Original Paper
Cerebrovasc Dis 2010;29:576583
DOI: 10.1159/000306645
Association of Platelet and Leukocyte
Counts with Delayed Cerebral Ischemia in
Aneurysmal Subarachnoid Hemorrhage
K.M.Kasius a C.J.M.Frijns a A.Algra a,b G.J.E.Rinkel a
a
Department of Neurology, Rudolf Magnus Institute of Neuroscience, and b Julius Center for Health Sciences and
Primary Care, University Medical Center, Utrecht, The Netherlands
Key Words
Subarachnoid hemorrhage Ischemia Platelets
Leukocytes
Abstract
Background and Purpose: A proinflammatory prothrom-
botic state may increase the risk of delayed cerebral ischemia
(DCI) after aneurysmal subarachnoid hemorrhage (SAH). We
studied the relationship of levels of leukocytes, platelets, C-
reactive protein (CRP), and erythrocyte sedimentation rate
(ESR) with the development of DCI and with clinical outcome
in patients with aneurysmal SAH. Methods: In 125 patients
admitted within 72 h after aneurysmal SAH, we dichoto-
mized initial blood levels at their median values and investi-
gated the prediction of DCI with Cox proportional hazard
analysis and of poor clinical outcome with logistic regression
analysis. We also analyzed concentrations before and after
onset of DCI with the paired-samples t test and compared
changes with those in patients without DCI. Results: During
the development of DCI (unrelated to treatment), patients
had a larger increase in counts of platelets (difference 49 !
109/l; 95% CI: 298) and leukocytes (difference 2.6 ! 109/l;
95% CI: 0.45.0) than patients without DCI during the same
period. CRP increased during DCI and decreased in patients
without DCI (difference 14 mg/l; 95% CI: 29 to 58). ESR in-
2010 S. Karger AG, Basel
10159770/10/02960576$26.00/0
Fax +41 61 306 12 34
E-Mail karger@karger.ch Accessible online at:
www.karger.com www.karger.com/ced
Received: August 4, 2009
Accepted: January 22, 2010
Published online: April 8, 2010
creased slightly in both groups (difference 3 mm/h; 95% CI:
15 to 20). None of the determinants at baseline predicted
the development of DCI. An increased risk of poor outcome
predicted by a high initial leukocyte count (OR 2.5; 95% CI:
1.15.7) decreased after adjustment for clinical variables (OR
2.1; 95% CI: 0.85.5). Conclusion: Counts of platelets and leu-
kocytes disproportionally increase during the occurrence of
DCI after aneurysmal SAH. Drugs with anti-thrombotic or
anti-inflammatory properties should be studied for preven-
tion and treatment of DCI. Copyright 2010 S. Karger AG, Basel
Introduction
In patients who survive the initial hours after aneurys-
mal subarachnoid hemorrhage (SAH) and in whom the
aneurysm is secured, delayed cerebral ischemia (DCI) is
a frequent neurological complication, and an important
contributor to death or poor functional outcome. The
pathogenesis of DCI is still incompletely understood. Va-
sospasm is partly responsible for its development, but
since 30% of patients with DCI have no detectable vaso-
spasm [1], other factors must also be involved [2]. Char-
acterization of predictors of DCI and poor outcome may
improve the identification of patients with an increased
K.M. Kasius, MD
Department of Neurology, Canisius Wilhelmina Hospital
PO Box 9015
NL6500 GS Nijmegen (The Netherlands)
Tel. +31 24 365 8765, Fax +31 24 365 7329, E-Mail kristel_kasius@hotmail.com
risk of DCI and may direct development of new preven-
tive therapies. Inflammatory processes and a prothrom-
botic state may increase the risk of secondary ischemia
after aneurysmal SAH.
The aim of this study was to investigate the relation-
ship of markers of a prothrombotic or proinflammatory
condition, both at admission and during the development
of DCI, with the occurrence of DCI and with poor func-
tional outcome.
Patients and Methods
Patients
The patients in this study were included in a previous study on
endothelial cell markers in SAH [3]. All were admitted to the Uni-
versity Medical Center Utrecht with the diagnosis of aneurysmal
SAH based on a characteristic medical history, clinical symp-
toms, a CT scan showing subarachnoid blood, and a CT angio-
gram or conventional angiography showing an aneurysm. Pa-
tients were not included if they were not admitted within 72 h
after the onset of symptoms, if death appeared imminent on ad-
mission, or if DCI had developed before admission.
Informed consent was obtained from all patients or their rep-
resentatives. The study was approved by the institutional review
board.
Patients were admitted to the Medium Care Unit, unless their
clinical condition necessitated admission to the Intensive Care
Unit. All patients were under continuous observation, with con-
tinuous monitoring of ECG, pulse rate, oxygen saturation, and
respiratory frequency. Every hour, pupillary reflexes and level of
consciousness were assessed by means of the Glasgow Coma Scale
(GCS) score [4]. Blood pressure was measured at least every 4 h,
and in the first few days or in case of unstable blood pressure every
hour. The temperature was taken at least every 6 h, and every 6 h
the fluid balance was assessed. Medical treatment consisted of
administration of oral nimodipine and intravenous fluids aimed
