INVESTIGATION ON CLINICAL FEATURES, BRAIN IMAGING, a NUMBER OF RIKS FACTORS AND THE VALUE OF d DIMER IN DIAGNOSIS OF CEREBRAL VENOUS THROMBOSIS

THE MINISTRY OF EDUCATION
AND TRAINING
THE MINISTRY OF
DEFENCE
MILITARY MEDICAL UNIVERSITY
LE VAN MINH
INVESTIGATION ON CLINICAL FEATURES, BRAIN
IMAGING, A NUMBER OF RIKS FACTORS AND THE
VALUE OF D-DIMER IN DIAGNOSIS OF CEREBRAL
VENOUS THROMBOSIS
Speciality: NEUROLOGY
Code: 62 72 01 47
MD-PhD. THESIS SUMMARY

HANOI - 2014
THE THESIS WAS COMPLETED AT MILITARY
MEDICAL UNIVERSITY
Supervisors:
A.Prof. Dr. PHAN VIET NGA
A. Prof. Dr.PHAM NGOC HOA
1st reviewer:
2nd reviewer:
3rd reviewer:
The thesis will be upheld before the University Grade Thesis
Examination Board
at: hour on day month year
.
The thesis can be found at:
- National library
- Military medical university’s library
FOREWORDS

Cerebral venous thrombosis (CVT) is estimated to
account for 0.5% of all strokes in adults. In recent years, with the
present of non-invasive and high accuracy diagnostic techniques
such as magnetic resonance imaging and computed tomography,
the rate of diagnosis of CVT also increases.
Congenital Thrombophilia states account a high rate in
white skin peoples and make up 15-30% cases of patients with
venous thrombosis. D-dimer test is an inexpensive and
commonly available method which has high performance in
diagnosis of CVT as proven by some prospective and meta-
analytic studies. Moreover, at present there is no published study
on the role of D-dimer test in the diagnosis of CVT in our
country.
In Vietnam, there have been many studies on aspects
of stroke. However, there are few studies on CVT, therefore
we decided to study CVT with the following objectives:
OBJECTIVES OF THE STUDY
1. To describe some clinical features and brain images of CVT.
2. To review some risk factors of CVT.
3. To determine the sensitivity, specificity, positive predictive
value and negative predictive value of D-dimer test in the
diagnosis of CVT.
1
NEW CONTRIBUTION OF THE THESIS
1. Our study described some clinical features and brain images of
CVT.
2. Our study found some genetic and acquired risk factors of
CVT.
3. Our study successfully determined the role of D-dimer test in
the diagnosis of CVT.

The thesis includes 115 pages, with 4 chapters, 43 tables, 16
charts, 1 diagram, 8 Vietnamese references and 83 English
references. 3 pages forewords, 35 pages overview, 19 pages
subjects and methods, 24 pages results, 31 pages discussion, 2
pages conclusion, 1 page recommendations.
Chapter 1 - OVERVIEW
1.1. GENERAL
1.1.1. Anatomy of cerebral veins and sinous
The veins of the brain have no muscular tissue in their
thin walls and possess no valves as compared to other types of
veins. They emerge from the brain and lie in the subarachnoid
space. They pierce the arachnoid mater and the meningeal
layer of the dura and drain into the cranial venous sinuses.
Blood from both brain is drained by cerebral veins into dural
sinuses and then into the jugular. Cerebral veins is divided
into 3 important groups which are superfical cerebral veins,
deep cerebral veins and posterior fossa veins.
1.1.3. Pathogenesis of cerebral venous thrombosis
2
Causes of CVT are diverse. According to the hypothesis
of Rudolf Vichow: there are 3 basic factors contributing in the
formation of venous thrombosis: venous congestion;
hypercoagulable state; vascular injuries. Most of the main factors
causing thrombosis are venous congestion and hypercoagulable
state.
1.2. CLINICAL
Clinical symptoms of CVT are diverse, non-specific and
the symptom onset in cerebral venous thrombosis is usually
subacute (2 days to 1 month), in some cases it can be acute (2
days or less) and simulate arterial stroke (20-30%). In patients

