De Gans J, Van De Beek D, European Dexamethasone in Adulthood Bacterial Meningitis Study Investigators. Dexamethasone in adults with bacterial meningitis.
Overall Study Question
The purpose of this study was to determine if adjuvant treatment with dexamethasone improves outcomes in adults with acute bacterial meningitis. Evidence supporting adjunctive treatment of acute bacterial meningitis with corticosteroids is controversial at best. A meta-analysis of studies conducted in children only showed benefit in reducing hearing loss in those with Haemophilus influenzae type b meningitis. Previously published reports involving adults have either been inconclusive, poorly designed, or inadequately powered.
This current study involved 301 patients from five European countries who were enrolled from 1993 to 2001. Adults with suspected meningitis were randomized to receive either dexamethasone or placebo in addition to empiric antibiotic therapy. The dexamethasone was given as a 10 mg dose administered intravenously 15 to 20 minutes prior to, or in combination with, the first dose of antibiotic and then every six hours for four days.
The primary study outcome was the Glasgow Outcome Scale (GOS) score. This was assessed by the patients’ physician at eight weeks following randomization. The secondary outcome measures were death, focal neurological abnormalities, hearing loss, gastrointestinal bleeding, fungal infection, herpes zoster infection and hyperglycemia. An unfavorable outcome was defined as a GOS score of 1 to 4, with a favorable outcome defined as a score of five (1-death; 2-vegetative state; 3-severe disability; 4-moderate disability; 5-mild or no disability).
What was the declared relationship between authors and sponsors for this study?
This study was supported, in part, by a grant from NV Organon, the suppliers of the study medication. No relationship between any member of the European Dexamethasone in Adulthood Bacterial Meningitis Study Investigators and the sponsors of this study was declared.
Are the Results of the Study Valid?
1. Was assignment of patients randomized?
2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?
All 301 patients were accounted for and included in the intention-to-treat analysis. The last recorded observation was carried forward for 32 patients (11 dexamethasone group; 21 placebo group) who died prior to the 8-week evaluation point, and seven patients (3 dexamethasone group; 4 placebo group) were lost to follow-up. In addition, 11 patients in each treatment group withdrew early from treatment but were included in the analysis.
3. Were patients, their clinicians, and study personnel ‘blind’ to treatment?
Yes. Treatment assignments were concealed from all investigators. Two emergencies occurred during the tenure of the trial and investigators were required to obtain access to patients’ treatment assignments.
4. Were the groups similar at the start of the trial?
The two groups appeared to be similar in terms of the measured baseline characteristics, although seizures occurred more frequently in the dexamethasone group than the placebo group (10% vs. 5%, respectively). Culture results and yield appeared similar between the two groups, however, statistical analysis was not described to assess similarity at baseline.
5. Aside from the experimental intervention, were the groups treated equally?
For both groups, empiric antimicrobial treatment consisted of amoxicillin (2 g IV every 4 hours) as monotherapy after which the antibiotic regimen could be tailored according to culture results. However, the protocol was amended at the time of the interim analysis to allow for individualized empiric antibiotic choice by the patient’s physician. Unfortunately, the elapsed time between patient arrival and initial treatment was not recorded for either group. A difference in time to treatment between groups could potentially affect the overall outcome since prompt treatment of bacterial meningitis is vital. The authors claim, however, that since other baseline characteristics were similar it is unlikely that the results were confounded.
What were the Results?
1. How large was the treatment effect?
At the 8-week evaluation point, an unfavorable outcome (GOS value of 1to 4) was observed in 15% of the dexamethasone treated patients and 25% of the placebo treated patients (relative risk reduction [RRR] 41%, P=0.03). Death occurred in 7% of dexamethasone treated patients, as compared to 15% of the placebo treated group (RRR 52%, P =0.04). Therefore the authors concluded that adjuvant therapy with dexamethasone can significantly reduce the risk of an unfavorable neurological outcome as well as provide a mortality benefit in adult patients with bacterial meningitis. Subgroup analyses suggested that these benefits may be more apparent in patients with pneumococcal meningitis, but this study was not powered to address this particular hypothesis. There was no statistically significant difference in the incidence of focal neurological abnormalities, hearing loss, or adverse events (i.e. gastrointestinal bleeding, hyperglycemia, herpes zoster and fungal infections) between the two treatment groups.
