Một số yếu tố dự báo sớm KMT lan rộng và giá trị của thang điểm “spot sign” trong tiên lượng chảy máu não tự phát giai đoạn cấp tt tiếng anh

MINISTRY OF EDUCATION AND TRAINING

MINISTRY OF HEALTH

HANOI MEDICAL UNIVERSITY

NGUYEN SONG HAO

SOME EARLY PREDICTIONS OF HEMATOMA
EXPANSION AND THE VALUE OF THE SPOT SIGN
SCORE IN THE PROGNOSIS OF ACUTE SPONTANEOUS
INTRACEREBRAL HEMORRHAGE

Field of study

: Intensive Care Medicine
Emergency and Clinical Toxicology

Code

: 62720122

SUMMARY OF MEDICAL DOCTORAL THESIS

HANOI – 2019


THE THESIS WAS COMPLETED AT:
HANOI MEDICAL UNIVERSITY

Scientific advisors:

1. Assoc.Prof. Dr. Nguyen Dat Anh
2. Assoc. Prof. Dr. Vu Dang Luu

Reviewer 1: Assoc.Prof. Dr. Mai Xuan Hien

Reviewer 2: Assoc.Prof. Dr. Nguyen Van Chi

Reviewer 3: Assoc.Prof. Dr. Nguyen Van Lieu

The thesis defense shall be held by the university-level Thesis
Assessment Board at Hanoi Medical University.
Time:

pm Date:

, 2019

The thesis can be found at:
- Library of Hanoi Medical University
- National Library


LIST OF OF THE AUTHOR’S SCIENTIFIC ARTICLES
RELATED TO THE THESIS

1.

Nguyen Song Hao, Nguyen Đat Anh, Vu Đang Luu (2019). A
reserarch on some predicting factors of hematoma expansion after
primary intracerebral hemorrhage. Vietnam medical journal, Vol

481 (1), 75-80.

2.

Nguyen Song Hao, Nguyen Đat Anh, Vu Đang Luu (2019).
Predictive value of CTA spot sign and the role of spot sign score
on hematoma expansion and clinical outcome in intracranial
hemorrhage patients. Vietnam medical journal, Vol 477 (1), 117122.


1
INTRODUCTION
1. The urgency of the study
Intracerebral hemorrhage (ICH) is a common neuropathy, with a
high mortality rate, severe sequelae, a major burden for families and
society. Around the world, about 2 million people get ICH every year.
Although there are many modern facilities applied in diagnosis, treatment
and resuscitation of stroke patients, the mortality rate in 30 days is still
high, up to 30-50%. About half of all deaths occur in the acute phase,
especially in the first 48 hours.
Hematoma expansion is a common and serious complication after
ICH. Although hematoma expansion is one of the major pathogenesis
mechanisms in phases of ICH, it is also a serious complication after the
acute phase. The mechanism of hematoma expansion during the acute
phase has not been clearly explained. The spread of hematoma is an
independent prognostic factor predicting the risk of death and adverse
outcomes in patients with ICH.
Many early predictions of hematoma expansion have been identified
as: initial hematoma volume, irregular hematoma shape, consciousness
disorder upon admission, time from onset to admission is short, using

anticoagulants, heterogeneous hematoma density. Some recent studies
show that a new point of early predictions of hematoma expansion is the
release of a contrast agent “extravasation contrast” or “spot sign” on
computed tomography angiography cerebrovascular (CTA) in acute ICH.
The “spot sign” is the extravasation of the contrast agent located in the
margin or center of the hematoma of the brain parenchyma, which can be
seen with the naked eye on CTA films. Delgado Almandoz (2010) have
stated that “spot sign” and spot sign score values are independent
prognostic factors, predicting the risk of death in inpatient treatment and
poor outcomes at patients survived.
In clinical practice, the scale helps clinicians evaluate patients
quickly and effectively to promptly take the best measures to manage
patients, making an important contribution to improving professional
quality, increases healing ability.
In the world there have been many researches on this issue. In
Vietnam, there are also some studies on some prognostic factors in ICH
by hypertension, but no research has fully evaluated the impact of risk
factors predicting hematoma expansion as well as the value of the spot


2
sign score to predict the severity, risk of death and invalid in patients
with ICH. Detecting risk factors that predict hematoma expansion, and
obtaining a truly valuable prognostic scale is an urgent requirement for
clinicians, especially in emergency stroke management.
From that fact, we carried out the research topic “Some early
predictions of hematoma expansion and the value of the spot sign score
in the prognosis of acute spontaneous intracerebral hemorrhage”, aiming
at two goals:
1. Describe some of the early predictions of hematoma expansion

in patients with acute spontaneous intracerebral hemorrhage.
2. Determine the value of the spot sign scorein the prognosis of
acute spontaneous intracerebral hemorrhage patients.
2. The layout of the thesis:
The thesis content consists of 137 pages with 40 tables (30 tables in
the results section), 7 charts (5 of the results section) with the layout:
Introduction (2 pages); Literature review (39 pages); Subjects and
research methods (18 pages); Research results (30 pages); Discussion (45
pages); Conclusion (2 pages); Recommendation (1 page); References:
125 documents (Vietnamese and English)
3. New contributions of the thesis
This is the first study in the country to fully address the predictive
factors of hematoma expansion in Vietnamese patients. Research results
do not overlap with other studies in the country and abroad.
The study identified the spot sign score value in prognosis of death,
disability outcomes in inpatient treatment, and after 3 months of
treatment in patients with ICH.
This research is a new contribution to clinical practice, helping
clinicians, especially doctors in emergency departments, but not
neurologists, to have a basis for assessing the prognosis. Since then, there are
early management strategies and are suitable for patients with acute ICH.
CHAPTER 1: LITERATURE REVIEW
1.1. OVERVIEW ABOUT HEMATOMA EXPANSION
1.1.1. Terminology and definitions
The term of hematoma expansion after spontaneous cerebral
bleeding has never been used uniformly in scientific literature. Therefore,
many authors have defined hematoma expansion as all forms of spatial


