Published Ahead of Print on January 22, 2016 as 10.1212/WNL.

0000000000002315

Community-acquired bacterial meningitis

in adults with cancer or a history of cancer

Joost M. Costerus, MD ABSTRACT

Matthijs C. Brouwer, Objectives: To study the incidence, clinical presentation, causative bacteria, and outcome of
MD, PhD community-acquired bacterial meningitis in adults with cancer.
Arie van der Ende, PhD
Methods: We evaluated incidence and characteristics of patients with cancer included in a nation-
Diederik van de Beek,
wide prospective cohort study of adults with community-acquired meningitis performed in the
MD, PhD
Netherlands from March 1, 2006, to September 31, 2014. All patients underwent a neurologic
examination at hospital discharge, and outcome was graded using the Glasgow Outcome Scale.
Correspondence to Results: Active cancer was identified in 68 of 1,351 episodes (5%) and a history of cancer in 87 (6%).
Prof. Dr. van de Beek: The annual incidence of community-acquired bacterial meningitis was 2.71-fold (95% confidence
[email protected]
interval [CI] 1.68–4.36, p , 0.001) increased for patients with cancer compared to patients without
cancer in 2010, and 3.52-fold (95% CI 2.16–5.73, p , 0.001) in 2013. The clinical presentation of
bacterial meningitis in patients with cancer compared to patients without cancer was similar. Patients
with active cancer presented with lower leukocyte count in blood (12.1 3 109 cells/L vs 17.3 3 109
cells/L, p , 0.001) and CSF (670 cells/mm3 vs 2,567 cells/mm3, p , 0.001) and were more likely to
be infected with Listeria monocytogenes (21% vs 5%, p , 0.001) than patients without cancer.
Active cancer was identified as an independent risk factor for unfavorable outcome in bacterial men-
ingitis (odds ratio 1.85, 95% CI 1.09–3.13).
Conclusions: One of 8 patients with community-bacterial meningitis was identified to have a his-
tory of cancer and cancer was considered active in half of these patients. Patients with active
cancer present with lower CSF leukocyte counts, are more likely to be infected with L monocy-
togenes, and are at high risk of unfavorable outcome. Neurology® 2016;86:1–7

GLOSSARY
CI 5 confidence interval; NRLBM 5 Netherlands Reference Laboratory for Bacterial Meningitis.

Bacterial meningitis may be acquired in the community setting or result from invasive procedures
or complicated head trauma.1,2 Common risk factors for community-acquired bacterial meningitis
are distant or continuous foci of infection (otitis, sinusitis, or pneumonia) or deficiencies of the
immune system.3 Immunodeficiency may either be inherited (genetic) or acquired.4 Acquired
causes of immunodeficiency predisposing to bacterial meningitis include splenectomy, alcoholism,
diabetes mellitus, the use of immunosuppressive medication, and cancer.5–10
Patients with cancer are thought to have a higher risk of CNS infections compared to the
general population, but prospective studies are lacking.9,10 Cancer may predispose to bacterial
meningitis in several ways. The risk of severe infections is increased due to poor condition of the
patient or a direct effect of cancer on the function of the immune system.11,12 Chemotherapy or
splenectomy, for a hematologic malignancy, can result in severe immunodeficiency, which
increases the risk of infections.5,13,14 Furthermore, patients receiving allogeneic hematopoietic
stem cell transplantation may require intensive immunosuppressive therapy for graft-vs-host
disease, a condition associated with high risk of infections, for which preventive antibiotic
treatment is given.15–18 Finally, tumors of the nasopharynx or skull may disrupt the protective
natural barriers surrounding the brain and create an entry site for bacteria.19
Supplemental data
at Neurology.org
From the Departments of Neurology (J.M.C., M.C.B., D.v.d.B.), Medical Microbiology (A.v.d.E.), and the Netherlands Reference Laboratory for
Bacterial Meningitis (A.v.d.E.), Academic Medical Center, Center of Infection and Immunity Amsterdam, the Netherlands.
Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.

