-
Adjunctive dexamethasone in bacterial meningitis: a
meta-analysis of individual patient data
Diederik van de Beeka, Jeremy J Farrard,e,h,*, Jan de Gansa,
Nguyen Thi Hoang Maid,Elizabeth M Molyneuxf, Heikki Peltolag, Tim E
Petoj, Irmeli Roinek, Mathew Scarboroughf,i,Constance Schultszb,e,
Guy E Thwaitesl, Phung Quoc Tuand,e, and AH Zwindermanc
aDepartment of Neurology, Centre of Infection and Immunity
Amsterdam, Academic Medical Center,Amsterdam, Netherlands bCentre
for Poverty-related and Communicable Diseases, AcademicMedical
Center, Amsterdam, Netherlands cDepartment of Clinical Epidemiology
and Biostatistics,Academic Medical Center, Amsterdam, Netherlands
dHospital for Tropical Diseases, Ho Chi MinhCity, Vietnam eOxford
University Clinical Research Unit, Ho Chi Minh City, Vietnam
fCollege ofMedicine, University of Malawi, Blantyre, Malawi
gHelsinki University Central Hospital, Hospital forChildren and
Adolescents, Helsinki, Finland hCentre for Tropical Medicine,
Oxford University,Oxford, UK iNuffield Department of Clinical
Laboratory Science, Oxford University, Oxford,UK jNuffield
Department of Medicine, John Radcliffe Hospital, Oxford, UK
kFaculty of HealthSciences, University Diego Portales, Santiago,
Chile lCentre for Molecular Microbiology andInfection, Imperial
College, London, UK
SummaryBackgroundDexamethasone improves outcome for some
patients with bacterial meningitis, butnot others. We aimed to
identify which patients are most likely to benefit from
dexamethasonetreatment.
MethodsWe did a meta-analysis of individual patient data from
the randomised, double-blind,placebo-controlled trials of
dexamethasone for bacterial meningitis in patients of all ages for
whichraw data were available. The pre-determined outcome measures
were death at the time of first follow-up, death or severe
neurological sequelae at 1 month follow-up, death or any
neurological sequelaeat first follow-up, and death or severe
bilateral hearing loss at first follow-up. Combined odds
ratios(ORs) and tests for heterogeneity were calculated using
conventional Mantel-Haenszel statistics. Wealso did exploratory
analysis of hearing loss among survivors and other exploratory
subgroupanalyses by use of logistic regression.
FindingsData from 2029 patients from five trials were included
in the analysis (833 [410%]aged
-
vs 443%; 092, 076111), death or any neurological sequelae or any
hearing loss (542% vs 574%;089, 074107), or death or severe
bilateral hearing loss (364% vs 389%; 089, 073169).However,
dexamethasone seemed to reduce hearing loss among survivors (241%
vs 295%; 077,060099, p=004). Dexamethasone had no effect in any of
the prespecified subgroups, includingspecific causative organisms,
pre-dexamethasone antibiotic treatment, HIV status, or age.
Poolingof the mortality data with those of all other published
trials did not significantly change the results.
InterpretationAdjunctive dexamethasone in the treatment of acute
bacterial meningitis does notseem to significantly reduce death or
neurological disability. There were no significant treatmenteffects
in any of the prespecified subgroups. The benefit of adjunctive
dexamethasone for all or anysubgroup of patients with bacterial
meningitis thus remains unproven.
FundingWellcome Trust UK.
IntroductionThe yearly incidence of bacterial meningitis is
estimated to be 2660 cases per 100000 inEurope and might be ten
times higher in less developed countries. Experimental models
haveshown that outcome is related to the severity of the
inflammatory process in the subarachnoidspace, and treatment with
corticosteroids results in a reduction of the inflammatory
responseand improved outcome. These findings have prompted several
randomised controlled trials ofcorticosteroids for bacterial
meningitis. Initial results suggested that the main beneficial
effectof the corticosteroid dexamethasone was to reduce the risk of
hearing loss in children withHaemophilus influenzae type b
meningitis. Additional data extended the likely benefit to
thosewith Streptococcus pneumoniae meningitis. In 2004, a
meta-analysis of five randomisedcontrolled trials showed that
treatment with corticosteroids reduced both mortality
andneurological sequelae in adults with bacterial meningitis,
without detectable adverse effects.Subsequently, a Cochrane
meta-analysis of data from 20 randomised controlled trials
andinvolving 2750 people showed an overall mortality benefit and a
reduction in neurologicalsequelae in patients treated with adjuvant
corticosteroids. However, three large randomisedcontrolled trials
published after this analysis showed conflicting results.
Adjunctivecorticosteroids seem to benefit some patients with
bacterial meningitis but not others, and howto select patients who
are likely to benefit is unclear. Our aim was to address this
question witha meta-analysis of data from five major trials for
which individual patient data were available.
MethodsStudy selection
Relevant trials were identified previously as part of a Cochrane
review (figure 1). Individualpatient data from five randomised,
double-blind, placebo-controlled trials of dexamethasonefor
bacterial meningitis published since 2001 were included in the
analysis; individual patientdata could not be acquired from the
older trials. The characteristics of the included studies areshown
in table 1.