at the maintenance of a neutral fluid balance. If clinical signs of
DCI developed, we administered a plasma expander (Gelo-
fusine) 500 ml 4 times per day. All patients received acetamino-
phen 1,000 mg four times a day for headache management. Non-
steroidal anti-analgesic drugs were not used. Compression stock-
ings were used to prevent deep vein thrombosis before occlusion
of the aneurysm; thereafter low molecular weight heparin was
given in a prophylactic dose.
Data Collection
All clinical, laboratory, and radiological data were collected
from the medical records and the hospital registry system, and an
anonymized medical history was made for each patient [3]. Dur-
ing the study period, standard laboratory investigations includ-
ing blood cell counts, C-reactive protein (CRP), and erythrocyte
sedimentation rate (ESR) (with modified Westergren assay)
were taken on admission and early in the morning on Mondays,
Wednesdays, and Fridays, according to our treatment protocol,
and more frequently on clinical indication. Since these data were
collected retrospectively, different numbers of blood samples
were analyzed per patient.
Platelets and Leukocytes in DCI after
SAH
Assessment of Secondary Ischemic Events
DCI was defined as a deterioration of consciousness (decrease
in GCS of 61 point) or focal signs that lasted for at least 1 h and
could not be explained by rebleeding, acute hydrocephalus, intra-
cerebral hemorrhage, epileptic seizures, or hemodynamic, respi-
ratory, or metabolic disturbances. These other causes were ex-
cluded with laboratory tests, and CT or MRI scanning of the brain
as soon as the deterioration was observed. We considered the pos-
sibility of epileptic seizures in any case of a decrease in conscious-
ness not otherwise explained. If an epileptic cause was consid-
ered, an EEG was performed. The clinical condition at admission
was classified according to the World Federation of Neurological
Surgeons Scale [5]. The amount of cisternal and ventricular blood
on the initial CT scan was assessed according to the Hijdra
score [6].
Radiological ischemic changes were defined as new ischemic
lesions if they were not visible on the initial scans and persisted
on follow-up scans. The occurrence of DCI was independently as-
sessed by 2 observers. All initial and follow-up scans were re-
viewed by the principal investigator (C.J.M.F.) and, in case of
doubt about ischemic lesions, also by a second observer (G.J.E.R.).
At the time of the clinical and radiological assessment of DCI, the
observers were not aware of the laboratory results. To exclude any
influence of treatment on the markers of a prothrombotic or pro-
inflammatory condition after DCI, we made a subgroup of pa-
tients with DCI unrelated to treatment, by excluding all instances
of DCI that had occurred during or within 24 h after an endovas-
cular procedure or clipping of the aneurysm.
Assessment of Poor Outcome
Functional outcome was evaluated at 3 months with the mod-
ified Rankin scale [7] by means of a personal or telephone inter-
view with the patients or their relatives by the same interviewer
(C.J.M.F.) in combination with data from rehabilitation clinics
and dichotomized into poor [dependent on help for activities of
daily living (Rankin 45) or death] or good outcome [able to live
at home (Rankin 03)].
Blood Sampling of Markers
The first blood sample was obtained within 72 h after the ini-
tial hemorrhage. The additional samples were obtained at least
thrice per week on alternating days. We assessed leukocyte count
(!109/l), platelet count (!109/l), CRP (mg/l), and ESR (mm/h).
Data Analysis: DCI
We analyzed both the relation between initial measurements
of the markers and development of DCI and the changes in levels
of markers during development of DCI. In these analyses, we in-
vestigated both DCI of any cause and DCI unrelated to treatment.
Patients with radiological ischemic changes without compatible
clinical symptoms of DCI were excluded from this analysis be-
cause accurate time of onset could not be determined.
For the relation between initial measurements and develop-
ment of DCI we used the first available blood level of each mark-
er and dichotomized these concentrations at their median values.
We calculated hazard ratios (HR) with corresponding 95% CI by
means of the Cox proportional hazard regression model to assess
the relationship between the markers at admission and the risk of
DCI. Patients were censored at the time of first rebleeding, death,
first cerebral ischemic event, or at the 21st day after onset, which-
Cerebrovasc Dis 2010;29:576583 577
ever came first. In a bivariable regression analysis we adjusted for
age, gender, World Federation of Neurological Surgeons Scale at
admission, loss of consciousness at the time of aneurysm rupture
(GCS !14, present or absent), signs of acute ischemia at time of
aneurysm rupture (i.e. loss of consciousness or focal signs in the
absence of an intracerebral hemorrhage, acute hydrocephalus or
other explanation), amount of cisternal blood, and amount of
ventricular blood. From the results of the bivariable regression
