The ability to control the delivery of ventilation to patients having the acute respiratory distress syndrome (ARDS) without encountering patient respiratory effort via the administration of neuromuscular blocking drugs has been a potentially appealing therapeutic option for decades (Light RW, et al.). This practice had been common in the late 20th century in order to avoid excessive tachypnea and appearance of patient discomfort with the collateral benefit of improving oxygenation and decreasing the fraction of inspired oxygen (FiO2) (Hansen-Flaschen JH, et al. JAMA. 1991;26:2870). Following the publication by the NIH-sponsored ARDS Network of the landmark low tidal volume lung protective ventilation trial, whereupon study subjects had been allowed to breathe up to 35 times per minute (ARDS Network, ) and additional concerns that neuromuscular blockade could potentially be associated with neuromuscular weakness, this practice fell out of favor.
Although the validity of using lung protective ventilation in ARDS, with a plateau pressure of less than 30 cm/H2O via delivery of a low tidal volume, has withstood the test of time, subsequent attempts to utilize methods that would further protect the lung with additional “rescue” approaches to mechanical ventilation led to a partial renaissance of the neuromuscular blockade (NMB) approach. For example, high frequency oscillatory ventilation, with its idiosyncratic delivery of minute volumes of ventilator gas, requires NMB in order to be used. However, the publication of two negative trials, including one demonstrating an increased mortality, sidelined this approach (Ferguson ND, et al.).
More notably, the use of NMB in patients with ARDS has been advocated during conventional mechanical ventilation to avoid the generation of large tidal volumes via ventilator asynchrony occurring during patient-triggered breaths. Ostensibly, wiping out any patient effort via NMB eliminates manifestations of asynchrony, such as double triggering, which can generate areas of regional tidal hyperinflation in the injured lung and thereby worsen ventilator-induced lung injury. The utilization of NMB early in the course of ARDS (less than 48 hours) resulted in less lung inflammation (Forel JM, et al.). Subsequently, the ACURASYS trial found that patients with moderately severe or severe ARDS treated with NMB had a mortality benefit comparable to that seen in the original ARDS low tidal volume trial (Papazian L, et al. N Engl J Med. 2010;369:980).
Several criticisms of ACURASYS led to the desire for a larger confirmatory trial be undertaken. The NIH-sponsored successor to the ARDS Network, the Prevention and Early Treatment of Acute Lung Injury (PETAL) Network, took this on straight away with its formation in 2014 (disclosure: the author is a Principal Investigator of one of the 13 PETAL Network Clinical Centers). This trial, called the Re-Evaluation of Systemic Early Neuromuscular Blockade, the ROSE trial, was published last year in the New England Journal of Medicine and failed to confirm a mortality benefit to NMB when used early in the course of ARDS, such as had been done earlier (Moss M, et al.).
What then, should clinicians consider the proper use of NMB in ARDS to be?
There has been a recent spate of large negative trials of once-promising interventions in critical care medicine (Laffey.). Among these were trials related to early mobility, vitamin D administration, transpulmonary pressure titrated positive end-expiratory pressure (PEEP), and of course, high frequency oscillatory ventilation, just to name a few disappointments. Recognition of heterogeneity of treatment effect (HTE), with some subgroups being more likely to respond to an intervention than others (Iwashyna. ), is cold comfort to the bedside clinician and all but the most dedicated health services researcher. At least to date, personalized medicine has fallen short of prospective validation in ARDS (Constantin et al. ).
The failure of the ROSE trial to demonstrate a mortality benefit to ARDS patients with a P/F ratio of less than 150 on at least 8 cm H2O treated with early NMB means the routine use of this approach in all such patients isn’t warranted. In a prescient nod to HTE, “a foolish consistency,” as Emerson said, “is the hobgoblin of little minds.” Importantly, there were several subtle but not necessarily irrelevant differences between ACURASYS and ROSE. ROSE used a high PEEP algorithm to titrate PEEP to FiO2, rather than the conventional low PEEP approach used in the original ARDS Network and ACURASYS trials. Potentially, the benefits of NMB on the injured lung in ARDS may have been mitigated by using higher PEEP levels. ROSE also failed to demonstrate a decrease in barotrauma as had been reported earlier. That said, it is difficult to ascribe the lack of benefit of NMB mechanistically to less asynchrony induced regional tidal hyperinflation in the NMB group at high PEEP, especially given the lighter sedation targets employed in both the NMB and the placebo group. Meanwhile, ROSE did confirm patients were not harmed by NMB by resulting in more neuromuscular weakness upon recovery.
Among patients with Berlin severe ARDS (ie. P/F less than 100 on at least 5 cm H2O PEEP) evaluated between publication of ACURASYS and ROSE, clinicians were far more inclined to use NMB than other rescue modalities, including prone ventilation (Duan, Ann Am Thorac Soc. 2017;12:1818). It seems unlikely the publication of ROSE will alter this. As rescue modalities go, NMB is relatively inexpensive, widely available and easily performed (Co, I and Hyzy RC, Crit Care Med. 2019 Dec 18.). Ultimately, though the question isn’t whether NMB will be used in ARDS patients with refractory hypoxemia early or even later, but whether prone ventilation should be simultaneously initiated at the time of, or even before the institution of NMB.
As in ACURASYS, patients in the landmark PROSEVA prone ventilation trial were treated with a low PEEP algorithm (Guérin C et al.). Prone ventilation has many salutary physiologic benefits, not the least of which is recruitment of areas of collapsed lung. Patients who are recruitable with PEEP, i.e. whose PaO2 increases with increasing PEEP in the face of an unchanged or minimally changed plateau pressure, may also demonstrate a mortality benefit (Goligher, EC et al. ). It remains unknown whether prone ventilation would remain of significant benefit should a high PEEP approach be employed.
Prone ventilation clearly has its adherents (Albert, RK,), although underutilization remains prevalent perhaps due to its somewhat cumbersome nature. While it might have been interesting had ROSE performed a simultaneous assessment of prone ventilation along with NMB via a factorial trial design, clinicians remain at the crossroads of how to escalate ventilator support in the ARDS patient with worsening, if not refractory hypoxemia. The use of NMB with a high PEEP approach often allows for recruitment and a concomitant lowering of FiO2 to acceptable levels in advance of the utilization of prone ventilation. Although some clinicians are able to successfully utilize prone ventilation without NMB, many are not, and NMB use was widespread in PROSEVA.
With no evidence of harm, the employment of NMB in the setting of Berlin severe ARDS is entirely justifiable, whether occurring early or late in the clinical course, regardless of, or potentially with the concomitant employment of prone ventilation. These two rescue modalities remain first line and, despite evidence to the contrary (Li, et al.
Dr. Hyzy is with the Division of Pulmonary and Critical Care Medicine, University of Michigan, Ann Arbor.