3

expansion of initial bleeding including; Increased volume of hematoma
in the brain parenchyma, flowing into the ventricle or into the
subarachnoid space adjacent to the original bleeding source and does not
include formation of cerebral edema around the hematoma.
The hematoma is determined to spread early when it occurs within
the first 24 hours after the onset of spontaneous cerebal bleeding.
1.1.2. Threshold determines hematoma expansion
The threshold for hematoma expansion has not been consistent
across studies. According to Steiner T (2010), the cut-off threshold is
used to determine a significant spread of> 33% or> 12.5 ml. According
to Wada R (2007), use the definition of> 30% or> 6ml and is supported
by previous studies of traumatic ICH showing the need for surgical
intervention when the hematoma increases by 5ml. According to Kazui
(1996), hematoma was determined to spread by an increase of 12.5 ml or
1.4 times, which is the optimal cut-off threshold for assessing the spread
of the hematoma with the naked eye.
1.1.3. Distinguish primary and secondary intracerebral hemorrhage
Non-traumatic cerebral bleeding is divided into 2 main subgroups of
primary ICH and secondary ICH:
Primary ICH accounts for about 78-88% of cases, resulting from
rupture of small blood vessels whose background is not clear due to
increased BP or powdered cerebrovascular disease. The main risk factors
for primary ICH are increased BP, powdered cerebrovascular disease,
low blood cholesterol, alcohol, and tobacco.
Secondary ICH occurs in lesser amounts, accounting for about 20%
of vascular abnormalities (cerebral vein malformations, aneurysms,
cerebrovascular cavernome), newborn brain tumors, coagulopathy,
trauma, drug abuse.
1.2. SOME PREDICTABLE FACTORS OF HEMATOMA
EXPANSION

1.2.1. Short time from onset to hospitalization
A retrospective study of Fujii Y (1994) or Kazui (1996)
demonstrated that patients who were hospitalized early, the time from
onset of symptoms to short-first CT scans had a higher risk of hematoma
on CT scan next time.
According to Fujii Y (1998), it has been shown that the short time
from onset of stroke to CT scan is the most powerful prognostic factor
for the spread of hematoma. As previous studies have reported, the


4
prevalence of hematoma expansion decreases as the time from onset to
hospitalization increases and spread of hematoma rarely occurs if the
time from onset to when hospitalized> 6 hours.
1.2.2. A disorder of consciousness when hospitalized
Fujii Y (1998) argues that the presence of consciousness disorder is
an independent predictable hematoma expansion, meaning that patients
with consciousness disorders are likely to spread hematoma after
hospitalize. No previous studies have found an association between
hematoma expansion and the degree of consciousness disorder. Although
it is unclear about the high prevalence of hematoma for inpatient patients
with conscious disorders. The disorder of consciousness is representative
of a number of factors including hematoma size.
1.2.3. Shape and density of hematoma
Bleeding that originates from a single point tends to appear enlarged
lesions with uniform edges, developing from the center of the hematoma,
the density of blood is more uniform. When a bleeding originates from
many points, there is almost an uneven edge lesions, and is developed
from the junction of hematoma with brain organization. The
inhomogeneous density on cranial CT scans may reflect ongoing

bleeding, more variable bleeding times and more points. Heterogeneous
blood flow is generated from many blood vessels, with low density spots,
dilute blood just flowing beside high density blood clots.
Barras C.D (2009), conducting research on the concept of abnormal
hematoma shape and density on cerebral CT scans may predict the risk of
hematoma expansion. This concept comes from the view that uneven and
heterogeneous hematoma may be the result of multiple bleeding points
leading to an increased risk of hematoma expansion later.
1.2.4. Low fibrinogen concentration
By Fujii Y (1998), multivariate analysis of hematoma expansion
prediction factors, has published five independent prognostic factors
related to hematoma expansion, in which low fibrinogen concentration is
an independent prognostic factor predicting hematoma expansion.
Therefore, decreased fibrinogen levels may be associated with the
decline of both endogenous and exogenous coagulation mechanisms.
Therefore, low fibrinogen levels are considered as a risk factor as well as
a prediction of hematoma expansion.


5
1.2.3. “spot sign” and prediction of hematoma expansion
The “spot sign” was first described in 1999 by Becker K.J. The
technique is carried out right after ICH diagnosis, detecting the escape of
contrast into the hematoma. The “spot sign” is correctly determined
when there is an internal density of contrast dye in the brain parenchyma
without an external blood vessel connection on vascular CT scans,
described such as drops, zigzag points, or many points. The “spot sign”
has a maximum density that is usually twice the density of the
hematoma, in a wide range of 100-200 Hounsfield units and size>
1.5mm. Although basic histopathology of the “spot sign” is unknown.

