|
|
Background Persistent acute respiratory distress syndrome (ARDS) is characterized by excessive fibroproliferation, ongoing inflammation, prolonged mechanical ventilation, and a substantial risk of death. Because previous reports suggested that corticosteroids may improve survival, we performed a multicenter, randomized controlled trial of corticosteroids in patients with persistent ARDS.
Methods We randomly assigned 180 patients with ARDS of at least seven days' duration to receive either methylprednisolone or placebo in a double-blind fashion. The primary end point was mortality at 60 days. Secondary end points included the number of ventilator-free days and organ-failure–free days, biochemical markers of inflammation and fibroproliferation, and infectious complications.
Results At 60 days, the hospital mortality rate was 28.6 percent in the placebo group (95 percent confidence interval, 20.3 to 38.6 percent) and 29.2 percent in the methylprednisolone group (95 percent confidence interval, 20.8 to 39.4 percent; P=1.0); at 180 days, the rates were 31.9 percent (95 percent confidence interval, 23.2 to 42.0 percent) and 31.5 percent (95 percent confidence interval, 22.8 to 41.7 percent; P=1.0), respectively. Methylprednisolone was associated with significantly increased 60- and 180-day mortality rates among patients enrolled at least 14 days after the onset of ARDS. Methylprednisolone increased the number of ventilator-free and shock-free days during the first 28 days in association with an improvement in oxygenation, respiratory-system compliance, and blood pressure with fewer days of vasopressor therapy. As compared with placebo, methylprednisolone did not increase the rate of infectious complications but was associated with a higher rate of neuromuscular weakness.
Conclusions These results do not support the routine use of methylprednisolone for persistent ARDS despite the improvement in cardiopulmonary physiology. In addition, starting methylprednisolone therapy more than two weeks after the onset of ARDS may increase the risk of death. (ClinicalTrials.gov number, NCT00295269
[ClinicalTrials.gov]
.)
Four trials of high-dose, short-course corticosteroids for early-phase ARDS failed to show improvements in survival.9,10,11,12 Several reports from small case series suggested a benefit of moderate-dose corticosteroids in patients with persistent ARDS.8,13,14,15,16,17 A single-center, randomized trial involving 24 patients who had had ARDS for seven or more days reported that moderate-dose corticosteroids improved lung function and survival.18
The risks associated with corticosteroids in these patients are unclear. Several studies involving patients with sepsis and ARDS have suggested that high-dose corticosteroids increase the risk of secondary infections,11,12,19,20,21 yet a meta-analysis of moderate-dose corticosteroids for sepsis did not substantiate this observation.22 Additional potential risks of corticosteroids include hyperglycemia, poor wound healing, psychosis, pancreatitis, and prolonged muscle weakness with impaired functional status.23
We conducted a multicenter randomized, controlled trial to determine the efficacy and safety of this therapy. Our hypothesis was that the administration of moderate-dose methylprednisolone to patients with persistent ARDS would improve clinical outcomes without significantly increasing complications.
Methods
Patients were enrolled from August 5, 1997, through November 17, 2003, at 25 hospitals of the National Heart, Lung, and Blood Institute (NHLBI) ARDS Clinical Trials Network (see the Appendix). The institutional review board of each hospital approved the trial. Written informed consent was obtained from patients or their surrogates. A complete description of the protocol and methods is available at www.ardsnet.org.
Patients
Patients who were intubated and receiving mechanical ventilation were eligible for enrollment 7 to 28 days after the onset of ARDS as previously defined.1 The ratio of the partial pressure of arterial oxygen (PaO2) to the fraction of inspired oxygen (FiO2) was adjusted for barometric pressure in Denver and Salt Lake City. Continuous mechanical ventilation and persistent bilateral infiltrates were required from the onset of ARDS to study entry. On the day of study entry, the PaO2:FiO2 ratio had to be less than 200. Exclusion criteria are listed in the Supplementary Appendix, available with the full text of this article at www.nejm.org.
