Polyclonal and Monoclonal Antibodies for the Treatment and Prevention of Influenza

Abstract

While at least 8 monoclonal and 3 polyclonal antibody products have been tested in clinical trials for the treatment of influenza, no products have been licensed, and most have stopped clinical development. The COVID-19 pandemic demonstrated that these approaches, especially monoclonal antibodies, may have unique potential in certain stages of disease and populations, especially in preventing severe disease in a population without preexisting immunity or in those with a limited capacity to mount an effective humoral immune response. This review summarizes past and ongoing efforts in using monoclonal and polyclonal antibodies for the treatment and prevention of influenza, focusing on products that have entered clinical trials and drawing lessons from COVID-19 to direct future efforts on these strategies.

The value of preformed antibodies to treat infectious diseases and toxin-mediated diseases has been recognized for >135 years, beginning when Behring and Kitasato demonstrated that sera from rabbits infected with Clostridium tetani conferred protection to naive mice against live Tetanus bacilli and against tetanus toxin [1, 2]. The use of polyclonal serum for the treatment of bacterial diseases largely lost favor with the discovery of antibiotics, but products did exist for several unique indications, such as the prevention of rabies, diphtheria, tetanus, and hepatitis B [3]. In 1986, the lymphocyte-targeting monoclonal antibody (mAb) OKT3 became the first mAb approved by the US Food and Drug Administration (FDA) for use in humans [4]. Thirteen years later in 1999, palivizumab, the first mAb for an infectious disease indication, was approved to prevent RSV infection [5]. In 2017, bezlotoxumab, a Clostridium difficile toxin B–binding mAb, became the second [6]. With the emergence of oseltamivir-resistant influenza [7] and the H1N1 influenza pandemic of 2009, there was renewed interest in the study of antibody therapeutics (immune plasma, immune globulin for intravenous use [IGIV], and mAbs) for the treatment of influenza [8]. The commercial success of palivizumab, coupled with the clinical need, led to the development of multiple influenza A–targeting monoclonal antibodies. However, as reviewed in this article, clinical trials with these agents have been negative or inconclusive, and no antibody-based therapeutic has gained FDA approval to date for the treatment of influenza, dampening enthusiasm for further investment [9].

As the field was undergoing this retrenchment, the COVID-19 pandemic dramatically changed perspectives. The COVID-19 pandemic saw the most extensive use of mAbs against a single infectious disease [10]. At least 8 anti–SARS-CoV-2 mAbs, alone or in combination, were tested in clinical trials. mAbs were most successful in an outpatient population with COVID-19 when the outcome of interest was reducing progression to hospitalization [10]. With COVID-19, when the virus was generally susceptible to the monoclonal therapy in question, treatment of outpatients was associated with 50%–80% reductions in 28-day rates of hospitalization or death. In the RECOVERY trial, treatment of COVID-19 for patients admitted to the hospital with casirivimab + imdevimab had a striking bimodal response: in patients without a detectable serologic response to COVID-19 at admission (32% of the population), the relative risk of death within 28 days was 0.79 (95% CI, .69–.91), an absolute mortality reduction from 30% to 24%. However, among patients who were seropositive at baseline (54% of the study population), there was no significant difference in mortality (19% vs 21%). Similarly, in the CANOPY trial, which evaluated prophylactic administration of pemivibart to patients with underlying moderate to severe immune compromise, the incidence of symptomatic SARS-CoV-2 infection (confirmed by reverse transcription–polymerase chain reaction [RT-PCR]), hospitalization, or death was reduced by 84% (95% CI, 61%–94%), an overall reduction from 11.9% among placebo recipients to 1.9% in pemivibart-treated patients.

There was little impact of any of these mAbs in the treatment of hospitalized patients with COVID-19 unless administered to individuals unable to mount a humoral immune response [11–13]. Tixagevimab plus cilgavimab also demonstrated efficacy in pre- and postexposure prophylaxis [14].

Within 2 years of the onset of the pandemic, the FDA granted emergency use authorization (EUA) to multiple mAbs for the prevention or treatment of COVID-19 in the outpatient setting [15]. While some mAbs granted EUA failed quickly due to the rapidly evolving landscape of SARS-CoV2 antigenic and strain diversity, certain mAbs or combinations of mAbs that were deliberatively selected to target conserved regions of the SARS-CoV-2 spike protein proved more resilient to viral escape variants [10]. It has been noted that combination mAbs had a longer shelf life than single mAbs due to their ability to create a higher barrier for resistance with 2 products [10]. In general, though, the rapid and ongoing evolution of SARS-CoV-2 throughout the pandemic proved too challenging for drug developers to overcome, despite the search for highly conserved and constrained target epitopes on spike proteins and their receptor-binding domains.

Antibody engineering extended the half-life of some of the mAbs in development. Tixagevimab/cilgavimab found clinical favor for prophylactic use in part due to its very long half-life of 90 days. Last, while widely touted, convalescent plasma failed to demonstrate clear efficacy [16], though an EUA still exists to allow the use of plasma for the treatment of COVID-19 in patients with immunosuppressive disease or immunosuppressive treatment.

