The New England Journal of Medicine APRIL 22, 2021

Tocilizumab in Hospitalized Patients with Severe Covid-19 Pneumonia (COVACTA)

Ivan O. Rosas, et al.

Bottom Line

In this randomized, double-blind, placebo-controlled phase 3 trial, tocilizumab failed to demonstrate a significant improvement in clinical status or 28-day mortality in patients hospitalized with severe COVID-19 pneumonia.

Key Findings

1. The primary endpoint, clinical status on a 7-category ordinal scale at day 28, showed no statistically significant difference between the tocilizumab group and the placebo group (P = 0.31).
2. Mortality at day 28 was 19.7% in the tocilizumab arm compared to 19.4% in the placebo arm, indicating no survival benefit (weighted difference 0.3 percentage points; 95% CI -7.6 to 8.2).
3. Serious adverse events were observed in 34.9% of the tocilizumab group versus 38.5% of the placebo group, suggesting no increased risk of serious safety signals.
4. While the trial did not meet its primary endpoint, secondary exploratory data suggested potential reductions in the time to hospital discharge and duration of intensive care unit stay, though these findings require cautious interpretation.

Study Design

Design
RCT
Double-Blind
Sample
452
Patients
Duration
28 days
Median
Setting
Multicenter, Global
Population Adults hospitalized with severe COVID-19 pneumonia (confirmed SARS-CoV-2 infection, oxygen saturation <=93% on room air or PaO2/FiO2 <300 mm Hg)
Intervention 8 mg/kg intravenous tocilizumab (maximum 800 mg) plus standard of care
Comparator Placebo plus standard of care
Outcome Clinical status on a 7-category ordinal scale at day 28

Study Limitations

The trial was not powered to detect a mortality difference, limiting its ability to rule out smaller but clinically relevant survival benefits.
The study population had varied baseline disease severity and standard-of-care treatments, which may have confounded the results.
The trial was conducted early in the pandemic, and the evolving standard of care (including widespread adoption of corticosteroids) limits the generalizability of these findings to current clinical practices.
The use of a 7-category ordinal scale as a primary endpoint may lack sensitivity to subtle clinical improvements in this specific patient population.

Clinical Significance

The COVACTA trial demonstrated that tocilizumab monotherapy does not provide a significant clinical or survival benefit in a broad, heterogeneous population of hospitalized patients with severe COVID-19. This study highlighted the importance of conducting robust, placebo-controlled trials during the COVID-19 pandemic, as initial observational studies suggesting a large benefit were not replicated, leading to a more nuanced understanding of patient selection and appropriate timing for immunomodulatory therapy.

Historical Context

During the early phases of the COVID-19 pandemic, the 'cytokine storm' hypothesis suggested that IL-6 inhibitors like tocilizumab could mitigate hyperinflammation and improve outcomes. Retrospective, observational data early in 2020 yielded promising results, leading to the rapid initiation of randomized trials like COVACTA. COVACTA was a pivotal study that tempered early enthusiasm, although later trials (such as RECOVERY) specifically targeting patients with systemic inflammation and oxygen requirements demonstrated a mortality benefit, ultimately establishing a role for tocilizumab in the later stages of care.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

Tocilizumab targets the interleukin-6 (IL-6) receptor; what is the physiological rationale for targeting this specific cytokine in the context of severe COVID-19 pneumonia?

Key Response

Severe COVID-19 is often characterized by a hyper-inflammatory state or 'cytokine storm.' IL-6 is a pleiotropic pro-inflammatory cytokine that mediates the acute phase response and is frequently elevated in patients with respiratory failure. Inhibiting its signaling via the IL-6 receptor aims to mitigate the secondary tissue damage and acute respiratory distress syndrome (ARDS) caused by the host's own immune overreaction.

Resident
Resident

The COVACTA trial did not show a mortality benefit for tocilizumab. How should this affect your management of a patient with severe COVID-19 who is currently on dexamethasone and showing rapidly increasing oxygen requirements?

Key Response

While COVACTA was a 'negative' trial for its primary endpoint, subsequent larger trials like RECOVERY and REMAP-CAP demonstrated that tocilizumab provides a survival benefit specifically when used in combination with corticosteroids in patients with rapid respiratory decompensation and high inflammatory markers (e.g., CRP > 75 mg/L). COVACTA's results are likely influenced by the low baseline use of corticosteroids (approx. 20%), which is now standard of care.