Study Protocol
Patients were randomly assigned with the use of permuted blocks to receive either intravenous methylprednisolone sodium succinate (methylprednisolone) diluted in 50 ml of 5 percent dextrose in water or placebo (50 ml of 5 percent dextrose in water), stratified according to hospital. A single dose of 2 mg of methylprednisolone per kilogram of predicted body weight24 was followed by a dose of 0.5 mg per kilogram of predicted body weight every 6 hours for 14 days, a dose of 0.5 mg per kilogram of predicted body weight every 12 hours for 7 days, and then tapering of the dose. Study drug was tapered over a period of 4 days if 21 days of treatment had been completed and the patient was unable to breathe without assistance for a period of 48 hours. Tapering occurred over a two-day period if disseminated fungal infection or septic shock developed or the patient was able to breathe for a period of 48 hours without assistance.
The protocol specified how patients were to be weaned from the ventilator. Patients were assessed daily for weaning readiness and weaned by pressure-support ventilation when acceptable arterial oxygenation could be maintained with the use of an FiO2 of 0.5 or less.
Data on demographic characteristics, physiological characteristics, radiographic features, coexisting conditions, and medication were recorded at study entry and on days 1, 2, 3, 4, 5, 7, 14, 21, and 28. Patients were followed until they died, were discharged home while breathing without assistance, or day 180, whichever came first.
Patients were monitored daily for cardiovascular, renal, and hepatic failure and coagulation abnormalities, as defined previously, for 28 days.24 According to our definition, organs were considered failure-free after patients had been discharged from the hospital.
Any positive culture of material from a normally sterile site, when clinically available (cerebrospinal fluid, blood, pleural fluid, urine [
105 colonies per milliliter]), as well as of bile and abdominal (peritoneal) fluid, was recorded. Data from weekly chart reviews were recorded if evidence of serious infections (e.g., nosocomial pneumonia, disseminated fungal infection, or sepsis) or other infections was present (a complete list of prospectively defined infections is provided in the Supplementary Appendix). Bronchoalveolar-lavage fluid and plasma were obtained at study entry and on day 7 and processed as previously described25 (details are provided in the Supplementary Appendix). Plasma samples were stored in EDTA-treated tubes.
Events of Neuromyopathy
Early in the course of the study, six adverse events of neuropathy, myopathy, or myositis were reported among patients in one treatment group. At the request of the data and safety monitoring board, we reviewed the charts of all 88 previously enrolled patients for evidence of neuromyopathy, as defined by the presence of the terms "myopathy," "myositis," "neuropathy," "paralysis," or "unexplained weakness" in the medical record. Clinically available information pertaining to neuromyopathy was collected from the charts of patients identified in this manner. The data and safety monitoring board and an NHLBI-appointed neurologist reviewed these data and recommended the continuation of the study. The charts of the final 91 patients who were enrolled were reviewed prospectively.
Statistical Analysis
The primary outcome was mortality 60 days after enrollment. Patients discharged home while breathing without assistance before 60 days were defined as survivors. The comparisons were conducted according to the intention-to-treat principle. We originally estimated that 400 patients would need to be enrolled for the study to have a statistical power of 85 percent to detect an absolute difference in mortality rates of 15 percent between groups (50 percent in the placebo group vs. 35 percent in the methylprednisolone group). After two years, we resized the study to detect a decrease in mortality from 40 percent to 20 percent because of survival data from our previous trial24 and because the low rate of enrollment precluded timely completion of the original study. The modified sample calculation of 180 patients, approved by the data and safety monitoring board, provided the study with 85 percent power to detect this reduction in mortality at a two-sided significance level of 5 percent. Interim analyses were performed by the data and safety monitoring board after the enrollment of successive groups of approximately 60 patients with the use of symmetric stopping boundaries (two-sided 
=0.05).
Secondary outcomes included the number of ventilator-free days (the number of days patients were alive and breathing without assistance) during the first 28 days,24,26 the number of days without organ failure during the first 28 days, infectious complications during the first 28 days, and changes in markers of inflammation and fibroproliferation on study day 7. The following a priori determined covariates were examined for a treatment interaction: PaO2:FiO2, respiratory-system compliance, minute ventilation, time of entry after onset of ARDS (greater than 13 days vs. 7 to 13 days), number of nonpulmonary organs that failed, sepsis-initiated ARDS (vs. ARDS from other causes), and baseline levels of procollagen peptide type III. Two other covariates, tidal volume (expressed as milliliters per kilogram of predicted body weight) and date of enrollment, were added after the completion of the ARDS Clinical Trials Network tidal volume study.24 Interactions were identified with the use of the Wald test and a logistic-regression model including main effects and interaction terms. Fatality rates were compared with use of Fisher's exact test. We calculated Wilson's confidence intervals.27
After the study was completed, an analysis revealed that 52 patients were still hospitalized and 12 patients were still receiving mechanical ventilation 60 days after enrollment. To obtain more complete information on outcomes, we reviewed patients' records to determine their hospital-discharge status up to 180 days after enrollment. This review was performed at each site by the site investigators without knowledge of patients' treatment assignments.
Results
A total of 4123 patients were screened for the study, 659 (16 percent) of whom never met eligibility criteria. Of the remaining 3464 eligible patients, 180 (5 percent) were enrolled between day 7 and day 28 after the onset of ARDS: 91 were randomly assigned to the placebo group and 89 to the methylprednisolone group. The most common reason for exclusion was immunosuppression, which included the prior use of corticosteroids (22 percent) (Figure 1). All patients had complete 180-day follow-up for survival status.
|
|
|
|
The methylprednisolone group had significantly more ventilator-free days than the placebo group during the first 28 days (Table 2) as well as at 180 days (Table 3). Patients given methylprednisolone were able to breathe without assistance sooner than patients given placebo (14.1 days vs. 23.6 days, P=0.006) (Figure 2). As compared with the placebo group, the methylprednisolone group also had significantly fewer days in the intensive care unit (ICU) during the first 28 days (Table 2), but not at day 180 (Table 3). Survivors in the placebo group spent more days receiving assisted ventilation than did survivors in the methylprednisolone group (P=0.006). There were no significant differences between groups in the median number of days spent in the ICU or the hospital at 180 days (Table 3). At 180 days after enrollment, 58 patients in the placebo group and 57 patients in the methylprednisolone group had been discharged home while breathing without assistance.
|
Significant improvements in PaO2:FiO2 were observed in the methylprednisolone group on days 3 and 14 after enrollment. Plateau pressure was significantly lower in the methylprednisolone group on days 4, 5, and 7 after enrollment, as compared with the placebo group, whereas respiratory-system compliance was significantly improved on days 7 and 14 (Figure 4 of the Supplementary Appendix).
As compared with the placebo group, the methylprednisolone group had significantly more days without shock during the first 28 days. There were no significant differences between groups in the number of days without coagulation abnormalities or hepatic and renal failure (Table 2).
The mean serum glucose level was not significantly different between groups at baseline but was significantly higher in the methylprednisolone group than the placebo group on days 1 (Figure 3), 2, and 4. The difference in means decreased from 77.1 mg per deciliter (4.3 mmol per liter) on day 1 (95 percent confidence interval, 52.1 to 102.1 mg per deciliter [2.9 to 5.7 mmol per liter]) to 20.2 mg per deciliter (1.1 mmol per liter) on day 4 (95 percent confidence interval, 2.4 to 37.9 mg per deciliter [0.1 to 2.1 mmol per liter]) (Figure 5 of the Supplementary Appendix).
|
Fifty-eight adverse events were reported: 25 in the placebo group and 33 in the methylprednisolone group (P=0.27). Serious adverse events of neuromyopathy were reported in nine patients, all of whom were in the methylprednisolone group (P=0.001 for the comparison with the placebo group). Chart review identified 48 patients with neuromyopathy: 21 in the placebo group and 27 in the methylprednisolone group (P=0.31). No significant difference in the incidence of neuromyopathy was detected by retrospective, as compared with prospective, chart review (26 of 88 patients, as compared with 22 of 92). Most patients (71 percent) with neuromyopathy had bilateral weakness of the arms and legs. Sensory deficits were reported in 17 percent of patients with neuromyopathy.
There was no significant difference in the percentage of patients exposed to neuromuscular-blocking agents in the placebo and methylprednisolone groups (49 percent and 42 percent, respectively; P=0.30). The duration of exposure to neuromuscular-blocking agents was generally short (median, 0 days; interquartile range, 0 to 2.5). The use of neuromuscular-blocking agents largely occurred before entry into the study: 34 percent of patients in the placebo group and 28 percent of patients in the methylprednisolone group received a neuromuscular blocker before study day 0. Exposure to neuromuscular-blocking agents was not significantly more common among patients who were identified as having neuromyopathy. Data on the association between treatment group and weakness with the duration of mechanical ventilation are given in Table 7 of the Supplementary Appendix.
As compared with the placebo group, the methylprednisolone group had significantly lower plasma interleukin-6 levels and neutrophil counts in bronchoalveolar-lavage fluid between day 0 and 7 (Figure 6 and Figure 7 of the Supplementary Appendix). Systolic and diastolic systemic arterial pressures were, for the most part, higher in the methylprednisolone group than in the placebo group beginning on study day 3 (See Figure 8 of the Supplementary Appendix). The changes in procollagen peptide type III levels between day 0 and 7 did not differ significantly between the groups. Data on the distribution of baseline levels of procollagen peptide type III in bronchoalveolar-lavage fluid are given in Figure 10 of the Supplementary Appendix.
Discussion
In this clinical trial of 180 patients with persistent ARDS, we found no beneficial effect of corticosteroids on survival in the hospital. Furthermore, the initiation of methylprednisolone 2 or more weeks after the onset of ARDS was associated with a significantly increased mortality rate at 60 and 180 days as compared with that in the placebo group.
However, corticosteroids improved cardiopulmonary physiology within three to seven days after their initiation and altered the course of ARDS. Treatment with methylprednisolone increased the number of ventilator-free days, ICU-free days, and shock-free days during the first 28 days. The number of days without assisted ventilation remained higher in the methylprednisolone group than in the placebo group at 180 days. The duration of hospitalization did not differ significantly between the groups at any time. Patients in the methylprednisolone group were able to breathe without assistance earlier than were patients in the placebo group, but they were also more likely to resume receiving assisted ventilation (28 percent vs. 9 percent, P=0.006).
The increased rate of return to assisted breathing appears to be an important reason why early physiological improvements in the methylprednisolone group did not translate to a survival advantage; 8 of the 20 methylprednisolone-treated patients who resumed receiving assisted ventilation died, as compared with 3 of 6 patients in the placebo group. Possible explanations for this effect include complications of corticosteroid therapy, such as neuromyopathy; complications related to the withdrawal of corticosteroids, including shock,30 after patients were breathing without assistance; and pulmonary parenchymal causes of hypoxemic respiratory failure, such as recrudescence of fibroproliferation as a result of corticosteroid withdrawal.30 Overall, shock was observed less often with methylprednisolone than with placebo, yet shock was common among patients who resumed receiving assisted ventilation. Neuromyopathy occurred at similar rates in the groups overall but also was common in patients who resumed receiving assisted ventilation. Given the small number of events, it is difficult to make causal inferences regarding why the resumption of assisted ventilation was more common in the methylprednisolone group.
An increased risk of secondary infections was reported in studies of high-dose corticosteroids for sepsis and early ARDS11,12,19,20 but not in studies of more moderate doses.22,31 We did not detect an increased rate of nosocomial infections in the methylprednisolone group. There were more cases of nosocomial pneumonia, septic shock, and positive blood cultures in the placebo group. The lower rates of pneumonia and septic shock in the methylprednisolone group lead us to conclude that infections themselves do not explain the higher rate of return to assisted ventilation in this group.
Corticosteroids affect muscle function, but the effects of moderate-dose corticosteroids in critically ill patients are not well understood. Recent studies of patients receiving assisted ventilation have found a strong association between corticosteroid treatment and muscle weakness.23,32 Although the overall rate of clinically suspected neuromyopathy was similar in our two groups, all nine reports of serious adverse events related to neuropathy or myopathy were in patients treated with methylprednisolone. One explanation for this disparity is that methylprednisolone increased the severity but not the incidence of neuromyopathy. The similar incidence in the two groups may also be explained by the presence of competing risks for neuromyopathy: corticosteroid-induced myopathy and hyperglycemia in the methylprednisolone group and protracted assisted ventilation in the placebo group.
We did not systematically assess nerve conduction or muscle function, and the study was underpowered to detect absolute differences in the incidence of neuromyopathy of less than 25 percent. Thus, the true rate and severity of neuromuscular abnormalities in either group are unknown. Despite these shortcomings, our results confirm that neuromyopathy is common in patients with persistent ARDS and raise concern that methylprednisolone may lead to an increase in clinically important neuromyopathy and its consequences.
The duration of ARDS before treatment significantly interacted with the effect of corticosteroids on survival. Patients who were enrolled at least 14 days after the onset of ARDS and who were randomly assigned to receive methylprednisolone had a significantly higher case fatality rate than similar patients who were assigned to receive placebo. Perhaps patients surviving at least 14 days of ARDS have less active fibroproliferation, as reflected by lower baseline lung levels of procollagen peptide type III (median, 0.4 vs. 0.9 U per milliliter; P=0.01), and are less responsive to corticosteroids. In prior studies purporting to show a benefit from corticosteroids, the median duration of ARDS was similar to the interval allowed before enrollment in our trial8,13,14,15,16,18,33 and only two patients in our study had had ARDS longer than 21 days before enrollment (23 days and 26 days) (see Table 5 and Table 6 of the Supplementary Appendix). Thus, the optimal timing, if it exists, of methylprednisolone therapy during the course of ARDS is unknown.
This study lasted seven years and spanned a period in critical care medicine in which there were substantial changes in practice. During this time, the use of low tidal volumes for ventilatory support in patients with ARDS,24 tight blood glucose control,34 corticosteroids for refractory septic shock,31 drotrecogin-alfa for severe sepsis,35 preventive strategies for ventilator-associated pneumonia,36 and protocolized sedation with daily awakening37 became part of evidence-based practice in the ICU. Although we cannot exclude the possibility that these interventions affected the outcome among patients enrolled in our study, we did not find an interaction between the time or baseline tidal volume and outcome with methylprednisolone. Average glucose levels were higher in the methylprednisolone group than in the placebo group at several points during the study. This disparity in glycemic control may have contributed to the mortality and neuromyopathy outcomes.
Although this study was not powered to detect relatively small treatment effects, we did not observe the magnitude of effect reported in a previous study.18 Methylprednisolone did not increase infectious complications but may have increased the risk of neuromyopathy associated with critical illness. Our results do not provide support for the routine use of methylprednisolone in patients with persistent ARDS and suggest that methylprednisolone therapy may be harmful when initiated more than two weeks after the onset of ARDS.
* The participating institutions are listed in the Appendix.
Source Information
The members of the Writing Committee (Kenneth P. Steinberg, M.D., Leonard D. Hudson, M.D., Richard B. Goodman, M.D., and Catherine Lee Hough, M.D., University of Washington, Seattle; Paul N. Lanken, M.D., University of Pennsylvania, Philadelphia; Robert Hyzy, M.D., University of Michigan, Ann Arbor; and B. Taylor Thompson, M.D., and Marek Ancukiewicz, Ph.D., Massachusetts General Hospital, Boston) of the National Heart, Lung, and Blood Institute ARDS Clinical Trials Network assume responsibility for the overall content and integrity of the article.
Address reprint requests to Dr. Steinberg at the Division of Pulmonary and Critical Care Medicine, Harborview Medical Center, Box 359762, 325 Ninth Ave., Seattle, WA 98104, or at steinkp{at}u.washington.edu.
References
Participants in the National Heart, Lung, and Blood Institute ARDS Clinical Trials Network were as follows (principal investigators and members of the steering committee are marked with an asterisk): Investigators: Cleveland Clinic Foundation — H.P. Wiedemann,* A.C. Arroliga, C.J. Fisher, Jr., J.J. Komara, Jr., P. Periz-Trepichio; Denver Health Medical Center — P.E. Parsons, Denver Veterans Affairs Medical Center — C. Welsh; Duke University Medical Center — W.J. Fulkerson, Jr.,* N. MacIntyre, L. Mallatratt, M. Sebastian, J. Davies, E. Van Dyne, J. Govert; Johns Hopkins Bayview Medical Center — J. Sevransky, S. Murray; Johns Hopkins Hospital — R.G. Brower, D. Thompson, H.E. Fessler, S. Murray; LDS Hospital — A.H. Morris,* T. Clemmer, R. Davis, J. Orme, Jr., L. Weaver, C. Grissom, F. Thomas, M. Gleich (deceased); McKay-Dee Hospital — C. Lawton, J. D'Hulst; MetroHealth Medical Center of Cleveland — J.R. Peerless, C. Smith; San Francisco General Hospital Medical Center — R. Kallet, J.M. Luce; Thomas Jefferson University Hospital — J. Gottlieb, P. Park, A. Girod, L. Yannarell; University of California, San Francisco — M.A. Matthay,* M.D. Eisner, J. Luce, B. Daniel, R. Kallet; University of Colorado Health Sciences Center — E. Abraham,* F. Piedalue, R. Jagusch, P. Miller, R. McIntyre, K.E. Greene; University of Maryland — H.J. Silverman,* C. Shanholtz, W. Corral; University of Michigan — G.B. Toews,* D. Arnoldi, R.H. Bartlett, R. Dechert, R. Hyzy, C. Watts; University of Pennsylvania — P.N. Lanken,* J.D. Christie, B. Finkel, B.D. Fuchs, C.W. Hanson III, P.M. Reilly, M.B. Shapiro; University of Utah Hospital — R. Barton, M. Mone; University of Washington–Harborview Medical Center — L.D. Hudson,* T. Berry-Bell, G. Carter, C.L. Cooper, A. Hiemstra, R.V. Maier, M.J. Neff, K.P. Steinberg; Utah Valley Regional Medical Center — T. Hill, P. Thaut; Vanderbilt University — A.P. Wheeler,* G.R. Bernard, B. Christman, S. Bozeman, L. Collins, T. Swope, L.B. Ware; Baylor College of Medicine–Ben Taub General Hospital—K. Guntupalli,* V. Bandi, C. Pope; Wake Forest University — R.D. Hite,* P. Morris, A. Howard; Clinical Coordinating Center: Massachusetts General Hospital, Harvard Medical School — D.A. Schoenfeld,* B.T. Thompson, M. Ancukiewicz, D. Hayden, F. Molay, N. Ringwood, C. Oldmixon, A. Korpak, R. Morse; National Heart, Lung, and Blood Institute Staff: D.B. Gail, A. Harabin,* P. Lew, M. Waclawiw; Steering Committee: G.R. Bernard (chair); Data and Safety Monitoring Board: R.G. Spragg (chair), J. Boyett, J. Kelley, K. Leeper, M. Gray Secundy, A.S. Slutsky, B. Turnbull, G. Corbie-Smith; Protocol Review Committee: J.G.N. Garcia (chair), S.S. Emerson, S.K. Pingleton, M.D. Shasby, W.J. Sibbald.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Corticosteroids in ARDS
Marik P., Pastores S., Annane D., Speich R., Schmid C., Stocker R., Okamoto H., DiNubile M. J., Steinberg K. P., Hudson L. D., Thompson B. T., the NHBLI ARDS Clinical Trials Network
Extract |
Full Text |
PDF
N Engl J Med 2006;
355:316-319, Jul 20, 2006.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
