Muscle relaxants in ARDS – The final verdict with the updated evidence

The ventilatory strategies in moderate-to-severe acute respiratory distress syndrome (ARDS) with PaO2: FiO2 150 or lower (arterial partial pressure of oxygen: fraction of inspired oxygen) are an area of active research. The European Society of Intensive Care Medicine (ESICM) and the American Thoracic Society (ATS) published their updated guidelines for ARDS in 2023.[1,2] The updated recommendations regarding the use of neuromuscular blocking agents (NMBAs) are interestingly opposite to each other. The ESICM recommended against the use of a continuous infusion of NMBA.[1] Meanwhile, the ATS suggested in favour of using infusion or boluses of NMBAs.[2] On the surface, these are seemingly contradictory statements. Since the evidence block cited in both documents is similar, it is critical to review said evidence to understand the reason behind the recommendations and position of NMBAs in the current management strategy.

Quintessentially, NMBAs are agents that induce pharmacological paralysis. The potential benefit of NMBAs in ARDS is achieved by reducing patient–ventilator dyssynchrony and the accompanying ventilator-induced lung injury and work of breathing.[1,2] Their use in ARDS was initially supported by data from controlled trials, chiefly the ACURASYS trial.[3,4,5] This trial evaluated the effect of an early initiation (<48 h from onset) of short-duration (<48 h) NMBA along with deep sedation on mortality and compared it to deep sedation alone as the control. It found a significant mortality benefit at 28 days, but not at 90 days.

However, NMBAs should not be used without first inducing sedation, commonly deep sedation. In the supporting trials, all participants were initially deeply sedated.[3,4,5] In the ACURASYS trial, the sedation level was defined by a Ramsay sedation score (RSS).[3] The target RSS was 6, or ‘no response to stimulus’, which is representative of deep sedation. However, since the ACURASYS was published, multiple studies have also compared deep to light sedation. A meta-analysis found that in comparison to deep sedation, light sedation lessens the risk of critical care-associated weakness, post-intensive care syndrome and delirium and is associated with better diaphragmatic function, more ventilator-free days and an overall lesser mortality.[6] The current practice has, therefore, evolved over time to target light rather than deep sedation.

As a logical consequence, NMBAs combined with deep sedation must now be compared to light sedation. The ROSE trial compared the effect of NMBAs in a similar ARDS population as the ACURASYS trial.[3,7] The intervention arm received cisatracurium along with deep sedation. However, the control arm received light rather than deep sedation, with target RSS: 2 or 3, Richmond agitation–sedation scale (RASS): 0 or -1 or Riker sedation–agitation scale (RAS): 3 or 4. The trial had to be stopped due to an interim analysis that revealed futility of the intervention. The observed 90-day mortality was 42.5% in the NMBA group and 42.8% in the light sedation group (95% confidence interval [CI]: -6.4 to 5.9, P = 0.9).[7] In addition, there was no difference in the number of ventilator-free days, incidence of critical care-associated weakness or the incidence of pneumothorax. The investigators, however, noted a significantly greater number of serious cardiac adverse events with NMBAs compared to light sedation.[7]

In a meta-analysis of seven trials including ACURASYS and ROSE, Tarazan et al.[8] evaluated data from 1598 patients. They noted a high degree of heterogeneity among the trials based on whether light or deep sedation was used in the control arm.[8] NMBA use significantly reduced the risk of barotrauma, but did not increase ventilator-free days.[8] Further, NMBAs along with deep sedation had a significant mortality benefit only when compared to deep sedation with as-needed boluses of NMBA (relative risk [RR]: 0.71, CI: 0.57 to 0.89; P = 0.003).[8] This benefit was, however, not noted when compared to light sedation (RR: 0.99, CI: 0.86 to 1.15; P = 0.9).[8] Another key issue is that the exact and singular level of sedation at which the benefit of NMBAs will be lost is unknown.[2,8] Further, since the current recommendation is to target light sedation, it is crucial to note that results from trials do not favour the routine use of NMBAs to reduce mortality in ARDS.