© 2016 American Academy of Neurology 1

ª 2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 We investigated the clinical and laboratory of meningitis or considered cured or inactive. Cancer was consid-
ered active if patients currently underwent chemo- or radiotherapy, if
features, causative bacteria, and prognosis of bac-
patients received palliative care for their malignancy, if patients with a
terial meningitis in patients with cancer who did diagnosis of cancer received no treatment at all, or if patients had a
not undergo neurosurgical intervention before metastasized malignancy. As chronic lymphocytic leukemia and
meningitis in a nationwide prospective cohort Waldenström macroglobulinemia are generally considered incurable,
patients with one of these diseases were included in the category of
study. active cancer. In multiple myeloma, a disease-free episode can be
achieved after intensive treatment. If this was the case, these patients
were included in the inactive category. We categorized cancer types
METHODS In this nationwide prospective cohort study, we
into hematologic malignancies and solid malignancies. Solid cancer
included patients older than 16 years who were listed in the data-
was defined as any cancer except hematologic cancer and basal-cell
base of the Netherlands Reference Laboratory for Bacterial Meningitis
carcinoma. Patients with both solid and hematologic cancer were
(NRLBM) from March 1, 2006, to September 31, 2014. This labo-
included in both groups.
ratory receives CSF from 90% of all patients with CSF culture-positive
All patients underwent a neurologic examination, and out-
bacterial meningitis in the Netherlands (population, 16.9 million).
come was graded using the Glasgow Outcome Scale at hospital
The NRLBM provided daily updates of the names of the hospitals
discharge.21 A favorable outcome was defined as a score of 5; an
where patients with bacterial meningitis had been admitted 2 to 6
unfavorable outcome was defined as a score of 1 to 4.
days previously. The treating physician was contacted. Physicians
The prevalence of cancer in the Netherlands was obtained from
could also contact the investigators without report of the NRLBM
the Dutch Cancer Registry, which systematically collects, analyzes,
for inclusion of patients. Episodes reported by physicians with
and reports data concerning people living with cancer in the
negative CSF cultures or PCR were only included if CSF results
Netherlands. Data on the Dutch population was obtained from
showed at least one individual predictor of bacterial meningitis
Statistics Netherlands (statline.cbs.nl) for the years 2010 and
(defined as a glucose level of less than 34 mg/dL [1.9 mmol/L], a
2013. The Mann–Whitney U test was used to identify differences
ratio of CSF glucose to blood glucose of less than 0.23, a protein
between episodes with and without (active) cancer regarding con-
level of more than 220 mg/dL, or a leukocyte count of more than
tinuous variables, and dichotomous variables were compared with
2,000/mL) and the clinical presentation was compatible with
use of the x2 test or Fisher exact. All tests were 2-tailed and a p value
bacterial meningitis.20 Episodes of meningitis in patients with a
,0.05 was considered significant. We used logistic regression anal-
neurosurgical device, with hospital-acquired meningitis, and with
ysis to calculate odds ratio and 95% confidence interval (CI) to
neurosurgery or neurotrauma within 1 month of the onset of
assess the strength of the association between potential risk factors
meningitis were excluded. Data on patient history, symptoms and
(including cancer) and outcome. Statistical analyses were performed
signs on admission, laboratory findings, radiologic examination,
with use of SPSS Statistics, version 20 (IBM Corp., Armonk, NY).
treatment, and outcome were prospectively collected by means of
an online Case Record Form. Standard protocol approvals, registrations, and patient
The Case Record Form included a standard question if the consents. The study was approved by the medical ethical commit-
patient had a history of cancer. All discharge letters of patients tee of the Academic Medical Centre, Amsterdam, the Netherlands.
with cancer were retrospectively evaluated to describe the type Informed consent was obtained from all participating patients or
of cancer, and to score if cancer was active during the episode their legally authorized representatives.

Table 1 Classification of malignancy in patients with bacterial meningitis and RESULTS Incidence. From March 2006 to September
cancera 2014, we included 1,351 episodes of bacterial meningitis
in the cohort: 168 (12%) meningitis episodes occurred
Solid malignancy (n 5 93) Total (active) Hematologic malignancy (n 5 67) Total (active)
in 168 patients with a history of cancer or active cancer.
Breast 25 (5) Multiple myeloma 20 (13) The annual incidence of community-acquired bacterial
Lung 13 (11) Non-Hodgkin lymphoma 20 (9) meningitis was 2.71-fold increased (95% CI 1.68–
Large intestine 11 (4) Chronic lymphoid leukemia 15 (15) 4.36, p , 0.001) for cancer patients compared to the
Urinary tract 9 (4) Hodgkin disease 6 (1)
normal population in 2010, and 3.52-fold (95% CI
2.16–5.73, p , 0.001) in 2013. Five patients with
Gynecologic 8 (1) Acute myeloid leukemia 3 (1)
cancer had experienced a previous episode of bacterial
Oto/nasal/laryngeal 7 (1) Waldenström macroglobulinemia 1 (1)
meningitis before the study period.
Prostate 5 (1) Acute lymphoid leukemia 1 (0)
Classification of malignancy. Data on type, localization,
Stomach 4 (1) Unspecified leukemia 1 (0)
and status of cancer could be retrieved for 163 of 168 ep-
Melanoma 3 (1) Unknown type (n 5 5) 5 (unknown)
isodes (97%; table 1). Noninvasive basal-cell carcinomas
Esophagus 2 (2) occurred in 8 patients, which were classified into the “no
Testis 2 (0) history of cancer” group. Active cancer was present in
Pancreas 1 (1) 68 patients (44%) and the remaining 87 patients
Liver 1 (1)
(56%) were considered cured from cancer, or there
were no signs of any active disease following treatment.
Primary brain tumor 1 (0)
A total of 93 patients with bacterial meningitis had been
Thyroid gland 1 (0)
diagnosed with a solid malignancy, most commonly
a
Solid and hematologic malignancy occurred in 5 patients simultaneously. breast (27%) and lung (14%) cancer. Of the 13