analysis, we selected the variables that changed the crude HR of
each marker by more than 5% to adjust for in the multivariable
analysis.
To assess changes during the development of DCI we com-
pared levels of markers in paired blood samples taken within
72h before and after the onset of DCI. To exclude the influence
of aneurysm treatment on the markers, the patients with DCI
were divided into 2 subgroups: (1) onset of DCI unrelated to treat-
ment of the aneurysm, or (2) onset of DCI within 24 h of clipping
or coiling of the aneurysm. Because the median day of DCI was
day 4, we compared the changes in concentrations in patients with
DCI with those in the same period in patients without DCI, i.e. in
blood samples taken between days 2 and 7 after onset of aneurys-
mal SAH with a minimum interval of 3 days between the 2 sam-
ples. These control patients without DCI were also divided into 2
subgroups: (1) no treatment of the aneurysm between the selected
blood samples, or (2) clipping or coiling between the blood sam-
ples. We used the paired t test for the detection of differences be-
tween serial concentrations, and the independent-samples t test
for comparisons between patients with and without DCI. Con-
centrations are reported as mean differences in concentrations
and their 95% CI. Values of p ! 0.05 were considered statistically
significant.
Because patients with SAH frequently develop infections, and
infections may influence levels of some of the markers, we also
analyzed the changes in serial concentrations after exclusion of
patients with a infection between the 2 samples, proven by micro-
biological cultures or chest radiographs. In case of leukocytosis or
an elevated CRP accompanied by fever or by clinical symptoms of
infections, a complete work-up was performed.
Data Analysis: Poor Outcome
To determine the relationship between the markers and poor
outcome, we used the first blood sample obtained after the hem-
orrhage. We dichotomized the concentrations of each marker at
the median value, and calculated crude and adjusted odds ratios
(OR) with corresponding 95% CI by means of logistic regression
analysis. The relevant variables adjusted for in the multivariable
regression analysis resulted from bivariable regression analysis in
the same manner as described for DCI.
Results
Patients
During the 2-year study period, 125 patients with SAH
consented to participate in the study. After the exclusion
of 18 patients with a non-aneurysmal cause of the hemor-
rhage including perimesencephalic hemorrhage, and 1
patient in whom severe cerebral ischemia occurred before
578 Cerebrovasc Dis 2010;29:576583
admission, 106 patients were included and analyzed for
functional outcome (fig.1). The clinical characteristics of
the patients are presented in table1. In 76 patients (71.7%),
the aneurysm was secured by surgical clipping on medi-
an day 4, and in 13 patients (12.3%) by endovascular coil-
ing on median day 3 after the onset of SAH. In the re-
maining patients, treatment of the aneurysm was not per-
formed due to poor clinical condition.
For the analyses with DCI, we additionally excluded
13 patients with only radiological ischemic lesions be-
cause of uncertainty about time of onset of DCI, and 2
patients who died within 3 days after the hemorrhage.
Thus, for DCI as outcome measurement, 91 patients re-
mained. This group of patients consisted of 33 patients
with DCI and 58 patients without DCI. Fourteen patients
met the criteria for DCI unrelated to treatment.
DCI
Prediction
There were no statistically significant associations of
initial levels of the investigated markers with the occur-
rence of DCI from any cause or with DCI unrelated to
treatment, both in the univariable and in the multivari-
able analysis in patients with symptomatic DCI (n = 33)
or in patients with both symptomatic and radiological
DCI (n = 29) (data not shown).
Changes during DCI
There was an increase in the platelet count in the entire
group of patients with DCI (46 ! 109/l, 95% CI 1476)
that was even more marked in the patients with DCI un-
related to treatment (93 ! 109/l, 95% CI 34153) (table2).
In the corresponding control group without DCI, there
was also an increase in the platelet count (44 ! 109/l, 95%
CI 2365). The difference in increases between the pa-
tients with DCI unrelated to treatment and the control
patients was 49 ! 109/l (95% CI 1.698).
The leukocyte count increased slightly in the entire
group of patients with DCI (0.8 ! 109/l, 95% CI 0.7 to
2.3) and decreased in the group of control patients (1.0
! 109/l, 95% CI 1.8 to 0.2). The difference between pa-
tients with and without DCI was 1.8 ! 109/l (95% CI 0.3
3.3). In the subgroup of patients with DCI unrelated to
treatment, this difference was more marked (2.6 ! 109/l,
95% CI 0.45.0).
CRP increased in the entire group of patients with
DCI of any cause by 34 mg/l (95% CI 859), and also in
the group of patients with DCI unrelated to treatment
(6mg/l, 95% CI 40 to 53), but decreased in the control
group of patients without intervention (8 mg/l, 95% CI
Kasius /Frijns /Algra /Rinkel
 n = 125