The study from South America used a 22 design to randomly assign
children with bacterialmeningitis to dexamethasone plus glycerol,
dexamethasone plus placebo, glycerol plusplacebo, or placebo plus
placebo. Data were available from children who were
assigneddexamethasone plus placebo or placebo only but not from
those who were given glycerol.During the study, the randomisation
schedule was altered from a ratio of two dexamethasoneper three
placebo (randomisation schedule 1) to one dexamethasone per one
placebo(randomisation schedule 2). Therefore, analyses from this
study were stratified according torandomisation schedule. The study
in Malawian adults used a 22 design to randomly assignpatients to
dexamethasone or placebo and to intravenous or intramuscular
ceftriaxone. In allstudies, patients were enrolled on the basis of
clinically suspected bacterial meningitis and CSF
van de Beek et al. Page 2
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
criteria. All the studies used computer-generated randomisation
to allocate patients todexamethasone or placebo. Treatment
concealment was adequate in all studies.
Definitions and outcome measuresThe members of the study group
met in October, 2006, and September, 2007, to discuss datasharing
and the analysis plan, including the definitions of subgroups,
which were specifiedbefore the data were collated, the final
database created, and the analysis started. The
principalinvestigators provided the raw data, which were checked by
a statistician (PQT).Inconsistencies and outlying data were
clarified with the principal investigators and resolvedfrom their
raw data before the analysis.
15 data fields for each patient were selected for the analyses.
The dataset included prognosticfactors for unfavourable outcome and
potential modifiers of the treatment effect ofdexamethasone, such
as antibiotic treatment before admission, HIV infection,
andmalnutrition. Definitions were agreed during the two study-group
meetings. Values forcontinuous variables were reassigned into
categories. Exposure to antibiotics beforerandomisation was defined
by administration of effective oral or intravenous antibiotics
within48 h before the first dose of study drug was received.
Malnutrition was defined by individualinvestigators: patients who
were not assessed were categorised according to the localprevalence
of malnutrition. HIV tests were not done on every patient and an
assessment wasmade of the likelihood of HIV infection based on
local epidemiology. All untested Malawianadults were defined as
likely to be HIV positive. No assumption was made for
untestedMalawian children. All other untested adults or children
were defined as likely to be HIVnegative. Impairment of
consciousness was categorised by use of the Glasgow coma scale
orthe Blantyre coma score (table 2). The causative pathogen was
defined by CSF microscopy,CSF or blood culture, PCR, or latex
agglutination.
The predetermined outcome measures were death at the time of
first follow-up; death or severeneurological sequelae (including
severe bilateral hearing loss) at 1 month follow-up; death orany
neurological sequelae (including any degree of hearing loss) at
first follow-up; and deathor severe bilateral hearing loss at first
follow-up. The number of studies that contributed toeach outcome is
shown by degrees of freedom (df=number of studies minus 1).
Additionally,as part of a post-hoc exploratory analysis and to
analyse every possible endpoint of interest,we analysed hearing
loss of any degree among survivors. The severity of neurological
sequelaein the adult studies was defined using the Glasgow outcome
score or the modified Rankinscale. In the paediatric studies,
severe neurological disability was defined as
blindness,quadraparesis, hydrocephalus requiring a shunt, or severe
psychomotor retardation. Hearingloss was categorised as moderate or
severe according to definitions used in the individualstudies.
Statistical analysisAll analyses were stratified according to
study site (including two strata from the SouthAmerican study) to
account for any possible centre effect, including differences in
mortalitybetween centres. If appropriate, analyses were also
stratified according to the baseline variableof interest. Combined
odds ratios (ORs) and tests for heterogeneity were calculated
usingconventional Mantel-Haenszel statistics. We also used
exploratory analyses with logisticregression. The main purpose of
the analysis was to establish whether dexamethasone had
adifferential effect in different subgroups of patients; hence,
heterogeneity between thesubgroups (I2 values) with significance
levels were calculated for each subgroup analysis. Testsfor
heterogeneity were calculated without allowing for multiple
comparisons, to increase thesensitivity of detecting any evidence
of between-subgroup heterogeneity. To maximise thepower of finding
significant heterogeneity, missing values were removed, except
where
van de Beek et al. Page 3
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
indicated, from the subgroup analyses. A continuity correction
was made for zero events.Significance tests, with the appropriate
degrees of freedom, were calculated to test for
possibleheterogeneity between studies for each subgroup
analysis.
To calculate the combined ORs for death from studies included in
the Cochrane reviews butnot otherwise included in the present
study, results available from the published literature werecombined
by use of conventional Mantel-Haenszel statistics. Calculation of
combined ORsand 95% CIs, tests of heterogeneity between studies,
and logistic regression analyses weredone by use of STATA version
10.
Role of the funding sourceThe study sponsors had no role in the
study design, collection, analysis, and interpretation ofthe data,
or the decision to submit the manuscript for publication. T E Peto
had full access toall data in the study. All authors approved and
were responsible for submission of themanuscript.
ResultsThe baseline characteristics were similar in placebo and
dexamethasone groups within the fivestudies (table 2). 1019 (502%)
patients received dexamethasone and 1010 (498%) patientsreceived
placebo. 833 (411%) patients were less than 15 years old, of whom
415 receiveddexamethasone and 418 received placebo. 1196 adults
(aged 15 years) were included, ofwhom 604 (505%) received
dexamethasone and 592 (495%) received placebo. The ages offive
patients were unknown.