As of this writing, only a single previously granted FDA EUA for COVID-19 is still in effect: pemivibart (Pemgarda) for preexposure prophylaxis in patients at high risk for severe disease (immune compromise or vaccine nonresponse; EUA granted March 2024 and renewed September 2024). Nonetheless, important lessons that may apply to the potential treatment of emergent influenza A strains with antibody-based therapeutics can be gleaned from this experience. Accordingly, effective monoclonals for life-threatening viral infection of the respiratory tract might be expected to

• Reduce hospitalization or death when administered to nonhospitalized patients with early disease who are at high risk

• Reduce symptomatic disease, hospitalization, or death when administered as preexposure prophylaxis to patients at high risk

• Reduce mortality when administered after hospital admission to patients at high risk who have not yet generated a COVID-19–specific antibody response or perhaps may not generate one due to an underlying condition

Additionally, these findings are most likely to be demonstrable in a disease such as COVID-19 with high morbidity and mortality. Insights from clinical trials of anti–SARS-CoV-2 mAbs may inform interpretation of some of the influenza therapeutic trials conducted to date. This review summarizes past and ongoing efforts in the use of monoclonal and polyclonal antibodies for the treatment and prevention of influenza, focusing on products that have entered clinical trials.

HIGH-TITER IMMUNE PLASMA FOR INFLUENZA A THERAPY

Studies from the 1918 influenza A/H1N1 pandemic reported that transfusion of influenza-convalescent human blood products reduced mortality in patients with influenza complicated by pneumonia [17]. A meta-analysis that reviewed reports from the 1918 influenza A/H1N1 pandemic suggested that early administration of convalescent blood products reduced the risk of death from pneumonia from 37% to 16% (95% CI, 15%–27%) [18]. With the emergence of high-pathogenicity avian influenza, there were reports of patients with severe influenza A/H5N1 or A/H7N9 receiving plasma therapy and achieving good outcomes [19, 20].

It was not until the 2009 A/H1N1 influenza pandemic that groups systematically studied this intervention. A prospective multicenter case-control study evaluated the use of convalescent plasma for severe influenza A/H1N1/pdm09 (Table 1). Ninety-three participants who required admission to the intensive care unit (ICU) while receiving oseltamivir therapy were enrolled, and all were offered immune plasma [21]. Twenty participants who accepted the intervention had a mortality of 20.0% as opposed to 54.8% in the 73 participants who declined (P = .01). This observation, while compelling, is confounded by the fact that mortality in the control arm was significantly higher than anticipated for similar severity of illness [22].

Table 1.

Polyclonal Antibody Preparations That Advanced to Clinical Trials for the Treatment of Influenza

Product	Latest Phase of Clinical Development	Clinical Studies
Immune plasma	Phase 3	NCT01306773
		NCT01052480
		NCT02572817
Immunoglobulin (IGIV)	Phase 3	NCT01617317
		NCT02008578
		NCT02287467
		NCT03315104
Equine F(ab')2 (FBF001)	Phase 1	NCT02295813

Abbreviation: IGIV, immune globulin for intravenous use.

This was followed by a randomized phase 2 study in which hospitalized participants with severe influenza A or B (defined as the presence of hypoxia or tachypnea) were assigned to either high-titer immune plasma (antibodies titers at least 1:40) plus standard care, most commonly oseltamivir therapy, or standard care alone. This trial had a nonsignificant trend toward benefit in the primary end point of normalization of respiratory status by day 28 (67% vs 53%, P = .069) [23]. Multiple secondary end points were suggestive of efficacy, including fewer days in the hospital (median, 6 vs 11; P = .13), fewer participants with ICU admissions (57% vs 69%; P = .097), fewer days undergoing mechanical ventilation (median, 0 vs 3; P = .14), and better clinical status at day 7 (among 6 options in descending order of severity: death, ICU level of care, hospitalized with oxygen, hospitalized without oxygen, not hospitalized but did not return to normal activities, or not hospitalized and did return to normal activities; P = .020). Participants receiving plasma therapy had fewer serious adverse events than recipients of standard care alone (20% vs 38%, P = .041). This trial was limited by its size and the fact that the intervention was unblinded.

A subsequent randomized, double-blind, phase 3 trial was conducted in which children and adults hospitalized with severe influenza A infection (H1N1 or H3N2) were randomized to receive either high-titer immune plasma (hemagglutination inhibition [HAI] titer of at least 1:80 toward the seasonal strains) or low-titer plasma (HAI ≤1:10) [24]. Almost all received concurrent neuraminidase inhibitor treatment. There was no demonstrated benefit to the high-titer immune plasma, with a proportional odds ratio (OR) of 1.22 (95% CI, .65–2.29; P = .54) for improved clinical status on day 7. There was no demonstrated benefit on multiple secondary end points, including duration of hospitalization or ICU admission, mechanical ventilation, and acute respiratory distress syndrome. Notably, 43% of all patients were already admitted to medical ICUs at enrollment and treatment, underscoring the difficulty of demonstrating an impact of antibody therapeutics in advanced influenza disease. The volume of plasma administered in both studies increased the antibody geometric mean HAI titer from 1:29 to 1:63 and from 1:26 to 1:46 [23, 24]. While an HAI titer of 1:40 is generally associated with 50% protection from clinical influenza [25], the increase in titer to >1:40 may have been insufficient to modify established severe disease.

