ORIGINAL ARTICLE Repeat lumbar puncture in adults with bacterial meningitis J. M. Costerus 1,3 , M. C. Brouwer 1,3 , A. van der Ende 2,3,4 and D. van de Beek 1,3 1) Department of Neurology, 2) Department of Medical Microbiology, 3) Centre of Infection and Immunity Amsterdam and 4) The Netherlands Reference Laboratory for Bacterial Meningitis, Academic Medical Centre, Amsterdam, The Netherlands Abstract In a prospective nationwide cohort study performed in the Netherlands from 2006 to 2014 we analysed clinical and laboratory characteristics of adults with community-acquired bacterial meningitis who underwent repeat lumbar puncture. Repeat lumbar puncture was performed in 124 of 1490 included episodes (8%), most commonly because of clinical deterioration (42%). Median cerebrospinal fluid (CSF) leucocyte count on admission was 1473 cells/mm 3 . Median CSF cell count showed a decrease of 19% when repeated within 2 days; of 84% within 3 – 7 days, of 93% within 8 – 14 days and of 98% within 15 – 21 days. Repeat lumbar puncture confirmed the diagnosis of meningitis in eight patients with normal initial CSF examination. Repeat CSF cultures yielded bacteria in nine patients, which led to identification of an underlying source of infection in two. We conclude that repeat lumbar puncture is performed in a small proportion of adults. Although it should not be seen as routine it can be useful in selected cases to confirm diagnosis, to exclude relapsing or persistent infection, or for therapeutic purpose in communicating hydrocephalus. © 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved. Keywords: Bacterial meningitis, cerebrospinal fluid, delayed cerebral thrombosis, hydrocephalus, lumbar puncture, repeat Original Submission: 7 October 2015; Revised Submission: 21 December 2015; Accepted: 21 December 2015 Editor: E. Bottieau Article published online: 21 January 2016 Corresponding author: D. van de Beek, Department of Neurology, Centre of Infection and Immunity Amsterdam (CINIMA), Academic Medical Centre, University of Amsterdam, PO Box 22660, 1100DD Amsterdam, The Netherlands E-mail: [email protected] Introduction Cerebrospinal fluid (CSF) analysis is essential to establish the diagnosis of bacterial meningitis, to identify the causative or- ganism and to undertake in vitro antibiotic susceptibility testing [1]. The clinical relevance of repeat lumbar puncture in bac- terial meningitis is unclear [2]. Although repeat lumbar punc- ture was formerly recommended to determine the length of antimicrobial therapy based on normalization of CSF leucocyte counts, CSF glucose concentration levels and CSF protein levels, it is now recommended in any patient who has not responded clinically after 48 hours of appropriate antimicrobial therapy; especially in patients with pneumococcal β-lactam- resistant strains or patients who are receiving adjunctive dexamethasone therapy [3,4]. However, there are other in- dications for repeat lumbar puncture, e.g. to reduce intracranial pressure in patients with communicating hydrocephalus, or in those who deteriorate after initial recovery; for example in delayed cerebral thrombosis [5 – 7]. Complications of a lumbar puncture are uncommon, but include post-puncture headache, spinal or epidural bleeding, damage of the spinal nerve roots and cerebral herniation [8 – 11]. Several studies have been performed to assess the yield of repeat lumbar puncture, of which most were retrospective, relatively small or addressed children only [3,12 – 17]. References values for CSF leucocyte count, glucose and protein levels in repeat CSF examination in adults with community-acquired bacterial meningitis are lacking. We analysed the clinical practice of repeat lumbar puncture in adults with community-acquired bacterial meningitis, providing reference values, and evaluated the impact of repeat lumbar puncture in clinical management of bacterial meningitis. Clin Microbiol Infect 2016; 22: 428 – 433 © 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved http://dx.doi.org/10.1016/j.cmi.2015.12.026
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Repeat lumbar puncture in adults with bacterial meningitisRepeat lumbar puncture in adults with bacterial meningitis
J. M. Costerus1,3, M. C. Brouwer1,3, A. van der Ende2,3,4 and D. van de Beek1,3
1) Department of Neurology, 2) Department of Medical Microbiology, 3) Centre of Infection and Immunity Amsterdam and 4) The Netherlands Reference
Laboratory for Bacterial Meningitis, Academic Medical Centre, Amsterdam, The Netherlands
Abstract
In a prospective nationwide cohort study performed in the Netherlands from 2006 to 2014 we analysed clinical and laboratory characteristics
of adults with community-acquired bacterial meningitis who underwent repeat lumbar puncture. Repeat lumbar puncture was performed in
124 of 1490 included episodes (8%), most commonly because of clinical deterioration (42%). Median cerebrospinal fluid (CSF) leucocyte
count on admission was 1473 cells/mm3. Median CSF cell count showed a decrease of 19% when repeated within 2 days; of 84% within
3–7 days, of 93% within 8–14 days and of 98% within 15–21 days. Repeat lumbar puncture confirmed the diagnosis of meningitis in
eight patients with normal initial CSF examination. Repeat CSF cultures yielded bacteria in nine patients, which led to identification of an
underlying source of infection in two. We conclude that repeat lumbar puncture is performed in a small proportion of adults. Although
it should not be seen as routine it can be useful in selected cases to confirm diagnosis, to exclude relapsing or persistent infection, or for
therapeutic purpose in communicating hydrocephalus.