2 Neurology 86 March 1, 2016

ª 2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 patients with lung cancer, 6 had intracerebral metastases. were seen in 10 patients with active cancer, and a pri-
An otolaryngeal malignancy was diagnosed in 7 patients, mary tumor in the nasopharynx in 2. Other abnormal-
of whom 6 underwent tumor resection 1 year or more ities were hydrocephalus in 5 patients, of whom one
before the meningitis episode. A primary intracerebral had active cancer, and brain abscess in 2 patients. We
malignancy occurred in one patient who had undergone identified 41 patients with cancer who were severely
resection of a medulloblastoma during childhood. The immunocompromised (defined as immunosuppressive
solid malignancy was considered active in 33 of 93 therapy or HIV infection), of whom 32 had active can-
patients (35%). A total of 67 patients with bacterial cer. Imaging was performed on admission in 28 of 41
meningitis had been diagnosed with a hematologic patients, of which in 21 (54%) before lumbar puncture.
malignancy, most frequently multiple myeloma In 13 of these 28 patients (46%), abnormalities were
(30%), non-Hodgkin lymphoma (30%), and chronic seen. These were metastasis in 6, a brain abscess in
lymphoid leukemia (22%). Forty patients (60%) with a one, generalized edema in 3, and old vascular lesions
hematologic malignancy were classified as having active in 3. Five of 41 patients (12%) died within 1 day after
cancer. A history of stem cell transplantation was found the lumbar puncture. In 4 of these 5, imaging was per-
in 11 patients, 3 in patients with active cancer and 8 in formed before the lumbar puncture, which showed no
patients with a history of cancer. Five patients with contraindication for lumbar puncture.
meningitis had been diagnosed with both a solid
Causative organism. Streptococcus pneumoniae was the
malignancy as well as a hematologic malignancy.
most common causative organism in all patients with
Cancer was newly diagnosed during the episode of
active cancer or a history of cancer (43 of 68 [63%]
bacterial meningitis in 10 patients. These patients
and 61 of 87 [70%]). Listeria monocytogenes was iden-
were diagnosed with colon carcinoma with liver
tified in 14 of 68 patients (21%) with active cancer,
metastasis, with a nephrotic syndrome due to
consisting of 9 of 33 patients (27%) with a solid
multiple myeloma, with gastric obstruction due to
tumor and 5 of 40 patients (13%) with a hematologic
esophageal cancer, non-Hodgkin lymphoma, multiple
malignancy. L monocytogenes was the causative orga-
myeloma, chronic lymphoid leukemia, or lung cancer
nism in 5 of 87 patients (6%) with inactive cancer.
with brain metastasis. In one patient, the result of the
Neisseria meningitidis caused bacterial meningitis in 6
nationwide screening for breast cancer came back during
patients, of whom one had active cancer.
the admission for bacterial meningitis, which showed
a ductal carcinoma. Nine patients had possible Treatment. Initial antibiotic treatment consisted of
disruption of the natural barriers of the brain: solid penicillin or amoxicillin in 15% of episodes, a third-
tumors in 6 and hematologic malignancy in 3. generation cephalosporin in 13%, and a combination
Nineteen patients with active cancer were of penicillin or amoxicillin with a third-generation
currently under chemotherapy and 16 were treated cephalosporin in 54%; other regimens were used in
with corticosteroids. Other (nonimmunosuppressive) 18% of episodes. Dexamethasone was given in 42
cancer-related therapies that patients with active of 68 patients with active cancer (62%). Of the 26
cancer were receiving when the meningitis episode patients who did not receive dexamethasone, 18
occurred were an angiogenesis inhibitor (n 5 2) and (69%) were being treated with immunosuppressive
tyrosine kinase inhibitors (erlotinib hydrochloride and therapy when the bacterial meningitis occurred.
imatinib, both n 5 1).
Cohort outcome. An unfavorable outcome occurred in
Clinical presentation. The majority of patients with 81 of 160 patients (51%) with bacterial meningitis
cancer presented with classic symptoms and signs of and cancer, and 42 patients (26%) died. Of 67 pa-
bacterial meningitis. Headache occurred in 78% of epi- tients (27%) with a hematologic malignancy, 18 died
sodes, neck stiffness in 69%, fever (temperature $38°C) following the bacterial meningitis episode, split up
in 75%, and a change in mental status (Glasgow Coma into 11 of 40 patients (28%) with an active hemato-
Scale score ,14) in 72%. logic malignancy and 7 of 27 patients (26%) with a
history of cancer. Of 93 patients with a solid malig-
CSF and imaging. All patients underwent lumbar
nancy, 24 (26%) died: 13 of 33 (39%) with active
puncture. CSF chemical results were available for
cancer and 11 of 60 (18%) with a history of cancer.
155 of 160 patients (97%) with bacterial meningitis
and cancer and showed at least one individual predictor Characteristics by cohort. We compared the 68 patients
of bacterial meningitis in 137 patients (88%): in 61 of 66 with active cancer to 1,278 patients with cured cancer,
patients with active cancer (92%) and in 72 of 84 pa- inactive cancer, or no history of cancer (table 2). Patients
tients with a history of cancer (86%).20 Cranial imaging with active cancer more often were male (69% vs 50%,
was performed on admission in 54 of 68 patients (79%) p 5 0.002), more often were on immunosuppressive
with active cancer, of which 42 (78%) were performed therapy (47% vs 6%, p , 0.001), and less often had
before lumbar puncture. Metastases in the brain or skull otitis or sinusitis (9% vs 36%, p , 0.001). Patients with