presenting isolated intracranial hypertension like brain tumor, the
onset of symptoms an be chronic. Depending on the location of
the thrombosis and the collateral blood flow, and patient’s age,
the range of clinical symptoms associated with cerebral
venous thrombosis is astonishingly varied, and its clinical
presentation is extremely variable and unspecific. The
progression of deep CVT may be gradually severe or variable
depending on the status of intracranial pressure with or without
accompanied seizures. Besides, CVT is often symmetry and
hemorrhagic transformation of the lesion is quite frequent
1.3. LABORATORYL
1.3.2. Magnetic resonance imaging
1.3.2.1. Magnetic resonance imaging (MRI)
3
Magnetic resonance imaging in combination with
magnetic resonance venography (MRV) imaging provide a best
non-invasive technique for the diagnosis of CVT which is
considered as a gold standard and can replace invasive
conventional cerebral angiography and is used as a leading mean
of diagnosis in case with clinical suspicion of cerebral venous
thrombosis. Advantages of magnetic resonance imaging are it
can survey the blood flow and allow us to see the direct imaging
of the venous sinous thrombosis and brain parenchymal injuries.
1.3.4. D-dimer in diagnosis of cerebral venous thrombosis
1.3.4.1. Biochemical characteristics of D-dimer
D-dimer, the final product of plasma in-mediated
degradation of fibrin-rich thrombi. It is a small protein chain in
blood after a clot is fibrinated. After the formation of a clot,
the hemolytic path is triggered with plasminogen is activated to
plasmin. Then plasmin cut fibrin at E-D bonds to form separate

D-D segments called D-dimer (2D).
1.4. RISK FACTORS
These risk factors are often related to Virchow's triad.
The markers such as: antithrombin, protein C, protein S, V
Leiden factor or activated protein C are considered as point
markers of congenital thrombophilia. The deficiency of
antithrombin (AT), protein C (PC) or protein S (PS) or the
present of V Leiden factor (FVL) will increase the risk of
thrombosis, each factor is listed hereunder.
4
Chapter 2: SUBJECTS AND METHODS OF THE STUDY
2.1. SUBJECTS OF THE STUDY
2.1.1. Group of patients with cerebral venous thrombosis
We chose patients ≥ 16 years old who had been
diagnosed with CVT at the Department of Neurology, Cho Ray
Hospital from January 2010 to May 2012 based on the following
inclusion criteria:
INCLUSION CRITERIA patients had all criteria as follows:
a) Clinically suspected to have CVT: when the patients got
one of the following criteria:
- Abnormal headache with acute, subacute or chronic onset
and a history did not relate to cluster headache, migraine
headache or muscle tension headache.
- Atypical headache accompanying vomiting and did not
reponse to conventional treatment methods.
- Clinical signs of brain injury (focal neurological signs,
seizures, disorders of consciousness).
- Images of ischemic stroke on computerized tomography
film showed atypically distribution along cerebral artery.
b) Found images of CVT by using routine magnetic resonance

imaging in combination with magnetic resonance imaging
of vein and/or routine magnetic imaging in combination
with computerized tomography of cerebral vein and/or
cerebral angiography with DSA technique.
EXCLUSION CRITERIA
1. Patients using anticoagulants.
5
2. Patients were diagnosed with deep vein thrombosis or
pulmonary embolism and/or calf vein thrombosis.
3. Patients with signs of disseminated intravascular
coagulation.
4. Patients had experienced major surgery or severe trauma
within 3 months.
5. Patients did not agree to participate in the study.
2.1.2. Control group
We chose 57 people who came for periodic health
examination at Cho Ray Hospital and they voluntarily
participated in the study. All these cases were carefully clinically
examined and tested. Protein S, Protein C, ATIII, V Leiden
factor, D-dimer.
2.2. METHODS OF THE STUDY
This is a prospective, cross-sectional descriptive study
with control group.
2.2.1. Clinical study
All patients were clinically evaluated risk factors
according to a common data collection form and were treated
according to a unique regime at the Department of Neurology,
Cho Ray Hospital.
2.2.2. Subclinical study
All patient with clinical signs of CVT were examined the