Excluded

n = 10 perimesencephalic
hemorrhage
n = 8 no detectable aneurysm
n = 1 cerebral ischemia before
admission

Analysis of
n = 106
outcome

Excluded

n = 12 only radiological ischemia,

timing of onset imprecise
n = 1 insufficient blood sampling
n = 2 died before day 3

Analysis of
n = 91
ischemic events

No ischemic event Ischemic events all causes

n = 58 n = 33

n = 14 unrelated to treatment
n = 13 <24 h after operation
n = 6 <24 h after coiling

Fig. 1. Flowchart of selection of patients

and reasons for exclusion.

28 to 12. The difference between patients with DCI un-
related to treatment and control patients without treat-
ment was 14 mg/l (95% CI 29 to 58). During DCI, ESR
increased by 31 mm/h (95% CI 1250) in the entire group
of patients with DCI of any cause and by 11 mm/h (95%
CI 5.9 to 29) in the subgroup of patients with DCI unre-
lated to treatment. In the entire control group, ESR in-
creased by 22 mm/h (95% CI 1429) and in the subgroup
of controls without intervention by 42 mm/h (95% CI 26
58). There was no relevant difference between patients
with and without DCI.
After exclusion of patients with an infection at the
time of the blood samples we used for the analysis of DCI,
all results remained essentially the same.
Clinical Outcome
The crude OR for poor outcome in patients with an
initial leukocyte count above the median (13.6 ! 109/l)
was 2.5 (95% CI 1.15.7) (table 3). In the multivariable
analysis, a non-significant trend remained with an OR of
2.1 (95% CI 0.85.5). The other markers did not predict
poor outcome.
Discussion
The main findings of the present study were increases
in both platelet and leukocyte counts during the develop-
ment of DCI which were significantly larger than in pa-
tients without DCI at the same period after onset of an-
eurysmal SAH. This disproportionate increase in platelet
count during the development of DCI has not been re-
ported previously. Our findings suggest that a prothrom-
botic and proinflammatory condition is involved in the
development of DCI. In a previous study in this same pa-
tient population, we found an increase in sP-selectin dur-