After looking at seemingly opposing recommendations from ESICM and ATS, and the underlying evidence, an uncertainty arises. How to use NMBAs effectively to achieve maximum benefit in ARDS with PaO2: FiO2 <150? The most fundamental rule to remember is that compared to light sedation, the routine use of NMBA in every patient with ARDS is not beneficial. NMBAs may, in fact, result in harm due to the concomitant use of deep sedation. The initial and most crucial step in the management of a mechanically ventilated patient of ARDS is the optimisation of ventilatory parameters.[1,2] In case of a failure to achieve the target PaO2, early (within the first 48 h of ARDS onset) protocolised prone positioning should be instituted without delay.[1,2] These two strategies have been found to achieve a significant mortality benefit, and thus form the crux of managing ARDS with PaO2: FiO2 <150.[1,2]

Despite these measures, some patients fail to achieve the target PaO2 (55–80 mmHg). One cause for this failure is patient–ventilator dyssynchrony. To mitigate this dyssynchrony, the level of sedation is often deepened instinctively. It is here that the use of NMBA can achieve a mortality benefit along with an improved PaO2: FiO2.[2,8] The resultant pharmacological paralysis eliminates any muscular effort by the patient and thus patient–ventilator dyssynchrony.

Initially, a trial of NMBAs can be initiated, with the intention to continue this intervention only if the trial was successful at improving PaO2: FiO2. The maximum duration of use should be up to 48 h.[2,8] Continuation beyond 48 h is not supported by trials and should be done with caution on a case-by-case basis. To avoid unnecessary delay to switch over to another strategy, NMBAs should be abandoned early if they are unsuccessful or failing. Since mortality is high, time is of essence in a patient with PaO2: FiO2 <150 who is unresponsive to optimised ventilatory strategy and prone positioning. Mortality rates have been observed to be greater than >50% when the PaO2: FiO2 falls below 80. NMBAs are unlikely to benefit significantly at this severity of ARDS, especially if unsuccessful. The ESICM recommends and the ATS suggests the use of venovenous extracorporeal membrane oxygenation (VV-ECMO) in ARDS with duration of mechanical ventilation <7 days when indications under the EOLIA trial are fulfilled.[1,2] These were: PaO2: FiO2 <80 for at least 6 h or <50 for at least 3 h, or the presence of hypercapnia (arterial partial pressure of carbon dioxide (PaCO2) >60 mmHg) with pH <7.25 for at least 3 h.[1,2] Extracorporeal membrane oxygenation (ECMO) may be the only strategy that can reduce mortality in severe ARDS refractory to other core management strategies.

For the choice of NMBA agent, the optimal pharmacological agent has not been established.[2] In published trials, cisatracurium has been commonly used and has, therefore, become the de facto choice.[8] However, there are no head-to-head trials comparing NMBA agents. In addition, cisatracurium has been shown to attenuate protease action in microvascular endothelial cells.[9] Further, the ATS suggests using either a continuous infusion or boluses as equally appropriate dosing regimens.[2] For the dosing amount, the ACURASYS protocol of 15 mg cisatracurium bolus followed by an infusion of 37.5 mg/h can be followed.[3]

In conclusion, despite opposing recommendations in new guidelines, the following strategy is the likely position of NMBAs in ARDS with PaO2: FiO2 <150. The routine use of NMBAs (and deep sedation) in every mechanically ventilated ARDS patient should be avoided. Manage mechanically ventilated patients using optimised ventilatory strategies and target light sedation. Failing this, early ARDS (time since onset <48 h) should be managed with prone positioning. If this is unsuccessful, identify the cause of failure. NMBAs can be used in the subset where the impediment to PaO2 target is patient–ventilator dyssynchrony. Initially, a trial of NMBA can be given. Although the optimal NMBA agent is not known, using cisatracurium may be desirable. A bolus of 15 mg, followed by an infusion of 37.5 mg/h is reasonable; however, the optimal dosing is not known and may need to be adjusted. If PaO2 responds, continue for preferably less than 48 h. Longer durations must entail caution and a case-by-case decision. If the trial is unsuccessful, evaluation and preparation for VV-ECMO (or transfer to an ECMO centre) should be initiated without unnecessary delays.