HIV co-infection was confirmed in 549 (415%) of 1322 patients
tested, of whom 391 (714%)were adults and 158 (288%) were children.
An HIV test was not done in 707 (348%) patientsbut, on the basis of
epidemiological risk, was judged likely to be positive in 31
untested adultsfrom Malawi and negative in adults from Europe and
children from South America. Noassumption was made about 139
untested children from Malawi. In total, 286 confirmed orlikely
HIV-infected patients received dexamethasone and 294 received
placebo.
The diagnosis of bacterial meningitis was microbiologically
confirmed in 1639 (808%)patients and was most frequently caused by
S pneumoniae (759 cases), H influenzae (297cases), and Neisseria
meningitidis (239 cases). The most common causative bacteria per
studywere as follows: Europe, N meningitis (38%); Malawi
(children), S pneumoniae (40%);Vietnam, Streptococcus suis (32%);
Malawi (adults), S pneumoniae (59%); and SouthAmerica, H influenzae
(47%). Mortality in the placebo groups differed substantially
betweenstudies: 15% in Europe, 31% in Malawian children, 12% in
Vietnam, 53% in Malawian adults,and 16% in South America.
Dexamethasone was not associated with a significant reduction in
death, death or severeneurological sequelae (including severe
bilateral hearing loss), death or any neurologicalsequelae
(including any hearing loss), or death or severe bilateral hearing
loss, if all patientswere included in the analysis (table 3).
However, hearing loss (of any severity) in survivorswas less common
in the dexamethasone group (162 [241%] of 672 vs 195 [295%] of
660;OR 077 [95% CI 060099], p=004).
The subgroup analyses for all outcome measures are shown in
figures 2 and 3, and thewebappendix. Duration of symptoms before
treatment, severity of coma at start of treatment,whether
dexamethasone was given before or after antibiotics, and HIV
infection status did notsignificantly influence treatment response.
Dexamethasone was more effective in patients agedolder than 55
years in analyses of death (OR 041 [95% CI 020084], p=001), death
or severe
van de Beek et al. Page 4
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
neurological sequelae (OR 053 [030084], p=003), and death or any
neurological sequelae(OR 056 [031100], p=005). However, there was
no clear evidence of heterogeneity betweenthe different age groups
(death, 2=69, 3 df, p=007, I2 545%; death or severe
neurologicalsequelae, 2=66, 3 df, p=009, I2=534%; death or any
neurological sequelae, 2=44, 3 df,p=023, I2=303%). Further
exploratory analyses, using age as a continuous variable, did
notshow any consistent interaction between age and a treatment
effect (data not shown). Therewas also no effect in a post-hoc
analysis that restricted the study to patients treated
withceftriaxone (webappendix).
The data were explored to identify evidence of heterogeneity
between the studies. 23 subgroupswere explored, each with five
different endpoints. In patients with moderate CNS impairmenton
admission, there was some evidence of heterogeneity between three
of the five endpoints.In the subgroup of patients with moderate CNS
impairment on admission, there was evidenceof benefit in death or
severe neurological sequelae or bilateral hearing loss in the
Europeanstudy (OR 019 [95% CI 004082], p=001), but also evidence of
harm in the study of childrenin Malawi (OR 370 [1361008], p=0006).
However, no evidence of heterogeneity wasobserved in patients with
either no or little CNS impairment or with severe CNS
impairment.Overall, there was no evidence of any difference in
outcome for any of the CNS subgroups inany of the five endpoints.
The effect of HIV was explored by adjustment with logistic
regressionanalysis and also by studying only patients with proven
HIV status. However, HIV status didnot have an effect on
dexamethasone treatment outcome (webappendix). We further
exploredthe relation between age, HIV status, and dexamethasone
treatment effect (table 4). In HIV-negative adults, dexamethasone
was associated with a reduction in death or severe
neurologicalsequelae, including severe bilateral hearing loss (OR
068 [95% CI 048095], p=002), deathor any neurological sequelae,
including any hearing loss (OR 067 [050091], p=001), anddeath or
severe bilateral hearing loss (OR 061 [042089], p=001). However,
this effect ofdexamethasone was not present in HIV-negative
children, or in HIV-positive children andadults.
Gastrointestinal bleeding was reported in all studies: 13 (13%)
of 1021 patients ondexamethasone and 19 (19%) of 1014 patients on
placebo (p=014). Hyperglycaemia andinfection by herpes simplex
virus and varicella zoster virus were reported in some but not
allstudies. Hyperglycaemia was recorded by the trials in Malawian
and European adults and wassignificantly associated with
dexamethasone treatment (79 of 390 [203%] on dexamethasonevs 60 of
376 [160%] on placebo; p=002). Neither infection with herpes
simplex virus (labialinfection in all) nor infection with varicella
zoster virus were significantly associated withdexamethasone
treatment.
Dexamethasone did not significantly affect mortality in a
combined analysis with the data fromother studies included in the
Cochrane analysis (OR 088 [95% CI 073104], p=014; figure4). 349
(180%) of 1944 patients who received dexamethasone died, compared
with 384(198%) of 1939 patients who received placebo. There was no
evidence of significantheterogeneity between the trials.
