31 / 74
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 25, No 5, September/October 2014
AFRICA
225
of thrombolytic therapy is significant, the inclusion of a risk-
prediction score to the protocol used for treatment may improve
the cost–benefit ratio, and optimise the allocation of scarce
resources.
The aim of this pilot study was to evaluate the performance
of both the SITS-SICH and SEDAN scores in predicting the risk
of SICH in a Groote Schuur thrombolysis stroke cohort.
Methods
Data were extracted from a prospective cohort comprising all
patients presenting to Groote Schuur Hospital (GSH) between
2000 and May 2012 with acute ischaemic stroke, who received
thrombolytic therapy with IV tPA according to the GSH Stroke
Unit protocol. Patients who received mechanical thrombolysis or
intra-arterial tPA were excluded.
Age, gender, past medical history and prior medication were
recorded. Admission data of vital signs, serum glucose levels,
stroke severity according to theNIHSS, disability according to the
modified Rankin score, and details of tPA administration (time
to onset and dose) were also recorded. For this study, all available
CT brain scans performed on admission (pre-thrombolysis) were
re-evaluated by a radiologist in training for signs of early infarct
or the dense middle cerebral artery (MCA) sign. The SITS-SICH
and SEDAN scores (see below) were calculated from these data
and patients were risk-stratified accordingly.
Both the SITS-SICH and SEDAN scoring systems (see
Tables 1, 2) were developed using multiple regression analyses,
and identified elevated serum glucose levels and high NIHSS
scores on admission as poor prognostic indicators. The risk
of SICH varied according to the SICH definition used: by the
SITS-MOST definition, the risk ranged from 0.2% (score of
0) to 9.2% (score
≥
9) while the risk of SICH by the ECASS
II definition ranged from 1.4% (score 0) to 23.2% (score
≥
9).
The SEDAN score revealed an increasing risk of SICH in the
external validation cohort, ranging from 0.01% (score 0) to
27.8% (highest score 6). The single largest risk factor identified
for the development of SICH in the SITS-MOST study was dual
antiplatelet therapy with both aspirin and clopidogrel.
12,13
The primary outcome was symptomatic intracranial
haemorrhage according to either the SITS-MOST and/or
ECASS II definitions. The SITS-MOST definition of SICH is
a local or remote type II parenchymal haemorrhage within 22
to 36 hours after treatment (or sooner) associated with a
≥
four-
point deterioration on the NIHSS score from baseline or from
the lowest score from baseline to 24 hours, or leading to death.
12
The ECASS II definition of SICH was any intracranial
haemorrhage on any post-treatment image, within seven days of
initiating treatment associated with a
≥
four-point deterioration
on the NIHSS score from baseline or from the lowest score in
seven days, or leading to death.
14
Other outcomes reported were
death, asymptomatic intracranial haemorrhage (AIH), and
extracranial haemorrhage (EH).
During most of the cohort period, the GSH Stroke Unit
protocol required post-thrombolysis CT brain scans to be
performed routinely within 48 hours of thrombolysis, or urgently
with any suspicion of an intracerebral haemorrhage. Post-
thrombolysis CT scans were reviewed for this study by a
radiologist trainee blinded to clinical outcomes for evidence of
intracranial haemorrhage.
Ethical approval was obtained from the UCT Groote Schuur
Hospital human research ethics committee (Ref: 499/2013).
Results
In total, 45 patients underwent thrombolysis for acute ischaemic
stroke at the GSH Stroke Unit from January 2000 to May 2012.
Four patients underwent mechanical thrombolysis with intra-
arterial tPA, and were excluded. The remaining 41 patients who
received IV tPA for acute ischaemic stroke were included (see
Table 3).
Five patients were older than 72 years, and two patients
were older than 75 years at stroke onset. Admission systolic
blood pressures (SBP) were above 180 mmHg in five patients
and greater than 220 mmHg in one patient. Of those taking
antiplatelet therapy at the time of admission, all were on aspirin
monotherapy. Time to onset of treatment with tPA was greater
than 180 min in 13 patients and none of these were treated
more than 4.5 hours after onset of symptoms. Pre- and post-
thrombolysis CT scans were available for review in 23 patients.
Two patients suffered SICH (ECASS II definition) post-
thrombolysis, comprising 4.9% of the cohort. CT scans were
available for review in one patient only and confirmed that the
haemorrhage fulfilled the SITS-MOST criteria (2.4%). Of the
four patients who died during their admission for stroke, one
patient suffered a fatal SICH, while three other patients died
Table 1. Components of SITS score and overall risk level
12
Category
Points (15)
Aspirin + clopidogrel therapy
2
Aspirin monotherapy
1
NIHSS
>
13
2
NIHSS 7–12
1
Blood glucose
≥
180 mg/dl*
2
Age
≥
72 years
1
Systolic BP
≥
146 mmHg
1
Weight
≥
95 kg
Onset-to-treatment time
≥
180 min
1
History of hypertension
1
*180 mg/dl
≈
10 mmol/l
Table 2. SEDAN score
13
Category
Total
6
Blood sugar (glucose) on admission
≤ 8 mmol/l
0
8.1–12 mmol/l
1
>
12 mmol/l
2
Signs of early infarction on admission CT*
No
0
Yes
1
Dense middle cerebral artery sign on
admission CT
No
0
Yes
1
Age (years)
≤ 75
0
>
75
1
NHSS score on admission
0–9 points
0
≥
10 points
1
*Signs of early infarction: hypo-attenuation of the middle cerebral
artery territory (
<
1/3), obscuration of the lentiform nucleus, cortical
sulcal effacement, focal hypo-attenuation, loss of the insular ribbon/
obscuration of the Sylvian fissure, loss of grey–white differentiation
in the basal ganglia, hypo-attenuation of the basal ganglia.