Unlike in most of the industrialized countries where the grip of bacterial meningitis (BM) of childhood has much loosened thanks to the large-scale Haemophilus influenzae type b (Hib) and Streptococcus pneumoniae (and Neisseria meningitidis group A in the meningitis belt of Africa) vaccinations, the global toll continues to be substantial. Reliable numbers are lacking, but estimated one million cases annually are hardly far from reality. Most cases are encountered in poor countries where mortality exceeds 30% and every second survivor is left with permanent sequelae. In a resource-poor setting, a deaf or otherwise neurofunctionally impaired child is a tragedy.
As newer antibiotics alone have not improved the prognosis, one has tried several “adjuvant” therapies. While dexamethasone (DXM) and, to a lesser extent, glycerol (GLY) have been studied among patients, some other promising approaches, such as the use of nonlytic antibiotics, are so far been tested in animal models.
DXM has clear positive biochemical effects1 if administered prior to or with the first antibiotic dose, but somewhat surprisingly, no pediatric study, in which the antibiotic has been optimal (third-generation cephalosporin) and different outcomes have been kept separated, has met statistical significance even in Hib meningitis with perfect timing of DXM.2,3 Keeping outcomes separated is important because the pathogenesis differs.
A shortcoming common to all earlier DXM studies is the sample size too small to show anything, positive or negative results. A pivotal early and much-cited study of Lebel et al.4 comprises 2 sets of children of whom 100 were treated with cefuroxime and 100 with ceftriaxone. Meager audiological benefit was found in the first but not second group.
To get a better picture of the entirety, 10 meta-analyses since 1989 have examined DXM in BM.3 One would expect all this information to have answered the burning question—Does DXM help the child with meningitis or not? So far it has not been answered conclusively. After combining 8 series from Western countries, one meta-analysis5 reached a significant reduction of hearing loss in Hib meningitis (N varied from 28 to 82)—odds ratio 0.31 (95% confidence interval: 0.14–0.69)—but when individual patient data of 833 children from Latin America and Malawi were scrutinized,2 no effect was found. Another meta-analysis6 reached significant reduction in mortality only after including a quasirandomized study from Egypt.7 A major heterogeneity of the data sources blurs a balanced interpretation.
Even more confusing is the fact that all meta-analyses, including the Cochrane review,8 have ignored the single most important covariate predicting death or neurological sequelae, the patient’s presenting condition.9 Of importance, it overrides even the role of etiology per se.9 In Latin America, <10% of children scoring between 15 and 13 on the Glasgow coma scale at arrival died, whereas mortality was around 70% if the score was ≤6. One simply cannot compare different studies unless the presenting condition is taken into account.
On the basis that adjuvant DXM has shown borderline effect in Hib meningitis, many opine that although perhaps of not much benefit, a few days course does not harm. A series of animal studies suggest otherwise. Adjuvant DXM aggravates neuronal injury by increasing apoptosis, programmed cell death, in the hippocampus of infant rats with pneumococcal meningitis and in rabbits with Escherichia coli meningitis.10,11 Because the hippocampus is the essential brain structure for learning and memory functions, adjuvant DXM led to decreased learning performance in infant rats with pneumococcal meningitis.10 Furthermore, because most children with meningitis arrive with high serum cortisol levels that associate with sequelae,12 how justified it is to increase the level by exogenous DXM?
Glycerol (glycerine, 1, 2, 3-propanetriol, GLY) is so far the only adjuvant medication that in a sufficiently powered (sample size 654), prospective, randomized and double-blind study has shown statistical significance in outcome of children treated with a third-generation cephalosporin.13 In Latin America, oral GLY reduced severe neurological sequelae with odds ratio of 0.31 (95% confidence interval: 0.13–0.76). When death and severe neurological sequelae were examined as a composite endpoint, highly significant difference in favor of GLY remained (odds ratio 0.44; 95% confidence interval: 0.25–0.76). No treatment improvement since chloramphenicol and ampicillin some 50 years ago has come even close to such an effect. Before this documentation, a pilot study14 had suggested salutary effects with the same nonsignificant strength as DXM had done in the underpowered trials on which the current recommendations base.