DiscussionThe aim of this analysis was to establish whether any
subgroups of patients with acute bacterialmeningitis might benefit
from adjunctive dexamethasone and thereby explain any
differencesbetween individual trial results. Extensive exploration
of 15 prespecified subgroups did notshow robust evidence that a
particular subgroup would benefit. The apparent benefit in
adultsaged over 55 years might have occurred by chance. However, it
is unclear whether it is morelikely to have occurred by chance than
the findings of no benefit in other subgroups.
van de Beek et al. Page 5
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
This analysis of 2029 patients from five trials showed that
treatment with adjunctivedexamethasone did not significantly reduce
mortality, neurological disability, or severe hearingloss in
bacterial meningitis. Combination of these results with those from
older published trials,for which the raw data were not obtainable,
did not show any evidence that dexamethasonewas significantly
effective in reducing these outcomes overall. However, a post-hoc
analysison the incidence of deafness among survivors suggested that
adjunctive dexamethasonetreatment reduced the rate of hearing loss
(OR 077 [95% CI 060099; p=004), irrespectiveof whether patients had
received antibiotics before dexamethasone treatment. The use
ofadjunctive dexamethasone treatment was not associated with an
increased risk of adverseevents.
Factors previously considered relevant to the decision to start
dexamethasone treatment inpatients with suspected or proven
bacterial meningitis could not explain differences in
resultsbetween the five trials. These factors include duration of
symptoms before treatment, severityof impaired consciousness at
start of treatment, whether dexamethasone was given before orafter
antibiotics, and HIV infection status. Because the results of the
prespecified analysis failedto show any significant heterogeneity,
extensive post-hoc analyses were done with theinclusion of an
additional deafness endpoint. Such analyses are usually considered
unreliable,particularly if no statistical allowance is made for
multiple comparisons, because of the highchance of a false-positive
result. However, the extra analyses were undertaken to allow
theidentification of subgroups of interest for further possible
study. These exploratory post-hocanalyses suggested a possible
overall effect on deafness among survivors and on death andsevere
neurological sequelae in the subgroup of HIV-negative adults (OR
068 [95% CI 054099], p=002). This apparent treatment effect ceased
to be significant after adjustment formultiple comparisons.
This meta-analysis is, as are all meta-analyses, limited by the
possibility that moreheterogeneity exists between the studies than
has been identified. If such heterogeneity wereto exist, combining
the studies would be inappropriate. Formal tests for heterogeneity
betweenstudies and between subgroups failed to show any convincing
evidence of heterogeneity.However, such tests are insensitive and
could miss important effects. We have thereforeexplored the data
exhaustively for relevant subgroups of patients that could reveal
possiblecauses of heterogeneity, although little such evidence was
found.
On the basis of previous meta-analyses, the administration of
dexamethasone to children withH influenzae type b meningitis before
the start of antibiotic therapy is thought to reduce theincidence
of deafness. However, we found no evidence of a benefit of
adjunctivedexamethasone in all children or in any subgroup of
children with this infection.
In summary, these data indicate that patients with bacterial
meningitis neither benefit from norare harmed by treatment with
adjunctive dexamethasone. Despite an individual patient
datameta-analysis of more than 2000 patients, we have been unable
to determine conclusivelywhether a subgroup of patients might
benefit. To establish with certainty whetherdexamethasone has a
place in the treatment of adult patients with bacterial meningitis,
a largemultinational randomised controlled trial would be
necessary. This represents a formidablechallenge and one that is
not likely to be met for many years. In the meantime, we suggest
thebenefit of adjunctive dexamethasone for all or any subgroup of
patients with bacterialmeningitis remains unproven and there is
little support for its routine use in the treatment ofthis
disease.
Contributors
van de Beek et al. Page 6
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
The study was conceived by JJF. All the authors contributed to
the study design and theselection of data for analysis. The
analysis was done by PQT, TEP, and AHZ. The paper waswritten by
DvdB, JJF, TEP, MS, and GET, with review and comment from all the
authors.
Conflicts of interest
We have no conflicts of interest.
Web Extra MaterialSupplementary Material1. Supplementary
webappendix.
AcknowledgmentsAcknowledgments
This work was supported by the Wellcome Trust UK. DvdB is
supported by grants from the Netherlands Organizationfor Health
Research and Development (NWO-Veni grant 2006 [916.76.023]) and the
Academic Medical Center (AMCFellowship 2008). TEP is supported by
the UK National Institute for Health Research, Biomedical Research
Centre,Oxford, UK. We thank Sarah Walker (Medical Research Council,
Clinical Trials Unit, London, UK) for independentstatistical
advice.
References1. van de Beek D, de Gans J, Spanjaard L, Weisfelt M,
Reitsma JB, Vermeulen M. Clinical features and
prognostic factors in adults with bacterial meningitis. N Engl J
Med 2004;351:18491859. [PubMed:15509818]
2. van de Beek D, de Gans J, Tunkel AR, Wijdicks EF.