The researchers suggested that the “spot sign” was formed from
pathological changes related to extravasation from primary or secondary
vascular injury.
Definition of “spot sign”: There are many definitions in clinical practice
about “spot signs”. However, according to Delgado Almandoz J.E (2010), the
“spot sign” is determined on CTA with the following four criteria:
- There is more than one point, the contrast stain is located in the
brain parenchyma.
- Density ≥ 120 HU.
- Intermittent from normal blood vessels or damaged blood vessels
adjacent to hematoma.
- Any size and morphology.
Distinguish “spot sign” and CTA techniques
The term “spot sign” changes in many studies and is often confused
between a “spot sign” or “Contrast Extravasation”. In order to elucidate
the difference between the escape signal of the contrast agent into the
hematoma on CTA scan and the CT scan after contrast injection through
the studies, the term “spot sign” is now reserved for drainage of vascular
contrast agent imaging into the hematoma on CTA capture, while the
terminology of “Contrast Extravasation” is described as the presence of
contrast agent on CT scans after injection.
The “spot sign” on CTA is divided into two groups: the “spot sign”
in the early phase or the arterial phase (fisrt- pass CTA) is done within 30
seconds after contrast injection. The “spot sign” in the late phase or
intravenous phase (second-pass CTA) is performed within 40 seconds - 2
minutes after contrast injection.
Distinguish images similar to “spot signs”: vascular and nonvascular diseases such as arteriovenous malformations (AVM), dural


6

arteriovenous fistulas, ruptured aneurysms, giant thrombotic aneurysms,
calcification in tumors, moyamoya disease is very Important because
each pathology has different management strategies. Other closely signs
are calcification of the mesangial plexus and can be distinguished by CT
scans without contrast.
1.2.4. Some other early predictable of hematoma expansion
Initial volume of hematoma: Dowlatshahi D (2011) retrospective
study on 496 patients; hematomas with volume less than 10ml are less
likely to cause hematomas to spread and have better outcomes, whereas
with hematoma > 30ml, it almost causes hematoma spread and ends bad.
Broderick J.P (1993) report that hematoma volume is best predicted for
mortality in the first 30 days for all bleeding sites. The initial hematoma
and Glasgow volume had a strong and easy-to-use effect intended to
report mortality and disability in ICH patients.
Hypertension problem: Hypertension is a common problem in the
acute phase of ICH, accounting for over 70% of ICH patients. Increased
BP may occur even in the absence of a previous history of hypertension
and is an independent prognostic factor with poor outcomes. Increasing
BP is a risk factor for hematoma expansion and increased mortality in
ICH patients. Prolonged systolic hypertension is associated with
increased cerebral edema around hematomas. Treatment to reduce BP
may reduce the risk of hematoma expansion. However, the evidence that
concluded this issue is not satisfactory.
In addition, the problem of coagulation and the use of
anticoagulants, alcohol abuse, internal factors were also mentioned by
some authors regarding the hematoma expansion after ICH.
1.3. SPOT SIGN SCORE AND SOME PROGNOSTIC SCALES
1.3.1. Rationale for forming spot sign score
Delgado Almandoz (2009), research and development of the spot
sign scoresystem, to find the most valuable characteristics to predict the

spread of hematoma, as well as the prognosis of mortality and bad
outcomes in ICH patients. The structure of the spot sign scoreconsists of
3 components:
"Spot" number: 1 -2 spot: 1 point; ≥ 3 spots: 2 points
Maximum horizontal diameter: 1 - 4mm: 0 point; ≥ 5mm: 1 point
The largest density: 120 -179 HU: 0 point; ≥ 180 HU: 1 point.


7
Table 1.1. Calculation of the Spot Sign Score
Spot Sign Characteristic
points
1 -2
1
Number of spot signs
≥3
2
Maximum horizontal
1 – 4mm
0
diameter
≥ 5 mm
1
120 -179 HU
0
The largest density
≥ 180 HU
1
Total points
0–4

1.3.2. Prognostic value of the spot sign score
Predictive hematoma expansion value of the spot sign score
increases with the point of the scale, when the value of the scale is 0, the
risk of hematoma expansion is 2% but the scale value is 4, the risk of
hematoma expansion is 100%.
Similarly, the spot sign score, which is an independent predictable
risk of death in hospitals and has poor outcomes among survival patients
when follow-up at the time 3 months.
1.3.5. Some other prognostic score
1.3.5.1. Intracerebral hemorrhage score (ICH score)
The ICH score is based on 5 clinical indicators: age> 80, Glasgow
scale, routine volume of CT hematoma at admission, hematoma location
(upper or under tent) and the presence of signs of intraventricular
bleeding. The highest ICH score is 6 points, the lowest is 0 points.
However, with ICH under the tent, no patient achieved a score of 6
because there was no volume of hematoma> 30ml.
1.3.5.2. FUNC score
The FUNC score was developed by Rost N.S in 2008, the structure
consists of five components: initial hematopoietic volume, age, ICH
position, glasgow point and cognitive deficiency before ICH. The lowest
score is 0 points, the highest is 11 points. The FUNC score is the
predictive outcome level in the acute phase, providing the most essential
basic guidelines for clinicians and patients' families who are faced with
decisions about treatment and care for patient and strategic options for
clinical trials.