Platelets and Leukocytes in DCI after Cerebrovasc Dis 2010;29:576583 579

SAH
Table 1. Clinical characteristics of patients and relationships with clinical outcome and delayed cerebral ischemia

Analysis of outcome Analysis of ischemic events

(n = 106) (n = 91)
good outcome poor outcome without DCI with DCI
(n = 66) (n = 40) (n = 58) (n = 33)

Age, years 51.4812.1 58.3812.7 53.7811.9 54.9814.4

Women 41 (62) 30 (75) 40 (69) 27 (82)
WFNS grade at admission
I 34 (52) 10 (25) 27 (47) 14 (43)
II 18 (27) 13 (33) 13 (22) 12 (36)
III 2 (3) 4 (10) 3 (5) 1 (3)
IV 10 (15) 8 (20) 11 (19) 5 (15)
V 2 (3) 5 (12) 4 (7) 1 (3)
Amount of blood1 > median 25 (38) 22 (55) 23 (40) 12 (36)
Time of blood sampling after 2 (03) 2 (03) 2 (03) 2 (04)
hemorrhage (median), days
Day of onset of DCI (median) 3 (013) 4 (116) 4 (116)
Treatment of aneurysm
Clipping 57 (86) 19 (48) 48 (83) 20 (61)
Coiling 9 (14) 4 (10) 6 (10) 5 (15)
No treatment (poor clinical condition ) 0 (0) 17 (42) 4 (7) 8 (24)
Day of operation (median) 4 (165) 3 (018) 9 (028) 3 (165)
Day of coiling (median) 3 (217) 4 (26) 3 (23) 3 (217)
Rebleeding 2 (3) 12 (30) 4 (7) 5 (15)
DCI
Clinical signs only 2 (3) 2 (5) 4 (12)
Clinical and radiological signs 12 (18) 19 (48) 29 (88)
Radiological signs only 6 (9) 6 (15)
Onset of DCI
No intervention 12 (18) 15 (38) 14 (42)
Intervention 8 (12) 11 (27) 19 (58)
Rankin score at 3 months
0 10 (15) 5 (9) 2 (6)
1 30 (46) 24 (41) 2 (6)
2 14 (21) 11 (18) 3 (9)
3 12 (18) 5 (9) 6 (18)
4 10 (25) 4 (7) 6 (18)
5 7 (17) 5 (9) 2 (6)
Death 23 (58) 4 (7) 12 (37)

Fig ures in parentheses contain percentages or ranges. WFNS = World Federation of Neurological Surgeons Scale [5].
1 Graded according to the Hijdra method (maximum score = 30) [6]; our median = 22.

ing the development of DCI [8]. At that time, we conclud-
ed that the increased levels of sP-selectin were most prob-
ably derived from platelets. The findings of the present
study are in line with that conclusion.
Platelet function has previously been implicated in the
pathogenesis of DCI after SAH. One study reported evi-
dence of platelet activation 4 days after the initial hemor-
rhage only in the patients who developed symptomatic va-
sospasm [9]. Two other studies also proposed that aug-
mented platelet function may be involved in the
pathogenesis of cerebral infarction after aneurysmal SAH
[10, 11]. Patients who developed DCI showed the highest
thromboxane release [11]. Nimodipine treatment has been
shown to increase platelet count and to decrease platelet
thromboxane B2 release [12]. Since all patients in our study
received nimodipine, irrespective of the development of
DCI, this cannot explain the difference in changes in
platelet count between patients with or without DCI. Yet

580 Cerebrovasc Dis 2010;29:576583 Kasius /Frijns /Algra /Rinkel

Table 2. Comparison of changes in serial concentrations after aneurysmal SAH in patients with or without DCI, and in subgroups ac-
cording to treatment of the aneurysm

DCI No DCI1 Difference DCI/no DCI

n before after change 95% CI n 1st 2nd change 95% CI difference 95% CI
DCI DCI sample sample