© 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
Keywords: Bacterial meningitis, cerebrospinal fluid, delayed cerebral thrombosis, hydrocephalus, lumbar puncture, repeat
Original Submission: 7 October 2015; Revised Submission: 21 December 2015; Accepted: 21 December 2015
Editor: E. Bottieau
Clin © htt
Corresponding author: D. van de Beek, Department of Neurology, Centre of Infection and Immunity Amsterdam (CINIMA), Academic Medical Centre, University of Amsterdam, PO Box 22660, 1100DD Amsterdam, The Netherlands E-mail: [email protected]
Introduction
Cerebrospinal fluid (CSF) analysis is essential to establish the
diagnosis of bacterial meningitis, to identify the causative or- ganism and to undertake in vitro antibiotic susceptibility testing
[1]. The clinical relevance of repeat lumbar puncture in bac- terial meningitis is unclear [2]. Although repeat lumbar punc-
ture was formerly recommended to determine the length of antimicrobial therapy based on normalization of CSF leucocyte counts, CSF glucose concentration levels and CSF protein
levels, it is now recommended in any patient who has not responded clinically after 48 hours of appropriate antimicrobial
Microbiol Infect 2016; 22: 428–433 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier p://dx.doi.org/10.1016/j.cmi.2015.12.026
therapy; especially in patients with pneumococcal β-lactam-
resistant strains or patients who are receiving adjunctive dexamethasone therapy [3,4]. However, there are other in- dications for repeat lumbar puncture, e.g. to reduce intracranial
pressure in patients with communicating hydrocephalus, or in those who deteriorate after initial recovery; for example in
delayed cerebral thrombosis [5–7]. Complications of a lumbar puncture are uncommon, but include post-puncture headache,
spinal or epidural bleeding, damage of the spinal nerve roots and cerebral herniation [8–11]. Several studies have been
performed to assess the yield of repeat lumbar puncture, of which most were retrospective, relatively small or addressed children only [3,12–17]. References values for CSF leucocyte
count, glucose and protein levels in repeat CSF examination in adults with community-acquired bacterial meningitis are lacking.
We analysed the clinical practice of repeat lumbar puncture in adults with community-acquired bacterial meningitis, providing
reference values, and evaluated the impact of repeat lumbar puncture in clinical management of bacterial meningitis.
Ltd. All rights reserved
Methods
In a nationwide prospective cohort study, we included patients
>16 years who were listed in the database of the Netherlands Reference Laboratory for Bacterial Meningitis (NRLBM) from 1 March 2006 to 31 May 2014. This laboratory receives CSF from
85%–90% of all patients with CSF culture-positive bacterial meningitis in the Netherlands (population, 16.9 million). The
NRLBM provided daily updates of the names of the hospitals where patients with bacterial meningitis had been admitted in
the previous 2–6 days. The treating physician was contacted, and informed consent was obtained from all participating pa-
tients or their legally authorized representatives. Physicians could also contact the investigators without report of the
NRLBM for inclusion of patients. Episodes reported by physi- cians with negative CSF cultures or PCR were only included if CSF results obtained from any lumbar puncture during admis-
sion showed at least one individual predictor of bacterial meningitis (defined as a glucose level <34 mg/dL (1.9 mmol/L), a
ratio of CSF glucose to blood glucose of <0.23, a protein level of >2.2 g/L, or a leucocyte count >2000/mm3) and the clinical
presentation was compatible with bacterial meningitis [18]. Patients with hospital-associated meningitis including neuro-
surgery patients, patients with a neurosurgical device, or neu- rotrauma within 1 month of the onset of meningitis were excluded. Data on patient history, symptoms and signs on
admission, laboratory findings, radiological examination, treat- ment and outcome were prospectively collected by means of a
Case Record Form (CRF), also evaluating if repeat lumbar puncture was performed. The discharge summaries of these
patients were retrospectively assessed for the indication and influence on treatment decisions of the repeat lumbar punc-
ture. Statistical analyses were performed with use of IBM SPSS Statistics, version 22. Dichotomous variables were compared
with use of the χ2 test and continuous variables were compared using the Mann–Whitney U-test. The study was approved by the medical ethical committee of the Academic Medical Centre,
Amsterdam, the Netherlands.
Results
Indication Frequency (n [ 124)
Clinical deterioration 41 Raised intracranial pressure/ hydrocephalus 19 Persistent or relapsing fever 12 Lack of clinical improvement despite treatment 7 Monitoring of response on treatment 7 Suspected herpes simplex encephalitis 4 Discrepancies in primary cerebrospinal fluid results 5 Othera 3 Not specified/ unknown 26
aHeadache, radiological or biochemical deterioration.
Repeat lumbar puncture was performed in 124 of 1490
included episodes (8%); 16 patients underwent more than two lumbar punctures. Distributions of causative pathogens in pa-
tients with repeat lumbar puncture were as follows: 80 pneu- mococcal meningitis, 13 listeria meningitis, five meningococcal
meningitis, 20 episodes were caused by a different microor- ganism, and in six patients no causative organism was identified.
© 2016 European Society of Clinical Microbiology
The median time between the first and second lumbar punc-
tures was 8 days (interquartile range (IQR) 4–15 days). The repeat lumbar puncture was performed in the first week of
admission in 59 patients (48%), in the second week in 32 pa- tients (26%), in the third week in 17 patients (14%) and after
more than 3 weeks in 14 patients (11%, interval unknown in two patients). Indications for performing repeat lumbar punc- ture were identified for 98 of 124 episodes (79%; Table 1) and
most commonly were clinical deterioration (42%), hydro- cephalus (19%) and persistent/relapsing fever (12%).
Results of initial and repeat CSF examinations are shown in Fig. 1 (on admission, days 0–2, 3–7, 8–14 and 15–21). On
admission, median CSF leucocyte count was 1473 cells/mm3
(IQR 237–3920), protein level 4.70 g/L (IQR 2.81–7.45) and
glucose level 0.20 mmol/L(IQR 0.00–1.75). At day 0–2, repeat examination showed higher CSF cell counts in 13 of 19 patients (68%) compared with examination at admission, while the
median remained similar (1200 cells/mm3 (IQR 188–2680)); median protein level decreased to 2.95 g/L (IQR 2.14–6.29) and
median glucose level increased to 2.00 mmol/L (IQR 0.40–3.90). After 3–7 days of treatment, median CSF leucocyte
count substantially decreased to 231 cells/mm3 (IQR 79–746 (decrease of 84%)), with protein and glucose levels of 1.18 g/L
(IQR 0.80–2.38) and 2.60 mmol/L (IQR 1.10–3.60). After 8–14 days, median CSF leucocyte count decreased to 104 cells/
mm3 (IQR 42–255), and after 15–21 days to 33 cells/mm3
(IQR 8–159). Fig. 2 illustrates indications for repeat lumbar puncture, CSF results and final diagnosis, stratified by time.