MRI tests:
6
- Investigation on features of CVT images using magnetic
resonance imaging.
- Investigation on rick factor protein S.
- Investigation on risk factor protein C.
- Investigation on ATIII risk factor.
- Investigation on V Leiden risk factor.
- Investigation on blood D-dimer concentration.
Chapter 3 - RESULTS OF THE STUDY
3.1. SOME GENERAL CHARACTERISTICS OF PATIENTS
There were 59 patients with cerebral venous thrombosis
(experimental group) and 57 people in the control group. The
male/female ratio in the experimental group was 1/0.78, and in
the control group was 1/0.72.
The percent of patients aged 21 to 50 accounted a
majority of both groups, wherein, the experimental group had 47
patients (82.46%), and the control group had 41 patients
(71.93%). The percent of patients aged less than 20 was mostly
equivalent between both groups (6.78% vs 7.02%). In the
experimental group, the percent of patients with subacute onset
was 79.7%, patients with acute onset was 11.8% and patients
with chronic onset was 8.5%.
3.2. CLINICAL CHARACTERISTICS
3.2.1. Symptoms
Table 3.7: Clinical symptoms on admission
7
Symptom Number of patients Percent (%)
Headache 58 98.31
Motor deficit 34 57.63

seizures 30 50.85
Altered consciousness 22 37.29
Cranial nerve palsy 19 32.2
Papilledema 18 30.5
Sensorial deficit 5 8.47
Aphasia 5 8.47
Meningeal signs 3 5.08
Vertigo 3 5.08
3.3. IMAGING CHARACTERISTICS
3.3.2. Location of cerebral venous thrombosis
Table 3.16: Characteristics on location of cerebral venous
thrombosis of 57 patients tested by magnetic resonance imaging
Location Number of
patients
Percent (%)
Superior sagittal sinus 43 75.44
Transverse sinus 37 64.91
Sigmoid sinus 36 63.1
Inferior sagittal sinus 3 5.26
Cavernous sinus 2 3.51
Straight sinus 8 14.04
Deep cerebral veins 4 7.02
Internal jugular veins 5 8.77
Cortical veins 23 40.35
≥ 2 veins had blood clots 41 71.93
Comment: The location of cerebral veins mostly affected
by thrombosis on the images obtained by magnetic resonance
8
imaging was superior sagittal sinuses with 43 cases (75.54%),
next to transverse sinuses with 37 cases (64.91%).

3.3.3. Brain parenchymal abnormalities on magnetic
resonance imaging
Among 57 patients with CVT, patients with hemorrhagic
transformation accounted the highest percent (36.84%), followed
by patients with infarction alone (22.81%), and hemorrhage
(21.05%). Subarachnoid hemorrhage accounted smallest percent
(10.53%).
3.4. RISK FACTORS OF CVT
3.4.1. Risk factors of primary thrombophilia
Table 3.20: Prevalence of primary thrombophilia
Risk factors Number of patients Percent (%)
Protein S
Protein C
ATIII
V Leiden factor
≥ 2 risk factors
≥ 3 risk factors
4 risk factors
16/53
14/57
11/57
23/47
29/57
5/56
0
30.2
24.5
19.3
48.9
33.3

8.9
0
3.4.2. Factors of acquired thrombophilia
Table 3.25: Status of administration with oral contraceptive drugs
in female patients
contraceptiv
e drugs
Number of patients Total
Experimental group Control group
Yes 10 3 13
9
No 23 28 51
Total 33 31 64
χ2 = 4.2; p = 0.04; OR = 4.05 (CI 95%, 0.88-25.1)
Table 3.26: Status of pregnancy in female patients
Pregnancy Number of patients Total
Experimental group Control group
Yes 1 0 1
No 32 27 59
Total 33 27 60
χ2 = 1.06 ; p = 0.361
Comment: There was no statistically significant difference on
CVT between pregnancy group and non-pregnancy group with p
= 0.361.
Table 3.27: Puerperium characteristics on female patients
Puerperium Number of patients Total
Experimental group Control group
Yes 8 0 8
No 25 27 52
Total 33 27 60