Although randomized controlled trials of convalescent plasma therapy for influenza have not demonstrated sufficient evidence for efficacy to lead to endorsement of this strategy, there may be clinical circumstances in which efficacy might be shown (eg, patients with IgG or IgG subclass deficiency). This potential utility would have to be weighed against the risks of plasma therapy, which may include allergic reaction, volume overload, and transfusion-related acute lung injury and thus be difficult to differentiate from influenza-associated acute respiratory distress syndrome [26].

HIGH-TITER IMMUNE GLOBULIN FOR INFLUENZA THERAPY

Polyclonal immune globulin for intravenous use (IGIV) has several benefits over plasma treatment: ABO blood matching is not required, and it provides a more uniform product as compared with individual plasma units, with differing antibody titers and epitope targeting. The plasma used for IGIV manufacturing can be selected for attributes such as high-titer neutralizing antibodies, referred to as high-titer immune IGIV.

A randomized controlled trial was conducted in participants (n = 35) hospitalized with severe H1N1 influenza in Hong Kong (Table 1; NCT01617317), which compared treatment with high-titer IGIV with nonimmune (prepandemic) IGIV [27]. Among those receiving IGIV, 29% died as compared with 24% of those receiving standard IGIV (not significant). Subgroup multivariate analysis of the 22 participants who received treatment within 5 days of symptom onset demonstrated that IGIV treatment was the only factor that independently reduced mortality (OR, 0.14; 95% CI, .02–.92; P = .04).

A randomized, double-blind, placebo-controlled clinical trial enrolled 313 participants with severe influenza A or B who were receiving treatment with neuraminidase inhibitor therapy, randomizing them to receive standard care plus either a single 500-mL infusion of high-titer IGIV (0.25 g/kg of body weight) or saline placebo (NCT02287467) [28]. Those receiving IGIV treatment had a similar clinical status ordinal outcome at day 7 (OR, 1.25; 95% CI, .79–1.97; P = .33). Notably, approximately 50% of enrolled patients had symptom onset >3 days prior to enrollment. The authors concluded that passive immunotherapy as an adjunctive therapy for adults hospitalized with severe influenza A does not provide clinical benefit. The authors noted a possible clinical benefit in influenza B (OR, 3.19; 95% CI, 1.21–8.42), but these findings should be interpreted with caution, as this benefit was mainly restricted to the Yamagata lineage (a lineage no longer circulating) and there was no increase in antibody titer to influenza B after IGIV administration as compared with placebo.

Another trial sponsored by an IGIV manufacturer was designed as a safety and exploratory efficacy trial (NCT03315104). This trial demonstrated that IGIV was safe (fewer moderate and severe adverse events) [29]. The exploratory efficacy end point of clinical status (by ordinal scale) on day 8 demonstrated that more participants receiving the high-titer IGIV were no longer hospitalized and had resumed normal activities than those receiving the placebo. However, this was statistically nonsignificant given the sample size (19–21 per arm).

EQUINE F(AB')₂ (FBF-001)

A different strategy for reducing reactions to animal proteins is the peptic digestion of equine IgG to remove the Fc portion of the molecule, which is thought to be responsible for some adverse reactions [30]. This results in F(ab')₂ fragments, which can be purified and administered intravenously. Polyclonal horse sera following H5N1 vaccination were used to produce a product (FBF-001) that demonstrated survival benefit efficacy in a ferret model of H5N1 infection and synergy with oseltamivir [31]. In a phase 1, placebo-controlled, single and multidose trial, 13 healthy volunteers received the anti-H5N1 FBF-001 [32]. In the multidose study, the half-life ranged from 7.6 to 15.7 hours, which was, as one would predict, significantly less than full human IgG. No current trials are listed on ClinicalTrials.gov or the company website.

MONOCLONAL ANTIBODY THERAPEUTICS

Monoclonal antibodies offer several advantages over polyclonal antibodies. Common epitopes can be targeted, such as those on all group 1 influenza hemagglutinins (HAs; including H1 and H5) and group 2 (including H3 and H7); alternatively, for selected mAbs (eg, MEDI8852, VIR2482, and VIS410), broadly active binding and inhibition of HA-mediated fusion may be achieved by targeting highly conserved and constrained HA stalk epitopes common to all influenza A HA subtypes (Figure 1). Production of mAbs at a commercial scale is expected to be less expensive than polyclonal options and may be more reliably and rapidly manufactured.

Figure 1.

Two-dimensional representation of an influenza A hemagglutinin prior to HA1/HA2 cleavage and fusion intermediate transformation. The HA homotrimer is comprised of three monomers. The globular head (top of image) comprised of three identical sequences derived from HA1, contains the sialic-acid binding residues. The coiled coil structure of α-helices derived from the three HA2 components of the trimer provide HA stem structure. Two of the three largely hidden fusion peptides of the HA2 N-termini are visible in this view, in the lower pole of the superimposed oval.Highly conserved sites identified in a bioinformatic analysis of over 26,500 HA sequences (GenBank and EpiFlu) were analyzed to determine site entropy and amino acid networking for H1N1 viruses. The most highly networked sites (in general, both conserved and constrained within and across HA subtypes) were identified in the HA stem region roughly identified by the superimposed oval. HA stem binding broadly neutralizing MAbs generally bind epitopes in this region. Analysis and figures from internal data, Visterra, Inc.