GLY was long used to reduce tissue pressure in conditions such as increased intracranial pressure, Ménière’s disease and glaucoma.13 There are likely more than one mechanism, but importantly, the small increase in serum osmolality15 recuperates rheology, speeds up cerebral circulation and probably improves brain oxygenation. Osmotic diuresis seems unimportant as the doses used (6 mL/kg/d qid, maximum 25 mL per dose) did not increase urinary output. It is, however, critical that GLY is not administered beyond 48 hours because then the gradient paradoxically reverses, changes toward brain tissue.16 The reasons of this surprising effect are not well understood, but the deleteriousness of prolonged GLY to patient was demonstrated decades ago.16
GLY, a naturally occurring trivalent alcohol found in the human cell membrane, is cheap, administrable orally, easily available and stable in hot climates. Therefore, great hopes were put on another prospective, randomized and double-blind study, now in Malawi. GLY, administered orally for 48 hours, did not work (Molyneux E., et al. PIDJ, in press), but adverse effects were not detected either.
Now we have a dilemma; how could 2 well-conducted prospective and sufficiently powered studies produce dissimilar results? The patient characteristics differed in many ways; for example, pneumococcal, not Hib meningitis (in which GLY performs best), prepondered in Malawi. The likely most important factor was, however, that the children in Africa were more ill (unpublished data). The critical first moments at the institution of treatment form a “window” period—children scoring between 12 and 7 on the Glasgow coma scale9—when an adjuvant may work. If you miss it, no adjuvant (or other) therapy rescues the moribund patient.
ANTIBIOTICS WITH REDUCED BACTERIOLYTIC POTENTIAL (DAPTOMYCIN)
Brain damage in BM is driven by the host’s own inflammatory reaction to the pathogens invading the subarachnoid space. β-lactams, the preferred antibiotics for BM, induce quick bacteriolysis, and in cerebrospinal fluid, a brisk accumulation of bacterial components—the main triggers of inflammatory reaction. As a result, an overshooting inflammation contributes to the brain damage.
This deleterious event might be avoided or mitigated if BM would be treated with an antibiotic with reduced bacteriolytic effect. The 2 most extensively studied agents in this regard are rifampin (inhibits protein synthesis by inhibition of the DNA-dependent RNA polymerase) and a cyclic lipopeptide daptomycin (disrupts the cell membrane function and cell wall metabolism).17 Their potential usefulness has been documented in experimental pneumococcal meningitis, a recent finding being that, compared with ceftriaxone, daptomycin substantially downregulates the inflammatory reaction. However, because daptomycin acts essentially only on Gram-positive bacteria, it needs to be combined with a broad-spectrum antibiotic for empirical therapy. The infant rat model has shown that pre-ceftriaxone daptomycin lowers the cerebrospinal fluid levels of proinflammatory mediators and reduces brain damage, hippocampal apoptosis and hearing loss.17 It is thus possible that starting treatment with daptomycin and instituting a broad-spectrum antibiotic thereafter would reduce inflammation and, ultimately, neurological sequelae in pneumococcal or other Gram-positive meningitides. This hypothesis is planned to be tested in a clinical trial once the pharmacokinetic data from an ongoing pediatric study in Switzerland (NCT01522105) are available.
OUR PRACTICAL APPROACH TO TREATMENT AND CONCLUSIONS
Unproven effectiveness of DXM in all pediatric studies2–6 and a significant effect of GLY in one13 but not the second study may incline to a nihilistic view that both adjuvants are useless. We disagree18; one cannot expect similar performance of an adjuvant in settings where mortality oscillates around 50% (some African countries) or 5–10% (industrialized world).
