SEATTLE—Although high-dose corticosteroid use in patients with sepsis or septic shock remains controversial, the need for affordable therapy and a better understanding of adrenal insufficiency have prompted a reexamination of the use of these agents in the ICU. At the annual meeting of the American Thoracic Society, experts discussed the evidence for including low-dose corticosteroids when treating patients with septic shock or the adult respiratory distress syndrome (ARDS).[1]

Current practice is not always driven by evidence, acknowledged Kenneth E. Wood, DO, Associate Professor of Medicine at the University of Wisconsin, Madison. For example, while rounding at one ICU, Dr. Wood noticed that all septic shock patients in the unit were receiving corticosteroids. He recounted, “I got a very quick response: ‘That’s the standard of care. Besides, it’s a lot cheaper’ than other current therapies.”

However, corticosteroids may be highly effective for one ICU patient but not for another. Even in those settings in which corticosteroid administration appears justified, questions remain.

SEPTIC SHOCK

By up-regulating inducible nitric oxide synthase (iNOS), lipopolysaccharide can impair contractile responses of the vasculature to norepinephrine (NE), thereby adversely affecting hemodynamics in septic shock patients. Because cortisol can block iNOS up-regulation, it is important for maintaining NE responsiveness, explained Djillali Annane, MD, PhD, Professor of Medicine at the University of Paris Ouest in Garches, France. Inflammatory cytokines can impair cortisol release; therefore, replacement may be beneficial in cases of adrenal insufficiency.

Dr. Annane tested this approach in nine septic shock patients and six healthy volunteers; adrenocorticotropic hormone (ACTH) measurement revealed that five of the septic shock patients had impaired adrenal reserve.[2] Compared with the healthy volunteers and the other septic shock patients, the five with adrenal insufficiency showed depressed responses to NE administration. Intravenous injection of 50 mg hydrocortisone improved NE responsiveness in all of the septic shock patients, but the effect was especially marked in those with impaired adrenal reserve. “When we treated these patients with hydrocortisone, the pressor response to NE was fully normalized,” Dr. Annane said. However, hydrocortisone had no effect on mean arterial pressure in the absence of NE administration.

These findings offer two important clinical insights. First, said Dr. Annane, “Patients who are not on vasopressors … may not benefit from hydrocortisone.” In other words, low corticosteroid doses may be inappropriate for sepsis patients who do not have shock. Second, “Only patients with a blunted response to the ACTH test have blunted responses of mean arterial pressure to NE” and therefore benefit from hydrocortisone.

Results of another experiment suggest that restoration of NE responsiveness in septic shock patients corresponds to changes in vessel contractility. Furthermore, Dr. Annane pointed out, hydrocortisone-induced increases in arterial pressure yield other benefits: Afterload increases, while heart rate and cardiac index decrease. Consequently, low-dose corticosteroids hasten weaning from vasopressors. Yet low-dose hydrocortisone does not produce pulmonary hypertension in septic shock patients.

These benefits translate to reduced mortality in ICU patients with adrenal insufficiency. In a multicenter one-week trial[3] that compared low-dose hydrocortisone and fludrocortisone with placebo in 300 septic shock patients, Dr. Annane and colleagues found “a significant improvement in one-month survival in patients who received steroids.” These results are in marked contrast to those of previous studies.

ARDS

Despite disappointing results of short (less than or equal to 24-h) trials of high-dose corticosteroids in ARDS patients, interest in these drugs has continued because of the hope that they might prove effective against the persistent fibroproliferation that sometimes develops in ARDS patients. New evidence suggests that prolonged treatment with low-dose corticosteroids may benefit patients with unresolving ARDS through multiple pathways. Said G. Umberto Meduri, MD, “I believe glucocorticoids do much more in patients with unresolving ARDS than just reversing the process of fibroproliferation.”

Dr. Meduri, Director of the Memphis Lung Research Program at University of Tennessee, outlined a cascade of anti-inflammatory effects triggered by activation of the glucocorticoid receptor. “NFkappaB is the central transcription factor that drives the inflammatory response,” he noted. “It is maintained in an inactive form in the cytoplasm by the inhibitory protein, IkappaB.” If activated, however, NFkappaB binds to the DNA to trigger transcription of genes including inflammatory cytokines, chemokines, adhesion molecules, iNOS, cyclooxygenase, and phospholipase A2.

“The glucocorticoid receptor is the most important down-regulator of NFkappaB … activity,” Dr. Meduri noted. It up-regulates IkappaB, which can interact with NFkappaB bound to nuclear DNA, thereby interfering with expression of inflammatory mediators. “Some believe this is the most important mechanism by which glucocorticoids block transcription of cytokines and chemokines,” he added. However, the activated glucocorticoid receptor can also interact directly with activated NFkappaB to block both factors’ effects on transcription.