Neurology 86 March 1, 2016 3

ª 2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 Table 2 Clinical and laboratory features in patients with bacterial meningitis and active cancer (n 5 68)
compared to patients with bacterial meningitis and cured, inactive, or no cancer (n 5 1,278)a

Characteristic Frequency (active cancer) Frequency (no/inactive cancer) p Value

Age, yr 67 (60–74) 60 (46–69) ,0.001

Male sex 47/68 (69) 634/1,278 (50) 0.002

Duration of symptoms <24 hr 33/67 (49) 577/1,222 (47) 0.74

Predisposing conditions

Immunocompromised stateb 39/68 (57) 308/1,278 (24) ,0.001

Otitis/sinusitis 6/68 (9) 452/1,272 (36) ,0.001

Pneumonia 11/64 (17) 107/1,224 (9) 0.02

Splenectomy 5/67 (7) 24/1,266 (2) 0.01

Immunosuppressive therapy 32/68 (47) 75/1,266 (6) ,0.001

Presenting symptoms

Headache 44/57 (77) 923/1,113 (83) 0.27

Nausea 34/55 (62) 654/1,056 (62) 1

Neck stiffness 41/63 (65) 901/1,204 (75) 0.08

Temperature ‡38°C 52/67 (78) 938/1,259 (75) 0.57

Score on GCSc 11 (8–14) 11 (9–14) 0.57

Altered mental status (GCS <14) 46/67 (69) 914/1,274 (72) 0.58

Coma (GCS <8) 10/67 (15) 168/1,274 (13) 0.70

Triad of fever, neck stiffness, altered 22/63 (35) 524/1,226 (41) 0.22
mental status

Blood chemical test resultsd

ESR, mm/hr 58 (23–93) 39 (21–69) 0.11

C-reactive protein, mg/L 230 (118–350) 188 (84–305) 0.22

Leukocyte count, 310 cells/L

9
12.1 (5.0–17.6) 17.3 (12.6–22.5) ,0.001

Leukocyte count £2.5 3 109 cells/L 12/66 (18) 8/1,259 (,1) ,0.001

Thrombocyte count, 109 cells/L 144 (98–213) 199 (152–254) ,0.001

Sodium, mmol/L 135 (132–137) 136 (134–139) 0.008

Potassium, mmol/L 3.9 (3.5–4.3) 3.5 (3.2–3.9) ,0.001

CSF examinatione

Leukocyte count, cells/mm3 670 (102–3,381) 2,567 (635–7,307) ,0.001

Leukocyte count <1,000/mm 3

34/60 (57) 370/1,174 (32) ,0.001

Leukocyte count <100/mm3 14/60 (23) 115/1,174 (10) ,0.001

Protein, g/L 3.61 (2.32–5.60) 3.95 (2.32–6.06) 0.56

Glucose, mmol/L 0.30 (0.00–2.00) 0.40 (0.00–2.50) 0.46

CSF to blood glucose ratio 0.03 (0.00–0.23) 0.04 (0.00–0.26) 0.51

CSF culture

Streptococcus pneumoniae 43/68 (63) 900/1,278 (70) 0.21

Neisseria meningitidis 1/68 (1) 133/1,278 (10) 0.02

Listeria monocytogenes 14/68 (21) 58/1,278 (5) ,0.001

Haemophilus influenzae 3/68 (4) 38/1,278 (3) 0.46

Otherf 5/68 (7) 109/1,278 (9) 0.74

Negative CSF culture 2/68 (3) 38/1,278 (3) 0.99

Abbreviations: ESR 5 erythrocyte sedimentation rate; GCS 5 Glasgow Coma Scale.

