β-agonists for ARDS: the dark side of adrenergic stimulation?

In The Lancet, Fang Gao Smith and the BALTI-2 study investigators¹ report the findings of their phase-3 randomised trial of intravenous salbutamol for acute respiratory distress syndrome (ARDS). The trial, based on reliable preclinical evidence and an encouraging phase-2 trial, was stopped early because of safety concerns.²

β-agonists had been an appealing therapeutic option for ARDS in view of extensive clinical experience, low cost, and an excellent safety profile when used for patients with obstructed airways. Stimulation of the β-2 receptor increases cyclic adenosine monophosphate and active transport of sodium and chloride across type-1 and type-2 alveolar epithelial cells, creating an osmotic gradient for reabsorption of water from the injured lung.³ Clearance of epithelial fluid is impaired in patients with ARDS, more so in fatal cases.⁴ A phase-2 study of intravenous salbutamol (BALTI-1) showed reductions in extravascular lung water, reduced alveolar capillary leak, and improved lung function in a single-centre trial with 40 patients, providing strong proof of concept.²

The BALTI-2 investigators designed a high-quality pragmatic multicentre trial of 1334 patients to determine if a 7-day continuous infusion of salbutamol would improve 28-day mortality versus placebo (0·9% sterile sodium chloride). The Data Monitoring and Ethics Committee (DMEC) recommended that the study stop after the second interim analysis of 273 patients because of a significant increase in mortality (risk ratio 1·55, 95% CI 1·07–2·24; p=0·02). The BALTI-2 Steering Committee accepted this recommendation. DMECs should balance the interests of participants, who are best served by stopping recruitment when harm is apparent, against the interest of society, which could be best served by continuation of a trial until the results are compelling enough to guide the care of future patients.⁵ Did the BALTI-2 DMEC and Steering Committee strike the right balance, and are the results sufficient to change practice?

The findings show consistent evidence for a harmful effect of intravenous salbutamol. The infusion was poorly tolerated because of tachycardia, lactic acidosis, and arrhythmias. The risk ratio for 28-day mortality was 1·47 (95% CI 1·03–2·08). Confidence limits for ventilator-free days and organ failure-free days excluded the null in the direction of harm. Although no previously defined subset seemed to show benefit, early termination limits the power to detect such interactions and is a major drawback of early stopping. Post-hoc attempts at identification of the cause of excess mortality with death certificates were unsuccessful; however, establishment of the proximate cause of death is difficult in patients who are critically ill. Mortality was lower than expected in the placebo group, but this finding parallels those in other contemporary ARDS trials and is probably due to improved supportive care practices, such as lower tidal volume ventilation and conservative fluid therapy.6, 7 Thus, the findings consistently indicate harm from intravenous salbutamol at the dose studied (15 μg/kg ideal weight per h), but do not provide a clear mechanism. The findings are in line with the acute lung injury (ALTA) trial⁸ of aerosolised salbutamol, which stopped early for futility with numerically lower ventilator-free days (p=0·087) and a significant reduction in days free of care in an intensive-care unit (p=0·023).

Injury to the epithelium during ARDS can render it unresponsive to β-agonists, leaving patients treated with salbutamol with little or no potential benefit. Possible explanations for harmful pulmonary effects have been reviewed.8, 9 Prolonged administration of β-agonists might downregulate β-receptors in lung epithelia and impair fluid removal. β-agonists could have a differential effect on the endothelium (harmful) versus the epithelium (beneficial if functional), and β-2 agonists could increase cardiac output thus aggravating alveolar capillary leak. Salbutamol activates the renin-angiotensin system and this might antagonise attempts at diuresis as part of conservative fluid management. More positive fluid balance was noted in the ALTA trial, in which conservative fluid management was protocolised in both groups.⁷ These potential adverse pulmonary effects could explain part of the adverse effect on ventilator-free days in BALTI-2.

© 2012 Science Photo Library

Endogenous or exogenous catecholamine exposure in critical illness, including ARDS, might have a dark side. The risk of increased adverse cardiac events in patients with cardiovascular disease is an obvious example, and salbutamol, although relatively β-2 selective, is a partial β-1 agonist.¹⁰ Plasma cathecholamine concentrations increase with criticall illness and might contribute to the pathogenesis of septic myocardial injury and cardiac failure during septic shock.¹¹ This hyperadrenergic state might cause pernicious adverse effects, including skeletal muscle catabolism, altered glucose homoeostasis and innate immunity, increased β-oxidation of lipids, altered mitochondrial and myocardial function, enhanced gastrointestinal bacterial translocation, and the promotion of bacterial growth.12, 13 β-blockade prevents or reverses many of these adverse effects in experimental animals and in patients with extensive burns, major cardiovascular surgery, and sepsis.13, 14

Much work needs to be done, but the BALTI-2 results should encourage research on approaches to reduce adrenergic upregulation and catecholamine exposure during critical illness and ARDS to identify if these approaches improve outcomes. For now, the results of the truncated BALTI-2 trial are sufficient to change practice. β-2 agonist treatment in patients with ARDS should be limited to the treatment of clinically important reversible airway obstruction and should not be part of routine care.