Use of Corticosteroids in Persistent ARDS

Abstract & Commentary

By James E. McFeely, MD, Medical Director Critical Care Units, Alta Bates Summit Medical Center, Berkeley, CA, is Associate Editor for Critical Care Alert.

Dr. McFeely reports no financial relationship to this field of study.

Synopsis: A prospective randomized trial carried out over a 6-year period of time enrolled 180 patients with ARDS of at least 7 days duration, and randomized them to receive either methylprednisolone or placebo. There was no significant difference noted in mortality at 60 days, though there was some improvement in ventilator-free and shock-free days during the first 28 days in patients treated with steroids. Steroids were also associated with an increased risk of death if started more than two weeks after the onset of ARDS.

Source: Steinberg KP, et al. N Engl J Med. 2006;354:1671-1684.

Corticosteroids have been studied in various phases in the treatment of Acute Respiratory Distress Syndrome (ARDS), which is characterized by an inflammatory injury to the lung resulting in acute hypoxemic respiratory failure. Four trials of steroids in early ARDS failed to show any improvement. However, several reports in small case series showed a possible benefit from moderate-dose steroids in patients with persistent ARDS, including a single-center study involving a small number of patients that suggested an improvement in lung function and survival. On this basis, the ARDS Network undertook a larger, multi-center, placebo-controlled randomized study of steroids in patients with persistent ARDS.

Patients were enrolled between August 1997 and November 2003 at 25 hospitals in the ARDS network. Out of 4,000 patients screened (3,464 of whom were eligible), 180 (5%) enrolled between day 7 and 28 after the onset of ARDS and were randomly assigned in a double-blind fashion to receive either methylprednisolone or placebo. The initial demographic variables were similar between the two study groups except for gender (more men were assigned placebo). Methylprednisolone was administered as a single dose of 2 mg/kg of predicted body weight, followed by a dose of 0.5 mg/kg every 6 hours for 14 days. The dose was subsequently reduced to 0.5 mg/kg every 12 hours for 7 days, then tapered. Tapering occurred more rapidly if fungal disease or septic shock developed, or if the patient improved to the point of having been extubated for at least 48 hours.

The study's primary end point was mortality at 60 days. Secondary end points included the number of ventilator-free and organ failure-free days, infectious complications, and markers of inflammation and fibroproliferation. The trial was initially designed to enroll 400 patients, but two years into the study sample size was changed based on low enrollment as well as results of the contemporaneous low tidal volume ARDS net study.1

Based on an intention-to-treat analysis, there was no significant difference in mortality between the 2 treatment groups, either at the 60-day time interval (28.6% placebo, 29.2% methylprednisolone) or at 180 days. The steroid group experienced significantly more ventilator-free days than the placebo group during the first 28 days (14 days vs 24 days); they also had more shock-free days and showed improvements in oxygenation and respiratory system compliance as compared with the placebo group. However, patients in the steroid group were more likely to require reintubation (28% vs 9%). Further, patients among the steroid group who had ARDS for more than 13 days before enrollment showed a significantly increased mortality rate.


The ARDS Network is to be commended for their perseverance in attempting to answer the question of steroid use in late ARDS. However, while the results of this paper might suggest that steroids (at least at the dose and timing used in the study) are unfortunately not going to be of use for the fibroproliferative phase of ARDS, problems with recruitment and the length of time it took to complete enrollment raise questions about the applicability of this result.

This study was open for enrollment at 25 hospitals for 2,296 days. During this time, the practice of critical care medicine was in rapid evolution. While the study was initially conducted in parallel with the ARDS low-tidal-volume study, subsequent results from the latter study1 led researchers to decrease enrollment and include additional covariate analyses. In addition, many changes in critical care practice have had a direct impact on the primary and secondary end points of the study, including the use of vasopressin for shock, tight glycemic control, early goal-directed therapy, use of activated protein C (Xigris) for sepsis, and development of protocols for weaning and use of sedatives. Each of these could be expected to have an effect on overall mortality of similar magnitude to steroid use in ARDS. Moreover, only 5% of patients who were eligible for the study (180/3464) were actually enrolled. While there were appropriate reasons for this, the exclusions do raise a question regarding the applicability of this result to a larger population.

What can be concluded from this study is that starting steroids more than two weeks after the onset of ARDS is probably a bad idea and likely increases the risk of death. This paper also documented high incidences of neuromyopathy. While there was no difference in frequency between the 2 groups, patients on steroids seem to have more severe myopathy. This was perhaps complicated by the increased frequency of hyperglycemia, also seen in the corticosteroid group. Fortunately, no significant increase in infectious complications was identified.

Reduction in the excessive inflammatory response to ARDS still seems to be a promising therapeutic target in ARDS. It may be that corticosteroids, however, are too blunt an instrument in this situation. In the 10 years since this trial was first designed, we've come a long way in identifying inflammatory and anti-inflammatory genes and mediators. This trial helps narrow the time window for optimal anti-inflammatory therapy. Hopefully, the next trial will be performed with a more targeted therapy applied over a broader range of patients, and will be brought to conclusion more quickly.


  1. The Acute Respiratory Distress Syndrome Network. N Engl J Med. 2000;342:1301-1308.