a
Data are no./no. evaluated (%) and continuous values are median (interquartile range).
b
Immunocompromised was defined as the use of immunosuppressive drugs, the presence of diabetes mellitus or
alcoholism, splenectomy, and patients with HIV.
c
Score on the GCS was determined in 67 and 1,274 patients.
d
ESR was determined in 32 and 592 patients, C-reactive protein in 65 and 1,218 patients, leukocyte count in 66 and
1,259 patients, and thrombocyte count in 65 and 1,220 patients.
e
CSF leukocyte count was determined in 60 and 1,174 patients, CSF protein in 65 and 1,212 patients, and CSF glucose
in 63 and 1,220 patients.
f
In active cancer: Escherichia coli, Streptococcus salivarius, Salmonella enterica (all n 5 1), and Staphylococcus aureus (n 5 2).

4 Neurology 86 March 1, 2016

ª 2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 active cancer had lower blood leukocyte count (median half of these patients. Few studies have described the
12.1 vs 17.3 3 109 cells/L, p , 0.001), lower blood incidence of cancer in patients with bacterial
thrombocyte count (median 144 vs 199 3 109 cells/L, meningitis. A large Spanish cohort study reported
p , 0.001), and lower CSF leukocyte count (median that 19 of 675 patients (3%) had a malignancy as
670 vs 2,567 3 109 cells/L, p , 0.001). Patients with predisposing factor, which is substantially lower than in
active cancer were more likely to be infected with L our cohort.22 Our prospective study was performed
monocytogenes (21% vs 5%, p , 0.001). Outcome in nationwide and therefore represents the actual incidence.
patients with active cancer was worse compared to those Of note, no other risk factor for meningitis than cancer
without active cancer: mortality was higher (22 of 68 itself could be identified in the majority of patients with
[32%] vs 205 of 1,278 [16%], p , 0.001), and more cancer. This concurs with an epidemiologic study
patients had an unfavorable outcome (40 of 68 [59%] vs showing the relative high risk of developing
461 of 1,278 [36%], p , 0.001; table 3). All 22 patients pneumococcal disease in patients with solid or
with active cancer who died had been treated with anti- hematologic cancer. 11 In our study, cancer was
biotics. In 6 patients, the treating physician stated with- independently associated with unfavorable functional
drawal of care as the cause of death. In multivariate outcome.
analysis including age, score on the Glasgow Coma The clinical presentation of patients with cancer was
Scale, causative organism, and dexamethasone treat- similar to that of patients without cancer. This is in con-
ment, active cancer was associated with unfavorable out- trast with a previous retrospective study.23 However,
come (odds ratio 1.85, 95% CI 1.09–3.13; table 4). In that study mainly included patients with nosocomial
tables e-1 and e-2 on the Neurology® Web site at meningitis who often present with atypical symptoms
Neurology.org, a comparison can be found of clinical and signs.2 We did observe differences in blood and
and laboratory features, complications, and outcome CSF chemical tests between bacterial meningitis pa-
between patients with inactive or cured cancer and tients with and without cancer, and a higher rate of
patients without cancer. infection by L monocytogenes in patients with cancer,
which is consistent with an immunocompromised state.
DISCUSSION Patients with cancer have an increased In almost a quarter of patients with active cancer, CSF
susceptibility for bacterial meningitis. We identified a white cell counts were below 100 cell/mm3, which may
3-fold-higher risk of meningitis as compared with potentially be misinterpreted as viral meningitis or men-
the normal population. One of 8 patients with ingitis carcinomatosa.24 In patients with active cancer,
community-bacterial meningitis was identified to have 12% had no individual CSF measure predictive of
a history of cancer and cancer was considered active in infection; therefore, antibiotic treatment should be con-
tinued until an alternative cause is found.
Cranial imaging before lumbar puncture was per-
Table 3 Complications and outcome in patients with bacterial meningitis and formed in 54% of patients with cancer and a severe
active cancer (n 5 68) compared to patients with bacterial meningitis immunocompromised state. Imaging in this population
and inactive or no cancer (n 5 1,278)a
identified abnormalities in 46%, including 7 patients
Frequency (active Frequency (no/inactive with space-occupying lesions. These findings suggest that
Characteristic cancer) cancer) p Value patients witch active cancer should undergo CT before
Systemic complications lumbar puncture, after blood has been drawn for culture,
Mechanical ventilation 16/65 (25) 417/1,222 (34) 0.11 and empirical antibiotic therapy has been initiated. One
Circulatory shock 8/62 (13) 130/1,212 (11) 0.59
should keep in mind that our study suffers from selection
bias since patients in whom lumbar puncture was not
Pneumonia 15/62 (24) 197/1,178 (17) 0.13
performed because of intracranial lesions were not
Neurologic complications
included in the study. In patients with a severe immuno-
Focal neurologic deficits 11/57 (19) 271/1,187 (23) 0.53
compromised state, cranial imaging before a lumbar
Seizures 12/62 (19) 159/1,219 (13) 0.15 puncture is advised as space-occupying lesions may be
Glasgow Outcome Scale present.25,26 To prevent treatment delay, empirical ther-
score
apy should be initiated in all patients with suspected
Death 22/68 (32) 205/1,278 (16) ,0.001 bacterial meningitis who undergo cranial imaging before
Vegetative state 0/68 (0) 2/1,278 (,1) lumbar puncture.
Severe disability 7/68 (10) 53/1,278 (4) We identified 10 patients in whom a diagnosis of
Moderate disability 11/68 (16) 201/1,278 (16)
cancer was made during hospitalization for bacterial
meningitis. Studies on long-term follow-up of pneu-
Minor or no disability 28/68 (41) 817/1,278 (64) ,0.001
mococcal and L monocytogenes meningitis patients
Time to death, d 3 (2–9) 7 (2–15) 0.054
showed a substantially increased risk of dying from
a
Data are no./no. evaluated (%) and continuous values are median (interquartile range). cancer following the meningitis episode, suggesting