8
1.3.5.3. Nine - Point Score
Brouwers H.B (2014), build a 9-point score based on data including:

using warfarin when onset of symptoms, initial volume of hematoma,time
from onset to initial CT scan, “spot sign” on CTA scan. The highest score is 9
points, the lowest is 0 point. From the data on the author successfully built a 9point score to predict the hematoma expansion.
CHAPTER 2: SUBJECTS AND METHOD
2.1. Participants
2.1.1. Inclusion criteria
- The time from onset to hospitalization ≤ 6 hours.
- Age ≥ 18 years old.
- Clinical: In accordance with AHA/ASA’s Stroke (2013): A stroke
is a sign that the brain’s dysfunction (localized or diffuse) progresses
rapidly, lasting more than 24 hours or leads to death with no apparent
cause other than of vascular origin. Localized neurological symptoms are
suitable for the brain area that damaged arteries are distributed, not due
to trauma.
- Subclinical (gold standard for ICH diagnosis): typically with
hematoma mass density 65 - 90 HU (with or without blood to the
ventricles) on CT scans without contrast.
2.1.2. Exclusion Criteria
- infratentorial intracerebral hemorrhage
- Secondary ICH (Cerebral aneurysm, cerebral vein malformations,
trauma, drug abuse etc)
- Severe renal failure, CVA sequelae with mRS ≥ 3, end-stage
cancer etc.
- Death in 24 hours, or surgery before the second CT scan.
- The patient or the patient's family does not agree to participate in
the study.
2.2. STUDY METHODS
2.2.1. Study design: prospective descriptive study, longitudinal tracking
2.2.2. Sample study
Formula to calculate sample size:

9
𝟐
𝐧 = 𝐙𝟏−∝/𝟐

𝐩 (𝟏 − 𝐩)
𝐝𝟐

Inside:
𝟐
• The level of statistical significance α = 0,05. So 𝐙𝟏−∝/𝟐
= (1,96)2 =
3,8416.
• p is the proportion of patients with hematoma expansion. According to
the results of the study by Demchuk A.M (2012), the proportion of
patients with hematoma expansion is 32%.
• d = 0.09. Is the desired accuracy.
• Apply to the formula n = (1.96)2 x 0.32(1-0.32)/(0.09)2 = 103.2
• Minimum sample size in theory is 103 patients. In this study, we
selected 126 eligible patients.
2.2.3. Method of sampling: convenience sampling
2.2.4. Location and time
The study was conducted at the Stroke Unit, ICU, Yen Bai
Provincal General Hospital during the period from November 2014 to the
end of December 2018.
2.2.5. Steps of the research
ICH patients meet the inclusion and exclusion criteria selected after
approval for participation in the study.
After asking diseases, examining and collecting data on age, gender,

medical history (hypertension, alcohol abuse, use of anticoagulants,
antiplatelet drugs, brain stroke, other pathologies), symptoms and onset
time, clinical symptoms (Glasgow score, pulse, blood pressure, focal
neurological signs) and ICH images on a CT scan of the skull at the first
time when hospitalized (period from the onset to the time of capture,
density, morphology and volume according to ABC/2 formula of
hematoma), the patient is given blood tests (CBC, basic coagulation,
hepatic and renal function) and computed tomography angiography
cerebrovascular (CTA) to determine the “spot sign” according to the
standard Delgado Almandoz, and to eliminate secondary ICH.
All study patients were treated according to the general regimen:
airway control, respiratory support, circulatory support, neurological


10
monitoring and evaluation, monitoring and examination at the Stroke
Unit, Yen Bai Provincal General Hospital. Steady patients have a CT
scan of the brain for the second time after 24 hours, or whenever the
patient shows signs of neurological impairment (Glasgow score
decreases by ≥ 2 points, and / or signs of new paralysis or increased
paralysis), the second CT scan was performed immediately. Evaluation
of clinical symptoms (Glasgow score, pulse, blood pressure, focal
neurological signs), time of imaging and image of ICH on the first brain
CTA scan (density, morphology and volume by ABC / 2 formula of
hematoma). The study patients were divided into two groups: Group I
had hematoma expansion; Group II has non-hematoma expansion.
Hematoma expansion was assessed according to Wada (2007) and Park
(2010) standards: ICH diagnostic criteria; volume of hematoma of
cerebral parenchyma increased> 30% or> 6ml on the second CT scan
compared with the first time (upon admission). Evaluate the shape and

density of hematoma according to the standards of Barras C.D (2009).
Comparing and analyzing the early predictive factors of hematoma
expansion and determining the value of the spot sign scoreaccording to
the research criteria.
2.2.3. Data processing and data analysis: The research data was
collected according to the form of researched medical record, processed
and analyzed on medical statistics software.
2.3. Ethical considerations
The patient or family member is explained about the purpose and
method of the study. Only patients or family members who represent
legal patients who agree to voluntarily participate are included in the
study. Patients or family members representing legal patients have the
right to discontinue participation in the study at any time.


11
CHAPTER 3: RESULTS
3.1. Sample characteristics
Table 3.1. Distribution characteristics of age and gender
p(‫)٭‬
Index
Value
65.1± 13.6
Average (year, 𝑋±SD)
Age
Highest (year)
96
<0.001
Lowest (year)
24

Gender (‫)٭‬
86
Male (percentage %)
(68.3%)
40
Female (percentage %)
(31.7%)
Comment: The average age of the study patients is 65.1 ± 13.6, the highest
is 96 years, the lowest is 24. The male rate is higher than the female rate. The
difference was statistically significant in the sex of the study patients (p
<0.001).
3.2. Some early predictable hematoma expansion of study patients
3.2.1. Glasgow score when hospitalized and the risk of hematoma
expansion
Table 3.2. Glasgow score when hospitalized and the risk of hematoma
expansion
Glasgow coma score when 3 - 8 score
9 - 12 score ≥ 13 score
hospitalized
n
%
n
%
n
%
11
27
Hematoma expansion
68,8
50,0 16 28,6

5
31,2
27
50,0 40 71,4
Non-hematoma expansion
Tổng số
16
100
54
100 56 100
p
0,006
Comment: The proportion of hematoma expansion lowest in the Glasgow
group when hospitalized ≥ 13 scores, the highest in the group with
Glasgow score of ≤8 scores. The difference was statistically significant
(with p = 0.006).