Initial CSF examination showed normal CSF cell count, defined as <7 cells/mm3, in 30 of 1382 patients included in the
cohort (2%; CSF cell count was available for 1382 of 1490 patients (93%)). Eight of these 30 patients underwent repeat
lumbar puncture (27%; Table 2) between day 1 and 18, revealing elevated CSF leucocyte counts in all (median 298 cells/
mm3 (IQR 41–810)). Gram staining of the CSF from the initial lumbar puncture showed bacteria in four of these eight patients (50%). Immunodeficiency was present in only one patient
(immunosuppressive medication), although two additional
and Infectious Diseases. Published by Elsevier Ltd. All rights reserved, CMI, 22, 428–433
FIG. 1. Cerebrospinal fluid (CSF) chemical results in repeat lumbar puncture. Median with interquartile range displayed. (a) CSF cell count. Numbers evaluated: at admission: n = 120, 0–2 days: n = 20,
3–7 days: n = 30, 8–14 days: n = 30, 15–21 days: n = 17. (b) CSF protein level. Numbers evaluated: at admission: n = 116, 0–2 days: n = 18, 3–7 days: n = 31, 8–14 days: n = 30, 15–21 days: n = 16. (c) CSF
glucose level. Numbers evaluated: at admission: n = 118, 0–2 days: n = 17, 3–7 days: n = 30, 8–14 days: n = 29, 15–21 days: n = 14.
430 ClinicalM
icrobiology and
Infection,V olum
e 22
N um
ber 5,M
ay 2016
–433
FIG. 2. Indication and results of repeat lumbar puncture, stratified by time. Cell count in cells/mm3. LP, lumbar puncture; ICP, intracranial pressure;
CSF, cerebrospinal fluid; BM, bacterial meningitis; DCT, delayed cerebral thrombosis.
CMI Costerus et al. Repeat lumbar puncture in bacterial meningitis 431
patients had a history of diabetes. Microbiological diagnosis was
made based on CSF culture of the initial CSF examination in seven of these eight patients and of the repeat CSF culture in
one: Streptococcus pneumoniae (n = 4), Streptococcus para- sanguinis (n = 2), Haemophilus influenzae (n = 1) and Strepto-
coccus agalactiae (n = 1). Repeat CSF culture showed bacteria in nine of 91 patients in
whom cultures were obtained (10%; Table 3). One patient had
a clinical relapse and was subsequently diagnosed with a recto- spinal fistula and one patient deteriorated at day 25 and was
diagnosed with ventriculitis; these findings led to a switch to broader spectrum antibiotics. The positive culture in the
additional seven patients was performed within 4 days after the start of therapy, and led to a change of antibiotic regimen in
two.
TABLE 2. Patients with an initial cerebrospinal fluid cell count <7 c
Sex/ age Relevant medical history CRP (g/L) Gram-
F/72 Diabetes mellitus 252 Pos M/64 Prednisolone for polymyalgia rheumatica 500 Pos M/59 Neutropenia due to chemotherapy 205 Pos M/72 None 235 NB F/73 None 319 Pos F/71 None 342 NB M/45 Pretreated with antibiotics for dental procedure 84 NB F/69 Pretreated with antibiotics for dental procedure 5 NB
CRP, C-reactive protein; CSF, cerebrospinal fluid; GOS, Glasgow Outcome Scale. aGram staining of initial CSF. Pos = Gram staining showed bacteria. NB = Gram staining sho bCSF cell count in cells/mm3, Δt = time interval between first and repeat lumbar puncture
© 2016 European Society of Clinical Microbiology
After 10 days of treatment, 25 of 50 patients (50%) in whom
lumbar puncture was repeated had a CSF cell count over 50 cells/mm3. This included one patient with relapsing bacterial
meningitis (day 19) and five with delayed cerebral thrombosis (DCT). In total, nine patients were diagnosed with DCT and all
had pleocytosis in the repeat lumbar puncture; median time to repeat lumbar puncture in DCT was 13 days (IQR 10–16). Patients with DCT had a higher median CSF cell count in the
repeat analysis compared with those who underwent lumbar puncture between days 8 and 14 without DCT (759 cells/mm3
(IQR 72–1464) versus 86 cells/mm3 (IQR 41–199); p 0.01). Unfavourable outcome occurred in 72 of 124 patients with
repeat lumbar puncture (58%) compared with 486 of 1366 patients with a single lumbar puncture (36%, p <0.001). The
mortality rate between patients with and without repeat
ells/mm3 a
staina Causative organism Repeat CSF cell count (Δt)b GOS
Streptococcus pneumoniae 1200 (1) 4 S. pneumoniae 810 (2) 5 S. pneumoniae ND (10) 3 Streptococcus parasanguinis 41 (4) 3 S. pneumoniae 486 (1) 5 Haemophilus influenzae 298 (5) 5 Streptococcus agalactiae 3 (18) 5 S. parasanguinis 56 (1) 5
wed no bacteria. in days.