Platelets, !109/l
All 27 247868 291894 46 14 to 76 49 239860 281873 42 25 to 59 4 35 to 29
No intervention 11 264872 357864 93 34 to 153 36 237859 281880 44 23 to 65 49 1.6 to 98*
Intervention 16 235865 246885 11 16 to 39 13 245867 281853 36 1.9 to 69 25 17 to 65
Leukocytes, !109/l
All 29 12.483.8 13.283.7 0.8 0.7 to 2.3 55 13.384.2 12.383.8 1.0 1.8 to 0.2 1.8 3.3 to 0.3**
No intervention 11 13.483.2 15.283.4 1.8 1.1 to 4.8 40 13.183.9 12.384.0 0.8 1.8 to 0.2 2.6 5.0 to 0.4**
Intervention 18 11.784.0 11.983.4 0.2 1.6 to 1.9 15 13.785.3 12.283.5 1.5 3.0 to 0.1 1.7 3.9 to 0.7
CRP, mg/l
All 29 51884 85893 34 8 to 59 55 58883 74896 16 8 to 38 18 54 to 18
No intervention 11 54867 60847 6 40 to 53 40 64888 56885 8 28 to 12 14 58 to 29
Intervention 18 49894 998111 50 20 to 81 15 43867 1208111 77 19 to 136 27 33 to 87
ESR, mm/h
All 11 31826 62828 31 12 to 50 39 24818 46828 22 14 to 29 9 27 to 6.8
No intervention 6 40832 51833 11 5.9 to 29 29 26819 40825 14 6.8 to 21 3 15 to 20
Intervention 5 21812 75817 54 30 to 79 10 21818 63829 42 26 to 58 12 38 to 13

Numbers of patients vary due to missing laboratory samples. * p = 0.05; ** p = 0.02.

1
Samples taken between days 2 and 7, with a minimum interval of 3 days.

Table 3. Crude and multivariable

adjusted OR for poor outcome (Rankin) n Crude OR Adjusted OR1
as predicted by dichotomized initial OR 95% CI OR 95% CI
concentrations (< or median)
Platelets 250 ! 109/l 104 1.3 0.62.8 1.0 0.42.5
Leukocytes 13.6 ! 109/l 105 2.5 1.15.7 2.1 0.85.5
CRP 9.8 mg/l 101 1.5 0.73.3 1.1 0.52.7
ESR 11 mm/h 80 2.3 0.95.9 2.1 0.77.0

Marker available in a total of n patients; numbers are <106 if initial laboratory values within 72 h
after subarachnoid hemorrhage are missing.
1Platelets adjusted for WFNS (World Federation of Neurological Surgeons Scale) grade at admis-

sion, loss of consciousness at the time of aneurysm rupture, signs of acute ischemia at the time of
aneurysm rupture, amount of cisternal blood, and sex.
Leukocytes adjusted for loss of consciousness at the time of aneurysm rupture, signs of acute isch-
emia at the time of aneurysm rupture, age, WFNS grade at admission, and amount of cisternal blood.
CRP adjusted for loss of consciousness at the time of aneurysm rupture, WFNS grade at admis-
sion, amount of ventricular blood, and signs of acute ischemia at the time of aneurysm rupture.
ESR adjusted for amount of ventricular blood, amount of intracisternal blood, age, loss of con-
sciousness at the time of aneurysm rupture, and sex.

another study found a more severe early decrease in plate-
let count in patients with symptomatic vasospasm com-
pared with those without [13]. This inconsistency between
the findings from this study and those in our patients may
be explained by the timing of the measurements which
was not related to the time of onset of the symptoms, or
by mixing patients with and patients without an operation
during the period of sampling in that study.
The mechanism behind the increased platelet count is
unknown. Reactive thrombocytosis occurs after major