12
3.2.2. The time from the onset to the first CT scan and the risk of
hematoma expansion
Table 3.3. The time from the onset to the first CT scan and the risk of
hematoma expansion
≤ 3 hours
(n=61)
n
%

> 3 hours
(n= 65)

n
%

Hematoma expansion

33

54.1

21

32.3

Non-hematoma expansion

28

45,9

44

67,7

Tổng

61

100

65

100

The time from the onset to
the first CT scan

p

0.014

Comment: The proportion of patients with hematoma expansion in
hospital admission group ≤ 3 hours was higher than that of patients
admitted after 3 hours (54.1% compared to 32.3%). There were
statistically significant differences (with p <0.05).
3.2.3. Some hematological and coagulation indicators and hematoma
expansion
Table 3.4. Some hematological and coagulation indicators and
hematoma expansion
NonHematoma
Total
hematoma
p
expansion
(n=126)
expansion
Index
(n1;
(n1 =54)
(n2 =72)
n2)

n
%
n
%
n
%
5
INR value
≥ 1,2 18 14.3 13 24.1
6.9
0.0101
< 1,2 108 85.7 41 75.9 67 93.1
11 8.7
7
Concentration < 2
13.0
4
5.6
0.1560
of fibrinogen
≥ 2 115 91.3 47 87.0 68 94.4
8
4
Platelet (G/l) <150 12 9.5
14.8
5.6
0.0910
≥150 114 90.5 46 85.2 68 94.4
Comment: The proportion of patients with INR ≥ 1,2 in the extensive
hematoma group was higher than that of the non-hematoma expansion

group (24.1% versus 6.9%). The difference was statistically significant
(with p <0.001). There was no difference in the value of fibrinogen and
platelets between the two groups of study patients.


13
3.2.4. Some biochemical indicators when hospitalized and hematoma
expansion
Table 3.5. Some biochemical indicators when hospitalized and
hematoma expansion
NonHematoma
Total
hematoma
expansion
p
(n =126)
expansion
(n1 =54) (‫)٭‬
Index
(n1;n2)
(n2 =72)
n
%
n
%
n
%
26
22
45.8

Glucose
≥8.3 48 38.1
54.2
0.046
(mmol/l)
(‫)٭‬
28
50
64.1
<8.3 78 61.9
35.9
13
9
AST value ≥ 40 22 17.5
24.1
12.5
0.095
(U/l)
41
75.9
63
87.5
< 40 104 82.5
21
18
ALT
≥ 40 39 31.0
38.9
25.0
0.972

Value (U/l) < 40 87 69.0
33
61.1
54
75.0
Comment: The rate of hematoma expansion in Glucose group ≥8,3mmol
/l was higher than in Glucose group <8.3mmol /l. The difference was
statistically significant with (p <0.05). There were no statistically significant
differences in AST and ALT enzymes between the two study groups.
3.2.5. Some image characteristics on the first CT scan and the risk of
hematoma expansion
Table 3.7. Shape of hematoma, Density of hematoma on the first CT
scan and the risk of hematoma expansion

Hematoma expansion

Shape of hematoma
Uneveness
Eveness
n
%
n
%
37
17
67.3
23.9

Non-hematoma expansion

18

32,7

54

76,1

Density of hematoma
Heterogeneity
Homogeneity
n

%

n

%

Hematoma expansion

40

61.5

14

23.0

Non-hematoma expansion

25

38,5

47

77,0

OR
KTC95%
p
6,52
(2,98-14,29)
<0.001
OR
KTC95%
p
5,37
(2,46-11,69)
<0.001


14
Comment: The proportion of hematoma expansion in the group with
irregular hematoma shape was higher than that of the regular hematoma
group. The difference was statistically significant with p <0.001
Table 3.8. Volume of hematoma on the first CT scan and the risk of
hematoma expansion
0.2-29.9 ml 30-59.9ml

> 60 ml
Volume of
p
%
hematoma
n
n
%
n
%
Hematoma
22 29.3 21 60.0 11 68.8
expansion
Non-hematoma
53 70,7 14 40,0
5
31,2 0,001
expansion
75
100 35 100
16
100
Total
Comment:The larger the volume of hematoma, the bigger hematoma
expansion rate. The difference was statistically significant with p <0.01.
3.2.6. Relationship between “spot sign”and hematoma expansion
Table 3.9. Relationship between “spot sign”and hematoma expansion
Yes
No
OR

“spot sign”
(n1 =42)
(n2 =84)
KTC 95%
p
n
%
n
%
35
19
6,52
Hematoma expansion
83.3
22.6
7
16,7
65
77,4 (2,98-14,29)
Non-hematoma
< 0.001
expansion
42
100
84
100
Total
Comment: The percentage of “spot sign”of the study patients was 33.3%.
The prevalence of hematoma expansion in the group with a “spot sign”
was higher than in the group with no “spot sign”.