Monoclonal antibodies may be derived from memory B cells of patients who are convalescent or from animals after direct inoculation with the virus. Animal-derived mAbs are subsequently “humanized” to reduce their immunogenicity, increase their serum half-life, and improve their effector functions if desired. Most clinically evaluated, broadly neutralizing anti–influenza A antibodies bind on the stem region of HA, locking it in its prefusion conformation and preventing the structural changes required for fusion with the cell membrane [33]. In addition, intact immunoglobulin molecules may engage Fc receptors on immune cells, enabling antibody-dependent cellular cytotoxicity and cellular phagocytosis, as well as complement activation [34].

Antibodies to neuraminidase, the other major influenza cell surface glycoprotein, contribute greatly to protection from and response to infection. These antibodies may function through direct inhibition of neuraminidase (sialidase) activity or through induction of antibody-dependent cellular cytotoxicity or cellular phagocytosis [35, 36]. Broadly protective human monoclonal antibodies that target the sialidase activity of influenza A and B viruses have been isolated and characterized [37] but have not advanced, to our knowledge, into the clinic.

CT-P27

CT-P27 (Celltrion) is a dual-antibody product containing CT149 and CT120 [33]. CT149 is reported to neutralize all group 2 viruses and some group 1 viruses [33]; CT120 has been shown to broadly neutralize all group 1 viruses but has no activity against group 2 viruses. When combined with CT-P27, this product has demonstrated the ability to neutralize viruses from 11 subtypes of group 1 and 2 influenza A viruses, including seasonal strains H1N1 and H3N2, as well as avian strains such as H5N1 and H7N9 [33].

A phase 1 dose-escalating safety and pharmacokinetic study of CT-P27 was conducted in 2013 (Table 2; South Korea Clinical Research Information Service KCT0001179). While no publication was found in peer-reviewed literature, the company announced that doses up to 20 mg/kg were safe and that CT-P27 had a half-life of 6 days [38]. This half-life is shorter than anticipated for a fully human mAb with no human targets. A phase 2 study of 10 or 20 mg/kg of CT-P27 was conducted with a controlled human infection model in healthy volunteers (NCT02071914), demonstrating reduced viral shedding as measured by quantitative PCR from nasopharyngeal swab in the CT-P27–treated arms.

Table 2.

Monoclonal Antibodies That Advanced to Clinical Trials for the Treatment of Influenza

Compound	Manufacturer	Latest Phase of Clinical Development	Prior Clinical Studies	Ongoing Clinical Trial (Yes/No)	Year of Most Recent Clinical Trial
CT-P27	Celltrion	Phase 2	KCT0001179	No	2019
			NCT02071914
			NCT03511066
			KCT0002211
CR6261	Crucell	Phase 2	NCT01406418	No	2018
			NCT02371668
CR8020	Crucell	Phase 2	NCT01756950	No	2014
			NCT01938352
FGI-101-1A6	Functional Genetics	Phase 1	NCT01299142	No	2011
MEDI8852	MedImmune	Phase 2	NCT02350751	No	2016
			NCT02603952
MHAA4549A	Genentech	Phase 2	NCT01877785	No	2017
			NCT02284607
			NCT01980966
			NCT02293863
			NCT02623322
TCN-032	Theraclone Sciences	Phase 2	NCT01390025	No	2012
			NCT01719874
VIR-2482	Vir Biotechnology	Phase 2	NCT04033406	No	2023
		Phase 3	NCT05567783
VIS410	Vistera	Phase 2	NCT02045472	No	2018
			NCT02468115
			NCT02989194
			NCT03040141

A phase 2 study of 228 participants with acute uncomplicated influenza (NCT03511066) demonstrated a 2-day shorter time to resolution of influenza symptoms and fever: 3.74 days (95% CI, 2.72–4.46) for CT-P27 at 90 mg/kg, 3.69 days (95% CI, 2.81–4.25) for CT-P27 at 45 mg/kg, and 5.72 days for placebo (95% CI: lower limit, 3.65; upper limit, not calculated due to small sample size) [39]. A phase 2B trial for CT-P27 (KCT0002211) is listed as ongoing but has not been updated since 2019. CT-P27 is no longer listed as in development on the company website.

