Key Point:

Whether ACTH stimulation test results can identify patients who might benefit from corticosteroids is presently unclear.

ORLANDO, FLA—Whether the results of an adrenocorticotropic hormone (ACTH) stimulation test should be used to guide corticosteroid therapy for adrenal insufficiency in patients with septic shock is a matter of debate. At the annual meeting of the Society of Critical Care Medicine, Jean-François Dhainaut, MD, argued that an ACTH test may be necessary, depending on the patient’s cortisol level.[1] His opponent in a pro-con debate, Charles L. Sprung, MD, countered that in most critical care settings, the ACTH test is of no use at all.

ARGUMENTS FOR THE TEST

To support his assertion, Dr. Dhainaut, who chairs the Department of Critical Care Medicine at Cochin Hospital in Paris, cited the recent New England Journal of Medicine review by Cooper and Stewart, which suggested that a cortisol level below 15 µg/dL indicates adrenal insufficiency and a level above 34 µg/dL tends to rule it out. Because these authors consider cortisol levels between 15 and 34 µg/dL to be inconclusive, they recommend that an ACTH test be used to check for adrenal insufficiency.

Dr. Dhainaut agrees with Cooper and Stewart that adrenal insufficiency is likely if the cortisol change after an ACTH test is less than 9 µg/dL. A larger cortisol response, he said, indicates one of two things:

• The risk of adrenal insufficiency is low.
• Steroid resistance has developed.

In patients with septic shock, these cutoffs, especially the ACTH test response, have been shown to correlate with mortality, related Dr. Dhainaut. They also appear to identify patients who may benefit from supplemental corticosteroids.

For example, in a phase III trial of low-dose hydrocortisone and fludrocortisone for severe septic shock, Annane et al stratified patients with septic shock according to their ACTH test response. Those whose cortisol levels rose by more than 9 µg/dL were considered responders; nonresponders were those with smaller cortisol changes. Among the responders, corticosteroid administration had no impact on 28-day mortality; in contrast, it lowered the relative risk of 28-day mortality by 46% in the nonresponders. Furthermore, corticosteroid use shortened the amount of time that vasopressors were needed in the nonresponders, but not in the responders.

In this study, 77% of the patients were nonresponders. Thus, three quarters of septic shock patients may benefit from corticosteroid therapy, Dr. Dhainaut believes.

He recommends that cortisol replacement for septic shock be initiated immediately after the ACTH test is administered if the results will not be available in a timely manner. Once the results are obtained, Dr. Dhainaut said, treatment can be continued for up to seven days in patients who are nonresponders; it should be stopped in responders.

ACTH TEST IS USELESS

Dr. Sprung countered that because the study by Annane et al did not include typical septic shock patients, its results cannot be used to identify those who are most likely to gain a survival benefit from corticosteroids. The 28-day mortality rate in the placebo cohort was 63%, he noted, which is unusually high. Conversely, the 23% rate of responders is unusually low; indeed, the small number of responders made it difficult to determine whether corticosteroids affected survival in this group.

Furthermore, Bollaert et al have clearly shown that mortality is reduced and shock is reversed when hydrocortisone is administered in late septic shock, regardless of a patient’s response to the ACTH test. This indicates that the test is unnecessary for predicting survival or response to treatment, said Dr. Sprung, who is Director of the General Intensive Care Unit at Hadassah Hebrew University Medical Center in Jerusalem.

Even if one accepts the hypothesis that ACTH nonresponders should be treated with corticosteroids, he noted, a problem remains: It is still quite difficult to identify them accurately. The results of a single ACTH test are highly variable, often poorly reflecting the 24-hour cortisol average, Dr. Sprung explained. In addition, there is a high degree of variability in test results between hospitals.

The ACTH test is also unlikely to be useful for determining relative adrenal insufficiency in septic shock patients because there is currently no consensus on how to define or diagnose that disorder. An even greater problem, said Dr. Sprung, is that in septic shock patients with high baseline cortisol levels, little or no change in those levels may be seen after the ACTH test—giving the false impression of adrenal hyporesponsiveness when, in fact, the patients actually have maximal adrenal stimulation. Because of the infrequency of this condition, Dr. Sprung believes that ACTH testing for absolute adrenal insufficiency in septic shock patients is probably not warranted.

—Timothy Begany

Reference
1. Dhainaut J-F, Sprung CL. Pro-con: Steroid therapy of septic shock should be guided by results of an ACTH stimulation test. Presented at: 33rd Critical Care Congress; February 22, 2004; Orlando, Fla.