Professional Documents
Culture Documents
Departments of a Radiology and b Neurology, Copenhagen University Hospital Bispebjerg and Copenhagen Stroke
Research Centre, Copenhagen, Denmark
Key Words
Spot sign Intracerebral haemorrhage Outcome
Computed tomography angiography
Abstract
Background: A spot sign on computed tomography angiography (CTA) is a potentially strong predictor of poor outcome
on ultra-early radiological imaging. The aim of this study was
to assess the spot sign as a predictor of functional outcome
at 3 months as well as long-term mortality, with a focus on
the ability to identify patients with a spontaneous, acceptable outcome. Methods: In a prospective, consecutive single-centre registry of acute stroke patients, we investigated
patients with spontaneous intracerebral haemorrhage (ICH)
admitted within 4.5 h after symptom onset from April 2009
to January 2013. The standard work-up in our centre included CTA for spot sign status, unless a contraindication was
present. Modified Rankin Scale (mRS) scores were assessed
at 3 months in the outpatient clinic or by telephone interviews. Long-term mortality was assessed by electronic chart
follow-up for up to 1,500 days. Results: Of the 128 patients,
37 (28.9%) had a spot sign on admission CTA. The presence
Inger Havsteen
Department of Radiology, Copenhagen University Hospital Bispebjerg and
Copenhagen Stroke Research Centre, Bispebjerg Bakke 23
DK2400 Copenhagen NV (Denmark)
E-Mail seestein@gmail.com
Downloaded by:
138.5.159.110 - 3/15/2015 6:16:14 AM
Spontaneous intracerebral haemorrhage (ICH) is regarded as the worst presentation of stroke, as 40% of patients die within the first month [1, 2], and only 20% of
patients achieve independent living at 6 months [3]. For
a long period of time, no new treatment options proven
effective in reducing mortality and morbidity in randomized clinical trials have emerged for ICH. Consequently,
the standard of care remains stroke unit rehabilitation
and prevention of medical complications. However, recently, blood pressure reduction has shown a small but
clinically significant benefit (INTERACT 2) [4], while
conclusive positive evidence is still lacking in surgical
treatment (STICH II) [5].
The spot sign is defined as contrast-enhancing foci within the haematoma observed on computed tomography angiography (CTA) source images [6]. This promising biomarker is an independent predictor of early haematoma
expansion and neurological deterioration [79] within the
first 6 h after symptom onset [1012]. Recent single-centre
case series [1315] and a multicentre study [16] have confirmed the spot sign to be an independent predictor of early haematoma expansion. However, the ability of spot sign
presence or absence to predict 3-month beneficial functional outcome and long-term mortality remains less established, and disagreement between studies exists [14, 16].
Robust and broadly validated biomarkers predicting outcome are essential in clinical decision-making as well as to
provide reliable information for patients and relatives.
Aim
The aim of this study was to assess the ability of the
spot sign to serve as a biomarker to predict 3-month functional outcome and long-term mortality in a consecutive,
prospective single-centre cohort of patients with ICH.
Patients
Copenhagen University Hospital Bispebjerg receives all patients from a well-defined catchment area consisting of the entire Capital Region of Denmark (approx. 1.7 million inhabit-
165
Downloaded by:
138.5.159.110 - 3/15/2015 6:16:14 AM
Introduction
Follow-Up
A follow-up interview was conducted 3 months after stroke and
included modified Rankin Scale (mRS) scores evaluated via a faceto-face interview or a telephone interview by a stroke neurologist
or a trained nurse [22]. For deceased patients, information was
retrieved from the national electronic patient chart system. Patients were prospectively followed up regarding mortality via the
national electronic patient chart system for up to 1,500 days after
admission.
Ethics Statement
This observational study of the long-term effects of best clinical
care and with no interventions was approved by the Danish Data
Protection Agency (file No. 2010-41-5205) in accordance with
Danish law.
Statistics
Endpoints were defined as (1) functional outcome measured
using the mRS 3 months after stroke onset, and (2) death from any
cause between the date of inclusion and July 1, 2013. Good functional outcome was defined as 3-month mRS scores of 02. Poor
outcome was defined as 3-month mRS scores of 56. Group differences in baseline characteristics were compared using Students
t test for continuous variables after testing for normality using the
Kolmogorov-Smirnov test. Non-normally distributed data were
analysed using the Mann-Whitney U test. Ordinal scale data were
analysed using the Mann-Whitney U test, and categorical variables
using the 2 test. Cohens was used to grade interrater agreement.
