Effects of bamlanivimab alone or in combination with etesevimab on subsequent hospitalization and mortality in outpatients with COVID-19: a systematic review and meta-analysis

Introduction

The crisis caused by the coronavirus disease 2019 (COVID-19), fueled by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), continues to be a substantial health threat. As of July 31, 2021, more than 200 million patients have been infected, resulting in more than 4 million deaths (Chang et al., 2020; Lai et al., 2020a; Johns Hopkins University & Medicine, 2021). The overall fatality rate of the initial COVID-19 variants is approximately 2–3%, however, this percentage decreased with the Omicron variants (Mathieu et al., 2020; Finelli et al., 2021). Those who are elderly, obese, or have underlying systemic diseases have a higher risk for unfavorable outcomes (Mathieu et al., 2020; Lai et al., 2020b; Yuki, Fujiogi & Koutsogiannaki, 2020). The pathophysiology of COVID-19 involves different phases of infection. The direct viral invasion is responsible for the initial injury and immune-mediated inflammatory processes result in subsequent damage to the host. The timely diagnosis and proper management of the disease, such as early quarantine, treatment, and prognosis, contributes to successful patient outcomes (Gandhi, Lynch & del Rio, 2020). Several pharmacotherapeutic medications, which use different mechanisms for treatment, have been used to combat COVID-19 in patients with varying disease severity and at different phases of the disease (Gandhi, Lynch & del Rio, 2020; Siemieniuk et al., 2020; Yang et al., 2021). Some antivirals, monoclonal antibodies, and convalescent plasma have been used to target viruses in their initial phase. Remdesivir is just such a medication, and it has been used to inhibit RNA-dependent RNA polymerase (Moreno et al., 2022). Some drugs are applied for their immune-modulatory effects in the inflammatory phase, including corticosteroids, janus kinase (JAK) inhibitor, and interleukin-6 (IL-6) inhibitors (Gautret et al., 2020; Lan et al., 2020). For patients with moderate to severe infections, monoclonal antibody treatments are ineffective, while IL-6 inhibitors, such as tocilizumab, are the mainstay treatment and contribute to better outcomes (Hariyanto, Hardyson & Kurniawan, 2021; Rosas et al., 2021). However, the effects of treatment are largely unclear and the optimal course of treatment is still under investigation.

COVID-19 has proven to be a highly contagious virus with moderate severity. The severity of Omicron infections was relatively mild, and most people were vaccinated when it appeared. Therefore, many countries adopted a co-existing policy, which allowed populations to return to a more normal life with eased social restrictions. However, the rapid increase in patients with COVID-19 caused a shortage of medical resources to the point where a medical system collapse could occur. Outpatient treatments with reduced subsequent hospitalization may help to preserve medical resources, thus, monoclonal antibodies were developed to ensure adequate and timely outpatient treatment. SARS-CoV-2 is composed by four major structural proteins including spike, envelope, membrane, and nucleocapsid proteins. Bamlanivimab (BAM) (Eli Lilly, former name: LY-CoV-555) is a type of neutralizing monoclonal antibody derived from the plasma of recovered patients (Cohen, 2021; Ganesh et al., 2021a). It binds to the epitopes in the receptor binding domain of the spike protein and provides rapid and passive immunity against SARS-CoV-2 (Jones et al., 2020; Taylor et al., 2021; Tuccori et al., 2021). Early monoclonal antibody treatment in patients with mild COVID-19 infection may block the viral replication and improve subsequent clinical outcomes. Etesevimab (ETE) (Eli Lilly, former name: LY-CoV-016) is another monoclonal antibody targeting a different epitope of the receptor binding domain. BAM alone or in combination with ETE (BAM/ETE) was emergency authorized to treat COVID-19 patients by the US Food and Drug Administration in November 2020. For non-hospitalized COVID-19 patients, BAM/ETE treatment decreased the risk of subsequent hospitalization and mortality (Cohen, 2021). A single infusion in an outpatient setting decreased the risk of subsequent hospitalization and helped to preserve medical capacity. However, this benefit was not observed in hospitalized patients. Therefore, our team conducted this systematic review and meta-analysis to investigate the effects of BAM ± ETE on patients with COVID-19.