Platelets and Leukocytes in DCI after Cerebrovasc Dis 2010;29:576583 581

SAH
surgery, in infections or chronic inflammation, but also
in situations of stress [14]. Our study cannot distinguish
between a rise in platelet count as a cause or as a conse-
quence of DCI. However, in vitro investigations suggest-
ed that platelets and CSF are both required to elicit cere-
bral vasospasm after SAH [15]. The association between
the presence of microemboli on transcranial Doppler and
the occurrence of DCI after aneurysmal SAH may also
serve as an argument for a causal relationship [16].
In accordance with our findings, one retrospective
study found a significant increase in leukocyte count at
the time of clinical manifestation of DCI, but no differ-
ence in leukocyte count at admission between patients
who later developed DCI and those who did not [17]. An-
other retrospective study found that patients who subse-
quently developed symptomatic vasospasm had a higher
peak leukocyte count within the first 5 days of aneurys-
mal SAH [18].
The combination of our findings suggesting a pro-
thrombotic and proinflammatory condition at the time
of the development of DCI, and the results from the lit-
erature suggest that treatment with low-dose aspirin may
not be sufficient for the prevention of DCI and that, rath-
er, anti-inflammatory aspirin dosages or other anti-in-
flammatory drugs (which also cause platelet inhibition)
are necessary for this purpose. In a recent Cochrane re-
view on anti-platelet agents in patients with SAH, a trend
towards a beneficial effect of these agents was found [19].
In the 7 included studies of 5 different agents, daily dos-
ages of acetylsalicylic acid ranged from 100 to 600 mg per
day. The lower dosages cause platelet inhibition, but are
too low to have a relevant anti-inflammatory effect [20].
The other drugs, such as ticlopidine and dipyridamole,
do not have an anti-inflammatory effect.
We found no independent prognostic markers for the
development of DCI or poor outcome. However, in the
univariable analysis, leukocyte counts were associated
with an increased risk of poor outcome. According to the
results of the present study, leukocyte count at admission
was not independently related to DCI or poor outcome
after SAH. We suggest that increased initial concentra-
tions of leukocytes may reflect the worse clinical condi-
tion of patients with an increased risk of a poor outcome.
However, a type 2 error cannot be excluded.
One of the strengths of this study is the choice of sim-
ple, reliable and easily repeatable laboratory assessments,
and the consistent and serial collection of the laboratory
data. For the analyses with respect to DCI, we carefully
selected concentrations before and after development of
DCI and from the same period after onset of aneurysmal
582 Cerebrovasc Dis 2010;29:576583
SAH in the control group. We also made a distinction
between patients who developed DCI related to treatment
and those who did not, to avoid confounding by the influ-
ence of treatment of the aneurysm on the plasma concen-
trations of the investigated markers. Finally, after exclu-
sion of patients who had an infectious disease at the time
of development of DCI, the results remained unchanged.
One of the shortcomings of this study is the small
number of patients, especially in the subgroups with DCI.
However, even despite the small number of patients with-
in each subgroup, we found statistically significant differ-
ences in platelet and leukocyte counts between patients
with DCI unrelated to treatment and patients without
DCI.
After adjustment for sex, clinical condition at onset
and on admission and amount of extravasated blood,
platelet and leukocyte counts were no longer statistically
significant related to poor outcome. We have therefore
not expanded the analyses with other predictors for poor
outcome, such as aneurysm size [21].
Our outcome event of interest was the occurrence of
DCI, not of radiological vasospasm. As vasospasm does
not always lead to cerebral ischemia, DCI is a more rele-
vant outcome measure than vasospasm from a clinical
point of view. Therefore, and also because of the limited
predictive value of transcranial Doppler and the potential
hazards of angiography, we did not routinely carry out
these investigations [1].
In conclusion, we found robust evidence of an associa-