CR6261, CR8020, AND CR9114

CR6261 (developed by Crucell, now part of Janssen Pharmaceuticals) binds a highly conserved epitope in the HA stem and broadly neutralizes group 1 influenza viruses [40, 41]. CR8020 also binds in the HA stem and has activity against group 2 viruses [42]. A phase 1 study of CR6261 at doses up to 50 mg/kg (Table 2; NCT01406418) was not published but reported to be safe [43]. A phase 2 placebo-controlled study examined CR6261 dosed intravenously at 50 mg/kg, 24 hours after H1N1 nasal inoculation in a controlled human infection (challenge) model (NCT02371668); it showed that CR6261 had no statistically significant effect on viral RNA area under the curve (AUC) by quantitative RT-PCR: 48.56 log copies/mL × days (IQR, 202) vs 25.53 (IQR, 155) for placebo (P = .315) [43]. There was no clinically important effect on influenza disease measures, including number of symptoms, duration of symptoms, or influenza patient-reported outcome scores. Based on these observations in healthy young adults administered a dose of virus predicted to result in influenza A infection of minimal to moderate severity in 60% of individuals, the authors concluded that monoclonal antistalk antibody therapy may be of limited efficacy for disease prevention. CR6261 appears to no longer be in development.

A phase 1 study of CR8020 at doses up to 50 mg/kg was completed in 2013 but has not been reported (NCT01756950). A phase 2 study of CR8020 at 15 mg/kg was completed in 2014 (NCT01938352). Twenty-two participants were administered 15 mg/kg, 2 days before a controlled human infection model with influenza H3N2 virus. Of 19 participants included in the efficacy analysis, 11 who were administered CR8020 had an AUC of 73 113 ± 155 589 (units not listed but typically reported as log-transformed copies/mL × days) as opposed to 28 ± 78 for the 9 participants receiving placebo [44]. Of those receiving CR8020, 54% had ≥4 qualitative PCR–positive nasopharyngeal swabs in contrast to 0% in the control arm, making the overall interpretation of this trial's results problematic. Symptom score AUC was also more pronounced in those receiving CR8020: 7.0 ± 7.9 vs 2.3 ± 3.7 (mean ± SD). The significant standard deviation in the CR8020 AUC suggests the presence of a few large outliers. A better understanding of the basis for these outliers and any relation to drug effect is critical to understanding the safety of CR8020.

CR9114 (Leyden Laboratories) is a broadly neutralizing IgG1 mAb that can bind HA stalk epitopes of influenza A groups 1 and 2 and influenza B viruses. Intranasally administered CR9114 protected ferrets from lethal H5N1 infection at doses as low as 4 µg [45]. Intriguingly, given the low doses demonstrated to be efficacious in this preclinical model, the drug's developers propose that it may have utility as an intranasally administered prophylactic agent in the setting of, for instance, an emergent human H5N1 pandemic to protect health care workers and populations at high risk in parallel with vaccine development and deployment.

FGI-101-1A6

A different approach to monoclonal therapeutics targets cellular markers on infected cells. Tumor susceptibility gene 101 (TSG101) is exposed at the surface of influenza-infected cells but remains intracellular in uninfected cells [46]. TSG101 expression interacts with M1 to facilitate release from infected cells, and binding of TSG101 with antibodies prevents the release of influenza virus from infected cells. FGI-101-1A6 (Functional Genetics) is a fully human IgG1 antibody targeting a universally conserved epitope of TSG-101.

A phase 1 study of FGI-101-1A6 in healthy volunteers included 6 ascending-dose cohorts from a starting single dose of 0.0017 mg/kg to 10 mg/kg (Table 2; NCT01299142). FGI-101-1A6 was well tolerated by all dose levels, although 3 participants had a grade 3 increase in serum lipase [47]. The mean half-life was 170 to 287 hours, shorter than a typical monoclonal with no other identified intrinsic targets, suggesting some binding in the absence of influenza infection. This monoclonal appears to be no longer under development.

MEDI8852

MEDI8852 (FI6; Medimmune/Astra Zeneca) is a broadly reactive antibody capable of binding HA group 1 and 2 viruses and optimized by affinity maturation to better bind H3 and H1 HA proteins [48]. A phase 1 study was conducted in 4 cohorts with doses up to 3000 mg or placebo control (Table 2; NCT02350751). The most frequently reported adverse events were headache and hypoglycemia, occurring in similar proportions in MEDI8852 and the placebo controls [49]. The terminal half-life ranged from 19.4 to 22.6 days.

A phase 2 treatment study was completed in December 2016 in which 126 low-risk adults with acute influenza were randomized to receive MEDI8852 (750 mg) plus oseltamivir, MEDI8852 (3000 mg) plus oseltamivir, MEDI8852 (3000 mg) alone, or a standard course of oseltamivir alone (NCT02603952). The median number of days that participants had symptoms was similar in all groups (10 days) [50]. Viral shedding by qualitative and quantitative PCR was similar in all cohorts on days 1, 3, 5, and 7. Overall rates of AEs were reported at 41.9% for all arms that included MEDI8852 vs 31.3% for oseltamivir alone, with bronchitis (11.8% vs 3.1%) and diarrhea (6.5% vs 0%) most commonly reported. Monotherapy with 3000 mg of MEDI8852 had equivalent efficacy to oseltamivir alone, but there was no clear additive efficacy when given in combination with oseltamivir. This study lacked a placebo arm.

Emergence of resistance mutations during treatment was not observed or reported in clinical trials (ie, HA stem amino acid changes that reduced MEDI8852 efficacy). Promising preclinical data support evaluation of MEDI8852 for avian influenza prophylaxis in humans, as described in the Discussion section. At the present time, this agent is no longer in clinical development.