and Infectious Diseases. Published by Elsevier Ltd. All rights reserved, CMI, 22, 428–433
TABLE 3. Characteristics of patients with bacterial meningitis and a positive second cerebrospinal fluid culture
Sex/ age First/ Second culture First CSF cell counta Second CSF cell count (Δt)a Additional diagnosis AB regimen expanded GOS
M/59 Streptococcus pneumoniae (2x) 33 140 (1) Multiple myeloma No 5 F/72 Negative/ S. pneumoniae 1 1200 (1) None No 4 M/60 Negative/ Streptococcus oralis 597 1125 (25) Ventriculitis Yes 1 F/69 Staphylococcus aureus (2x) ND ND (1) None No 5 F/85 Listeria monocytogenes (2x) 263 855 (2) Hydrocephalus No 3 M/81 S. aureus (2x) ND ND (2) Hydrocephalus No 5 M/30 Contaminated/ L. monocytogenes 3100 1140 (3) Hydrocephalus Yes 4 M/82 Negative/ Escherichia coli 11377 52033 (5) Recto-spinal fistula Yes 1 M/72 Streptococcus parasanguinis/ S. pneumoniae 3 14 (4) Hydrocephalus Yes 3
CSF, cerebrospinal fluid; AB, antibiotic; ND, not determined; GOS, Glasgow Outcome Scale Score. aCSF cell count in cells/mm3, Δt = time interval in days.
432 Clinical Microbiology and Infection, Volume 22 Number 5, May 2016 CMI
lumbar puncture was similar (25 of 124 (20%) versus 229 of
1137 (20%)).
Discussion
Our study shows that repeat lumbar puncture is performed in a
small proportion of adults with bacterial meningitis (8%), most commonly because of clinical deterioration or hydrocephalus.
This rate is much lower than previously reported. One retro- spective US single centre cohort reported a rate of repeat
lumbar puncture of 68% in the period 1969 and 1980 [14]. A more recent retrospective single centre ICU study in France reported a rate of 47% in the period 2000–11 [3]. The latter
study concluded that the yield of repeat lumbar puncture in adults with pneumococcal meningitis treated with adequate
antibiotic therapy is low, even in patients infected by strains resistant to penicillin or cephalosporin [14,19]. Our study was
nationwide and, therefore, we were able to study a represen- tative sample of adults with acute bacterial meningitis. The low
rate of repeat lumbar puncture in our study concurs with current guidelines recommending repeat lumbar puncture in any patient who has not responded clinically after 48 h of
appropriate antimicrobial therapy to identify penicillin-resistant or cephalosporin-resistant strains [4].
Our data suggest that repeat lumbar puncture can be useful in selected cases. First, repeat lumbar puncture is indicated in
cases of diagnostic uncertainty after first CSF examination. We identified a small number of patients with suspected bacterial
meningitis but normal CSF examination; in these patients repeat lumbar puncture confirmed the diagnosis [20]. Second, repeat
lumbar puncture is needed to rule out relapsing infection in patients who deteriorate after 48 h of adequate antibiotic therapy. We identified two patients with relapsing infection. A
small proportion of patients with repeat lumbar puncture was diagnosed with delayed cerebral thrombosis after clinical
deterioration (< 1%); these patients had increased CSF white cell counts but negative bacterial cultures at repeat CSF
© 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier
examination. Third, repeat lumbar puncture can be used
therapeutically in patients with hydrocephalus. Hydrocephalus occurs in about 5% of patients and has been associated with
unfavourable outcome and death [5]. Acute obstructive hy- drocephalus in bacterial meningitis is thought to be very rare, and these rare cases usually occur in young children [5,21,22].
In patients with obstructive hydrocephalus lumbar puncture should not be performed [23]. Finally, compliant with current
guidelines, repeat lumbar puncture can be considered in any patient who has not responded clinically after 48 h of appro-
priate antimicrobial therapy [4]. Our study provides reference values for repeat CSF exam-
ination in adults with community-acquired bacterial meningitis. The distribution of causative pathogens among patients with
repeat lumbar puncture is similar to that in the total group of patients with community-acquired bacterial meningitis [24]. Median CSF cell count in the first 2 days after start of treatment
remained stable, followed by a quick decrease between 3 and 14 days and a subsequent relative plateau phase after 14 days of
treatment. Median CSF protein level stabilized after a decrease of 75% within 3–7 days of treatment. CSF glucose levels
showed a quick increase within 2 days of treatment. In patients with an abnormal clinical course and a CSF cell count >50 cells/
mm3 after 10 days, recurrent infection or an immunological phenomenon (delayed cerebral thrombosis) should be consid- ered. This is in contrast with a retrospective paediatric study
from 1972 to 1977 showing CSF cell count >50 cells/mm3 in 9 of 27 children (33%) with bacterial meningitis who all recovered
well [12]. Differences between findings can be explained by confounding by indication or adjunctive treatment with dexa-
methasone. In the childhood study repeat lumbar puncture was performed in all patients, whereas in our study this was done
on indication only (8%). In our study 77% was treated with adjunctive dexamethasone but this was not routine practice in
in the period 1972–77. Our study has several limitations. We did not have infor-
mation on indication for repeat lumbar puncture in 20% of
cases. This was mainly caused by the research design as the
Ltd. All rights reserved, CMI, 22, 428–433
CMI Costerus et al. Repeat lumbar puncture in bacterial meningitis 433
reason for repeat lumbar puncture was not prospectively
collected in the case record form. Second, only patients who underwent lumbar puncture and who had a positive CSF cul-
ture were included. Negative CSF cultures occur in 11–30% of patients with bacterial meningitis [25]. No significant differences
in clinical presentation have been reported between patients with culture-positive bacterial meningitis and those with culture-negative bacterial meningitis. Therefore, it is unlikely
that this factor confounded our results. Finally, rates of antibi- otic resistance among meningococcal and pneumococcal iso-
lates in the Netherlands are low [26]. Therefore, our results cannot automatically be extrapolated to regions with high
antibiotic resistance rates. However, a previous observational study concluded that repeat lumbar punctures in patients with
pneumococcal meningitis were not very helpful as routine, also not in an area of high penicillin-resistance [3].