χ2 = 7.5 ; p = 0.006
Comment: There was a statistically significant difference
on number of patients with CVT between Puerperium group and
control group.
3.5. CHARACTERISTICS OF D-DIMER TEST
3.5.1. Comparison of mean D-dimer concentration between
two groups
Table 3.29: D-dimer concentration (µg/L) of patients
D-dimer(µg/L) Experimental group Control group
10
Mean value 1890.92 146.19
Standar deviation 309.03 15.55
Number of
investigated patients
58 57
KTC 95% 1272.5 - 2509 115.03 - 177.35
Non-uniform variance T-test t = -5.63; p = 0.000
3.5.3. Identify the cut off of D-dimer test
Table 3.31: Cut off of D-dimer concentration
Cut off (µg/L) ≥256 ≥280 ≥302 ≥424 ≥502 ≥604
Sensitivity 93.1 93.1 91.33 79.3 74.1
4
62.07
Specificity 80.7 82.46 89.47 98.25 98.25 98.25
PPV (%) 83.08 84.38 89.83 97.87 97.73 97.3
NPV (%) 92 92.16 91.07 82.35 78.87 71.79
Area under ROC 0.869 0.877 0.90
4
0.887 0.861 0.801
Chapter 4 - DISCUSSION

4.1.SOME GENERAL CHARACTERISTICS OF PATIENTS
In all 59 cases with CVT, we diagnosed based on clinical
criteria and had CVT images obtained using magnetic resonance
imaging, computerized tomography and DSA.
Sex
In our study, the percent of female patients with CVT
was higher than percent of male patients with CVT (55.93% vs
44.7%), with the female/male ratio was 1/0.78. According to the
11
study of Tanislav, among 39 patients with CVT, the percent of
female patients was significantly higher than the percent of male
patients (71% vs 29%). According to the study of Khealani,
among 109 patients, the percent of female patients was higher
than the percent of male patients (53% vs 47%).
Overall, our study was similar to other studies of other
authors with the percent of female patients with CVT was higher
than the percent of male patients, especially female patients in
childbearing age, this suggested that the risk of CVT may relate
to pregnancy, puerperium and birth control drugs.
Age
Mean age of patients in our study was 37.8, wherein
percent of patients aged from 21 to 50 was 82.46%, percent of
patients aged > 50 was 13.56%.
According to the study of Nguyen Ngoc Hung on 37
patients with CVT, the mean age was 38.7. A retrospective study
on 48 patients with CVT of Terazzi et al. in 2004 showed that the
mean age of CVT patients was 44.8. According to the study of
Ferro, in 624 patients with CVT, there was 92.8% of patients <
65 years old.
Thus, according to above stated studies, the mean age of

adult patients with CVT was 32.7 to 44.8. Result of mean age in
our study was also withing this range and was not significantly
different to other studies of other authors worldwide.
Onset time and diagnosis time
12
In our study, the number of patients with subacute onset
accounted the highest percent (79.7%) followed by acute onset.
According to Paciaroni, onset of CVT is commonly subacute (2
days to 1 month) with 50-80% of all cases, but sometimes it
suddenly occurs with acute onset (< 2 days) like stroke (20-30%
of all cases). In some cases, the clinical symptoms of CVT are
similar to cerebral tumors, and there are few cases CVT occurs
with clinical signs are increased intracranial pressure, and chronic
symptoms at the onset (>1 month), 10-20%. According to the
study of Terazzi, in 48 patients with CVT, the percent of patients
with acute onset was 44%, subacute onset was 35% and chronic
onset was 21%. In general, the onset characteristics of CVT are
varied, but the majority of patients have cubacute onset and our
study data was equivalent to the result of Paciaroni's study.
4.2. CLINICAL CHARACTERISTICS
4.2.1. Symptoms and onset time
Headache: In our study, the headache was most common
with a rate of 98.31% of all cases. This result is similar to the
result of the study of Le Van Thinh and Trinh Tien Luc with
headache accounted 88%. In general, it is like other studies of
other authors, headache was most common and presented with
highest rate in these studies.
Paraparesis: Similar to results of other studies of other
authors, our study found that the symptom of paraparesis ranked
the 2nd position after headache with a percent of 57.63%. Signs