Predictors of good and poor outcome 3 months after stroke
were analysed using multivariable stepwise logistical regression.
Besides the spot sign, we entered consistent predictors of functional outcome and mortality from the literature into the models:
age [23], prestroke mRS score, admission ICH volume [1], intraventricular extension [23], subarachnoid extension [24], admission plasma glucose level [25], admission National Institutes of
Health Stroke Scale (NIHSS) score [12] and warfarin treatment
[26]. Areas under the receiver operating characteristic curve (c statistics) were assessed to determine the calibration and discriminative ability of the models.
Non-adjusted survival (death) analysis of patients with and
without spot sign was undertaken using Kaplan-Meier curves
stratified for the presence of a spot sign. Difference between survival curves was tested for using the log rank test. Factors related
to survival during the entire follow-up time were assessed using the
Cox proportional hazards model. The proportional hazards assumptions were evaluated using Schoenfeld residuals. A general
significance level of p < 0.05 was employed. Statistical analyses
were performed by SPSS statistical software version 20 (IBM Corp.,
Armonk, N.Y., USA).
166
Results
Downloaded by:
138.5.159.110 - 3/15/2015 6:16:14 AM
A consultant neuroradiologist (A.F.C.) and a senior neuroradiological fellow (I.H.) blinded to all clinical data systematically
reviewed all images. Disagreement was resolved by consensus.
Agreement between the observers was substantial ( = 0.72). A
spot sign was defined as 1 or more 1- to 2-mm foci of enhancement
within the haematoma on CTA source images [6]. The CTA images were analysed in bone window (width 2,0002,500, level 400
500) and spot window settings (width 200, level 110) [20] to differentiate between spot sign and calcification. Haematoma volumes were calculated using the ABC/2 method [21].
Spot sign
Number of patients
Gender male, n
Mean age SD, years
Medical history
Median prestroke mRS score
Prior stroke, n
Atrial fibrillation, n
Hypertension, n
Coronary heart disease, n
Diabetes, n
Hyperlipidaemia, n
History of smoking, n
Excessive use of alcohola, n
Antiplatelet treatment, n
Warfarin treatment, n
Admission values
Mean time from onset to CT scanning SD, min
Median admission NIHSS score
Mean BP SD, mm Hg
Systolic
Diastolic
Mean blood glucose SD, mmol/l
Mean platelet count SD, 109/l
Mean INR SD
Mean APTT SD, s
Haematoma characteristics
Median initial volume, ml
IVH, n
SAH, n
Lobar, n
Basal ganglia, n
Posterior fossa, n
No spot sign
37
22 (60%)
7212
91
52 (58%)
6912
0.71
0.09
0 (01)
9 (25%)
8 (22%)
28 (78%)
1 (2.8%)
3 (8.3%)
17 (74%)
12 (41%)
7 (23%)
11 (31%)
8 (22%)
0 (00)
11 (12%)
10 (11%)
68 (76%)
4 (4.4%)
7 (7.8%)
49 (70%)
34 (45%)
7 (8.9%)
15 (17%)
9 (10%)
0.01
0.11
0.28
0.81
0.66
1.00
1.00
0.83
0.06
0.09
0.08
11749
19 (1223)
11248
10 (616)
0.70
<0.0001
192.029.6
99.920.0
7.43.0
225.279.5
1.40.8
28.15.5
38.0 (18.078.0)
16 (43%)
14 (38%)
16 (43%)
19 (51%)
2 (5.4%)
185.729.6
103.023.8
7.12.0
225.171.1
1.20.7
28.06.4
12.0 (5.024.0)
24 (27%)
19 (21%)
28 (31%)
58 (64%)
4 (4.4%)
0.25
0.53
0.58
0.91
0.10
0.70
<0.0001
0.10
0.08
0.23
0.17
0.81
come and a negative predictive value (NPV) of 95% (positive predictive value, PPV, 40%). Spot sign presence
showed 86% specificity (sensitivity: 60%) and an NPV of
81% (PPV 68%) for poor outcome.
The median follow-up time was 594 days (range:
01,500 days). During the entire follow-up period, 22
spot sign-positive patients (59.5%) and 15 spot sign-negative patients (16.5%) died. The Kaplan-Meier curves
show significantly different survival patterns between
groups (p< 0.0001; fig.1). The presence of a spot sign
yielded an NPV of 88% and a PPV of 57% for 3-month
mortality (sensitivity and specificity: 66 and 83%, respectively).