2. How precise was the estimate of the treatment effect?
The 95% confidence interval (CI) around the RR of the primary endpoint (unfavorable GOS) of 0.59 was 0.37 – 0.94. The 95 percent CI around the RR of death of 0.48 was 0.24 – 0.96. Therefore, one unfavorable outcome can be prevented for every 10 patients treated (95% CI 6-100) and one death (95% CI 7-100) can be prevented for every 13 patients treated with dexamethasone versus placebo treated patients.
Will the Results Help Me in Caring for My Patients?
1. Can the results be applied to my patient care?
The clinical applicability of this study may be influenced by antimicrobial resistance patterns at individual institutions. The empiric antimicrobial therapy in this study was based on microbiological susceptibility data from the Netherlands where penicillin-resistant Streptococcus pneumoniae (PRSP) is less prevalent than in North America. Prior to the protocol amendment, monotherapy with amoxicillin was the chosen empiric antibiotic. Overall, the most common empirically prescribed antibiotics in the study were amoxicillin or penicillin in 77% of patients, a third-generation cephalosporin in 8% and a penicillin or amoxicillin combined with a cephalosporin in 8% of patients. Empiric antimicrobial treatment in Canada often includes vancomycin since the prevalence of PRSP is approximately 15%. This raises the question of whether administration of steroid therapy could theoretically reduce the permeability of the blood brain barrier and compromise the delivery of a large molecule like vancomycin. Although this concept is controversial and there are conflicting reports in the literature, the potential consequences of inadequate antimicrobial delivery may negate the benefits of adjuvant steroid therapy for meningitis. At our institution empiric antimicrobial therapy for bacterial meningitis usually consists of high dose ampicillin, vancomycin and a third-generation cephalosporin. Since low-level penicillin resistance can often be overcome with high dose penicillin or ampicillin, the need for vancomycin is controversial and thus steroid therapy will likely be of benefit to our patient population.
2. Were all clinically important outcomes considered?
3. Are the likely treatment benefits worth the potential harms and costs?
Yes. There was no statistically significant difference in the rates of adverse events between the dexamethasone and placebo treated groups. The cost of dexamethasone 100 mg intravenous solution is approximately $16.00 (CAD).
In adults 18 to 50 years of age, the most common pathogens responsible for bacterial meningitis are Streptococcus pneumoniae and Neisseria meningitides, and were found to be the most common pathogens isolated in this study (36% and 32% respectively). All of the S. pneumoniae isolates tested for susceptibility to penicillin (72%) and all but one of the N. meningitides isolates tested (83%) were sensitive to penicillin. The prevalence of PRSP is approximately 15% in Canada and 24% in the United States therefore empiric antibiotic regimens containing vancomycin may be warranted. The likelihood of steroid therapy compromising the penetration of vancomycin across the blood brain barrier is controversial and was not addressed by this study. Future studies examining the efficacy of adjuvant corticosteroids with empiric vancomycin treatment are needed.
Another issue that warrants consideration is the timing of antibiotic and steroid administration. It may not always be reasonable to delay antimicrobial therapy until the results of the CSF Gram’s stain are reported. Inevitably dexamethasone will be given to patients with non-bacterial meningitis and currently the outcome in this population is unknown. Given the potential mortality benefit of early steroid administration and the relatively low risk of toxicity, it would be reasonable to empirically administer steroids if the suspicion of bacterial meningitis is high and to discontinue the drug if the Gram’s stain suggests a higher likelihood of a viral pathogen. Also, administration of dexamethasone to patients already receiving appropriate antimicrobials cannot be recommended at this time.
The role of dexamethasone therapy for patients who present with both evidence of bacterial meningitis and hemodynamic compromise is unclear. While lower, ‘stress doses’ of corticosteroids have been shown to be beneficial in septic shock, higher doses have been shown in meta-analyses to be either of no benefit or have a trend towards increased mortality. It is possible that the beneficial effects of dexamethasone in this study may have been confounded by its effect in patients with refractory hypotension or septic shock. The rates of hypotension or septic shock were not reported however hypotension was found to be a predictor of an unfavourable outcome (p=0.03) and significantly fewer patients treated with dexamethasone developed cardiorespiratory failure (10% vs 20%; p=0.02). Until further research is done in this area, lower doses of hydrocortisone should be recommended over high doses of dexamethasone (as used in this study) for patients with both bacterial meningitis and refractory hypotension.