ROSE

Early Neuromuscular Blockade in the Acute Respiratory Distress Syndrome

The National Heart, Lung and Blood Institute PETAL Clinical Trials Network. NEJM 2019; 380: 1997-2008. doi: 10.1056/NEJMoa1901686

Clinical Question

  • In patients with moderate to severe ARDS, does a strategy of early neuromuscular blockade with heavy sedation, as opposed to usual care with lighter sedation targets, result in a lower 90 day mortality?

Background

  • The acute respiratory distress syndrome (ARDS) remains a syndrome with high morbidity and mortality
  • Neuromuscular blockade is proposed to help the management of these patients by promoting ventilator synchrony, reducing work of breathing and reducing alveolar fluid accumulation, but may come at the cost of acquired weakness
  • The ACURASYS trial performed in 2010 showed a statistically significant reduction in mortality in patients with moderate-to-severe ARDS treated with cisatracurium, compared with deeply sedated control patients, albeit after an adjusted analysis of the primary end point
  • Neuromuscular blockade has been weakly recommended in the management of ARDS, and uptake by clinicians has been variable
  • The management of critically ill patients has altered over the past decade, and deep sedation is no longer routine current practice
  • With so many questions remaining, the Prevention and Early Treatment of Acute Lung injury (PETAL) group designed a phase III trial

Design

  • Multicentre, unblinded randomised control trial
  • Permutational block design stratified by site, assigning patients in 1:1 ratio
  • For 90% power with two sided alpha 0.05, 1408 patients required (based on control group mortality 35% and intervention group 27%)
  • Intention to treat analysis
  • No stopping rule for futility
  • Designed to be stopped for superiority using symmetrical group sequential stopping boundaries
  • Wald test for proportions used to compare primary outcomes between groups
  • Risk difference scales used to compare secondary outcomes
  • Weighted Poisson regression used to compare adverse events

Setting

  • 48 ICUs in the United States
  • Jan 2016-April 2018

Population

  • Inclusion:
    • P/F ratio <150 mmHg (22 kPa)
    • PEEP>8 cm H2O
    • Bilateral opacities on CXR/CT that could not be explained by effusions, pulmonary collapse, or nodules
    • Respiratory failure that could not be explained by cardiac failure or fluid overload
    • Onset of illness within one week of known clinical insult or new/worsening respiratory symptoms
  • Exclusion:
    • Refusal of consent
    • Neuromuscular blockade at prior to enrolment
    • Pregnant
    • ECMO at time of assessment
    • Chronic respiratory failure (defined as pCO2 >8 kPa in outpatients)
    • Home ventilation
    • BMI >100
    • Severe chronic liver disease
    • Bone marrow transplant within last year
    • Expected mandatory ventilation duration <48 hours
    • Decision to withhold life-sustaining therapy
    • Moribund and not expected to survive 24 hours
    • Diffuse alveolar haemorrhage from vasculitis
    • >70% BSA Burns
    • Unwillingness/unable to stick to ARDSnet 6mL/kg protocol
    • Hypersensitivity to cisatracurium
    • Neuromuscular conditions that may potentiate NMB/inhibit spontaneous ventilation
    • Neurological conditions undergoing treatment for intracranial hypertension
    • Enrolment in an interventional ARDS trial with direct impact on neuromuscular blockade/PEEP
    • P/F ratio >200 mmHg (32 kPa) after meeting inclusion criteria and before randomisation
    • Ventilation >5 days
    • On lung transplant list
  • 4848 patients screened, 1006 patients included in primary analysis
  • Patient demographics very similar between groups, except shock, at baseline

Intervention

  • Deep sedation within four hours of randomisation
  • 15 mg bolus of cisatracurium, followed by 37.5 mg/h infusion

Control

  • Light sedation defined by
    • RASS 0 to -1
    • Riker sedation scale 3-4
    • Ramsay sedation score 2-3
  • Institutions were allowed to use their normal sedation scoring modality

Management common to both groups

  • Low tidal volume ventilation within two hours of randomisation, with high PEEP for up to five days
  • Lower PEEP was allowed for
    • Hypoxia
    • Hypotension
    • Pplat>30
    • Acidosis (pH <7.15) despite reducing tidal volumes, fluid boluses, or increasing respiratory rate

Outcome

  • Primary outcome: In-hospital death from any cause at 90 days
    • 42.5% (intervention) vs 42.8% (control)
    • Between group difference -0.3% (95% CI -6.4 to 5.0)
    • P value 0.93
  • Secondary outcome:
    • No significant difference seen in any secondary outcomes except for serious cardiovascular effects
      • 14 (intervention) vs 4 (control)
      • P value = 0.02
    • Other pre-specified secondary outcomes were
      • ICU acquired weakness
      • Plasma IL-6 concentration (removed at final analysis)
      • 28-day mortality
      • Ventilator-free days at day 28
      • Organ failure-free days till day 28
      • ICU-free days at day 28
      • Hospital-free days at day 28
      • Physiologic measures
      • Long term assessments of quality of life
      • Use of rescue procedures e.g ECMO
      • Awareness/recall

Authors’ Conclusions

  • “Among patients with moderate-to-severe ARDS who were treated with a strategy involving a high PEEP, there was no significant difference in mortality at 90 days between patients who received an early and continuous cisatracurium infusion and those who were treated with a usual-care approach with lighter sedation targets”

Strengths

  • Randomised
  • Intention to treat analysis
  • Registered at clinicaltrials.gov
  • High compliance with study protocols
  • Large cohort
  • Use of high PEEP and lighter sedation targets in the control arm more closely reflect current clinical practice

Weaknesses

  • Unblinded
  • Large loss of patients at screening- the largest reasons for exclusion were patients improving between screening and inclusion (658) and patients having already received neuromuscular blockade (655)
  • The 655 patients who received neuromuscular blockade, and thus were excluded from the trial, may have biased the results in favour of the control group
  • Depth of neuromuscular blockade not measured- the authors make a strong case for the weakness of traditional “train-of-four” measurements, however it could be argued that this was not a standard of care
  • Stopped for futility, in the absence of a pre-specified stopping rule- this was decided by a steering committee independent of the authors; however this renders ROSE an underpowered study
  • The use of rescue therapies, in particularly proning, was not standardised
    • In ACURASYS, 50% of patients were proned, as opposed to 16% in this study
    • Proning has been shown to improve mortality- that may have affected the results of both ACURASYS and this trial

The Bottom Line

  • In this trial, deep sedation combined with an infusion of cisatracurium did not improve 90-day mortality when compared with a light sedation strategy in a cohort of patients recruited early in the course of moderate-to-severe ARDS
  • This does not change my practice of using neuromuscular blockade in selected patients to improve ventilator synchrony

External Links

Metadata

Summary author: Segun Olusanya
Summary date: 27th June 2019
Peer-review editor: Charlotte Summers

One comment

  • manofonezerg

    I don’t think your conclusions regarding prone positioning in the two trials is correct. In ACURASYS 28% of patients in the NMBA arm were proned, compared to 27% in the control arm (see ACURASYS Supplementary Appendix Table 8: co-interventions during the study period), so proning rates were no different between the intervention and control arms.

    In ROSE about 15% of patients in both intervention and control groups were proned. Although ROSE had a lower rate of proning, both trials had similar proportions of patients proned in intervention and control groups, so the effect of proning on mortality should not have affected the trial outcomes.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.