Neurology 86 March 1, 2016 5

ª 2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 screened for cancer, but patients may have been missed.
Table 4 Multivariate analysis for effect on unfavorable outcome
We also excluded patients with a neurosurgical opera-
Variable Odds ratio (95% CI) tion in the last month and patients with a permanent
Age (per increasing year) 1.03 (1.02–1.03)
neurosurgical device, e.g., a CSF catheter. These factors
could have led to an underestimation of the incidence
Active cancer 1.85 (1.09–3.13)
of cancer in patients with meningitis.
GCS score at admission (per increasing point) 0.87 (0.83–0.90)
Finally, the association between unfavorable out-
Dexamethasone 0.52 (0.39–0.69) come and cancer could be caused by withdrawal of
Causative organism care because of bad prognosis. However, in only 3 pa-
Streptococcus pneumoniae 1.65 (0.82–3.31) tients with cancer, antibiotics were stopped directly
Neisseria meningitidis 0.79 (0.33–1.92) after admission because of fulminant meningitis in
combination with advanced cancer.
Abbreviations: CI 5 confidence interval; GCS 5 Glasgow Coma Scale.
We conclude that cancer is an important risk fac-
Odds ratios are calculated in 10-year increments for age.
tor for bacterial meningitis and is associated with
death and unfavorable functional outcome. Patients
potential (subclinical) presence of cancer during the with active cancer present with similar clinical charac-
meningitis episode.27,28 The low overall incidence (10 teristics, but the hematologic and CSF laboratory
of 1,352 [,1%]) in our cohort shows that routine measures of infection are attenuated compared to
screening for unknown cancer in patients with bacte- the normal population and patients with a history
rial meningitis is not expected to be useful. of cancer. Patients with active cancer are at risk of
The most important causative pathogens were S pneu- infection with L monocytogenes.
moniae and L monocytogenes. S pneumoniae is the most
common cause of bacterial meningitis and is associated AUTHOR CONTRIBUTIONS
with an unfavorable outcome.3 The 3-fold-higher risk of J.M.C.: substantial contribution to conception and design, acquisition of
data, analysis and interpretation of data, drafted the manuscript, final
meningitis emphasizes the importance of pneumococcal approval of the version to be published. M.C.B.: substantial contribution
vaccine coverage in patients with cancer.29 L monocyto- to conception and design, acquisition of data, analysis and interpretation
genes is known to cause meningitis predominantly in of data, revised the manuscript for important intellectual content, final
approval of the version to be published. A.v.d.E.: substantial contribution
elderly and immunocompromised patients.30,31 A study
to acquisition of data, revised the manuscript for important intellectual
involving 88 patients with cancer and bacterial meningitis content, final approval of the version to be published. D.v.d.B.: substan-
identified L monocytogenes as the most common causative tial contribution to conception and design, acquisition of data, analysis
pathogen in patients with cancer.10 However, this study and interpretation of data, revised the manuscript for important intellec-
tual content, final approval of the version to be published.
was performed between 1971 and 1974, and therefore
does not represent the current situation, especially regard-
ACKNOWLEDGMENT
ing modern cancer treatments. Nevertheless, empirical The authors are indebted to all the Dutch physicians who participated in
treatment for patients with cancer and suspected bacterial the study.
meningitis should always cover L monocytogenes.
Our study has several limitations. Patients with a STUDY FUNDING
negative CSF culture were only included if the treat- This study was supported by the Netherlands Organization for Health
Research and Development (ZonMw; NWO-Veni grant [916.13.078]
ing physician reported the patient to the investigators.
to M.C.B., NWO-Vidi grant [016.116.358] to D.v.d.B.), the Academic
Patients with cancer are at higher risk of infections Medical Center (AMC Fellowship to D.v.d.B.), and the European
and so generally there is a low threshold to start anti- Research Council (ERC Starting Grant to D.v.d.B.). The Netherlands Ref-
biotics, or the patient may receive prophylactic antibi- erence Laboratory is funded by the National Institute of Public Health and
the Environment. D.v.d.B. is supported by grants from the Netherlands
otic treatment. Therefore, CSF and blood cultures Organization for Health Research and Development (ZonMw; NWO-
may be negative in a larger proportion of patients Vidi grant 2010 [016.116.358]), and the European Research Council
compared to the general bacterial meningitis popula- (ERC Starting Grant 281156). M.C.B. is supported by a grant from the
Netherlands Organization for Health Research and Development (ZonMw;
tion. Furthermore, patients with a hematologic malig-
NWO-Veni grant 2012 [916.13.078]).
nancy or those receiving chemotherapy may have
thrombocytopenia, which is a contraindication for per- DISCLOSURE
forming a lumbar puncture. In these patients, the lum- The authors report no disclosures relevant to the manuscript. Go to
bar puncture is often postponed awaiting thrombocyte Neurology.org for full disclosures.
transfusion or may be deferred. CSF cultures are more
likely to be negative in these patients and therefore they Received September 9, 2015. Accepted in final form November 5, 2015.

