated patients with high antibody titers may have little to no neutralizing activity to emerging variants.23-26 Given this context, the results from this study may support the use of REGEN-COV in hospitalized patients with Covid-19, regardless of serostatus. Future studies are needed to further explore the potential clinical benefit in seropositive patients.
REGEN-COV is the first monoclonal antibody therapy, and the first SARS-CoV-2 antiviral, that significantly lowers viral load and reduces mortality in hospitalized patients with Covid-19. Very few treatments have demonstrated a mortality benefit in hospitalized Covid-19 patients. Most of these treatments are designed to modulate the immune response late in the disease course after damage has occurred, rather than to clear SARS-CoV-2. The corticosteroid dexamethasone showed a 17% improvement in 28-day mortality in the RECOVERY trial, with the greatest benefit in patients receiving mechanical ventilation.21 Baricitinib, a Janus kinase inhibitor, improved 28-day mortality by 38% in hospitalized patients.27 Recently, interleukin-6 inhibitors such as tocilizumab and sarilumab were recommended by the World Health Organization for use in hospitalized patients, in which they reduced mortality by 13%.28,29 The Food and Drug Administration-approved medication remdesivir has shown neither a reduction in viral load nor a mortality benefit in hospitalized patients.30 REGEN-COV’s mechanism of action and safety profile should allow combination approaches with any or all of these other agents.
The safety profile of REGEN-COV in patients on low-flow or no supplemental oxygen was consistent with that observed previously in outpatients and hospitalized patients with Covid-19,10,13 showing low rates of infusion-related and hypersensitivity reactions, as expected for a fully human antibody against an exogenous target. Overall, REGEN-COV was well tolerated. The placebo group experienced a greater frequency of SAEs and adverse events leading to death than the REGEN-COV group, consistent with the clinical benefit of treatment.
The respiratory status of the population in this manuscript includes those receiving low-flow or no supplemental oxygen, as the study did not enroll sufficient numbers of patients on high-intensity oxygen or mechanical ventilation prior to pausing of these cohorts due to potential safety imbalances, which were not observed in the much larger RECOVERY trial.21 The absence of full representation across the spectrum of hospitalized patients on varying degrees of oxygen support is a limitation of this study. Additionally, this study was prematurely terminated due to slow recruitment prior to the current surge associated with the emergence of the Delta variant. As a result of the smaller sample size, key analyses pooled the two patient cohorts as well as the two doses. Sensitivity analyses did not reveal major efficacy differences across the cohorts or doses; minor variability in the magnitude of risk reductions, with greater effects for the 2.4 g dose compared to the 8.0 g dose, was likely due to small numbers within each group suggesting that either dose can be utilized in hospitalized individuals requiring low-flow or no supplemental oxygen. Overall, trends for treatment benefit on mortality and other efficacy endpoints extend the observations in the larger RECOVERY trial, where benefit was seen in all patients regardless of respiratory status.
Authorized options for intervention with a monoclonal antibody anti-viral treatment have been restricted to outpatients with Covid-19 or as post-exposure prophylaxis.5-7 Taken together with reports from the RECOVERY trial, these data support REGEN-COV representing a well-tolerated and compelling treatment option to reduce the risk of mortality in hospitalized Covid-19 patients and across the disease continuum of SARS-CoV-2, from prevention to hospitalization.
Data Availability
Qualified researchers may request access to study documents (including the clinical study report, study protocol with any amendments, blank case report form, and statistical analysis plan) that support the methods and findings reported in this manuscript. Individual anonymized participant data will be considered for sharing once the indication has been approved by a regulatory body, if there is legal authority to share the data and there is not a reasonable likelihood of participant re-identification. Submit requests to https://vivli.org/.
SUPPORTED BY
Supported by Regeneron Pharmaceuticals, Inc. Certain aspects of this project have been funded in whole or in part with Federal funds from the Department of Health and Human Services; Office of the Assistant Secretary for Preparedness and Response; Biomedical Advanced Research and Development Authority, under OT number HHSO100201700020C.
FINANCIAL DISCLOSURE
Disclosure forms provided by the authors are available with the full text of this article.
DATA SHARING STATEMENT
Qualified researchers may request access to study documents (including the clinical study report, study protocol with any amendments, blank case report form, and statistical analysis plan) that support the methods and findings reported in this manuscript. Individual anonymized participant data will be considered for sharing once the indication has been approved by a regulatory body, if there is legal authority to share the data and there is not a reasonable likelihood of participant re-identification. Submit requests to https://vivli.org/.
ACKNOWLEDGMENTS
We thank the patients who participated in this study as well as their families; the study investigators; the members of the independent data monitoring committee; Kaitlyn Scacalossi, Ph.D. and Caryn Trbovic, Ph.D. from Regeneron Pharmaceuticals for assistance with development of the manuscript; and Prime, Knutsford, UK, for formatting and copy-editing suggestions.
Footnotes
↵‡ Former employee of Regeneron Pharmaceuticals, Inc.