3.2.7. The role of the “spot sign” and the spot sign score in the early
forecast of hematoma expansion through multivariate analysis
Table 3.10. The role of the “spot sign” and some early predictable of
hematoma expansion
Risk factor
p
OR
95%CI
Fibrinogen (<2 g/l; ≥ 2g/l)
1.01- 19.28
0.049 4.4
“Spot sign” (Yes; No)
0.000 11.15 3.51- 35.44
Comments: “spot sign” on CTA capture and fibrinogen levels are two
independent prognostic factors that predict early hematoma expansion.


15
Table 3.11. The role of the spot sign scoreand some early predictable of
hematoma expansion
Risk factor
p
OR
95%CI
Glasgow score when hospitalized
2.09
1.01- 4.37
0.048
(3-8; 9-12; ≥13 điểm)
Fibrinogen (<2 g/l; ≥ 2g/l)

4.5
1.1 -18.51
0.037
spot sign score
3.62
1.64-7.99
0.002
Comment: There are three independent prognostic factors that predict
the spread of hematoma in a multivariate regression model: spot sign
score; Glasgow score upon admission and fibrinogen concentration.
3.3. The value of the spot sign score in the prognosis of acute phase
bleeding patients
3.3.1. The ability to predict early the hematoma expansion of the spot
sign score

Chart 3.1. The ability to predict early the hematoma expansion of the
spot sign score
Comment: The higher the point of the spot sign score, the greater the risk
of hematoma expansion with an area under the ROC curve of 0.779 with
p <0.001.
3.3.2. Prognostic ability of death of the spot sign score in inpatient
treatment

Chart 3.2. Prognostic ability of death of the spot sign scorein inpatient
treatment


16
Comment: The higher the point of the spot sign score, the higher the
mortality rate with the area under the ROC curve of 0.791, p <0.001.

3.3.3. The ability to predict severe disability outcomes of the spot sign
score after 3 months

Chart 3.3. The ability to predict severe disability outcomes of the spot
sign score after 3 months
Comment: The higher the point of the spot sign score, the greater the rate of
heavy disability with the area under the ROC curve is 0.674, p <0.001.
3.3.4. The spot sign score role to predict mortality and disability of
study patients through multivariate analysis
Table 3.12. Some several factors related to prognosis of death in inpatient
treatment
Prognostic factor
p
OR
95%CI
Glasgow score when hospitalized 0.000
6.89 2.77- 17.13
(3-8; 9-12; ≥13 score)
Spot sign score
2.72
1.52- 4.87
0.001
Comment: Glasgow score when admitted to hospital and spot sign score,
are two independent prognostic factors predicting death in Hospital.
Table 3.13. Some several factors are related to the prognosis of severe
disability after 3 months
Prognostic factor
p
OR
95%CI

Glasgow score when hospitalized 0.000
6.37
2.48- 16.36
(3-8; 9-12;≥13 score)
Intraventricular bleeding
3.67
1.15- 11.70
0.028
(Yes: No)
Initial volume of hematoma (<30; 0.002
4.40
1.72- 11.25
30-60; ≥ 60ml)
Spot sign score
4.30
1.67- 11.07
0.003


17
Comment: Glasgow score upon admission; Intraventricular bleeding;
Initial hematoma volume and spot sign score are 4 independent
prognostic factors that predict severe disability outcomes after 3 months.
3.3.4. Compare the value of the spot sign score with some scales in
the prognosis of research patients
3.3.4.1. Compare the spot sign score value and 9- point score in early
forecasts of hematoma expansion

Chart 3.4. ROC curve for predictive hematoma expansion value of the
spot sign score and 9- point score

Comment: The spot sign score has an area under the curve (AUC) of 0.779;
95% CI: 0.70 - 0.85 and the 9- point scorehas an AUC of 0.787; 95% CI:
0.71 - 0.85. The difference was not statistically significant for AUC between
the two scales (with p = 0.7212).
3.3.4.2. Compare the value of the spot sign score and the ICH score in
the prognosis of death in inpatient treatment

Chart 3.5. ROC curve for the prognostic value of death in inpatient
treatment of the spot sign score and ICH scale
Comment: The spot sign score has AUC of 0.791 (95% CI: 0.71 - 0.86)
and the ICH score has AUC of 0,800 (95% CI: 0.72 - 0.87). The
difference was not statistically significant about AUC between the two
scales (p = 0.8609).


18
CHAPTER 4. DISCUSSION
4.1. Sample characteristics
Average age: In the study patients was 65.1 ± 13.6 years, the highest
was 96 years, the lowest was 24 years old. The results of our study are
similar to that of Dalgado Almaldoz. JE (2010) average age is 66.7 years,
or Wada R (2007), median is 64 years old (31-81), higher than research
by Nguyen Huu Tin (2004) and Ngo Thi Kim Trinh (2013) (57.24 ±
11.85 and 55.37 ± 11.78, respectively).
Gender: We meet men more than women, the male / female ratio is 2.15
/ 1. However, this rate changes through studies, our results are similar to
the research results of Tang Viet Ha (2008) is 1.98 / 1, Wada R (2007) is
2/1, or Li N (2011) was 2.15 / 1 and Dalgado Almaldoz JE (2010) male /
female ratio was 1.19 / 1.
4.2. Some early predictors of hematoma expansion of study patients