“Two cytokines—TNF-alpha [tumor necrosis factor alpha] and IL-1ß [interleukin-1ß]—are the most important proximal mediators of the host defense,” said Dr. Meduri. But, “release of TNF and IL-1 in the circulation from the local site of inflammation activates the hypothalamic-pituitary axis in a graded manner.” The resulting cortisol production suppresses release of inflammatory mediators.

Dr. Meduri and colleagues[4] examined these mediators in 28 patients with ARDS. The researchers divided the 24 patients who survived at least one week into two groups based on changes in lung injury scores by day 8. The seven patients whose scores had dropped were termed “improvers,” whereas the 17 patients whose scores increased or remained the same were “nonimprovers.”

TNF-alpha and IL-1 levels steadily declined in improvers, but elevated cytokine levels persisted in nonimprovers. In improvers, glucocorticoid receptor binding to the DNA progressively increased, whereas NFkappaB binding declined; this suggests that the former down-regulates the latter. In nonimprovers, glucocorticoid receptor activity was steady, but NFkappaB activity progressively increased; this increase was even more dramatic in the four patients who died within the first week.

Expression of TNF-alpha and IL-1 paralleled NFkappaB activity. Transcription of the anti-inflammatory cytokine IL-10 increased in the patients who improved, but it remained unchanged in those who failed to improve.

“There are two types of ARDS, based on lung injury score: resolving and unresolving,” Dr. Meduri postulated. “Patients with unresolving ARDS may have endogenous glucocorticoid inadequacy and/or excessive systemic inflammation-induced peripheral glucocorticoid resistance.” In patients with resolving ARDS, responses are driven by the glucocorticoid receptor. NFkappaB drives responses in those with unresolving ARDS; in these patients, ACTH and cortisol levels remain elevated, but they fail to quell systemic and pulmonary inflammation, yielding persistent elevation in lung injury scores.

“If acquired glucocorticoid resistance plays a role in the pathogenesis of unresolving ARDS,” Dr. Meduri speculated, then corticosteroid administration should produce “a reduction in inflammatory cytokines over time, a reduction in ACTH and cortisol, and an improvement in organ dysfunction.” He and colleagues tested this idea in 17 ARDS patients with lung injury that persisted after seven days of conventional treatment; the patients received 2 mg/kg/d methylprednisolone or placebo.[5]

Methylprednisolone reduced the extent of both lung injury and multiple organ dysfunction, significantly improving ICU and hospital survival, while reducing the duration of mechanical ventilation. It also lowered circulating levels of ACTH and cortisol as well as of fibroproliferation markers in plasma and bronchoalveolar lavage. Although both TNF-alpha and IL-1 concentrations declined in control patients, levels decreased significantly more in the patients given methylprednisolone. Cells exposed to plasma from patients receiving prolonged corticosteroid treatment showed enhanced binding of the glucocorticoid receptor to DNA and consequent transcription of IL-10, with reduced NFkappaB activation and transcription of TNF-alpha and IL-1ß.[6]

LESS IS MORE

A take-home message, said Dr. Wood, is that lower doses of steroids have what appears to be benefit, whereas higher doses seem to fail. Nevertheless, uncertainties remain. Optimum dosage levels are unclear, and there is no agreement on when to start treatment. It is also not yet certain how relative adrenal insufficiency should be diagnosed in critically ill patients, particularly now that the ACTH test is no longer available in the United States. “There is no real, clinically available tool, other than clinical response, to assess the adrenal receptor phenomenon,” Dr. Wood pointed out. However, for patients with septic shock or ARDS who have not responded to other therapies, a trial of corticosteroids may be warranted.

—Mimi Zucker, PhD

References
1. Meduri U, Van Den Berghe G, chairs. New understanding of metabolic disturbances in the critically ill patient. Presented at: annual meeting of the American Thoracic Society; May 20, 2003; Seattle, Wash.
2. Annane D, Bellissant E, Sebille V, et al. Impaired pressor sensitivity to noradrenaline in septic shock patients with and without impaired adrenal function reserve. Br J Clin Pharmacol. 1998;46:589-597.
3. Annane D, Sébille V, Charpentier C, et al. Effect of treatment with low doses of hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA. 2002;288:862-871.
4. Stentz F, Tolley EA, Chrousos G, Meduri GU. Mechanisms of NF-kB and glucocorticoid receptors in activation and regulation of systemic inflammation in ARDS. Am J Respir Crit Care Med. 2001;163:A450.
5. Meduri GU, Headley AS, Golden E, et al. Effect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: a randomized controlled trial. JAMA. 1998;280:159-165.
6. Meduri GU, Tolley EA, Chrousos G, Stentz F. The effects of prolonged methylprednisolone treatment on dysregulated systemic inflammation in patients with unresolving ARDS: time dependent correction of immune system glucocorticoid resistance. Am J Respir Crit Care Med. 2002;165:983-991.

Return to table of contents