Materials and Methods

Results

Included studies and demographic characteristics

A total of 14 studies investigating BAM/ETE were identified in the literature by July 31, 2021. These studies included 28,984 total participants (Fig. 1, Table 1) (Alam et al., 2021; Bariola et al., 2021; Chen et al., 2021; Corwin et al., 2021; Destache et al., 2021; Dougan et al., 2021; Ganesh et al., 2021b; Gottlieb et al., 2021; Karr et al., 2021; Koehler et al., 2021; Kumar et al., 2021; Lundgren et al., 2021; Piccicacco et al., 2021; Webb et al., 2021). Three publications reported on BLAZE-1 phase 2 and phase 3 studies and some participants were found to be overlapping (Chen et al., 2021; Dougan et al., 2021; Gottlieb et al., 2021). One study was conducted in Germany, one in Denmark, and one in Singapore, with the majority of the studies being conducted in the United States. Four studies were randomized-controlled trials and the others were observational cohort studies. Two studies involved hospitalized patients, while the remaining studies focused on non-hospitalized patients (Koehler et al., 2021; Lundgren et al., 2021). Monotherapy with BAM was used in most of the studies and combination therapy with ETE was investigated in only three studies. The COVID variants of interest in the studies included 21C (Epsilon), 20E (EU1), and B.1.1.7 (Alpha). Seven additional studies were identified in January 2023, and these were included in our analysis (Chilimuri et al., 2022; Cooper et al., 2021; Iqbal et al., 2021; Leavitt et al., 2021; Murillo et al., 2022; Rainwater-Lovett et al., 2021; Rubin et al., 2021).

Table 1:

Clinical characteristics and demographic data of included studies.

Study	Country	Study period (prevalent strain)	Study design	Severity	Participants	Gender, female	Age (years)	Intervention	Control	Outcomes	Safety outcome
BLAZE-1, monotherapy	USA	June 2020–August 2020 (non-variant)	Phase 2, randomized, double-blind, placebo-controlled	Outpatients, mild/moderate	452	249	45	LY-CoV555, 700, 2,800, 7,000 mg for single dose	Placebo	Virological endpoints, clinical outcomes	Adverse events
BLAZE-1, phase 3	USA	September 2020– December 2020 (21C, Epsilon)	Phase 3, randomized, double-blind, placebo-controlled	>= 12 y/o outpatients, mild/moderate	1,035	538	53.8	2,800 mg of bamlanivimab and 2,800 mg of etesevimab	Placebo	Virological endpoints, clinical outcomes	Adverse events
BLAZE-1, combination	USA	June 2020–September 2020 (non-variant)	Phase 2/3, randomized, double-blind, placebo-controlled	Outpatients, mild/moderate	577	315	44.7	Bamlanivimab (700, 2,800, 7,000 mg), the combination treatment (2,800 mg of bamlanivimab and 2,800 mg of etesevimab)	Placebo	Virological endpoints, clinical outcomes	Adverse events
ACTIV-3	USA, Denmark, Singapore	August 2020–October 2020 (non-variant and 21C)	Randomized, double-blind, placebo-controlled	Hospitalized patients without end-organ failure	314	137	61	LY-CoV555 7,000 mg for single dose	Placebo	Pulmonary ordinal outcome, Time to sustained recovery, Time to hospital discharge	Infusion reaction, Composite safety outcome, Composite safety outcome, organ dysfunction, or serious Coinfection, Death
Alam	USA	November 2020–January 2021 (21C, Episilon)	Retrospective case-control study	Long-term care facility, mild/moderate	246	129	82.4	LY-CoV555 700 mg for single dose	Not treated	Clinical outcomes (death and hospitalization)	4.37% side effects
Bariola	USA	December 2020–March 2021 (21C Episilon and Alpha)	Propensity-matched observational study	Non-hospitalized, mild/moderate	1,392	650	67	700 mg of bamlanivimab	Not treated	1. 28-day hospitalization, all-cause mortality 2. Hospitalization or ER without hospitalization
Chilimuri	USA	November 2020–Mar 2021 (21C Episilon and Alpha)	Retrospective observational study	Mild to moderate	38	22	65	700 mg of bamlanivimab or casirivimab/imdevimab	Not treated	30-day hospitalization and mortality
Cooper	USA	November 2020–May 2021 (21C Episilon and Alpha)	Propensity-matched observational cohort study	Outpatients	5,758	3,159	60	Bamlanivimab, bamlanivimab-etesevimab, casirivimab/imdevimab	Not treated	1. 14- and 28-day hospitalization 2. Intensive care unit admission, mortality
Corwin	USA	November 2020–January 2021 (21C)	Cohort study	Outpatients who have BMI ≥35 and/or are age ≥65 years old, mild/moderate	6,117	3,456	62.6, 56.7	700 mg of bamlanivimab	Not treated	Admission, emergency department visit, death	5.4% adverse effects
Destache	USA	November 2020–December 2020 (21C)	Propensity-matched retrospective cohort study	Outpatients, mild/moderate	234	124	72	700 mg of bamlanivimab	Not treated	Clinical outcomes
Ganesh	USA	November 2020–February 2021 (21C, Alpha)	Propensity-matched retrospective cohort study l	Outpatients, mild/moderate	4,670	2,306	63	700 mg of bamlanivimab	Not treated	Clinical outcomes: Admission, intensive care unit admission, mortality	0.81% adverse events
Iqbal	USA	November 2020–January 2021 (21C)	Observational study	Outpatients	284	149	68	Bamlanivimab	Not treated	14- and 30-day hospitalization and mortality
Karr	USA	December 2020–January 2021 (21C)	Retrospective, cohort study	Outpatients, mild/moderate	46	17	69	700 mg of bamlanivimab	Not treated	Clinical outcomes, Emergency department visits/admission	10% adverse effects
Koehler	Germany	February 2021–April 2021 (Alpha and 20E)	Cross-sectional, retrospective study	Inpatients	43	28	71.1	700 mg of LY-CoV555 (eight cases) and 2,400 mg of REGN-CoV-2 (casirivimab/imdevimab) (three cases).	Not treated	Clinical and therapeutical features and outcomes	0%
Kumar	USA	November 2020–Jaunuary 2021 (21C)	Retrospective case-control study	Outpatients, mild/moderate	376	193	64	700 mg of bamlanivimab	Not treated	Clinical outcomes
Leavitt	USA	December 2020–January 2021 (21C)	Retrospective observational study	Outpatients	279	149	63, 69	Bamlanivimab	Not treated	7-, 14-, and 28-day hospitalizations and mortality	None
Murillo	USA	November 2020–December 2020 (21C)	Retrospective study	Residents of nursing home, mild/moderate	107	61	76	700 mg of bamlanivimab	Not treated	Hospitalizations, mortality	5% adverse effects
Piccicacco	USA	November 2020–Jaunuary 2021 (21C)	Retrospective cohort study	Outpatients, mild/moderate	400	207	63.2	Bamlanivimab 700 mg (152 cases) or casirivimab/imdevimab 2,400 mg (48 cases)	Not treated	Clinical outcomes	3.2–3.9% adverse effects
Rainwater-Lovett	USA	January 2021 (21C)	Retrospective study	Outpatients	598	366	62.3	Bamlanivimab	Not treated	30-day medical visit
Rubin	USA	December 2020– February 2021 (21C)	Retrospective case-control study	Outpatients with BMI ≥35 or age ≥65	1,257	707	64.2	Bamlanivimab	Not treated	30-day hospitalizations or mortality
Webb	USA	July 2020–January 2021 (21C)	Cohort study	Outpatients with risk score ≥7.5 points	13,534	6,064	61	Bamlanivimab 700 mg (479 cases) or casirivimab/imdevimab 2,400 mg (115 cases)	Not treated	Clinical outcomes	1.2% adverse events

DOI: 10.7717/peerj.15344/table-1

Meta-analysis of bamlanivimab treatment on primary and secondary outcomes

The quality assessment of the identified studies was summarized in Fig. 2. A low risks of bias in the randomized-controlled trials was observed in all domains of RoB-2. The concern of bias risk was noted in item 1 in the selection of all cohort studies. A total of 10 studies had potential risk of comparability, but the overall scores of the assessed studies were high (6 to 7 points). All of the studies were qualified to be enrolled in subsequent meta-analysis.