Funding
This study was funded by the Netherlands Organization for Health Research and Development (ZonMw; NWO-Veni- Grant (916.13.078) to MB, NWO-Vidi-Grant (016.116.358)
to DvdB), the Academic Medical Centre (AMC Fellowship to DvdB), and the European Research Council (ERC Starting
Grant (281156) to DvdB). The Netherlands Reference Labo- ratory is funded by the National Institute of Public Health and
the Environment. Other authors: no financial support.
Acknowledgements
We are indebted to all the Dutch physicians who participated in the MeninGene study.
Transparency Declaration
All authors have stated that there are no conflicts of interest.
References
[1] Brouwer MC, Thwaites GE, Tunkel AR, van de Beek D. Dilemmas in the diagnosis of acute community-acquired bacterial meningitis. Lancet 2012;380:1684–92.
[2] van de Beek D, de Gans J, Tunkel AR, Wijdicks EF. Community-ac- quired bacterial meningitis in adults. N Engl J Med 2006;354:44–53.
[3] Denneman L, Vial-Dupuy A, Gault N, Wolff M, van de Beek D, Mourvillier B. Repeated lumbar puncture in adults with pneumococcal meningitis: an observational study. J Infect 2013;67:350–3.
© 2016 European Society of Clinical Microbiology
[4] Tunkel AR, Hartman BJ, Kaplan SL, Kaufman BA, Roos KL, Scheld WM, et al. Practice guidelines for the management of bacterial meningitis. Clin Infect Dis 2004;39:1267–84.
[5] Kasanmoentalib ES, Brouwer MC, van der Ende A, van de Beek D. Hydrocephalus in adults with community-acquired bacterial meningitis. Neurology 2010;75:918–23.
[6] Schut ES, Brouwer MC, de Gans J, Florquin S, Troost D, van de Beek D. Delayed cerebral thrombosis after initial good recovery from pneumococcal meningitis. Neurology…
J. M. Costerus1,3, M. C. Brouwer1,3, A. van der Ende2,3,4 and D. van de Beek1,3
1) Department of Neurology, 2) Department of Medical Microbiology, 3) Centre of Infection and Immunity Amsterdam and 4) The Netherlands Reference
Laboratory for Bacterial Meningitis, Academic Medical Centre, Amsterdam, The Netherlands
Abstract
In a prospective nationwide cohort study performed in the Netherlands from 2006 to 2014 we analysed clinical and laboratory characteristics
of adults with community-acquired bacterial meningitis who underwent repeat lumbar puncture. Repeat lumbar puncture was performed in
124 of 1490 included episodes (8%), most commonly because of clinical deterioration (42%). Median cerebrospinal fluid (CSF) leucocyte
count on admission was 1473 cells/mm3. Median CSF cell count showed a decrease of 19% when repeated within 2 days; of 84% within
3–7 days, of 93% within 8–14 days and of 98% within 15–21 days. Repeat lumbar puncture confirmed the diagnosis of meningitis in
eight patients with normal initial CSF examination. Repeat CSF cultures yielded bacteria in nine patients, which led to identification of an
underlying source of infection in two. We conclude that repeat lumbar puncture is performed in a small proportion of adults. Although
it should not be seen as routine it can be useful in selected cases to confirm diagnosis, to exclude relapsing or persistent infection, or for
therapeutic purpose in communicating hydrocephalus.
© 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
Keywords: Bacterial meningitis, cerebrospinal fluid, delayed cerebral thrombosis, hydrocephalus, lumbar puncture, repeat
Original Submission: 7 October 2015; Revised Submission: 21 December 2015; Accepted: 21 December 2015
Editor: E. Bottieau
Clin © htt
Corresponding author: D. van de Beek, Department of Neurology, Centre of Infection and Immunity Amsterdam (CINIMA), Academic Medical Centre, University of Amsterdam, PO Box 22660, 1100DD Amsterdam, The Netherlands E-mail: [email protected]
Introduction
Cerebrospinal fluid (CSF) analysis is essential to establish the
diagnosis of bacterial meningitis, to identify the causative or- ganism and to undertake in vitro antibiotic susceptibility testing
[1]. The clinical relevance of repeat lumbar puncture in bac- terial meningitis is unclear [2]. Although repeat lumbar punc-
ture was formerly recommended to determine the length of antimicrobial therapy based on normalization of CSF leucocyte counts, CSF glucose concentration levels and CSF protein
levels, it is now recommended in any patient who has not responded clinically after 48 hours of appropriate antimicrobial
Microbiol Infect 2016; 22: 428–433 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier p://dx.doi.org/10.1016/j.cmi.2015.12.026
therapy; especially in patients with pneumococcal β-lactam-
resistant strains or patients who are receiving adjunctive dexamethasone therapy [3,4]. However, there are other in- dications for repeat lumbar puncture, e.g. to reduce intracranial
pressure in patients with communicating hydrocephalus, or in those who deteriorate after initial recovery; for example in
delayed cerebral thrombosis [5–7]. Complications of a lumbar puncture are uncommon, but include post-puncture headache,
spinal or epidural bleeding, damage of the spinal nerve roots and cerebral herniation [8–11]. Several studies have been
performed to assess the yield of repeat lumbar puncture, of which most were retrospective, relatively small or addressed children only [3,12–17]. References values for CSF leucocyte
count, glucose and protein levels in repeat CSF examination in adults with community-acquired bacterial meningitis are lacking.
We analysed the clinical practice of repeat lumbar puncture in adults with community-acquired bacterial meningitis, providing
reference values, and evaluated the impact of repeat lumbar puncture in clinical management of bacterial meningitis.