MHAA4549A

MHAA4549A (Genentech) is a mAb that binds to a highly conserved epitope on the stalk of HA and can neutralize H1, H2, H3, H5, and H7 viruses [51]. Two phase 1 studies were conducted with weight-based doses up to 45 mg/kg (Table 2; NCT01877785) or a fixed dose of 10 800 mg (eg, 135 mg/kg for an 80-kg participant; NCT02284607). The most common adverse events were headaches, occurring in similar proportions in those receiving MHAA4549A and placebo controls [52]. MHAA4549A was demonstrated to have a mean half-life of 22.5 to 23.7 days.

A phase 2 study was conducted in 100 participants inoculated intranasally with influenza A/H3N2 (A/Wisconsin/67/2005) and administered a single intravenous dose of MHAA4549A (400, 1200, or 3600 mg), standard oral dose of oseltamivir, or placebo, starting 24 hours after inoculation. In the infected population as compared with the placebo group, 3600 mg of MHAA4549A had a 97.5% reduction in AUC quantitative PCR (11 vs 458 log₁₀ copies [viral particles]/mL × h, P = .005) [53]. However, the 1200-mg cohort had only a 3% reduction (444 log₁₀ copies/mL × h, P = .902), and the 400-mg cohort had a 46% reduction (247 log₁₀ copies/mL × h, P = .046). Oseltamivir had an 87% reduction as compared with placebo (57 log₁₀ copies/mL × h, P = .059). The composite symptom score was 37.3 in the 3600-mg cohort (82% reduction, P = .29), 192.1 for the 1200-mg cohort (8% reduction, P = .87), 87.5 for the 400-mg cohort (58% reduction, P = .20), 8.1 for the oseltamivir cohort (96% reduction, P = .09), and 207 for placebo.

A phase 2 study of 168 participants hospitalized with influenza A was conducted comparing MHAA4549A (3600 mg) + oseltamivir, MHAA4549A (8400 mg) + oseltamivir, and oseltamivir alone + placebo control [54]. The primary efficacy end point was time to normalize respiratory function (ie, no longer requiring supplemental oxygen with saturations >95%). The MHAA4549A (8400 mg) + oseltamivir cohort had a median time to respiratory normalization of 2.7 days vs 4.3 days for oseltamivir alone (P = .20) and 2.8 in the 3600-mg cohort (P = .61). However, 30-day mortality was 9.1% in the 8400-mg cohort, 7.7% in the 3600-mg cohort, and 5.6% in the oseltamivir monotherapy control [54].

An additional phase 2 study of uncomplicated influenza was conducted in a low-risk, nonelderly, ambulatory population [55]. Outpatient participants with a confirmed diagnosis of influenza (n = 124) were randomized 1:1:1 to receive a placebo or MHAA4549A at a single intravenous dose of 3600 or 8400 mg. No participants required hospitalization, and no mortality was observed. Median time to alleviate influenza symptoms was shorter in placebo recipients (117 hours) than MHAA4549A recipients (154 and 146 hours for the 3600- and 4800-mg dosing groups, respectively). The Genentech website indicates no evidence of further development of MHAA4549A.

TCN-032

TCN-032 (Theraclone Sciences) targets the ectodomain of the matrix protein 2 (M2e). Amino acid residues 1 to 9 and 239 to 252 of M2e are encoded by overlapping nucleotides in the same reading frame of M1 and thus are very stable across human influenza viruses; additionally, infection and current vaccines do not induce antibodies to M2e, minimizing antigen drift pressures [56]. In preclinical experiments, influenza A matrix protein 2 ectodomain escape mutants have emerged under antibody selection pressure, raising some concern with this strategy [57]. Additionally, the M2e is well conserved among influenza viruses, but avian influenza viral M2e differs from that of human by several amino acids [58]. The impact on neutralization of avian variants, however, is not known.

A phase 1 study of TCN-032 was conducted at doses up to 40 mg/kg. Adverse events were similar in those receiving TCN-032 and placebo; the pharmacokinetic evaluation demonstrated a half-life of approximately 15 days [59]. A phase 2 study was conducted in a controlled human infection with influenza A/H3N2 (Wisconsin/67/2005), and TCN-032 was administered 24 hours after infection [60]. The primary end point (ie, the percentage of participants with any grade ≥2 influenza symptom or pyrexia between days 1 and 7) was similar in the TCN-032 (35%) and placebo (48%) cohorts (P = .14) [60]. Participants treated with TCN-032 showed a 35% reduction in median total symptom AUC (P = .047) and a 2.2-log reduction in median viral load quantitative PCR AUC (P = .09) as compared with placebo control. While promising, these results did not meet the trial's prespecified primary end point. TCN-032 has not had any further development. Multiple additional M2e candidate monoclonals have been identified, but none have advanced into clinical trials at this time [61].