For assessing the risk of subsequent hospitalization, patients with BAM/ETE treatment had a significantly lower risk of hospitalization with moderate heterogeneity (18 trials, OR: 0.37, 95% CI [0.29–0.49], I²: 69%; p < 0.01) (Fig. 3). A sensitivity test was conducted and the results remained unchanged by trials of individual studies. A funnel plot showed certain asymmetric distributions of the enrolled studies and was suggestive of publication bias (File S2). Another contour-enhanced funnel plot and Egger’s test demonstrated the presence of publication bias (Files S3, S4; Egger’s test, t = −2.66, p = 0.0172). Despite these results, BAM/ETE still had a significantly lower risk of subsequent hospitalization (eight trials, OR: 0.39, 95% CI [0.3–0.49], I²: 32%; p = 0.17) after removing those studies with publication bias.

There was a significant reduction in mortality with BAM/ETE treatment and with low heterogeneity (15 trials, OR: 0.27, 95% CI [0.17–0.43], I²: 0%; p = 0.85) (Fig. 4), however, some asymmetry was observed in the funnel plot (File S5). After removing studies with potential publication bias, the beneficial effects were similar (eight trials, OR: 0.24, 95% CI [0.11–0.51], I²: 11%; p = 0.35). Furthermore, we conducted a subgroup analysis with BAM monotherapy which showed similar benefits (risk of subsequent hospitalization, 16 trials, OR: 0.43, 95% CI [0.34–0.54], I²: 57%; p = 0.01) (Fig. 5). An additional funnel plot showed mild asymmetry and the Egger’s test showed a low risk of publication bias (Files S6, S7; t = −2.2, p = 0.0448). BAM monotherapy was also shown to decrease subsequent mortality with low heterogeneity (14 trials, OR: 0.28, 95% CI [0.17–0.46], I²: 0%; p = 0.9). Asymmetry was also observed in this funnel plot.

Two studies that investigated hospitalized patients showed no benefit of BAM treatments (Koehler et al., 2021; Lundgren et al., 2021). Adverse events ranged from 0–10%, including immediate hypersensitivity reactions and infusion related discomforts, which were tolerable. The hypersensitivity reactions that developed during infusion were reported as mild in severity and were without a dose-dependent relationship (Gottlieb et al., 2021; Zhao et al., 2022).

Discussion

Based on our systematic review and meta-analysis, non-hospitalized patients receiving BAM with or without ETE had a significantly lower risk of subsequent hospitalization and mortality (OR: 0.37 for hospitalization and 0.27 for mortality). This benefit was also observed in patients treated with BAM monotherapy (OR: 0.43 for hospitalization and 0.28 for mortality). The adverse events of both of these treatments were tolerable. A single BAM/ETE infusion provided in an emergency department or infusion center was a reasonable and feasible treatment for non-hospitalized patients. This treatment led to decreased subsequent hospitalization and mortality and preserved medical resources during the study period when the Alpha variant was the predominant circulating strain.

