Determinants of passive antibody efficacy in SARS-CoV-2 infection

Abstract

SummaryBackgroundA large number of studies have been carried out involving passive antibody administration for the treatment and prophylaxis of COVID-19 and have shown variable efficacy. However, the determinants of treatment effectiveness have not been identified. Here we aimed to aggregate all available data on randomised controlled trials of passive antibody treatment for COVID-19 to understand how the dose and timing affect treatment outcome.MethodsWe analysed published studies of passive antibody treatment from inception to 7 January 2022 that were identified after searching various databases such as MEDLINE, Pubmed, ClinicalTrials.gov. We extracted data on treatment, dose, disease stage at treatment, and effectiveness for different clinical outcomes from these studies. To compare administered antibody levels between different treatments, we used data on in vitro neutralisation of pseudovirus to normalise the administered dose of antibody. We used a mixed-effects regression model to understand the relationship between disease stage at treatment and effectiveness. We used a logistic model to analyse the relationship between administered antibody dose (normalised to the mean convalescent titre) and outcome, and to predict efficacy of antibodies against different Omicron subvariants.FindingsWe found that clinical stage at treatment was highly predictive of the effectiveness of both monoclonal antibodies and convalescent plasma therapy in preventing progression to subsequent stages (p<0.0001 and p=0.0089, respectively, chi-squared test). We also analysed the dose-response curve for passive antibody treatment of ambulant COVID-19 patients to prevent hospitalisation. Using this quantitative dose-response relationship, we predict that a number of existing monoclonal antibody treatment regimens should maintain clinical effectiveness in infection with currently circulating Omicron variants.InterpretationEarly administration of passive antibody therapy is crucial to achieving high efficacy in preventing clinical progression. A dose-response curve was derived for passive antibody therapy administered to ambulant symptomatic subjects to prevent hospitalisation. For many of the monoclonal antibody regimens analysed, the administered doses are estimated to be between 7 and >1000 fold higher than necessary to achieve 90% of the maximal efficacy against the ancestral (Wuhan-like) virus. This suggests that a number of current treatments should maintain high efficacy against Omicron subvariants despite reduction in in vitro neutralisation potency. This work provides a framework for the rational assessment of future passive antibody prophylaxis and treatment strategies for COVID-19.FundingThis work is supported by an Australian government Medical Research Future Fund awards GNT2002073 and MRF2005544 (to MPD, SJK), MRF2005760 (to MPD), an NHMRC program grant GNT1149990 (SJK and MPD), and the Victorian Government (SJK). SJK is supported by a NHMRC fellowship. DC, MPD, ZKM and EMW are supported by NHMRC Investigator grants and ZKM and EMW by an NHMRC Synergy grant (1189490). DSK is supported by a University of New South Wales fellowship. KLC is supported by PhD scholarships from Monash University, the Haematology Society of Australia and New Zealand and the Leukaemia Foundation. TT, HW and CB are members of the National COVID-19 Clinical Evidence Taskforce which is funded by the Australian Government Department of Health.Research in contextEvidence before this studyWe identified randomised controlled trials (RCTs) evaluating the effectiveness of SARS-CoV-2-specific neutralising monoclonal antibodies, hyperimmune immunoglobulin and convalescent plasma in the treatment of participants with a confirmed diagnosis of COVID-19 and in uninfected participants with or without potential exposure to SARS-CoV-2. The RCTs were identified from published searches conducted by the Cochrane Haematology living systematic review teams.A total of 37 randomised controlled trials (RCT) of passive antibody administration for COVID-19 were identified. This included 12 trials on monoclonal antibodies, 21 trials of convalescent plasma treatment, and 4 trials of hyperimmune globulin. These trials involved treatment of individuals either prophylactically or at different stages of infection including post-exposure prophylaxis, symptomatic infection, and hospitalisation. The level of antibody administered ranged from a 250 ml volume of convalescent plasma through to 8 grams of monoclonal antibodies. Data for analysis was extracted from the original publications including dose and antibody levels of antibody administered, disease stage and timing of administration, primary outcome of study and whether they reported on our prespecified outcomes of interest, which include protection against symptomatic infection, hospitalisation, need for invasive mechanical ventilation (IMV) and death (all-cause mortality at 30 days).Added value of this studyOur study included data across all 37 RCTs of passive antibody interventions for COVID-19 and aggregated the studies by the stage of infection at initiation of treatment. We found that prophylactic administration or treatment in earlier stages of infection had significantly higher effectiveness than later treatment. We also estimated the dose-response relationship between administered antibody dose and protection from progression from symptomatic ambulant COVID-19 to hospitalisation. We used this relationship to predict the efficacy of different monoclonal antibody treatment regimes against the Omicron subvariants BA.1, BA.2, and BA.4/5. We also used this dose-response relationship to estimate the maximal efficacy of monoclonal antibody therapy in the context of pre-existing endogenous neutralising antibodies.Implications of all the available evidenceThis work identifies that both prophylactic therapy and treatment in the early stages of symptomatic infection can achieve significant protection from infection or hospitalisation respectively. The dose-response relationship provides a quantitative means to predict the change in efficacy of different monoclonal antibodies against new variants and in semi-immune populations based on in vitro neutralisation data. We predict a number of existing monoclonal antibodies will be effective for preventing severe outcomes when administered early in BA.4/5 infections. It is likely that these therapies will provide little protection in individuals with high levels of endogenous neutralising antibodies, such as healthy individuals who have recently received a third dose of an mRNA vaccine.