might not be included in this study, leading to a selec-

REFERENCES
tion in patients included in our cohort. Of note, for the 1. van de Beek D, de Gans J, Tunkel AR, Wijdicks EF.
interpretation of a history of cancer, we depended on Community-acquired bacterial meningitis in adults. N
the including physician. All discharge letters were Engl J Med 2006;354:44–53.

6 Neurology 86 March 1, 2016

ª 2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

 2. van de Beek D, Drake JM, Tunkel AR. Nosocomial bac- 18. Wingard JR, Vogelsang GB, Deeg HJ. Stem cell transplanta-
terial meningitis. N Engl J Med 2010;362:146–154. tion: supportive care and long-term complications. Hematol-
3. van de Beek D, de Gans J, Spanjaard L, Weisfelt M, ogy Am Soc Hematol Educ Program 2002:422–444.
Reitsma JB, Vermeulen M. Clinical features and prognos- 19. Huang CR, Lu CH, Chien CC, Lee PY, Chang WN. High
tic factors in adults with bacterial meningitis. N Engl J incidence of gram-negative bacillary infection and high mor-
Med 2004;351:1849–1859. tality in adult patients with bacterial meningitis and naso-
4. Brouwer MC, Read RC, van de Beek D. Host genetics and pharyngeal carcinoma. Eur J Clin Microbiol Infect Dis
outcome in meningococcal disease: a systematic review and 2003;22:509–511.
meta-analysis. Lancet Infect Dis 2010;10:262–274. 20. Spanos A, Harrell FE Jr, Durack DT. Differential diagno-
5. Adriani KS, Brouwer MC, van der Ende A, van de Beek D. sis of acute meningitis: an analysis of the predictive value
Bacterial meningitis in adults after splenectomy and hypo- of initial observations. JAMA 1989;262:2700–2707.
splenic states. Mayo Clin Proc 2013;88:571–578. 21. Jennett B, Teasdale G, Braakman R, Minderhoud J, Knill-
6. Mourtzoukou EG, Pappas G, Peppas G, Falagas ME. Vac- Jones R. Predicting outcome in individual patients after
cination of asplenic or hyposplenic adults. Br J Surg 2008; severe head injury. Lancet 1976;1:1031–1034.
95:273–280. 22. Cabellos C, Verdaguer R, Olmo M, et al. Community-
7. Muller LM, Gorter KJ, Hak E, et al. Increased risk of acquired bacterial meningitis in elderly patients: experience
common infections in patients with type 1 and type 2 over 30 years. Medicine 2009;88:115–119.
diabetes mellitus. Clin Infect Dis 2005;41:281–288. 23. Safdieh JE, Mead PA, Sepkowitz KA, Kiehn TE,
8. Nelson S, Kolls JK. Alcohol, host defence and society. Nat Abrey LE. Bacterial and fungal meningitis in patients with
Rev Immunol 2002;2:205–209. cancer. Neurology 2008;70:943–947.
9. Chernik NL, Armstrong D, Posner JB. Central nervous system 24. Pruitt AA. CNS infections in patients with cancer. Con-
infections in patients with cancer. Medicine 1973;52:563–581. tinuum 2012;18:384–405.
10. Chernik NL, Armstrong D, Posner JB. Central nervous 25. Hasbun R, Abrahams J, Jekel J, Quagliarello VJ. Com-
system infections in patients with cancer: changing pat- puted tomography of the head before lumbar puncture in
terns. Cancer 1977;40:268–274. adults with suspected meningitis. N Engl J Med 2001;
11. Kyaw MH, Rose CE Jr, Fry AM, et al. The influence of 345:1727–1733.
chronic illnesses on the incidence of invasive pneumococ- 26. Brouwer MC, Thwaites GE, Tunkel AR, van de Beek D.
cal disease in adults. J Infect Dis 2005;192:377–386. Dilemmas in the diagnosis of acute community-acquired
12. Hoffman MJ, Stosor V. Central nervous system infections bacterial meningitis. Lancet 2012;380:1684–1692.
in cancer patients and hematopoietic stem cell transplant 27. Roed C, Engsig FN, Omland LH, Skinhoj P, Obel N.