Ltd. All rights reserved
Methods
In a nationwide prospective cohort study, we included patients
>16 years who were listed in the database of the Netherlands Reference Laboratory for Bacterial Meningitis (NRLBM) from 1 March 2006 to 31 May 2014. This laboratory receives CSF from
85%–90% of all patients with CSF culture-positive bacterial meningitis in the Netherlands (population, 16.9 million). The
NRLBM provided daily updates of the names of the hospitals where patients with bacterial meningitis had been admitted in
the previous 2–6 days. The treating physician was contacted, and informed consent was obtained from all participating pa-
tients or their legally authorized representatives. Physicians could also contact the investigators without report of the
NRLBM for inclusion of patients. Episodes reported by physi- cians with negative CSF cultures or PCR were only included if CSF results obtained from any lumbar puncture during admis-
sion showed at least one individual predictor of bacterial meningitis (defined as a glucose level <34 mg/dL (1.9 mmol/L), a
ratio of CSF glucose to blood glucose of <0.23, a protein level of >2.2 g/L, or a leucocyte count >2000/mm3) and the clinical
presentation was compatible with bacterial meningitis [18]. Patients with hospital-associated meningitis including neuro-
surgery patients, patients with a neurosurgical device, or neu- rotrauma within 1 month of the onset of meningitis were excluded. Data on patient history, symptoms and signs on
admission, laboratory findings, radiological examination, treat- ment and outcome were prospectively collected by means of a
Case Record Form (CRF), also evaluating if repeat lumbar puncture was performed. The discharge summaries of these
patients were retrospectively assessed for the indication and influence on treatment decisions of the repeat lumbar punc-
ture. Statistical analyses were performed with use of IBM SPSS Statistics, version 22. Dichotomous variables were compared
with use of the χ2 test and continuous variables were compared using the Mann–Whitney U-test. The study was approved by the medical ethical committee of the Academic Medical Centre,
Amsterdam, the Netherlands.
Results
Indication Frequency (n [ 124)
Clinical deterioration 41 Raised intracranial pressure/ hydrocephalus 19 Persistent or relapsing fever 12 Lack of clinical improvement despite treatment 7 Monitoring of response on treatment 7 Suspected herpes simplex encephalitis 4 Discrepancies in primary cerebrospinal fluid results 5 Othera 3 Not specified/ unknown 26
aHeadache, radiological or biochemical deterioration.
Repeat lumbar puncture was performed in 124 of 1490
included episodes (8%); 16 patients underwent more than two lumbar punctures. Distributions of causative pathogens in pa-
tients with repeat lumbar puncture were as follows: 80 pneu- mococcal meningitis, 13 listeria meningitis, five meningococcal
meningitis, 20 episodes were caused by a different microor- ganism, and in six patients no causative organism was identified.
© 2016 European Society of Clinical Microbiology
The median time between the first and second lumbar punc-
tures was 8 days (interquartile range (IQR) 4–15 days). The repeat lumbar puncture was performed in the first week of
admission in 59 patients (48%), in the second week in 32 pa- tients (26%), in the third week in 17 patients (14%) and after
more than 3 weeks in 14 patients (11%, interval unknown in two patients). Indications for performing repeat lumbar punc- ture were identified for 98 of 124 episodes (79%; Table 1) and
most commonly were clinical deterioration (42%), hydro- cephalus (19%) and persistent/relapsing fever (12%).
Results of initial and repeat CSF examinations are shown in Fig. 1 (on admission, days 0–2, 3–7, 8–14 and 15–21). On
admission, median CSF leucocyte count was 1473 cells/mm3
(IQR 237–3920), protein level 4.70 g/L (IQR 2.81–7.45) and
glucose level 0.20 mmol/L(IQR 0.00–1.75). At day 0–2, repeat examination showed higher CSF cell counts in 13 of 19 patients (68%) compared with examination at admission, while the
median remained similar (1200 cells/mm3 (IQR 188–2680)); median protein level decreased to 2.95 g/L (IQR 2.14–6.29) and
median glucose level increased to 2.00 mmol/L (IQR 0.40–3.90). After 3–7 days of treatment, median CSF leucocyte
count substantially decreased to 231 cells/mm3 (IQR 79–746 (decrease of 84%)), with protein and glucose levels of 1.18 g/L
(IQR 0.80–2.38) and 2.60 mmol/L (IQR 1.10–3.60). After 8–14 days, median CSF leucocyte count decreased to 104 cells/
mm3 (IQR 42–255), and after 15–21 days to 33 cells/mm3
(IQR 8–159). Fig. 2 illustrates indications for repeat lumbar puncture, CSF results and final diagnosis, stratified by time.
Initial CSF examination showed normal CSF cell count, defined as <7 cells/mm3, in 30 of 1382 patients included in the
cohort (2%; CSF cell count was available for 1382 of 1490 patients (93%)). Eight of these 30 patients underwent repeat
lumbar puncture (27%; Table 2) between day 1 and 18, revealing elevated CSF leucocyte counts in all (median 298 cells/
mm3 (IQR 41–810)). Gram staining of the CSF from the initial lumbar puncture showed bacteria in four of these eight patients (50%). Immunodeficiency was present in only one patient
(immunosuppressive medication), although two additional
and Infectious Diseases. Published by Elsevier Ltd. All rights reserved, CMI, 22, 428–433
FIG. 1. Cerebrospinal fluid (CSF) chemical results in repeat lumbar puncture. Median with interquartile range displayed. (a) CSF cell count. Numbers evaluated: at admission: n = 120, 0–2 days: n = 20,
3–7 days: n = 30, 8–14 days: n = 30, 15–21 days: n = 17. (b) CSF protein level. Numbers evaluated: at admission: n = 116, 0–2 days: n = 18, 3–7 days: n = 31, 8–14 days: n = 30, 15–21 days: n = 16. (c) CSF
glucose level. Numbers evaluated: at admission: n = 118, 0–2 days: n = 17, 3–7 days: n = 30, 8–14 days: n = 29, 15–21 days: n = 14.