Community-acquired bacterial meningitis inadults. N Engl J Med
2006;354:4453. [PubMed: 16394301]
3. Saez-Llorens X, McCracken GH. Bacterial meningitis in
children. Lancet 2003;361:21392148.[PubMed: 12826449]
4. Scarborough M, Thwaites GE. The diagnosis and management of
acute bacterial meningitis inresource-poor settings. Lancet Neurol
2008;7:637648. [PubMed: 18565457]
5. Scheld WM, Dacey RG, Winn HR, Welsh JE, Jane JA, Sande MA.
Cerebrospinal fluid outflowresistance in rabbits with experimental
meningitis: alterations with penicillin and methylprednisolone.J
Clin Invest 1980;66:243253. [PubMed: 6995482]
6. Tauber MG, Khayam-Bashi H, Sande MA. Effects of ampicillin
and corticosteroids on brain watercontent, cerebrospinal fluid
pressure, and cerebrospinal fluid lactate levels in
experimentalpneumococcal meningitis. J Infect Dis 1985;151:528534.
[PubMed: 3973406]
7. van de Beek D, Weisfelt M, de Gans J, Tunkel AR, Wijdicks EF.
Drug insight: adjunctive therapiesin adults with bacterial
meningitis. Nat Clin Pract Neurol 2006;2:504516. [PubMed:
16932615]
8. van de Beek D, de Gans J, McIntyre P, Prasad K.
Corticosteroids for acute bacterial meningitis.Cochrane Database
Syst Rev 2007;1 CD004405.
9. Havens PL, Wendelberger KJ, Hoffman GM, Lee MB, Chusid MJ.
Corticosteroids as adjunctivetherapy in bacterial meningitis: a
meta-analysis of clinical trials. Am J Dis Child
1989;143:10511055.[PubMed: 2773883]
10. McIntyre PB, Berkey CS, King SM. Dexamethasone as adjunctive
therapy in bacterial meningitis: ameta-analysis of randomized
clinical trials since 1988. JAMA 1997;278:925931.
[PubMed:9302246]
11. van de Beek D, de Gans J, McIntyre P, Prasad K. Steroids in
adults with acute bacterial meningitis:a systematic review. Lancet
Infect Dis 2004;4:139143. [PubMed: 14998499]
12. Peltola H, Roine I, Fernandez J. Adjuvant glycerol and/or
dexamethasone to improve the outcomesof childhood bacterial
meningitis: a prospective, randomized, double-blind,
placebo-controlled trial.Clin Infect Dis 2007;45:12771286. [PubMed:
17968821]
van de Beek et al. Page 7
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
13. Nguyen TH, Tran TH, Thwaites G. Dexamethasone in Vietnamese
adolescents and adults withbacterial meningitis. N Engl J Med
2007;357:24312440. [PubMed: 18077808]
14. Scarborough M, Gordon SB, Whitty CJ. Corticosteroids for
bacterial meningitis in adults in sub-Saharan Africa. N Engl J Med
2007;357:24412450. [PubMed: 18077809]
15. Molyneux EM, Walsh AL, Forsyth H. Dexamethasone treatment in
childhood bacterial meningitisin Malawi: a randomised controlled
trial. Lancet 2002;360:211218. [PubMed: 12133656]
16. de Gans J, van de Beek D. Dexamethasone in adults with
bacterial meningitis. N Engl J Med2002;347:15491556. [PubMed:
12432041]
17. Belsey MA, Hoffpauir CW, Smith MH. Dexamethasone in the
treatment of acute bacterial meningitis:the effect of study design
on the interpretation of results. Pediatrics 1969;44:503513.
[PubMed:5346629]
18. Bademosi O, Osuntokun BO. Prednisolone in the treatment of
pneumococcal meningitis. TropGeograph Med 1979;31:5356.
19. Bhaumik S, Behari M. Role of dexamethasone as adjunctive
therapy in acute bacterial meningitis inadults. Neurol India
1998;46:225228.
20. Ciana G, Parmar N, Antonio C, Pivetta S, Tamburlini G,
Cuttini M. Effectiveness of adjunctivetreatment with steroids in
reducing short-term mortality in a high-risk population of children
withbacterial meningitis. J Trop Pediatr 1995;41:164168. [PubMed:
7636936]
21. deLemos RA, Haggerty RJ. Corticosteroids as an adjunct to
treatment in bacterial meningitis: acontrolled clinical trial.
Pediatrics 1969;44:3034. [PubMed: 4893958]
22. Girgis NI, Farid Z, Mikhail IA, Farrag I, Sultan Y,
Kilpatrick ME. Dexamethasone treatment forbacterial meningitis in
children and adults. Pediatr Infect Dis J 1989;8:848851. [PubMed:
2626285]
23. Kanra GY, Ozen H, Secmeer G, Ceyhan M, Ecevit Z, Belgin E.
Beneficial effects of dexamethasonein children with pneumococcal
meningitis. Pediatr Infect Dis J 1995;14:490494.
[PubMed:7667053]
24. Lebel MH, Freij BJ, Syrogiannopoulos GA. Dexamethasone
therapy for bacterial meningitis. Resultsof two double-blind,
placebo-controlled trials. N Engl J Med 1988;319:964971. [PubMed:
3047581]
25. Odio CM, Faingezicht I, Paris M. The beneficial effects of
early dexamethasone administration ininfants and children with
bacterial meningitis. N Engl J Med 1991;324:15251531.
[PubMed:2027357]
26. Qazi SA, Khan MA, Mughal N. Dexamethasone and bacterial
meningitis in Pakistan. Arch Dis Child1996;75:482488. [PubMed:
9014599]
27. Thomas R, Le TY, Bouget J. Trial of dexamethasone treatment
for severe bacterial meningitis inadults: adult Meningitis Steroid
Group. Intensive Care Med 1999;25:475480. [PubMed: 10401941]
28. Jennett B, Bond M. Assessment of outcome after severe brain
damage. Lancet 1975;1:480484.[PubMed: 46957]
29. UK-TIA Study Group. United Kingdom transient ischaemic
attack (UK-TIA) aspirin trial: interimresults. BMJ 1988;296:316320.
[PubMed: 2894232]
30. Bennett IL, Finland M, Hamburger M, Kass EH, Lepper M,
Waisbren N. The effectiveness ofhydrocortisone in the management of
severe infections. JAMA 1963;183:462465.
31. King SM, Law B, Langley JM, Heurtler H, Bremner D, Wang EE.
Dexamethasone therapy forbacterial meningitis: better never than
late? Can J Infect Dis 1994;5:210215.
32. Lebel MH, Hoyt J, Waagner DC, Rollins NK, Finitzo T,
McCracken GH. Magnetic resonance imagingand dexamethasone therapy
for bacterial meningitis. Am J Dis Child 1989;143:301306.
[PubMed:2644815]
33. Wald ER, Kaplan SL, Mason EO. Dexamethasone therapy for
children with bacterial meningitis.Pediatrics 1995;95:2128.
[PubMed: 7770303]
34. Kilpi T, Peltola H, Jauhiainen T. Oral glycerol and
intravenous dexamethasone in preventingneurologic and audiological
sequelae of childhood bacterial-meningitis. Pediatr Infect Dis
J1995;14:270278. [PubMed: 7603807]
35. Schaad UB, Lips U, Gnehm HE, Blumberg A, Heinzer I, Wedgwood
J. Dexamethasone therapy forbacterial meningitis in children.
Lancet 1993;342:457461. [PubMed: 8102428]
van de Beek et al. Page 8
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
36. van de Beek D, de Gans J. Dexamethasone in adults with
community-acquired bacterial meningitis.Drugs 2006;66:415427.
[PubMed: 16597160]
37. Tunkel AR, Hartman BJ, Kaplan SL. Practice guidelines for
the management of bacterial meningitis.Clin Infect Dis
2004;39:12671284. [PubMed: 15494903]
38. Fitch MT, van de Beek D. Drug insight: steroids in CNS
infectious diseasesnew indications for anold therapy. Nat Clin
Pract Neurol 2008;4:97104. [PubMed: 18256681]
39. Greenwood BM. Corticosteroids for acute bacterial
meningitis. N Engl J Med 2007;357:25072509.[PubMed: 18077815]
van de Beek et al. Page 9
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
Figure 1.Literature search
van de Beek et al. Page 10
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
Figure 2.Subgroup analyses for deathBM=bacterial meningitis.
OR=odds ratio.
van de Beek et al. Page 11
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
Figure 3.Subgroup analyses for deathBM=bacterial meningitis.
OR=odds ratio.
van de Beek et al. Page 12
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
Figure 4.Effect of adjunctive dexamethasone therapy on
deathTrials included in the rest of this study and other studies
included in the Cochrane systematicreview are shown. OR=odds ratio.
*Study 1 in Lebel. Study 2 in Lebel.
van de Beek et al. Page 13
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
-
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
van de Beek et al. Page 14
Table 1
Characteristics of the five studies included in the analysis
Study period Patients (n) Age Inclusion criteria Dexamethasone
dose Empirical antibiotic* Primary outcome
Europe 19922001 301 >16 years Clinicallysuspected BMplus CSF
criteria
10 mg four times dailyfor 4 days
Intravenousamoxicillin 2 g every 4h (77% of patients)
Unfavourableoutcome (definedby a Glasgowoutcome score of14) at 8
weeks
Malawi (child) 19972001 598 2 monthsto 13years
Clinicallysuspected BMplus CSF criteria
04 mg/kg twice dailyfor 2 days
Intravenousbenzylpenicillin 200000 IU/kg every 24 hplus
chloramphenicol100 mg/kg every 24 h
Death at 1 month
Vietnam 19962005 429 >14 years Clinicallysuspected BMplus CSF
criteria
04 mg/kg twice dailyfor 4 days
Intravenousceftriaxone 2 g every12 h
Death at 1 month
Malawi (adult) 20022005 465 >15 years Clinicallysuspected
BMplus CSF criteria
16 mg twice daily for4 days
Intravenous orintramuscularceftriaxone 2 g every12 h
Death at 1 month
South America 19962003 236 2 monthsto 16years
Clinicallysuspected BMplus CSF or bloodcriteria
015 mg/kg four timesdaily for 2 days
Intravenousceftriaxone 80100mg/kg every 24 h
Death, severeneurologicalsequelae, oraudiologicalsequelae
athospital discharge
BM=bacterial meningitis.
*Dexamethasone was given before or with the first dose of
per-protocol parenteral antibiotic in all five studies.
23% of patients received other antibiotic treatment.
22 design with patients randomly assigned to dexamethasone or
placebo and to intravenous or intramuscular ceftriaxone.
22 design with patients randomly assigned to dexamethasone plus
glycerol, dexamethasone plus placebo, placebo plus glycerol, or
placebo plus
placebo; patients assigned to receive glycerol with either
dexamethasone or placebo were excluded from the individual patient
data meta-analysis;data from this trial were analysed as two strata
according to randomisation schedule.
