Background: Acute respiratory distress syndrome (ARDS) is defined as severe hypoxemic respiratory failure resulting from diffuse lung injury and secondary to direct and indirect insults. Despite advances, mortality remains as high as 40-60%. Neuromuscular blocking agents (NMBAs) are used to facilitate mechanical ventilation in patients with ARDS and have been shown to improve arterial partial pressure of oxygen. However, the association between NMBAs and mortality is unclear. Furthermore, morbidity concerns exist, particularly regarding a putative role in intensive care unit (ICU)-acquired weakness.
Objective: The purpose of this study was to compare survival in adult patients with early ARDS who were randomized to receive either a 48-hr infusion of the NMBA, cisatracurium, or a placebo.
Design and setting: This study was a multicentre double-blinded randomized controlled trial involving 20 ICUs in France from March 2006 to March 2008.
Patients: Eligible patients were > 18 yr with an intubated trachea and ventilated lungs for acute hypoxemic respiratory failure. Their PaO(2)/F(i)O(2) ratio was < 150 at a tidal volume of 6-8 mL·kg(-1) ideal body weight and a positive end-expiratory pressure (PEEP) ≥ 5 cm H(2)O for < 48 hr. Additional inclusion criteria were radiographic evidence of bilateral pulmonary infiltrates and the absence of left atrial hypertension. Exclusion criteria included patients already receiving NMBA at enrolment; those who had increased intracranial pressure, severe chronic respiratory disease, or severe chronic liver disease; those who had received a bone marrow transplant or had chemotherapy-induced neutropenia; those who had a pneumothorax; and those who were expected to require mechanical ventilation for < 48 hr or were enrolled in another trial within 30 days.
Intervention: Three hundred twenty-six patients were screened, and 340 of these underwent randomization in blocks of four and received either a 48-hr infusion of cisatracurium (15 mg bolus followed by 37.5 mg·hr(-1)) or a volume equivalent placebo. One hundred and seventy-eight patients received a cisatracurium infusion, and one patient withdrew leaving 177 patients included in the analysis. One hundred and sixty-two patients received the placebo infusion. Prior to either infusion, patients were sedated to a Ramsay sedation score of 6. Patients' lungs were ventilated by a volume assist-controlled mode according to the ARDS Clinical Network Mechanical Ventilation Protocol (http://www.ardsnet.org/) with the goal SpO(2) of 88-95% (or PaO(2) 55-80 mmHg) and goal plateau pressure ≤ 35 cm H(2)O. Open-label boluses of cisatracurium 20 mg (maximum of two per 24-hr period) were allowed if plateau pressures remained > 32 cm H(2)O despite increased sedation and despite decreased PEEP and decreased tidal volumes. Monitoring of paralysis via peripheral nerve stimulation was not permitted.
Measurements: The primary outcome was death before hospital discharge and within 90 days of study enrolment. It was determined a priori that this would be adjusted for imbalance in key risk factors at baseline, as derived from Cox regression. Secondary outcomes included 28-day mortality, number of ventilator-free days, number of days outside of ICU, number of days without organ system failure, rate of barotrauma, and rate of ICU-acquired paresis (as defined by a Medical Research Council [MRC] score < 48) on day 28 and at ICU discharge.
Main results: With regard to the primary outcome, crude 90-day mortality was 31.6% in the cisatracurium group vs 40.7% in the placebo group. This outcome did not reach statistical significance (P = 0.08). However, post hoc analysis found a reduction in 90-day mortality in the cisatracurium group compared with placebo (95% confidence interval 0.48 to 0.98; P = 0.04). Results suggest that the reduction in 90-day mortality in the cisatracurium group was confined to those patients with a PaO(2)/F(i)O(2) ratio < 120. Additionally, 28-day mortality was significantly lower in the cisatracurium group (absolute difference -9.6%; P = 0.05). The cisatracurium group also had significantly more ventilator-free days, more days outside of the ICU, and more days free of organ-failure. Similarly, pneumothorax developed more often and earlier in the placebo group than in the cisatracurium group. The rate of ICU-acquired weakness at day 28 or at ICU discharge did not differ significantly between the two groups.
Conclusions: Treatment in early severe ARDS with the NMBA, cisatracurium, for 48 hr was associated with lower adjusted 90-day mortality. It was also associated with decreased morbidity, which included increased ventilator-free days, increased ICU-free days, and increased organ failure-free days. These benefits occurred without increasing the incidence of ICU-acquired weakness.