Fellow
Fellow

In COVACTA, the primary endpoint was clinical status at day 28 based on a 7-category ordinal scale. Why might this endpoint fail to capture a benefit that was later seen in trials using 'hospital discharge' or 'days free of organ support' as primary outcomes?

Key Response

Ordinal scales can be insensitive if the treatment effect is not uniform across all levels (violating the proportional odds assumption) or if the sample size is insufficient to detect shifts between adjacent categories. Additionally, COVACTA's follow-up of 28 days might have been too short to capture the full recovery trajectory of the most critically ill patients, a factor that trials with longer windows or different composite endpoints addressed.

Attending
Attending

COVACTA highlights the 'moving target' problem in pandemic research. How do you teach trainees to interpret a high-quality RCT that yields neutral results when the standard of care (SOC) changed significantly during the trial's enrollment period?

Key Response

This is a key lesson in external validity. In COVACTA, the emergence of dexamethasone as the SOC changed the baseline inflammatory profile of the participants. A neutral result in an RCT where the SOC is obsolete by the time of publication must be contextualized; it doesn't necessarily mean the drug is ineffective, but rather that its efficacy in the 'new' SOC environment was not the primary question tested.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

Critique the statistical power and sample size of COVACTA (n=438) compared to the RECOVERY trial (n=4116) in the context of detecting a treatment effect in a highly heterogeneous disease like COVID-19.

Key Response

COVACTA was likely underpowered to detect a modest but clinically significant mortality benefit, especially given the heterogeneity of severe COVID-19 (ranging from simple oxygen mask to ECMO). Small trials are susceptible to imbalances in baseline characteristics that can obscure treatment effects. The RECOVERY trial's massive scale allowed for the detection of a 4% absolute mortality reduction that COVACTA's design simply could not reliably identify.

Journal Editor
Journal Editor

As a peer reviewer, how would you evaluate the impact of the 20% 'crossover' or use of non-trial immunomodulators in the placebo arm of a trial like COVACTA on the reported Hazard Ratios for clinical improvement?

Key Response

Contamination or the use of potent rescue therapies in the control arm generally biases the results toward the null (making the treatment and control look more similar). A reviewer would flag this as a threat to internal validity, potentially masking a real treatment effect. Editors must decide if the 'Intent-to-Treat' analysis remains robust enough to support the study's conclusions despite these confounding factors.

Guideline Committee
Guideline Committee

Current NIH and IDSA guidelines recommend tocilizumab for certain hospitalized patients despite the neutral findings of COVACTA. What specific evidence-based criteria do these committees use to override the 'negative' results of an early Phase 3 trial?

Key Response

Guideline committees perform meta-analyses across multiple trials. While COVACTA was neutral, the pooled data from RECOVERY, REMAP-CAP, and others showed a clear benefit. Guidelines now specify 'phenotype-targeted' use: patients must be on corticosteroids and have evidence of systemic inflammation (CRP >= 7.5 mg/dL) or be within 24-48 hours of ICU admission, criteria that were not strictly enforced in the broader COVACTA inclusion set.

Clinical Landscape

Noteworthy Related Trials

2021

RECOVERY Trial

n = 4116 · Lancet

Tested

Tocilizumab

Population

Hospitalized COVID-19 patients with hypoxia and systemic inflammation

Comparator

Usual care

Endpoint

28-day mortality

Key result: Tocilizumab significantly reduced mortality and increased the probability of discharge from hospital among patients with COVID-19.
2021

REMAP-CAP Trial

n = 803 · NEJM

Tested

Tocilizumab or Sarilumab

Population

Critically ill COVID-19 patients requiring respiratory or cardiovascular support in ICU

Comparator

Usual care

Endpoint

Organ support-free days

Key result: Treatment with IL-6 receptor antagonists improved outcomes and reduced mortality in critically ill patients.
2021

EMPACTA Trial

n = 389 · NEJM

Tested

Tocilizumab

Population

Hospitalized patients with COVID-19 pneumonia not receiving mechanical ventilation

Comparator

Placebo

Endpoint

Cumulative proportion of patients who progressed to mechanical ventilation or death

Key result: Tocilizumab reduced the likelihood of progression to mechanical ventilation or death in hospitalized patients with COVID-19 pneumonia.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis