Key Point

If started 14 days after disease onset, corticosteroids are not effective for ARDS, but if started early and tapered slowly, they may be beneficial.

PHOENIX—ARDS usually arises from various injuries to the lung that accompany critical illness. The treatment for ARDS differs according to time of onset and severity and duration of illness. Corticosteroids have been used with varying degrees of success in ARDS but are still a point of contention among intensivists. At this year’s Critical Care Congress, Leonard D. Hudson, MD, and Gianfranco U. Meduri, MD, discussed the findings of their respective research on corticosteroids in ARDS patients.¹

ARDSNet

“What is the role of steroids in ARDS?” asked Dr. Hudson, Professor of Medicine at the University of Washington in Seattle. He presented the preliminary results of a multicenter, prospective, randomized, double-blind, controlled clinical trial undertaken by ARDSNet—a network initiated by the National Heart, Lung, and Blood Institute, NIH, to carry out multicenter clinical trials of ARDS treatments.

Patients in the study had had ARDS for at least seven but not more than 24 days before study entry, had persistent bilateral infiltrates, and were on continuous mechanical ventilation. The primary outcome of ARDSNet was the percentage of patients alive 60 days after study entry. Secondary outcomes were the number of ventilator-free days at 28 days and organ failure–free days at 28 days.

The treatment group received methylprednisolone starting with a 2-mg/kg IV bolus, followed by 2 mg/kg/d in four divided doses for 14 days and then 1 mg/kg/d in two divided doses for the next seven days. Thus, there was rapid tapering of corticosteroids after 21 days.

Patients treated with corticosteroids were taken off the ventilator after a mean of 14 days, compared with 23 days for the placebo group. Dr. Hudson noted that this was highly significant. Patients treated with corticosteroids also had improvement in oxygenation and static compliance.

However, once they were taken off mechanical ventilation, patients in the corticosteroid group were put back on the ventilator more frequently than were those in the placebo group (20 patients vs six, respectively).

The primary outcome (60-day, in-hospital mortality) was the same for both groups. Of note was the observation that patients who had ARDS for more than 14 days prior to study entry fared much worse when treated with corticosteroids. Dr. Hudson concluded that very late treatment of ARDS with corticosteroids was less effective than earlier treatment.

As far as secondary outcomes were concerned, the number of cardiovascular organ failure–free days was significantly greater in the corticosteroid group. There were also significantly more suspected or confirmed cases of pneumonia in the control group than in the corticosteroid group (14 vs 5.6, respectively). Likewise, septic shock occurred more frequently in the control group. On the other hand, central nervous system and neuromuscular events occurred in 10 patients in the corticosteroid group but none of the control group.

Dr. Hudson concluded that “ARDSNet cannot recommend methylprednisolone for persistent ARDS.” He noted that there are several practical interpretations of the study’s findings. “One could be don’t use steroids after 14 days. Another could be don’t use steroids at all in persistent ARDS.” He added that corticosteroids could be useful for their morbidity benefits (ie, ventilator-free days) and could also be appropriate for selected patients.

TIMING AND DURATION CRITICAL

“What affects response to prolonged glucocorticoids?” asked Dr. Meduri, a Professor of Medicine at the University of Tennessee Health Science Center in Memphis. “The answer is timing of therapy and duration of treatment.” He and his coworkers conducted a study very similar in design to that of the ARDSNet study, except that instead of rapid tapering of corticosteroids, the patients’ doses were tapered very slowly. His study found a reduction in mortality for the corticosteroid-treated group.

One practice included in Dr. Meduri’s study was infection surveillance. Because of the blunted febrile response in patients receiving corticosteroids, he explained, infections may go undetected. At study entry, bronchoscopy with bilateral bronchoalveolar lavage was performed; it was repeated weekly thereafter.

Dr. Meduri noted that prior to randomization, 50% of patients in the corticosteroid group had infections, compared to 25% in the placebo group. However, “weekly bronchoalveolar lavage helped avoid ventilator-associated pneumonia by identifying it before steroids were started,” he said.

Dr. Meduri concurred with Dr. Hudson’s finding that delaying initiation of corticosteroid treatment may mean that patients won’t respond. According to the literature, he continued, reasons for not responding to corticosteroids include cellular fibrosis with a loss of alveolar architecture, accelerated fibroproliferation, a higher level of inflammation in the early phase of ARDS, and possible liver failure.

In both the ARDSNet and Dr. Meduri’s study, the findings after 14 days were similar.

Dr. Meduri outlined his argument for prolonging treatment with corticosteroids. He noted that:

Short-term therapy not only increases the risk of rebound complications, but the patient is also at risk if infection develops. Short-term steroids actually amplify the inflammatory response.

So, can treatment with corticosteroids prevent progression to ARDS? The answer is not clear. In any case, said Dr. Meduri, treatment should be started early because in late ARDS, corticosteroids are ineffective. In addition, infection surveillance should be performed throughout the treatment course.

—Gale Jurasek

Reference
1. Slutsky A. Acute lung injury. Presented at: annual meeting of the Society for Critical Care Medicine; January 18, 2005; Phoenix, Ariz.

Return to table of contents