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
-
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
van de Beek et al. Page 15
Tabl
e 2
Bas
elin
e ch
arac
teris
tics o
f pat
ient
s inc
lude
d in
the
anal
ysis
Eur
ope(
n=30
1)M
alaw
i (ch
ild)(n
=598
)V
ietn
am(n
=429
)M
alaw
i (ad
ult)(
n=46
5)So
uth
Am
eric
aT
otal
(n=2
029)
Dex
amet
haso
ne (n
=101
9)Pl
aceb
o (n
=101
0)
Ran
dom
isat
ion
sche
dule
1(n
=126
)R
ando
mis
atio
n sc
hedu
le 2
(n=1
10)
Age
(yea
rs)
55
102
010
616
00
224
112
112
Unk
now
n0
11
00
35
23
Sex
Men
169
(56%
)33
7 (5
6%)
315
(73%
)23
0 (5
0%)
73 (5
8%)
63 (5
7%)
1187
(58%
)60
1 (5
9%)
586
(58%
)
Sym
ptom
s 20
122
(40%
)31
(5%
)24
8 (5
8%)
72 (1
6%)
53 (4
2%)
46 (4
2%)
572
(28%
)28
4 (2
8%)
288
(29%
)
Unk
now
n10
(3%
)92
(15%
)2
(05
%)
61 (1
3%)
9 (7
%)
3 (3
%)
177
(9%
)87
(8%
)90
(9%
)
CSF
pro
tein
(mg/
dL)
250
195
(65%
)41
2 (6
9%)
191
(45%
)34
1 (7
3%)
23 (1
8%)
33 (3
0%)
1195
(59%
)58
8 (5
8%)
607
(60%
)
-
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
van de Beek et al. Page 17
Eur
ope(
n=30
1)M
alaw
i (ch
ild)(n
=598
)V
ietn
am(n
=429
)M
alaw
i (ad
ult)(
n=46
5)So
uth
Am
eric
aT
otal
(n=2
029)
Dex
amet
haso
ne (n
=101
9)Pl
aceb
o (n
=101
0)
Ran
dom
isat
ion
sche
dule
1(n
=126
)R
ando
mis
atio
n sc
hedu
le 2
(n=1
10)
Oth
er a
erob
ic g
ram
neg
ativ
e ba
cilli
3 (1
%)
38 (6
%)
21 (5
%)
22 (5
%)
3 (2
%)
1 (1
%)
88 (4
%)
51 (5
%)
37 (4
%)
Oth
er27
(9%
)11
(2%
)68
(16%
)5
(1%
)3
(2%
)5
(4%
)11
9 (6
%)
52 (5
%)
67 (7
%)
Def
inite
ly n
ot b
acte
rial m
enin
gitis
00
11 (3
%)
38 (8
%)
1 (1
%)
050
(25
%)
28 (3
%)
22 (2
%)
Unk
now
n (p
roba
ble
bact
eria
lm
enin
gitis
)40
(13%
)74
(12%
)78
(18%
)10
2 (2
2%)
27 (2
1%)
19 (1
7%)
340
(17%
)17
6 (1
7%)
164
(16%
)
Bac
teria
l dia
gnos
is c
onfir
med
mic
robi
olog
ical
ly26
1 (8
7%)
524
(88%
)34
0 (7
8%)
325
(70%
)98
(78%
)91
(83%
)16
39 (8
1%)
815
(80%
)82
4 (8
1%)
Tre
atm
ent a
lloca
tion
Allo
cate
d to
dex
amet
haso
ne tr
eatm
ent
157
(52%
)30
5 (5
1%)
215
(50%
)23
3 (5
0%)
50 (4
0%)
59 (5
4%)
1019
(50%
)..
..
Dat
a ar
e nu
mbe
r (%
).
* HIV
sero
stat
us w
as n
ot a
vaila
ble
in p
atie
nts i
n th
e Eu
rope
an o
r Sou
th A
mer
ican
tria
ls; p
atie
nts i
n th
ese
trial
s wer
e as
sum
ed to
be
HIV
neg
ativ
e. In
the
Vie
tnam
tria
l, fo
ur u
ntes
ted
patie
nts w
ere
assu
med
to b
e H
IV n
egat
ive
and
in th
e M
alaw
i adu
lt tri
al, 3
1 un
test
ed p
atie
nts w
ere
assu
med
to b
e H
IV p
ositi
ve. I
n th
e M
alaw
i pae
diat
ric tr
ial,
139
(23%
) pat
ient
s wer
e no
t tes
ted
and
no a
ssum
ptio
n w
as m
ade
abou
t the
ir se
rost
atus
. Pos
itive
val
ues o
nly
incl
ude
patie
nts t
este
d an
d no
t tho
se a
ssum
ed to
be
posi
tive.
Gla
sgow
com
a sc
ale
is c
ateg
oris
ed in
adu
lts a
s 15=
norm
al, 1
114
=mild
impa
irmen
t, 8
10=m
oder
ate
impa
irmen
t, 3
7=se
vere
impa
irmen
t. B
lant
yre
com
a sc
ore
is c
ateg
oris
ed a
s 5=n
orm
al, 3
4=m
ild im
pairm
ent,
2=m
oder
ate
impa
irmen
t, 0
1=se
vere
impa
irmen
t.