Although vaccination plays a crucial role in the battle against COVID-19, effective anti-viral agents are indispensable (Chiu et al., 2021; Chiu et al., 2022). The stage and severity of the disease may affect the strategy and effects of treatment. There are two significantly different stages in the pathophysiology of COVID-19: virus replication and subsequent inflammation. Monoclonal antibodies, convalescent plasma, and antiviral agents block viral entry and abrogate viral replication for patients with no symptoms or only mild illness, as seen in the early stage (Gandhi, Lynch & del Rio, 2020). Corticosteroid and immunomodulatory agents may be applied to mitigate the inflammatory cascade for patients with moderate or severe illness in the later stages of the disease (Hariyanto, Hardyson & Kurniawan, 2021). Therefore, the use of BAM/ETE on hospitalized patients with moderate or severe infection is ineffective (Borrok et al., 2018). BAM/ETE targets the spike protein of SARS-CoV-2, and early BAM/ETE treatment will block the virus’s entry, which is beneficial for the successful treatment of COVID-19 (Cohen, 2021; Jones et al., 2020). However, BAM/ETE is not advantageous for hospitalized patients. For these patients, viral replication is robust and a high viral load and inflammatory cytokines are typically observed. Providing passive immunity to inhibit the viral spike protein dos not modify the subsequent inflammatory processes. Similar findings were reported in remdesivir, an RNA-dependent RNA polymerase inhibitor. The use of remdesivir has been shown to be beneficial to patients receiving oxygen; however, the outcomes were not significantly different in ventilated patients (Beigel et al., 2020). Therefore, the treatment timing is imperative for BAM/ETE in the clinical setting. For patients with mild illness and risk factors for severe infection, including older age, obesity, and other underlying systemic diseases, timely BAM/ETE administration is crucial for successful treatment.

In addition to disease timing and severity, rapidly mutating variants of the virus affect the effects of pharmacologic treatment, and drug resistance is the main barrier for the clinical application of monoclonal antibodies (Tuccori et al., 2021). BAM is not effective against the variants B.1.351., P.1., B.1.6.7.2 (Delta), and B.1.1.529 (Omicron). Pseudovirus neutralizing antibody testing showed that BAM was effective against variants with the N501Y mutation (B.1.1.7., Alpha), but there was a significant decrease in sensitivity against variants with the E484K mutation (B.1.351, P.1, B.1.526, and B.1.617), and the L452R mutation (B.1.427 and B1.429) (Hoffmann et al., 2021). Resistance was also confirmed by a neutralizing antibody test of sera and a significantly-reduced sensitivity to the B1.6.7.2 (Delta) variant was observed (Planas et al., 2021). Mutation in the N-terminal domain and the receptor binding domain of the spike protein may be responsible for the observed immunological evasion and escape. However, ETE, casirivimab, and imdevimab remained sensitive to the Delta variant but not Omicron. As Delta and Omicron variants become the predominant strains in many countries, monoclonal antibody treatment with BAM/ETE should be avoided. Our study found a significantly reduced risk of hospitalization and mortality in both BAM monotherapy and a combination therapy with BAM and ETE. The prevalent strains and variants of interest were 21C (Epsilon), 20E (EU1). However, the variant of concern, B.1.1.7 (Alpha), was prevalent in the included studies. The clinical application of BAM/ETE is limited by the prevalence of rapidly mutating variants. Moreover, the widespread use of monoclonal antibodies may accelerate the emergence of COVID mutants (Focosi et al., 2022) since the monoclonal antibody targets single epitopes and the spike protein escape mutation may occur under specific selective pressure. For example, the simultaneous resistance to BAM and ETE with the spike mutation Q943R was reported in a patient given BAM/ETE treatment (Focosi et al., 2021). Therefore, genomic surveillance is recommended, especially for patients with poor responses to monoclonal antibody treatment.

The COVID-19 pandemic is novel and there is a battle between viral mutations and available treatments. Several monoclonal antibodies have been developed, similar to BAM/ETE, including casirivimab plus imdevimab, sotrovimab, tixagevimab plus cilgavimab, and bebtelovimab. The use of BAM/ETE and casirivimab/imdevimab was halted due to the resistance of the Omicron mutants, which have been the predominant strains worldwide since late 2021. Sotrovimab is a medication that was initially identified in a patient with SARS-CoV infection. It was shown to target an epitope conserved in both SARS-CoV and SARS-CoV-2. Sotrovimab is effective against the Omicron variant and has been authorized to treat mild-to-moderate COVID-19 in non-hospitalized patients at high risk for progressing to severe disease. Bebtelovimab is a broadly neutralizing monoclonal antibody targeting relatively conserved epitopes. It is effective against the current variants, including those of Omicron (Dougan et al., 2022; Westendorf et al., 2022). This treatment has received emergency use authorization from the Food and Drug Administration to treat mild to moderate COVID-19 patients (US Food & Drug Administation, 2020).