tion between a disproportionate increase in counts of
both platelets and leukocytes and the development of
spontaneous DCI after aneurysmal SAH. We suggest that
a combined prothrombotic and proinflammatory state is
involved in the development of this serious complication
of aneurysmal SAH, and that future trials with agents
with both an anti-platelet and an anti-inflammatory ac-
tion, such as NSAIDs or high-dose aspirin, should be
considered.
References 1 Rabinstein AA, Friedman JA, Weigand SD,
McClelland RL, Fulgham JR, Manno EM,
Atkinson JL, Wijdicks EF: Predictors of ce-
rebral infarction in aneurysmal subarach-
noid hemorrhage. Stroke 2004; 35: 1862
1866.
2 Dumont AS, Dumont RJ, Chow MM, Lin
CL, Calisaneller T, Ley KF, Kassell NF, Lee
KS: Cerebral vasospasm after subarachnoid
hemorrhage: putative role of inflammation.
Neurosurgery 2003;53:123135.
Kasius /Frijns /Algra /Rinkel
 3 Frijns CJM, Fijnheer R, Algra A, van Mourik
JA, van Gijn J, Rinkel GJE: Early circulating
levels of endothelial cell activation markers
in aneurysmal subarachnoid haemorrhage:
associations with cerebral ischaemic events
and outcome. J Neurol Neurosurg Psychiatry
2006;77:7783.
4 Teasdale G, Jennett B: Assessment of coma
and impaired consciousness: a practical
scale. Lancet 1974;2:8184.
5 Drake CG: Report of World Federation of
Neurological Surgeons Committee on a Uni-
versal Subarachnoid Hemorrhage Grading
Scale. J Neurosurg 1988;68:985986.
6 Hijdra A, Brouwers PJAM, Vermeulen M,
van Gijn J: Grading the amount of blood on
computed tomograms after subarachnoid
hemorrhage. Stroke 1990; 21:11561161.
7 Bamford JM, Sandercock PA, Warlow CP,
Slattery J: Interobserver agreement for the
assessment of handicap in stroke patients.
Stroke 1989; 20:828.
8 Frijns CJM, Kasius KM, Algra A, Fijnheer R,
Rinkel GJE: Endothelial cell activation
markers and delayed cerebral ischaemia in
patients with subarachnoid haemorrhage. J
Neurol Neurosurg Psychiatry 2006; 77: 863
867.
Platelets and Leukocytes in DCI after
SAH
9 Ohkuma H, Suzuki S, Kimura M, Sobata E:
Role of platelet function in symptomatic ce-
rebral vasospasm following subarachnoid
hemorrhage. Stroke 1991; 22:854859.
10 Juvela S: Cerebral infarction and release of
platelet thromboxane after subarachnoid
hemorrhage. Neurosurgery 1990; 27: 929
935.
11 Juvela S, Hillbom M, Kaste M: Platelet
thromboxane release and delayed cerebral
ischemia in patients with subarachnoid
hemorrhage. J Neurosurg 1991;74:386392.
12 Juvela S, Kaste M, Hillbom M: Effect of ni-
modipine on platelet function in patients
with subarachnoid hemorrhage. Stroke
1990;21:12831288.
13 Hirashima Y, Hamada H, Kurimoto M, Ori-
gasa H, Endo S: Decrease in platelet count as
an independent risk factor for symptomatic
vasospasm following subarachnoid hemor-
rhage. J Neurosurg 2005;102:882887.
14 Schafer AI: Thrombocytosis. N Engl J Med
2004;350:12111219.
15 Pyne GJ, Cadoux-Hudson TAD, Clark JF:
Platelets play an essential role in the aetiol-
ogy of cerebral vasospasm after subarach-
noid haemorrhage. Med Hypotheses 2003;
60:525530.
Cerebrovasc Dis 2010;29:576583
16 Romano JG, Rabinstein AA, Arheart KL,
Nathan S, Campo-Bustillo J, Koch S, Korteza
AM: Microemboli in aneurysmal subarach-
noid hemorrhage. J Neuroimaging 2008; 18:
396401.
17 Spallone A, Acqui M, Pastore FS, Guidetti B:
Relationship between leukocytosis and isch-
emic complications following aneurysmal
subarachnoid hemorrhage. Surg Neurol
1987;27:253258.
18 McGirt MJ, Mavropoulos JC, McGirt LY, Al-
exander MJ, Friedman AH, Laskowitz DT,
Lynch JR: Leukocytosis as an independent
risk factor for cerebral vasospasm following
aneurysmal subarachnoid hemorrhage. J
Neurosurg 2003;98:12221226.
19 Dorhout Mees SM, van den Bergh W, Algra
A, Rinkel GJE: Antiplatelet therapy for aneu-
rysmal subarachnoid haemorrhage. Co-
chrane Database Syst Rev 2007:006184.
20 Patrono C, Coller B, Dalen JE, FitzGerald
GA, Fuster V, Gent M, Hirsch J, Roth G:
Platelet-active drugs: the relationship among
dose, effectiveness, and side effects. Chest
2001;119:39S63S.
21 Roos EJ, Rinkel GJE, Velthuis BK, Algra A:
The relation between aneurysm size and out-
come in patients with subarachnoid hemor-
rhage. Neurology 2000;54:23342336.
583
Copyright: S. Karger AG, Basel 2010. Reproduced with the permission of S. Karger AG, Basel. Further
reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright
holder.