4.2.1. Glasgow score when hospitalized
The lower the glasgow score when hospitalized, the higher the rate
of hematoma, specifically in the glasgow score group 3-8 scores, the rate
of spareading hematoma was 68.8% higher than that of Glasgow group
9-12 scores, 50% and group ≥ 13 scores was 28.6%, the difference was
statistically significant with p <0.05. According to Fujii Y (1998), the
rate hematoma expansion in high consciousness disorders group when
hospitalized has signification than the group with no consciousness
disorders (20% versus 8%). However, Nguyen Huu Tin (2004), showed
that the group of patients with glassgow score ≤ 8 had proportion of
hematoma expansion of 26.5% and glassgow score group> 8 of 14.6%,
the difference not statistically significant. In our opinion, this difference
is due to the input variables of different studies.
4.2.2. The time of onset until the first CT scan and the hematoma
expansion
The rate of hematoma expansion in the hospitalized group within 3
hours after stroke accounted for 54.1%, higher than the hospital
admission after 3 hours accounted for 32.3%, the difference was
statistically significant (p < 0.05). Brott T (1997), prospective studies on
103 patients admitted to hospital within the first 3 hours after onset of CT
scans immediately, taken again 1 hour and 20 hours, showed that
hematoma expansion 38% in the first 24 hours after onset (in which 26%
of patients had hematoma expansion in the first 3 hours, the remaining
12% had hematoma expansion in the next 21 hours. Nguyen Huu Tin
also said that the shorter the time from onset to CT scan is the most
powerful prognostic factor for hematoma expansion.


19
4.2.3. Blood glucose levels on admission

Many studies have demonstrated that there is a relationship between
hyperglycemia at hospitalization and the risk of death and disability in
stroke patients, regardless of whether or not the patient has diabetes
before. We analyzed a single variable showing that the rate of hematoma
expansion in blood sugar group ≥8,3mmol / l was higher than blood
sugar group <8.3mmol / l (54.2% compared to 35.9%). Similarly, Qureshi AI
(2011), a multicenter analysis study evaluated the effect of hyperglycemia on
admission on outcomes in patients with cerebral bleeding hospitalized within 6
hours after initiation showed that the risk of hematoma expansion in patients
with hyperglycemia was 2.59 times higher than in the group without
hyperglycemia. However, the relationship between hyperglycemia and
hematoma expansion is still controversial.
4.2.4. Some image characteristics on the first CT scan and hematoma
expansion
The shape of hematoma: Some studies have shown that the rate of
hematoma expansion in cases, that has uneven bank shape of hematoma
on the first CT skull scan, is much higher than the group has even round
bank hematoma. The results of our study showed that the proportion of
hematoma expansion in the uneven bank hematoma shape group higher
than that of the regular shaped group (67.3% and 23.9%). Our results are
similar to those recorded by Nguyen Huu Tin (2004) and Ngo Thi Kim
Trinh (2013). To explain this problem, the authors argue that the shape of
the hematoma is uneven due to the simultaneous bleeding in many
arterioles into the hematoma.
The density of hematoma: The proportion of hematoma expansion in
group with a higher heterogeneous density has significant compared to
the group with homogeneous proportions (61.5% and 23%, p <0.001).
Barras CD (2009), when the prospective study of 90 patients with acute
ICH received an initial CT scan after a 3-hour onset, it was concluded
that hematomas with the large volumes had more abnormal hematoma

shapes and more heterogeneous hematoma density, as well as higher
hematoma expansion risk.
Volume of hematoma on the first time CT scan: The results of our
study showed that the rate of hematoma expansion in the group with
hematoma volume <30ml accounted for 29.3%, in the group of 3059.9ml hematoma volume accounted for 60% and in the group> 60ml
was 68.8%, the difference was statistically significant with p <0.01. This
shows that the larger the volume of hematoma, the higher the risk of
hematoma. Delgado Almandoz J.E (2009), when investigating the


20
relationship between hematoma volume on first CT scan with the risk of
hematoma expansion, it was announced that the risk of hematoma
expansion increases with increasing volume of hematoma. Fujii Y
(1998), when analyzing 259 patients with ICH lenticular nucleus, the rate
of hematoma expansion increased as the volume of hematoma on the first
CT scans increased. Although there was diffirent between the criteria
divisions the hematoma volume group on the study and the criteria
choices patients on the the study, most studies have demonstrated that
the initial hematoma volume is associated with an increased risk of
hematoma expansion.
“Spot Sign”: in our study is 33.3%, the rate of hematoma expansion on
the group with higher “spot sign” had more significative than the group
with non-“spot sign” (83.3% compared with 22.6%, p <0.001). Li N
(2011), the study studied 139 patients, the presence of “spot sign”
accounted for 21.6%, the group of patients with “spot sign” had a greater
risk of hematoma expansion, longer inpatient treatment time, higher
mortality rates and worse outcomes than those without “spot sign”. To
explain this difference, we thought it may be due to the time from the
onset of stroke to take CTA had difference through studies. On the other

hand, the process of taking and models of CT scanners in different
studies leads to different "spot" sign rates. However, the presence of a
“spot sign” was related to the spread of hematoma through studies.
4.2.5. The role of the “spot sign” and the spot sign score in the early
forecast hematoma expansion through multivariate analysis
Initially, there were eight statistically significant variables in
univariate analysis and were added to variables (AST, ALT, SBP,
HATB, platelets, fibrinogen) into the multivariate analysis model. The
results showed that: “spot sign” and fibrinogen levels at hospitalization
were two independent prognostic factors that predict the spread of
hematoma early (Table 3.10). Next, we put the spot sign score into the
multivariate logistic regression model. The “spot sign” is a feature of the
spot sign score, so we didn't include it in a multivariate analysis model at
the same time. The results of the study noted: spot sign score has an
independent prognostic value, predicting the spread of hematoma early.
There is also a glasgow score on admission and Fibrinogen levels are two
independent prognostic factors that predict hematoma expansion early
(Table 3.11).
Delgado Almadoz JE (2009), when the “spot sign” is included in the
logistic multivariate regression model, the “spot sign” is the predictive
independent predictor of hematoma expansion. and two other factors are