In a stepwise regression model, a spot sign independently predicted poor 3-month functional outcome (OR
3.40; 95% CI: 1.1010.5; table2). Finally, a spot sign was
inversely associated with good 3-month functional outcome (OR 0.17; 95% CI: 0.030.88). Both final models
derived from the logistic regression analysis provided excellent discriminative ability towards good and poor outcome (c statistics; table2). Significant predictors of death
during the follow-up period were spot sign (HR 3.04; 95%
CI: 1.456.34), admission NIHSS score (HR 1.11; 95% CI:
1.061.17 per unit) and increasing age (HR 1.30; 95% CI:
1.131.51 per 5 years). The same parameters as in table2
were entered in the Cox regression analysis.
167
Downloaded by:
138.5.159.110 - 3/15/2015 6:16:14 AM
Values in parentheses denote IQR unless specified otherwise. BP= Blood pressure; INR= international normalized ratio; APTT= activated partial thromboplastin time; IVH= intraventricular haemorrhage; SAH= subarachnoid haemorrhage.
aWeekly alcohol consumption of more than 168 g for females and 252 g for males.
100
p < 0.0001
80
60
40
20
0
0
300
600
OR
95% CI
0.82
0.68
0.21
0.17
0.740.90
0.530.87
0.050.83
0.030.88
1.21
1.70
1.02
3.40
1.101.32
1.262.31
1.001.04
1.1010.5
Gender, age, haematoma location, admission NIHSS score, admission ICH volume, prestroke mRS score, subarachnoid haemorrhage, intraventricular haemorrhage, warfarin treatment, plasma
glucose level and spot sign were entered in both models. First model (mRS scores 02): c statistics= 0.88 (95% CI: 0.820.94); second
model (mRS scores 56): c statistics= 0.92 (95% CI: 0.870.97); c
statistics indicate a good discriminative power for both models.
Discussion
1,200
1,500
14
75
9
57
7
35
4
16
1
2
Downloaded by:
138.5.159.110 - 3/15/2015 6:16:14 AM
900
Time (days)
Conclusion
Disclosure Statement
No conflicts of interest declared.
References
169
Downloaded by:
138.5.159.110 - 3/15/2015 6:16:14 AM
170
24 Maas MB, Nemeth AJ, Rosenberg NF, Kosteva AR, Guth JC, Liotta EM, Prabhakaran S,
Naidech AM: Subarachnoid extension of primary intracerebral hemorrhage is associated
with poor outcomes. Stroke 2013; 44: 653
657.
25 Lee SH, Kim BJ, Bae HJ, Lee JS, Lee J, Park BJ,
Yoon BW: Effects of glucose level on early and
long-term mortality after intracerebral haemorrhage: the Acute Brain Bleeding Analysis
Study. Diabetologia 2010;53:429434.
26 Dowlatshahi D, Butcher KS, Asdaghi N, Nahirniak S, Bernbaum ML, Giulivi A, Wasserman JK, Poon MC, Coutts SB: Poor prognosis
in warfarin-associated intracranial hemorrhage despite anticoagulation reversal. Stroke
2012;43:18121817.
27 Brouwers HB, Falcone GJ, McNamara KA,
Ayres AM, Oleinik A, Schwab K, Romero JM,
Viswanathan A, Greenberg SM, Rosand J,
Goldstein JN: CTA spot sign predicts hematoma expansion in patients with delayed presentation after intracerebral hemorrhage.
Neurocrit Care 2012;17:421428.
28 Brouwers HB, Greenberg SM: Hematoma expansion following acute intracerebral hemorrhage. Cerebrovasc Dis 2013;35:195201.
29 Brouwers HB, Goldstein JN, Romero JM,
Rosand J: Clinical applications of the computed tomography angiography spot sign in
acute intracerebral hemorrhage: a review.
Stroke 2012;43:34273432.
30 Mayer SA, Brun NC, Begtrup K, Broderick J,
Davis S, Diringer MN, Skolnick BE, Steiner T:
Recombinant activated factor VII for acute
intracerebral hemorrhage. N Engl J Med
2005;352:777785.
31 Mayer SA, Brun NC, Begtrup K, Broderick J,
Davis S, Diringer MN, Skolnick BE, Steiner T:
Efficacy and safety of recombinant activated
factor VII for acute intracerebral hemorrhage.
N Engl J Med 2008;358:21272137.
Havsteen/Ovesen/Christensen/Hansen/
Nielsen/Christensen
Downloaded by:
138.5.159.110 - 3/15/2015 6:16:14 AM