13
of this symptom were mainly 4th degree (25.5%), 3rd degree
(13.5%), 2nd degree (10.6%). According to the study of Tanislav
et al., this symptom also ranked the 2nd position after headache
(69%) with a rate of 44%. In general, the percent of patients with
paraparesis symptom in our study was equivalent to results from
others studies of other authors and the similar point is this
symptom ranked that 2nd or 3rd position among other clinical
signs.
Seizures: In our study, the percent of seizures symptom
ranked a 3rd position after headache and paraparesis. According
to the study of Le Van Thing and Trinh Tien Luc, seizures
symptom accounted 32%, ranked the 3rd position after headache
and cranial nerve paralysis. According to the study of Ferro,
seizures symptom accounted 39% in patients < 65 years old, and
45% in patients >= 65 years old.
Consciousness disorders: In our study, percent of patients
with consciousness disorcers accounted 37.29%, ranked the 4th
position among other symptoms. Wherein, mild consciousness
disorders was most common with 30.5% while severe
consciousness disorders accounted 6.8%. According to the study
of Bruijin in 59 patients with CVT (equal to the number of
patients in our study), the percent of patients with consciousness
disorders was 39%. Wherein, mild consciousness disorders
accounted 23.7%, and severe symptom accounted 15.3%.
Overall, the percent of CVT patients with consciousness
14
disorders symptom in the studies of domestic and foreign authors
was not as high as other symptoms, from 21% to 39%, the
percent of our study was within this range and similar to results

of other authors.
Cranial nerve palsy: In our study, the percent of patients
with cranial nerve palsy was 32.2%. According to the study of Le
Van Thinh and Trinh Tien Luc, the percent of patients with
cranial nerve palsy was 48%. According to the study of
Paciaroni, the percent of patients with cranial nerve palsy was
12%, this symptom commonly occurred in III, IV, V, VI, VII,
VIII, IX, X and XI.These nerves could be injured separately or
link to other nerves. Almost authors did not recognize cranial
nerve analysis in their studies, might be due to small sample size
then they only recognized that patients had focal neurological
dificits, such as in the study of Fink, this author only general
recorded that 60% of patients had focal neurological dificits.
Overall, the cranial nerve palsy symptom on patients with CVT
was not as common as other symptoms and accounted from 12%
to 48%, the percent in our study was 32.2%.
Papilledema: According to the results of our study, the
percent of patients with papilledema was 30.5%. This symptom
varied depending on studies of other authors (table 4.39).
According to the study of Stolz, papilledema accounted 40%.
According to the study of Ferro, among 624 patients there was
only 29% of patients < 65 years old and 14% of patients ≥ 65
15
years old had this symptom, ranked after headache and
positioning nerve sign and seizures. Overall, the percent of
patients with papilledema in our study was smaller than results
from studies of other authors, this could be due to the percent of
patients exhibited increased cranial pressure in our study was
higher than results from studies of other authors.
4.3. IMAGING CHARACTERISTICS

4.3.2. The location of the venous sinous thrombosis
Because in our study patients were diagnosed by
magnetic resonance imaging then we only discuss and compare
our study results to other authors based on this group (57
patients). The location of the venous sinous thrombosis system
which was highest percent was superior sagittal sinuses
(75.54%), followed by transverse sinuses (64.91%), sigmoid
sinuses (63.1%), and cortical veins (40.35%). The remaining
veins and sinus were less common such as: straight sinuses
14.4%, juglar veins 8.77%, deep cerebral veins 7.02%, inferior
sagittal sinuses 5.26% (3 cases), cavernous sinuses 2% (2 cases).
The number of patients with ≥ 2 veins or sinouses thrombosis
accounted 41/57 cases (71.93%).
According to the study of Zubkov, among 59 patients
diagnosed by routine magnetic resonance imaging in combination
with magnetic resonance veinography, there were 34% (19 cases)
of all patients had images of brain parenchymal injuries on the
magnetic resonance images and this author compared the percent
16
of patients with thrombisis on 2 groups of patients. Among
patients with brain parenchymal injuries, thrombosis were most
common in transverse sinuses with 78.9% (15 cases), followed by
superior sagittal sinuses 57.8% (11 cases), sigmoid sinuses 47.3%
(9 cases), straight sinuses 26.3% (5 cases), Galen veins 26.3% (5
cases), other position thrombosis were not significant. Among 37
patients without cerebral injuries, blood clots were most common
in veins of transverse sinuses with 86.4% (32 cases), sigmoid
sinuses 43.2% (16 cases), superior sagittal sinuses 24.3% (9
cases), straight sinuses 8.1% (3 cases).
Overall, we recognized that: Location of thrombosis