21
the time from the onset of stroke to the time of CTA capture and the
blood glucose level when hospitalized. However, when included in the
multivariate regression model, the spot sign score became the most
powerful prediction of hematoma expansion (with p <0.0001). According
to Dinh Vinh Quang (2015), logistic multivariate regression analysis
shows that there were two independent prognostic factors that predict the

spread of hematoma: (1) The shape of hematoma with irregular edges
(OR = 0.19; p = 0.005), and (2) There are signs of contrast staining (spot
sign) on cranial CT scans (OR = 2.41; p = 0.044).
In our opinion, there were differences in the number of independent
prognostic factors, but most authors share the same opinion, the “spot
sign” and the spot sign score is a strong prognostic factor that predicts
hematoma expansion in acute ICH patients.
4.3. The value of the spot sign score in the prognosis of acute ICH phase
4.3.1. The ability to predict the hematoma expansion of the spot sign
score in researched patients
The results of our study showed that the higher the point of the spot
sign score, the greater the risk of hematoma expansion. When the point
of the spot sign score is 0, the rate of hematoma expansion to only
22.6%, the point of the spot sign score is 1, the rate of hematoma
expansion has increased to 80% and when the score of spot sign score is
4, the rate of hematoma expansion to 100% and AUC is 0.779. Compare
the value of the spot sign score in our research with Delgado Almadoz
J.E (2009), also published similar results: when the score of spot sign
score is 0 then the rate of hematoma expansion is only 2%, the point of
the spot sign score is 1, the spread rate of hematoma has increased to
33% and when the score spot sign score is equal to 4, the rate of
hematoma expansion to 100% and AUC of 0.93 with p <0.001.
However, Huynh T.J (2013), when evaluating primary outcomes,
shows: The point of the spot sign score is 0, the rate of hematoma
expansion is 21.6%, when the score of scale is 1, the rate of hematoma
expansion in the study is 56%, but when the point of the spot sign score
increases to 4, the spread rate of hematoma to only 50%, with the AUC
being only 0, 68 lower than our research.
4.3.2. The ability to predict mortality of the spot sign score in inpatient
treatment

Delgado Almadoz J.E (2010), when applying the spot sign scoreto
predict mortality in inpatient treatment shows, the higher the point of the
spot sign score, the higher the risk of death. The point of the spot sign
score is 0, 1, 2, 3 and 4, respectively, mortality rates will be 24%, 41%,


22
59%, 61% and 64%, and the AUC is 0.64 (with p = 0.0001). In our
research results, the point of the spot sign score is 0, 1, 2, 3 and 4, the
corresponding mortality rate is 11.9%, 75%, 62.5%, 60% and 100%, and
AUC is 0.791, with p = 0.001. Although the AUC is higher than that of
the Delgado Almadoz J.E (0.791 vs. 0.64), we think this difference is
probably due to the so small sample size of our research, when the score
spot sign scoreis 4, there is only 1 case and this case died in inpatient
treatment, so it does not reflect fully the risk of death.
4.3.3. The ability to predict severe disability outcomes of the spot
sign score after 3 months.
The severe disability rate gradually increases with the point of the
spot sign score, when the point of the spot sign score is 0 with the rate of
severe disability accounting for 29.8% and when the point of the spot
sign score≥ 3, the disability rate is up to 100% with p <0.001 and AUC is
0.674. According to Romero J.M (2013), when assessing severe
disability outcomes (mRS≥3) at 3 months time, the spot sign score is an
independent prognostic factor predicting severe disability outcomes in
patients with ICH and the risk is 3 times higher.
So far there are not many foreign and domestic studies on the value
of the spot sign score to predict severe disability outcomes for acute ICH.
Delgado Almadoz J.E (2010), 393 patients survived, were included in the
analysis of clinical outcomes with severe disability outcomes determined
when mRS≥4. The results show that the higher the point of the spot sign

score, the more severe the incidence of disability is with the AUC of
0.56. Comparing our AUC higher than that of the Delgado Almadoz J.E
(2010) is 0.674 compared to 0.56. To explain this difference, according
to us, due to the selection of different study patients, the time of our CTA
scan was earlier so our “spot sign” rate was higher.
4.3.3. The role of the spot sign score to predict mortality in inpatient
treatment through multivariate analysis
Delgado Almadoz JE (2010) noted that the spot sign score is an
independent predictor of the risk of death in inpatient with an odds ratio
of OR = 1.5 (95% CI: 1, 2 -1,9; p <0.0002). Romero J.M (2013) shows
that there are five independent prognostic factors predicting the risk of
death in hospitals: (1) age, (2) the rate of hematoma expansion, (3)
spot sign score, (4) intaventricular bleeding, (5) initial hematoma
volume. Among the variables evaluated by multivariate regression
analysis, initial hematoma volume was arguably the most powerful
predictor of death risk.