VIR-2482

VIR-2482 (Vir Biotechnology) was derived from MEDI8852 and includes introduction of the “LS mutation” (M428L/N434S) into the Fc region of MEDI8852 to extend elimination half-life through increased neonatal Fc receptor–mediated antibody recirculation. VIR-2482 targets a conserved HA stem epitope with broad reactivity to HA proteins from all 18 influenza A subtypes. A phase 1 pharmacokinetic trial in healthy volunteers demonstrated 57- to 71-day serum elimination half-life after intramuscular injection, with nasopharyngeal concentrations of 2% to 5% that of serum [62]. Overall tolerability was acceptable. Recognizing the difficulty of achieving timely delivery of mAb therapies for established influenza infection, VIR-2482 was advanced, instead, into a phase 3 prophylaxis trial (PENINSULA) in which unvaccinated healthy adults (n = 2977) were randomized to receive VIR-2482 at doses of 450 mg or 1200 mg or placebo via intramuscular injection [63]. The primary efficacy end point was the proportion of participants with RT-PCR positivity for influenza A and protocol-defined influenza-like illness. These proportions did not differ among treatment arms (2.54%, 2.45%, and 2.12% for placebo, 450-mg, and 1200-mg arms, respectively), with a nonsignificant relative risk reduction of 15.9% for the 1200-mg VIR-2482 arm vs placebo. Promisingly, in a predefined subgroup analysis based on criteria defined by the Centers for Disease Control and Prevention for influenza-like illness (temperature >37.8 °C with sore throat or cough and RT-PCR confirmation of infection), the relative risk reduction for the 1200-mg VIR-2482 group vs placebo was 57.2% (95% CI, −2.5 to 82.2). In addition, exploratory analyses indicated that the higher dose of VIR-2482 was associated with reduced time to resolution of protocol-defined influenza-like illness symptoms (median, 85.8 vs 111.9 hours in placebo).

Emergence of resistance mutations during treatment was not observed or reported in clinical trials (ie, HA stem amino acid changes that reduced VIR2482 efficacy). This monoclonal is not progressing in clinical development at this time.

VIS410

VIS410 (Visterra) is a broadly neutralizing mAb engineered to bind to HA stem epitopes of group 1 and 2 influenza A strains, inhibiting HA-mediated membrane fusion and inducing antibody-dependent cell-mediated cytotoxicity. In murine H3N2 and H7N9 infection models, VIS410 alone or in combination with oseltamivir mitigated disease severity, reducing pulmonary viral load and improving acute respiratory distress syndrome pathophysiology [64, 65].

A phase 1 study of VIS410 with doses up to 50 mg/kg showed that drug exposure was dose proportional, with a mean half-life of 12.9 days. Mean VIS410 C_max levels in the upper respiratory tract were 20.0 and 25.3 μg/mL at the 30- and 50-mg/kg doses, respectively. Similar proportions of those who received VIS410 and placebo overall reported adverse events (67% vs 64%), with more gastrointestinal events in participants who received VIS410 (33%; mainly diarrhea) as opposed to 0% for placebo (P < .05) [66]. While the etiology of the gastrointestinal events was unclear, pretreatment with single-dose oral diphenhydramine (50 mg) combined with 600 mg of ibuprofen mitigated these effects.

In a phase 2 controlled human infection study with influenza A(H1N1)pdm09, there was a 76% decrease in the virus AUC by quantitative RT-PCR, a 91% reduction in median viral load AUC by TCID₅₀ (50% tissue culture infectious dose; P = .019), and a >2-log₁₀ reduction in the mean peak virus load by quantitative RT-PCR and TCID₅₀ measurement [67]. The median viral load was undetectable sooner, and the median time to resolution of symptoms was shorter by around 2 days.

A phase 2, double-blind, randomized controlled enrolled 150 participants with acute uncomplicated influenza (H3N2 in 92% of confirmed infections), with 1:1:1 randomization to receive a placebo or VIS410 at a 2000- or 4000-mg dose by intravenous infusion [68]. Treatment-emergent adverse events, most commonly diarrhea of mild severity, were dose related (occurring in 20%, 35%, and 55%, respectively). VIS410 was associated with reduced median nasopharyngeal viral load TCID₅₀ AUC_Day7 (3.66 pooled VIS410 vs 4.78 placebo, P = .08), and the difference was more pronounced in the subset of patients with baseline HAI titers ≤40 (4.2 pooled VIS410 vs 6.2 placebo, P = .009). Kaplan-Meier estimate of the time to resolution of viral shedding was reduced (1.9 vs 3.6 days, P = .03) in patients treated with VIS410. Baseline influenza patient-reported outcome symptom scores were balanced among groups, with lower mean scores by days 3 and 4 in the pooled VIS410 treatment groups (P = .023) [68].