We discourage all treatments that lack a documentation of a change in outcome and, even more so, if there is a theoretical risk of a harm.10,11 This said, if data show a benefit in settings not too far from ours, we use that treatment. While fully recognizing the currently still wavering information, we welcome all openings that aim to treat this devastating disease better.
1. Odio CM, Faingezicht I, Paris M, et al. The beneficial effects of early dexamethasone administration in infants and children with bacterial meningitis. N Engl J Med. 1991;324:1525–1531
2. van de Beek D, Farrar JJ, de Gans J, et al. Adjunctive dexamethasone in bacterial meningitis: a meta-analysis of individual patient data. Lancet Neurol. 2010;9:254–263
3. Borchorst S, Møller K. The role of dexamethasone in the treatment of bacterial meningitis - a systematic review. Acta Anaesthesiol Scand. 2012;56:1210–1221
4. Lebel MH, Freij BJ, Syrogiannopoulos GA, et al. Dexamethasone therapy for bacterial meningitis. Results of two double-blind, placebo-controlled trials. N Engl J Med. 1988;319:964–971
5. McIntyre PB, Berkey CS, King SM, et al. Dexamethasone as adjunctive therapy in bacterial meningitis. A meta-analysis of randomized clinical trials since 1988. JAMA. 1997;278:925–931
6. Geiman BJ, Smith AL. Dexamethasone and bacterial meningitis. A meta-analysis of randomized controlled trials. West J Med. 1992;157:27–31
7. Girgis NI, Farid Z, Mikhail IA, et al. Dexamethasone treatment for bacterial meningitis in children and adults. Pediatr Infect Dis J. 1989;8:848–851
8. Brouwer MC, McIntyre P, de Gans J, et al. Corticosteroids for acute meningitis. Cochrane Database Syst Rev. 2010:CD004405
9. Roine I, Peltola H, Fernández J, et al. Influence of admission findings on death and neurological outcome from childhood bacterial meningitis. Clin Infect Dis. 2008;46:1248–1252
10. Leib SL, Heimgartner C, Bifrare YD, et al. Dexamethasone aggravates hippocampal apoptosis and learning deficiency in pneumococcal meningitis in infant rats. Pediatr Res. 2003;54:353–357
11. Spreer A, Gerber J, Hanssen M, et al. Dexamethasone increases hippocampal neuronal apoptosis in a rabbit model of Escherichia coli
meningitis. Pediatr Res. 2006;60:210–215
12. Singhi SC, Bansal A. Serum cortisol levels in children with acute bacterial and aseptic meningitis. Pediatr Crit Care Med. 2006;7:74–78
13. Peltola H, Roine I, Fernández J, et al. Adjuvant glycerol and/or dexamethasone to improve the outcomes of childhood bacterial meningitis: a prospective, randomized, double-blind, placebo-controlled trial. Clin Infect Dis. 2007;45:1277–1286
14. Kilpi T, Peltola H, Jauhiainen T, et al. Oral glycerol and intravenous dexamethasone in preventing neurologic and audiologic sequelae of childhood bacterial meningitis. The Finnish Study Group. Pediatr Infect Dis J. 1995;14:270–278
15. Singhi S, Järvinen A, Peltola H. Increase in serum osmolality is possible mechanism for the beneficial effects of glycerol in childhood bacterial meningitis. Pediatr Infect Dis J. 2008;27:892–896
16. Rottenberg DA, Hurwitz BJ, Posner JB. The effect of oral glycerol on intraventricular pressure in man. Neurology. 1977;27:600–608
17. Grandgirard D, Burri M, Agyeman P, et al. Adjunctive daptomycin attenuates brain damage and hearing loss more efficiently than rifampin in infant rat pneumococcal meningitis. Antimicrob Agents Chemother. 2012;56:4289–4295
18. Peltola H, Singhi S, Roine I. Glycerol in meningitis should not be condemned so hastily. Lancet Infect Dis. 2011;11:897–898