recipients. Cancer Treat Res 2014;161:253–298. Long-term mortality in patients diagnosed with pneumo-
13. Crawford J, Dale DC, Lyman GH. Chemotherapy- coccal meningitis: a Danish nationwide cohort study. Am J
induced neutropenia: risks, consequences, and new direc- Epidemiol 2010;172:309–317.
tions for its management. Cancer 2004;100:228–237. 28. Roed C, Engsig FN, Omland LH, Skinhoj P, Obel N.
14. Lukes SA, Posner JB, Nielsen S, Armstrong D. Bacterial Long-term mortality in patients diagnosed with Listeria
infections of the CNS in neutropenic patients. Neurology monocytogenes meningitis: a Danish nationwide cohort
1984;34:269–275. study. J Infect 2012;64:34–40.
15. Wingard JR, Hsu J, Hiemenz JW. Hematopoietic stem cell 29. Robin C, Beckerich F, Cordonnier C. Immunization in
transplantation: an overview of infection risks and epidemiol- cancer patients: where we stand. Pharmacol Res 2015;92:
ogy. Hematol Oncol Clin North Am 2011;25:101–116. 23–30.
16. Mihu CN, Schaub J, Kesh S, et al. Risk factors for late Staph- 30. Koopmans MM, Brouwer MC, Bijlsma MW, et al. Liste-
ylococcus aureus bacteremia after allogeneic hematopoietic ria monocytogenes sequence type 6 and increased rate of
stem cell transplantation: a single-institution, nested case- unfavorable outcome in meningitis: epidemiologic cohort
controlled study. Biol Blood Marrow Transpl 2008;14: study. Clin Infect Dis 2013;57:247–253.
1429–1433. 31. Brouwer MC, van de Beek D, Heckenberg SG,
17. Baden LR, Bensinger W, Angarone M, et al. Prevention Spanjaard L, de Gans J. Community-acquired Listeria
and treatment of cancer-related infections. J Natl Compr monocytogenes meningitis in adults. Clin Infect Dis
Canc Netw 2012;10:1412–1445. 2006;43:1233–1238.

Neurology 86 March 1, 2016 7

ª 2016 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Community-acquired bacterial meningitis in adults with cancer or a history of cancer
Joost M. Costerus, Matthijs C. Brouwer, Arie van der Ende, et al.
Neurology published online January 22, 2016
DOI 10.1212/WNL.0000000000002315

This information is current as of January 22, 2016

Updated Information & including high resolution figures, can be found at:
Services http://www.neurology.org/content/early/2016/01/22/WNL.0000000000
002315.full.html
Supplementary Material Supplementary material can be found at:
http://www.neurology.org/content/suppl/2016/01/22/WNL.000000000
0002315.DC1.html
Subspecialty Collections This article, along with others on similar topics, appears in the
following collection(s):
All Oncology
http://www.neurology.org//cgi/collection/all_oncology
Clinical neurology examination
http://www.neurology.org//cgi/collection/clinical_neurology_examinati
on
Cohort studies
http://www.neurology.org//cgi/collection/cohort_studies
Critical care
http://www.neurology.org//cgi/collection/critical_care
Meningitis
http://www.neurology.org//cgi/collection/meningitis
Permissions & Licensing Information about reproducing this article in parts (figures,tables) or in
its entirety can be found online at:
http://www.neurology.org/misc/about.xhtml#permissions
Reprints Information about ordering reprints can be found online:
http://www.neurology.org/misc/addir.xhtml#reprintsus

Neurology ® is the official journal of the American Academy of Neurology. Published continuously since
1951, it is now a weekly with 48 issues per year. Copyright © 2016 American Academy of Neurology. All
rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.