430 ClinicalM
icrobiology and
Infection,V olum
e 22
N um
ber 5,M
ay 2016
–433
FIG. 2. Indication and results of repeat lumbar puncture, stratified by time. Cell count in cells/mm3. LP, lumbar puncture; ICP, intracranial pressure;
CSF, cerebrospinal fluid; BM, bacterial meningitis; DCT, delayed cerebral thrombosis.
CMI Costerus et al. Repeat lumbar puncture in bacterial meningitis 431
patients had a history of diabetes. Microbiological diagnosis was
made based on CSF culture of the initial CSF examination in seven of these eight patients and of the repeat CSF culture in
one: Streptococcus pneumoniae (n = 4), Streptococcus para- sanguinis (n = 2), Haemophilus influenzae (n = 1) and Strepto-
coccus agalactiae (n = 1). Repeat CSF culture showed bacteria in nine of 91 patients in
whom cultures were obtained (10%; Table 3). One patient had
a clinical relapse and was subsequently diagnosed with a recto- spinal fistula and one patient deteriorated at day 25 and was
diagnosed with ventriculitis; these findings led to a switch to broader spectrum antibiotics. The positive culture in the
additional seven patients was performed within 4 days after the start of therapy, and led to a change of antibiotic regimen in
two.
TABLE 2. Patients with an initial cerebrospinal fluid cell count <7 c
Sex/ age Relevant medical history CRP (g/L) Gram-
F/72 Diabetes mellitus 252 Pos M/64 Prednisolone for polymyalgia rheumatica 500 Pos M/59 Neutropenia due to chemotherapy 205 Pos M/72 None 235 NB F/73 None 319 Pos F/71 None 342 NB M/45 Pretreated with antibiotics for dental procedure 84 NB F/69 Pretreated with antibiotics for dental procedure 5 NB
CRP, C-reactive protein; CSF, cerebrospinal fluid; GOS, Glasgow Outcome Scale. aGram staining of initial CSF. Pos = Gram staining showed bacteria. NB = Gram staining sho bCSF cell count in cells/mm3, Δt = time interval between first and repeat lumbar puncture
© 2016 European Society of Clinical Microbiology
After 10 days of treatment, 25 of 50 patients (50%) in whom
lumbar puncture was repeated had a CSF cell count over 50 cells/mm3. This included one patient with relapsing bacterial
meningitis (day 19) and five with delayed cerebral thrombosis (DCT). In total, nine patients were diagnosed with DCT and all
had pleocytosis in the repeat lumbar puncture; median time to repeat lumbar puncture in DCT was 13 days (IQR 10–16). Patients with DCT had a higher median CSF cell count in the
repeat analysis compared with those who underwent lumbar puncture between days 8 and 14 without DCT (759 cells/mm3
(IQR 72–1464) versus 86 cells/mm3 (IQR 41–199); p 0.01). Unfavourable outcome occurred in 72 of 124 patients with
repeat lumbar puncture (58%) compared with 486 of 1366 patients with a single lumbar puncture (36%, p <0.001). The
mortality rate between patients with and without repeat
ells/mm3 a
staina Causative organism Repeat CSF cell count (Δt)b GOS
Streptococcus pneumoniae 1200 (1) 4 S. pneumoniae 810 (2) 5 S. pneumoniae ND (10) 3 Streptococcus parasanguinis 41 (4) 3 S. pneumoniae 486 (1) 5 Haemophilus influenzae 298 (5) 5 Streptococcus agalactiae 3 (18) 5 S. parasanguinis 56 (1) 5
wed no bacteria. in days.
and Infectious Diseases. Published by Elsevier Ltd. All rights reserved, CMI, 22, 428–433
TABLE 3. Characteristics of patients with bacterial meningitis and a positive second cerebrospinal fluid culture
Sex/ age First/ Second culture First CSF cell counta Second CSF cell count (Δt)a Additional diagnosis AB regimen expanded GOS
M/59 Streptococcus pneumoniae (2x) 33 140 (1) Multiple myeloma No 5 F/72 Negative/ S. pneumoniae 1 1200 (1) None No 4 M/60 Negative/ Streptococcus oralis 597 1125 (25) Ventriculitis Yes 1 F/69 Staphylococcus aureus (2x) ND ND (1) None No 5 F/85 Listeria monocytogenes (2x) 263 855 (2) Hydrocephalus No 3 M/81 S. aureus (2x) ND ND (2) Hydrocephalus No 5 M/30 Contaminated/ L. monocytogenes 3100 1140 (3) Hydrocephalus Yes 4 M/82 Negative/ Escherichia coli 11377 52033 (5) Recto-spinal fistula Yes 1 M/72 Streptococcus parasanguinis/ S. pneumoniae 3 14 (4) Hydrocephalus Yes 3
CSF, cerebrospinal fluid; AB, antibiotic; ND, not determined; GOS, Glasgow Outcome Scale Score. aCSF cell count in cells/mm3, Δt = time interval in days.
432 Clinical Microbiology and Infection, Volume 22 Number 5, May 2016 CMI
lumbar puncture was similar (25 of 124 (20%) versus 229 of
1137 (20%)).
Discussion
Our study shows that repeat lumbar puncture is performed in a
small proportion of adults with bacterial meningitis (8%), most commonly because of clinical deterioration or hydrocephalus.
This rate is much lower than previously reported. One retro- spective US single centre cohort reported a rate of repeat
lumbar puncture of 68% in the period 1969 and 1980 [14]. A more recent retrospective single centre ICU study in France reported a rate of 47% in the period 2000–11 [3]. The latter
study concluded that the yield of repeat lumbar puncture in adults with pneumococcal meningitis treated with adequate
antibiotic therapy is low, even in patients infected by strains resistant to penicillin or cephalosporin [14,19]. Our study was
nationwide and, therefore, we were able to study a represen- tative sample of adults with acute bacterial meningitis. The low
rate of repeat lumbar puncture in our study concurs with current guidelines recommending repeat lumbar puncture in any patient who has not responded clinically after 48 h of
appropriate antimicrobial therapy to identify penicillin-resistant or cephalosporin-resistant strains [4].