Furthermore, like the use of the highly active antiretroviral therapy to overcome the resistance of human immunodeficiency virus, cocktail therapies with different kinds monoclonal antibodies have been developed to treat COVID-19 variants. These cocktails include combinations of bebtelovimab, BAM, and ETE (Dougan et al., 2022). Additional cocktail combinations using monoclonal antibodies and antiviral agents have also been studies. BAM/ETE has been shown to be ineffective against the Delta and Omicron variants and its clinical application is limited at present. More options are required when facing rapidly mutating viruses. Many older drugs have been repurposed in an attempt to treat COVID-19 in its different stages and severity. The use of BAM/ETE is a reminder of the importance of genomic surveillance it may be repurposed as a component of a cocktail regimen for treating future variants.

Monoclonal antibody treatment is an attractive treatment option because it may be provided as a single infusion without hospitalization. For example, peramivir is an anti-influenza viral agent that has shown success with a single dose that can be provided in an outpatient setting (Hsieh et al., 2021). A decrease in subsequent hospitalizations is valuable for its ability to preserve vital medical capacity necessary during the pandemic. Furthermore, BAM may also used for prophylaxis after close contact with an infected person (Cohen et al., 2021). Residents in nursing and assisted living facilities are at higher risk of contracting the disease during a COVID-19 outbreak. BAM prophylaxis could significantly decrease the risk of COVID-19 infection in residents and staff (BAM vs. placebo: 8.5% vs. 15.2%; odds ratio, 0.43 95% CI [0.28–0.68]) and its administration is feasible (Dale et al., 2021). However, the COVID variant will affect the prophylactic effects of BAM and more evidence is required to determine its benefits.

Our study is subject to certain limitations. First, the virus has been shown to mutate rapidly and the prevalent strain at present is different from those of the study periods. Although our study shows the significant benefits of early BAM/ETE treatment, in vitro studies showed a decreased sensitivity of BAM/ETE against some variants of concern. Most hospitals do not have the ability to test for specific COVID variants, thus the use of BAM/ETE is limited. BAM/ETE is ineffective against variants of concern, including Delta and Omicron. However, mutation occurs rapidly and unexpectedly, so repurposing old drugs may provide a potential treatment option when facing a new variant. Second, funnel plots and Egger’s tests indicated the presence of publication bias. Studies with negative or inconclusive results may not be published. Finally, the benefits of BAM/ETE are obvious in non-hospitalized patients, but the criteria of hospitalization may vary in different health care units. Further objective saturation or laboratory testing may assist primary care physicians to identify the patients who would be most appropriate for BAM/ETE treatment.

Conclusions

In conclusion, we found a significant reduction of subsequent hospitalization and mortality in non-hospitalized patients treated with BAM, with or without ETE treatment (OR: 0.37 for subsequent hospitalization, moderate heterogeneity, and 0.27 for mortality, low heterogeneity). Adverse events were not frequent and were tolerable to the 28,577 patients identified in our systematic review and meta-analysis. A single infusion with BAM with or without ETE in an outpatient setting was feasible for patients with risk factors and mild illness. This has been shown to be a reasonable treatment strategy to decrease subsequent hospitalization and to preserve medical capacity which ensure timely and adequate medical resources for patients with severe COVID-19. However, the effectiveness of BAM/ETE is challenged by the rapidly emerging variants of SARS-CoV-2. The application course of BAM/ETE is a reminder of the importance of genomic surveillance, and the repurposed use of BAM/ETE may provide a potential treatment regimen in treating future variants.

Supplemental Information