Effects of potent neutralizing antibodies from convalescent plasma in patients hospitalized for severe SARS-CoV-2 infection

The study addresses whether convalescent plasma (ConvP) containing neutralizing antibodies improves outcomes in hospitalized patients with severe COVID-19. Despite advances such as dexamethasone reducing mortality after 7 days of symptoms and in patients requiring oxygen or ICU care, effective and widely available antiviral therapies remain limited. Neutralizing antibodies targeting the SARS-CoV-2 spike receptor-binding domain can block virus entry and are a promising therapeutic modality. ConvP is scalable and showed signals of benefit in previous coronavirus outbreaks and protective effects in preclinical SARS-CoV-2 models. However, clinical evidence for ConvP in COVID-19 has been inconclusive, with mixed findings and meta-analyses suggesting benefit only in subsets. The study hypothesized that administering high-titer ConvP would improve clinical outcomes, hasten viral clearance, and modulate inflammatory markers in hospitalized COVID-19 patients.

Prior work demonstrated mortality benefits from dexamethasone in hospitalized COVID-19 patients; remdesivir showed benefit in some trials but availability is limited. Neutralizing monoclonal antibodies are under investigation. Historical use of convalescent plasma in SARS and MERS suggested possible benefit, and preclinical SARS-CoV-2 models indicated protection with high neutralizing titers. Clinical studies of ConvP in COVID-19 have yielded inconclusive results, with some benefits reported when administered later in disease in specific cohorts and meta-analyses indicating potential benefit in subsets. NIH guidelines stated a lack of conclusive evidence for ConvP in hospitalized patients, despite emergency use authorization by the FDA. The Placid trial found no reduction in progression to severe disease or death and no reduction in hospital stay, though donors were not pre-screened for neutralizing titers; retrospective analysis showed lower donor titers compared to those selected in the present study.

Design: ConV-001 was a multicenter, open-label randomized clinical trial conducted in 14 Dutch secondary and academic hospitals. Enrollment began April 8, 2020. Eligible participants were adults (≥18 years) with RT-PCR-confirmed SARS-CoV-2 infection within the previous 96 hours and hospitalized for symptomatic COVID-19. Patients were non-mechanically ventilated or ventilated for no longer than 96 hours at inclusion. Key exclusions included prior plasma therapy or contraindications; patients in terminal phase or significantly improved at screening were not included. Randomization and intervention: Participants provided informed consent, had blood drawn, and were randomized 1:1 via ALEA to either standard of care (SoC) or SoC plus 300 mL intravenous convalescent plasma on day 1. Patients without clinical response and with persistent RT-PCR positivity could receive a second plasma dose after 5 days. Concomitant use of locally approved COVID-19 therapies (e.g., remdesivir, hydroxychloroquine, lopinavir/ritonavir) was recorded. Donor selection: Recovered COVID-19 donors (RT-PCR-confirmed) were recruited by Sanquin Blood Supply and screened per blood safety legislation. Donors completed questionnaires and provided plasma. Donor selection prioritized high neutralizing titers assessed by plaque-reduction neutralization test (PRNT50); a minimum PRNT50 ≥80 was required for eligibility, and many selected donors had PRNT50 ≥320. ELISAs for total Ig and IgM to RBD supported screening but PRNT guided selection due to limited predictive value of ELISA at high OD ratios. Sampling and endpoints: Clinical status using the 8-point WHO COVID-19 disease severity scale was assessed at days 1, 5, 10, 15, and 30. Serum and nasopharyngeal swabs were collected at baseline and days 3, 7, 10, and 14. Immunological assays included ELISAs for RBD total Ig and IgM, nucleocapsid IgM/IgG, and PRNT50 for neutralization titers, as available. Cytokines (IL-6, TNFα, IFNγ, IL-1β, IL-2, IL-4, IL-10, IL-12p70) were measured at baseline, day 7, and day 14 in subsets. Virology included RT-PCR quantification of SARS-CoV-2 RNA from nasopharyngeal swabs and viral culture on Vero cells when feasible. Primary endpoint: Overall mortality up to 60 days after enrollment or hospital discharge. Secondary endpoints: Improvement on WHO 8-point severity scale at day 15 and day 30, time to hospital discharge, SARS-CoV-2 shedding dynamics, impact on humoral immunity, inflammatory biomarkers, and safety (transfusion reactions, deaths). Statistical analysis: Planned sample size was 426 patients but the trial was terminated early; 86 patients were randomized (43 per arm). Analyses included multivariable logistic regression for mortality adjusted for known predictors (age, sex, ICU admission, CRP, lymphocyte count, bilirubin, FiO2), proportional odds ordinal logistic regression for WHO scale outcomes (with inverse propensity score weighting), Fine-Gray competing risks model and cause-specific Cox models for time-to-discharge, mixed-effects models (random intercepts/slopes) for log10 viral load over time, and nonparametric tests for immunological comparisons. Zero viral load values had 0.001 added prior to log transformation. Safety events were summarized descriptively.

Participants: 204 screened; 86 enrolled and randomized (SoC n=43, ConvP n=43). Baseline: 72% male; median age 63 years (IQR 56–74); median 10 days (IQR 6–15) from symptom onset to inclusion; median 2 days (IQR 1–3) since hospital admission. Thirteen were ICU-admitted and mechanically ventilated (≤96 h at inclusion). Clinical outcomes: No mortality benefit. Unadjusted mortality OR for ConvP vs SoC: 0.47 (95% CI 0.15–1.38). Adjusted mortality OR at day 60: 0.95 (95% CI 0.20–4.67; p=0.95). WHO 8-point severity scores were similar between arms at day 15 (p=0.58) and day 30 (p=0.67). Proportion improved by day 15 was similar (adjusted IRR 1.30; 95% CI 0.52–3.32; p=0.58). Time to discharge showed no difference (adjusted HR 0.88; 95% CI 0.49–1.60; p=0.68); median length of stay 8 days in both groups. Inflammation and cytokines: No significant differences in highest CRP or ferritin or lowest lymphocyte counts within 14 days. In subsets with serial cytokines, IL-6, TNFα, and IFNγ levels and fold changes over 14 days were similar between arms; no differences for IL-1β, IL-2, IL-4, IL-10, or IL-12p70. Virology: No evidence that ConvP accelerated upper respiratory viral clearance. Baseline viral loads differed between arms but patterns over time remained without appreciable ConvP effect; adjusted mixed models did not support faster decline with ConvP. Viral culture in a small subset showed no signal favoring ConvP. Immunology and donor titers: At baseline, most patients already had SARS-CoV-2 antibodies: 80% total Ig to RBD positive; 77% IgM to RBD positive. Neutralizing antibodies (PRNT50 ≥20) were detected in 79% of tested patients at enrollment, with median titers comparable to donors. Among 115 donors, 96% had detectable neutralizing antibodies (median PRNT50 160; IQR 80–640). Nineteen donors selected for transfusion had high titers (at least 12/19 with PRNT50 ≥320). Patients’ baseline ELISA OD ratios for RBD antibodies were comparable to selected donors. Safety: No serious adverse events related to ConvP were observed.

Administration of high-titer convalescent plasma did not improve mortality, disease severity trajectories, hospital length of stay, inflammatory markers, cytokine responses, or viral clearance in hospitalized COVID-19 patients roughly 10 days after symptom onset. The most plausible explanation is that most patients had already mounted robust autologous neutralizing antibody responses by admission, leaving little incremental benefit from passive antibody transfer. These findings align with trials such as Placid, which also found no clinical benefit, though donor neutralizing titers in that study were lower. The data suggest timing and antibody dose are critical: antibody-based therapies are more likely to be effective when given early—ideally before day 7 post symptom onset and before endogenous humoral responses develop—and with sufficiently high neutralizing titers. Reliance on ELISA alone may be insufficient for donor selection, underscoring the importance of neutralization assays or standardized hyperimmune preparations. Future trials should target earlier disease stages, consider seronegative or immunocompromised patients, and ensure high-titer products, while coordinating to achieve adequate power to detect clinically meaningful effects.

In this multicenter randomized trial, convalescent plasma containing potent neutralizing antibodies conferred no clinical, immunological, or virological benefit in patients recently hospitalized with COVID-19, likely due to the presence of high baseline autologous neutralizing antibody titers. Convalescent plasma therapy should be targeted earlier in the disease course, before robust endogenous humoral immunity develops, and prioritized for patients at high risk of progression or those unable to mount humoral responses. Future research should focus on early outpatient administration, serology-guided patient selection, standardized high-titer products, and adequately powered collaborative trials.

The trial was terminated prematurely, limiting power to detect modest effects and precluding definitive conclusions. Donor and patient sampling/logistical constraints during the pandemic led to missing immunological data for some participants. Potential non-neutralizing immunomodulatory effects of plasma are unlikely at the doses used but cannot be entirely excluded. Corticosteroid use was not recorded, which could confound outcomes, although national guidelines at the time did not recommend steroids for non-ConvP patients. Baseline imbalances and differences in initial viral loads existed but adjusted analyses did not show ConvP benefit. Generalizability may be limited to hospitalized patients approximately 10 days post symptom onset, among whom most already have neutralizing antibodies.