Our data suggest that repeat lumbar puncture can be useful in selected cases. First, repeat lumbar puncture is indicated in
cases of diagnostic uncertainty after first CSF examination. We identified a small number of patients with suspected bacterial
meningitis but normal CSF examination; in these patients repeat lumbar puncture confirmed the diagnosis [20]. Second, repeat
lumbar puncture is needed to rule out relapsing infection in patients who deteriorate after 48 h of adequate antibiotic therapy. We identified two patients with relapsing infection. A
small proportion of patients with repeat lumbar puncture was diagnosed with delayed cerebral thrombosis after clinical
deterioration (< 1%); these patients had increased CSF white cell counts but negative bacterial cultures at repeat CSF
© 2016 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier
examination. Third, repeat lumbar puncture can be used
therapeutically in patients with hydrocephalus. Hydrocephalus occurs in about 5% of patients and has been associated with
unfavourable outcome and death [5]. Acute obstructive hy- drocephalus in bacterial meningitis is thought to be very rare, and these rare cases usually occur in young children [5,21,22].
In patients with obstructive hydrocephalus lumbar puncture should not be performed [23]. Finally, compliant with current
guidelines, repeat lumbar puncture can be considered in any patient who has not responded clinically after 48 h of appro-
priate antimicrobial therapy [4]. Our study provides reference values for repeat CSF exam-
ination in adults with community-acquired bacterial meningitis. The distribution of causative pathogens among patients with
repeat lumbar puncture is similar to that in the total group of patients with community-acquired bacterial meningitis [24]. Median CSF cell count in the first 2 days after start of treatment
remained stable, followed by a quick decrease between 3 and 14 days and a subsequent relative plateau phase after 14 days of
treatment. Median CSF protein level stabilized after a decrease of 75% within 3–7 days of treatment. CSF glucose levels
showed a quick increase within 2 days of treatment. In patients with an abnormal clinical course and a CSF cell count >50 cells/
mm3 after 10 days, recurrent infection or an immunological phenomenon (delayed cerebral thrombosis) should be consid- ered. This is in contrast with a retrospective paediatric study
from 1972 to 1977 showing CSF cell count >50 cells/mm3 in 9 of 27 children (33%) with bacterial meningitis who all recovered
well [12]. Differences between findings can be explained by confounding by indication or adjunctive treatment with dexa-
methasone. In the childhood study repeat lumbar puncture was performed in all patients, whereas in our study this was done
on indication only (8%). In our study 77% was treated with adjunctive dexamethasone but this was not routine practice in
in the period 1972–77. Our study has several limitations. We did not have infor-
mation on indication for repeat lumbar puncture in 20% of
cases. This was mainly caused by the research design as the
Ltd. All rights reserved, CMI, 22, 428–433
CMI Costerus et al. Repeat lumbar puncture in bacterial meningitis 433
reason for repeat lumbar puncture was not prospectively
collected in the case record form. Second, only patients who underwent lumbar puncture and who had a positive CSF cul-
ture were included. Negative CSF cultures occur in 11–30% of patients with bacterial meningitis [25]. No significant differences
in clinical presentation have been reported between patients with culture-positive bacterial meningitis and those with culture-negative bacterial meningitis. Therefore, it is unlikely
that this factor confounded our results. Finally, rates of antibi- otic resistance among meningococcal and pneumococcal iso-
lates in the Netherlands are low [26]. Therefore, our results cannot automatically be extrapolated to regions with high
antibiotic resistance rates. However, a previous observational study concluded that repeat lumbar punctures in patients with
pneumococcal meningitis were not very helpful as routine, also not in an area of high penicillin-resistance [3].
Funding
This study was funded by the Netherlands Organization for Health Research and Development (ZonMw; NWO-Veni- Grant (916.13.078) to MB, NWO-Vidi-Grant (016.116.358)
to DvdB), the Academic Medical Centre (AMC Fellowship to DvdB), and the European Research Council (ERC Starting
Grant (281156) to DvdB). The Netherlands Reference Labo- ratory is funded by the National Institute of Public Health and
the Environment. Other authors: no financial support.
Acknowledgements
We are indebted to all the Dutch physicians who participated in the MeninGene study.
Transparency Declaration
All authors have stated that there are no conflicts of interest.
References
[1] Brouwer MC, Thwaites GE, Tunkel AR, van de Beek D. Dilemmas in the diagnosis of acute community-acquired bacterial meningitis. Lancet 2012;380:1684–92.
[2] van de Beek D, de Gans J, Tunkel AR, Wijdicks EF. Community-ac- quired bacterial meningitis in adults. N Engl J Med 2006;354:44–53.
[3] Denneman L, Vial-Dupuy A, Gault N, Wolff M, van de Beek D, Mourvillier B. Repeated lumbar puncture in adults with pneumococcal meningitis: an observational study. J Infect 2013;67:350–3.
© 2016 European Society of Clinical Microbiology
[4] Tunkel AR, Hartman BJ, Kaplan SL, Kaufman BA, Roos KL, Scheld WM, et al. Practice guidelines for the management of bacterial meningitis. Clin Infect Dis 2004;39:1267–84.
[5] Kasanmoentalib ES, Brouwer MC, van der Ende A, van de Beek D. Hydrocephalus in adults with community-acquired bacterial meningitis. Neurology 2010;75:918–23.
[6] Schut ES, Brouwer MC, de Gans J, Florquin S, Troost D, van de Beek D. Delayed cerebral thrombosis after initial good recovery from pneumococcal meningitis. Neurology…
Related Documents