The
use
of u
rine
reag
ent s
trips
in th
e tri
als f
rom
Mal
awi p
rovi
ded
sem
i-qua
ntita
tive
estim
ates
of C
SF g
luco
se a
nd p
rote
in. P
rote
in a
nd g
luco
se c
once
ntra
tions
wer
e m
easu
red
with
a u
rine
dips
tick
(Mul
tistix
8SG
Bay
er),
whi
ch p
rovi
ded
colo
ur-c
oded
resu
lts o
f (pr
otei
n/gl
ucos
e)ne
gativ
e (0
/0 m
g/dL
); tra
ce (20
00/>
1000
mg/
dL).
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
-
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
van de Beek et al. Page 18
Tabl
e 3
Prim
ary
endp
oint
s for
eac
h st
udy
and
for a
ll pa
tient
s ass
igne
d to
ster
oid
ther
apy
Eur
ope
Mal
awi (
child
)V
ietn
amM
alaw
i (ad
ult)
Sout
h A
mer
ica
Ove
rall
Eve
nts/
tota
l (%
)T
est f
or h
eter
ogen
eity
Ran
dom
isat
ion
sche
dule
1R
ando
mis
atio
n sc
hedu
le 2
Dex
amet
haso
nePl
aceb
o2
(5 d
f)p
I2
Dea
th0
44 (0
20
096
, 00
3)0
96 (0
70
140
, 09
6)0
82 (0
45
151
, 05
3)1
16 (0
80
167
, 04
3)1
46 (0
63
337
, 03
7)0
74 (0
27
200
, 05
5)0
97 (0
79
119
, 07
5)27
0/10
19 (2
65%
)27
5/10
10 (2
72%
)6
50
2622
7%
Dea
th o
rse
vere
neur
olog
ical
sequ
elae
or
bila
tera
lse
vere
deaf
ness
060
(03
41
11, 0
07)
120
(08
71
66, 0
28)
075
(04
81
17, 0
20)
102
(06
91
50, 0
93)
074
(03
51
55, 0
42)
074
(03
31
67, 0
52)
092
(07
61
11, 0
39)
424/
1003
(42
3%)
439/
992
(44
3%)
65
026
232
%
Dea
th o
r any
neur
olog
ical
sequ
elae
or
any
hear
ing
loss
049
(02
80
84, 0
01)
102
(07
41
42, 0
89)
081
(05
51
18, 0
27)
103
(06
71
56, 0
91)
129
(06
02
77, 0
51)
084
(03
91
79, 0
65)
089
(07
41
07, 0
23)
541/
999
(54
2%)
567/
988
(57
4%)
71
022
293
%
Dea
th o
rse
vere
bila
tera
lhe
arin
g lo
ss
055
(03
10
99, 0
04)
103
(07
31
45, 0
86)
064
(03
81
08, 0
09)
108
(07
31
58, 0
70)
107
(04
92
32, 0
87)
070
(02
91
69, 0
43)
089
(07
31
69, 0
23)
343/
942
(36
4%)
363/
934
(38
9%)
62
028
198
%
Any
hea
ring
loss
insu
rviv
ors
075
(03
41
67, 0
48)
080
(05
11
28, 0
35)
077
(04
91
21, 0
26)
080
(04
41
45, 0
45)
059
(02
11
65, 0
31)
081
(03
02
14, 0
66)
077
(06
00
99, 0
04)
162/
672
(24
1%)
195/
660
(29
5%)
03
100
00%
Dat
a ar
e O
R (9
5% C
I, p
valu
e) u
nles
s oth
erw
ise
stat
ed. O
R v
alue
s bel
ow 1
sugg
est a
ben
efic
ial e
ffec
t of s
tero
ids.
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.
-
Sponsored Docum
ent Sponsored D
ocument
Sponsored Docum
ent
van de Beek et al. Page 19
Table 4
Exploratory analyses of the influence of age and HIV infection
on the treatment effect of dexamethasone
HIV negative HIV positive Overall Test for heterogeneity
Adult Child Adult Child 2 (3 df) p I2
Death 066 (042102, 006) 143 (096212, 007) 119 (081175, 036) 054
(028103, 006) 099 (080123, 099) 107 001 720%
Death orsevereneurologicalsequelae orbilateralseveredeafness
068 (048095, 002) 109 (077155, 062) 110 (073166, 067) 077
(036166, 044) 090 (073110, 029) 48 019 374%
Death or anyneurologicalsequelae orany hearingloss
067 (050091, 001) 109 (077156, 062) 115 (073182, 054) 077
(035171, 053) 088 (072107, 018) 60 011 501%
Death orseverebilateralhearing loss
061 (042089, 001) 116 (080167, 043) 113 (075170, 055) 062
(030129, 020) 089 (072109, 026) 81 004 285%
Any hearingloss insurvivors
076 (052113, 017) 067 (042107, 009) 087 (046163, 066) 109
(037319, 087) 077 (059099, 006) 09 083 00%
Data are OR (95% CI, p value) unless otherwise stated. Adults
were defined as 15 years. HIV negative includes patients who tested
negative or werelikely to be negative. HIV positive includes those
who tested positive or were likely to be positive.
Published as: Lancet Neurol. 2010 March ; 9(3): 254263.