The Use of Corticosteroids in Treating Infectious Diseases
Abstract & Commentary
By David J. Pierson, MD, Editor, Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, is Editor for Critical Care Alert.
Synopsis: Although clinicians have historically assumed that corticosteroid therapy is harmful for patients with serious infections, this comprehensive review of the evidence on the subject shows that, with a few unusual exceptions and in the presence of appropriate antimicrobial therapy, corticosteroids are not harmful but actually beneficial in most settings.
Source: McGee S, et al. Arch Intern Med. 2008;168(10):1034-1046.
Systemic corticosteroids are immunosuppressive and associated with serious long-term side effects. Primarily for these reasons, clinicians have long avoided their use in patients with known infections. To determine whether such avoidance is appropriate in the era of effective antimicrobials, McGee and Hirschmann performed a comprehensive review of randomized clinical trials of corticosteroids in patients with serious infections. Using rigorous methods, they searched for all English-language reports of randomized, double-blind trials comparing corticosteroids and placebo in infections. Except for some trials of viral infections, sore throat, and cerebral cysticercosis, all patients in the included studies also received appropriate antimicrobial agents. Excluded were studies in which corticosteroids were not administered systemically, and also trials in patients with septic shock. One hundred ninety trials met the authors' criteria and were evaluated. Except for those in patients with Pneumocystis jirovecii (carinii) pneumonia (PCP), the studies generally excluded patients with underlying immunosuppression.
Patients with acute herpes zoster experienced more rapid clinical improvement with systemic corticosteroid therapy, particularly if this was initiated within a few days of symptom onset, although there was no effect on the incidence of postherpetic neuralgia. Corticosteroids produced modest clinical improvement in infectious mononucleosis but had no effect on the duration of illness or time lost from school or work. Treatment of viral hepatitis with corticosteroids (primarily hepatitis B) led to worse outcomes as compared to placebo, including increased viral replication, delayed biochemical remission, more frequent relapses, and higher mortality. In children with acute laryngotracheobronchitis (croup), steroid treatment was effective regardless of illness severity, shortening the duration of illness, reducing the need for hospitalization and for intubation, and shortening the duration of intubation when the latter was required.
Corticosteroids were effective in patients with non-tuberculous bacterial infections. For bacterial meningitis, in 15 randomized trials including more than 2000 patients in all age groups, there were some differences in effectiveness in different examined variables, but in general both survival and clinical course were positively affected, particularly in adult patients. No trials found corticosteroid therapy to be harmful when administered prior to or shortly after initiation of antibiotic therapy. In bacterial pneumonia, few studies have examined the effects of corticosteroids; 2 trials in patients with pneumococcal pneumonia from the 1950s showed accelerated clinical improvement; more recent studies in more severely ill patients have been small but have demonstrated no significant benefits. In one study of 123 children with septic arthritis, corticosteroid therapy accelerated clinical improvement and was associated with considerably less residual joint dysfunction at 1 year. Available trials in acute pharyngitis and peritonsillar abscess showed faster symptom resolution and shorter duration of illness with corticosteroids. Similarly, in a single trial in patients hospitalized with mainly lower extremity cellulitis, the durations of fever, erythema, and hospitalization were shorter with 8 days of corticosteroid therapy. In severe tetanus, the one available randomized trial showed decreased mortality.
In studies of patients with different forms of tuberculosis, corticosteroids were also shown to be beneficial, although the effects on mortality varied. Numerous studies in patients with pulmonary tuberculosis, in both HIV-infected and non-HIV-infected patients, have shown more rapid clinical improvement (such as defervescence, improvement in cough, rapidity of clearing acid-fast bacilli from sputum by smear, radiographic resolution, and weight gain), but no differences in mortality. Patients with pleural tuberculosis generally experienced more rapid clinical improvement, but no differences in radiographic or pulmonary function outcome or in mortality, with corticosteroids. In contrast, studies in patients with tuberculous meningitis and tuberculous pericarditis demonstrate reduced mortality and improved long-term clinical outcomes in addition to more rapid short-term improvement when corticosteroids are used. No studies of military tuberculosis, or of pulmonary tuberculosis specifically causing respiratory failure, were included.
Finally, corticosteroids have been shown to be beneficial in HIV-infected patients with PCP that is moderate-to-severe. In 6 randomized studies of 489 patients, administration of corticosteroids for 10 to 21 days along with appropriate antimicrobial therapy accelerated the resolution of fever, tachypnea, and dyspnea, and reduced both the development of respiratory failure and mortality.
The table summarizes the effects of corticosteroid therapy in 15 of the 25 infections evaluated by McGee and Hirschmann that are most relevant to ICU practice.
Commentary
Many of the studies included in this exhaustive review are decades old. However, the authors have applied rigorous criteria for applicability to current practice, and have also included all available more recent trials. The bottom line is that, except in patients with viral hepatitis or cerebral malaria, corticosteroids are either not harmful or frankly beneficial in patients with serious infections—assuming that appropriate antimicrobial therapy is also being administered. Corticosteroid therapy for septic shock is a different matter not addressed by this study.
As the authors point out, based on strong support from the literature, patients with bacterial meningitis, tuberculous meningitis, tuberculous pericarditis, moderate or severe PCP, or tetanus should receive steroid therapy. Although the evidence base is less extensive, they also recommend this therapy for patients with septic arthritis. One can also make a good case for adding corticosteroids—at least in the short term—for patients with infections in the "improved symptoms" column in the accompanying table, regardless of whether reduced mortality or other outcomes have been shown. McGee and Hirschmann caution that courses of corticosteroids longer than 3 weeks should be withheld from HIV-infected patients with low CD4 counts, because of the risk of opportunistic infection.
Systemic corticosteroids are immunosuppressive and associated with serious long-term side effects. Primarily for these reasons, clinicians have long avoided their use in patients with known infections.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content