This study was followed by a global phase 2 randomized controlled trial evaluating VIS410 (2000 or 4000 mg) or placebo in combination with oseltamivir in hospitalized adults with influenza A requiring oxygen therapy (NCT03040141) [69]. The trial aimed to enroll 150 participants, but funding restrictions forced an early termination, with enrollment of 89 patients (H3N2 in 54%). Time since onset of influenza was >72 hours in 52% of enrolled participants. At baseline, the VIS410 arms had more participants in the ICU undergoing mechanical ventilation (3.6% placebo vs 12.3% pooled VIS410) and more patients in the ICU not undergoing mechanical ventilation (10.7% placebo vs 24.6% pooled VIS410), a highly significant disparity (P = .037). The day 7 ordinal scale OR, the primary end point of clinical status, was 1.4 favoring placebo (P = .435). Time to normal oxygenation (room air O₂ saturation >94%) was equivalent between treatment arms (97 vs 86 hours, P = .88; pooled VIS410 vs placebo). In a subgroup analysis of those hospitalized and receiving oxygen therapy but not admitted to the ICU, VIS410 recipients demonstrated a faster time to normal oxygenation (46.4 vs 82.5 hours, P = .133). Among participants with a positive result on baseline nasopharyngeal viral culture, day 3 culture negativity rates were significantly improved in pooled VIS410 recipients (78% vs 53%, P = .047). Treatment-emergent adverse events occurred in 13.8% of VIS410 recipients vs 6.7% among placebo recipients, with the majority of events being mild to moderate diarrhea or nausea among VIS410 recipients [69].

Emergence of resistance mutations during treatment was not observed (ie, HA stem amino acid changes that reduced VIS410 epitope binding or antiviral efficacy), nor were emergent potential resistance-conferring HA stem sequence variants observed in global surveillance sequence monitoring. VIS410 is not in active clinical trials at this time.

FUTURE DIRECTIONS FOR INFLUENZA ANTIBODIES

Our effort herein has been to review the status of polyclonal and monoclonal antibody therapeutics to determine if they might be of benefit in the context of the next influenza pandemic. The answer, at present, cannot be more positive than “perhaps.” However, as informed by insights gained from the COVID-19 pandemic, the answer may become “almost certainly” if we select the proper context for their development and indication for use.

We have learned from multiple phase 2 studies of hospitalized patients with moderate to severe influenza that mAb therapeutics are of limited utility if administered more than 72 to 96 hours after the onset of symptoms and that observed benefits in subgroup analyses are greater in patients with disease that has not advanced to ICU admission or mechanical ventilation at the time of enrollment. These observations are generally consistent with findings in COVID-19 trials. This poses the challenge of how antibody therapeutics that are administered intravenously can be delivered efficiently early in the treatment of disease outside a hospital setting. This problem has not been solved in the clinical trials reported to date.

An alternative strategy is to reserve the use of influenza monoclonals for the early treatment of patients at high risk (immunosuppressed or elderly) for whom clinical algorithms for management might be adopted in a pandemic setting that are not practical during an interpandemic period. Even more attractive is the potential for prophylaxis of infection in patients at high risk. This strategy was tested recently in a lethal H5N1 avian influenza challenge study in macaques [70], in which the broadly neutralizing mAb MEDI8852 was demonstrated to provide complete protection from lethality and a substantial reduction of disease severity when administered intravenously at a dose of 30 mg/kg, 3 days prior to virus challenge. Based on pharmacodynamic analyses, the authors predicted that a single 30-mg/kg infusion of MEDI8852 in macaques would provide approximately 8 weeks of protection from severe disease or death. These data are consistent with prior observations of MEDI8852 prophylaxis in H5NI and H7N9 avian flu challenge studies in mice and ferrets [71].

The mAb prophylaxis strategy was evaluated with VIR-2482, as described earlier. The trial was conducted, however, in unvaccinated healthy young adults in whom flu severity, if the infection was contracted, was predicted to be mild. Indeed, severe disease was neither expected nor observed. In addition, serum and nasopharyngeal mAb concentrations were low as compared with the MEDI8852 exposures in the macaque H5N1 challenge study. Thus, the relevance of this trial is limited when considering the potential utility of an intravenously administered mAb to protect highly at-risk populations in a pandemic setting.

Herein lies the challenge. Drug developers typically invest only in products with a predictable future market (eg, oseltamivir or baloxavir for seasonal influenza). In the absence of robustly supportive clinical trial data to date, such a scenario has not emerged for the use of mAbs for seasonal influenza in a nonpandemic setting, explaining the current paucity of clinical trials. However, experience with COVID-19 coupled with compelling preclinical prophylactic efficacy data for influenza suggests the potential to address unmet needs in the protection of highly at-risk patients in a pandemic setting. This may be achievable with intravenously administered therapy (particularly with monoclonals enhanced with Fc alterations that increase serum half-life) and plausibly with new medical devices that can deliver large volumes of antibody preparations subcutaneously, without the need for patient or staff to perform injections [72]. These data suggest that it would be prudent to keep therapeutic mAb strategies viable for future use against emergent influenza strains of high pathogenicity in human populations.

Notes

Acknowledgments. We thank Andrew Wollacott of Visterra for Figure 1.

Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health.

Supplement sponsorship. This article appears as part of the supplement “Advances in Influenza Therapeutics,” sponsored by Flu Lab, World Health Organization, the International Society for Respiratory Viruses, F. Hoffmann-La Roche AG, Shionogi & Co., Ltd., Cidara Therapeutics, Inc., Eradivir Inc, Leyden Laboratories, Gilead Sciences, and the International Federation of Pharmaceutical Manufacturers & Associations.