Meningococcal meningitis, dexamethasone, and Class III evidence.

Neurology 2012;79:1528–1529 In bacterial meningitis, lysis of bacteria during treatment triggers the release of cytokines that increase intrathecal inflammation. Studies with animals have shown that adjunctive treatment with glucocorticoids reduces both the inflammatory response and neurologic sequelae.1 Such treatment in humans, however, has been controversial for decades despite 44 clinical trials (29 randomized) and 10 meta-analyses.1–4 The most recent Cochrane review concluded that adjunctive dexamethasone significantly reduced mortality in meningitis due to Streptococcus pneumoniae but not Neisseria meningitidis, with benefit in adults and children in high-income but not lowincome countries.5 Dexamethasone was not associated with serious adverse effects, and the authors recommended a 4-day regimen of 0.6 mg/kg/day given before or with the first dose of antibiotic. Studies that failed to show significant benefit in patients with meningococcal meningitis often showed a trend toward reduced mortality or neurologic complication; failure to reach statistical significance was plausibly attributable to insufficient numbers of patients. To achieve adequate statistical power would require a study much larger than any conducted to date, an unlikely prospect, especially with a declining incidence of meningococcal meningitis since the introduction of Type C vaccine. Moreover, dexamethasone should be given prior to or at the time of the first antibiotic dose before bacterial lysis occurs. Most patients with meningococcal meningitis will therefore receive dexamethasone at least until the organism is identified. The 2010 Cochrane review did not exclude N meningitidis from its recommendation to treat “acute bacterial meningitis” with adjunctive corticosteroids. Conversely, some American and European guidelines, apparently unwilling to base recommendations on a trend (while also considering potential hazards of corticosteroids), advise that dexamethasone be discontinued should N meningitidis be identified as the organism.6,7 In this issue of Neurology, Heckenberg and coworkers8 compare 2 cohorts of patients with meningococcal meningitis, 1 during 1998 –2002 before treatment with adjunctive dexamethasone became widely used and 1 during 2006 –2011 after it became standard therapy. Two years ago the same authors compared similar cohorts of patients with pneumococcal meningitis and reported reduced mortality and unfavorable outcome among those treated during 2006 –2009, compared to those treated during 1998 –2002.9 The present study found that despite recommendations to stop dexamethasone once N meningitidis is diagnosed, 90% of patients with meningococcal meningitis during 2006 –2011 received it. Differences between cohorts, however, were less dramatic than in the S pneumoniae study. The major finding was that dexamethasone was not harmful. There was no significant reduction in death or neurologic complication in the 2006 –2011 cohort compared to the 1998 –2002 cohort. The only significant benefit, in fact, was a reduced likelihood of immune-mediated arthritis. As with earlier randomized studies, there was “a favorable trend [toward reduced rates] for death and hearing loss” between the cohorts, and that trend, plus evidence of no harm, allowed the authors to conclude, “... there is no need to discontinue empiric treatment with adjunctive dexamethasone in patients with culture-proven meningococcal meningitis.” These recommendations are based on Class III level of evidence, but as Gross and Johnston10 stressed in a Neurology editorial, Class III or IV evidence is not inherently flawed. Although their data are less than definitive, Heckenberg and coworkers have added to cumulative evidence supporting the use of adjunctive dexamethasone in bacterial meningitis, including meningococcal. The findings of this report will hardly resolve the controversy, but at the present time the burden of proof is on those who would withhold adjunctive dexamethasone in patients with suspected or proven meningococcal meningitis.