occurred with highest percent in superior sagittal sinuses (36% to
97.8%); followed by transverse sinuses and sigmoid sinuses (43%
to 83%); cortical veins accounted a high varied percent range
(3% to 43.2%), this might be due to the previous studies did not
apply GRE pulse sequence then the possibility of false negative
results was relatively high. Percent of straight sinuses with blood
clots ranged from 8.1% to 34.4%; deep cerebral veins ranged
from 7.02% to 25.3%. Other rare veinous sinouses thrombosis
were not mentioned in studies of other authors are cavenous
sinuses, inferior sagittal sinuses and jugular veins. In most cases,
when occurred CVT, cases with just one vein or sinous
thrombosis were rare, percent of cases with ≥ 2 position
thrombosis ranged from 54.7% to 71.93%. Thus, percent of
17
venous sinous thrombosis in our study was within the range of
other authors.
4.3.3. Brain parenchymal abnormalities on magnetic
resonance imaging
Based on the images obtained by magnetic resonance
imaging, we found that the percent of patients with CVT
accompanied parenchymal injuries was 84.21%, hemorrhagic
infarction accounted a highest percent 36.84%, followed by
infarction (22.81%); parenchymal hemorrhage 21.05%,
subarachnoid hemorrhage 10.53%.
According to the results from the study of Terazzi in 48
patients with CVT, among 44 patients were tested by
computerized tomography, there were 9 cases (21%) suspected to
have CVT (empty Delta sign, rope sign), 11 patients (25%) were
completely normal on imeges of computerized tomography, 24
patients (55%) had signs of cerebral hemorrhage or ischemic

injuries. Among 19 patients were tested again by computerized
tomography, there were 5 cases had signs of CVT. Also, in this
study, there were 42 cases diagnosed with CVT by magnetic
resonance imaging, wherein 12 cases (29%) had signs of CVT
alone, 15 cases (36%) related to parenchymal hemorrhage and
infarction. 12 cases (29%) had parenchymal hemorrhage or
infarction alone.
Overall, by comparing to results of studies of domestic
and foreign authors, we found that our study results were
18
relatively equivalent to results of other authors. Wherein, percent
of patients with infarction ranged from 19.4% to 70%;
parenchymal hemorrhage ranged from 10.2% to 57%;
subarachnoid hemorrhage ranged from 8% to 10.53%. However,
percent of patients with subarachnoid hemorrhage in our study
was higher than results from other studies, this might be due to
the diversity of injuries in CVT.
4.4. RISK FACTORS OF CEREBRAL VENOUS
THROMBOSIS
4.4.1.Factors of primary thrombophilia
According to results of our study, in patients with
primary thrombophilia, V Leiden factor accounted the highest
percent, followed by decreased protein S, decreased protein C,
decreased ATIII. Compared to control group, we found that OR
of protein C was highest, followed by V Leiden factor, ATIII. We
could not calculate OR of protein S because there was no case
with decreased protein S in control group.
According to the study of Bombeli [10], when analyzed
51 patients with CVT among 260 patients with venous
thrombosis (cerebral veins, portal veins, retinal veins, upper limb

veins, lower limb vein) and 120 healthy controls, he recognized
that V Leiden factor accounted the highest percent 13.7% with
OR 2.1 (CI 95%: 0.7-6), percent of decreased protein S was 2%
with OR 2.4 (CI 95%: 0.1-38.3); percent of decreased protein C
and ATII was 2%, he could not calculate OR of protein C and
19
ATIII due to limitation of sample size. According to the study of
Martinelli in 121 patients with CVT, the percent of V Leiden
factor was highest with OR = 4.7 (CI 95%: 1.8-11.8), followed by
decreased protein C, decreased protein S and decreased ATIII.
Due to limitation of sample size then the author calculated
general OR = 6 for 3 factors decreased protein C, decreased
protein S and decreased ATIII.
Thus, percents of factors of primary thrombophilia in our
study were higher than results of other studies of domestic and
foreign authors, this may relate to recial characteristics of Asians.
If compare each factor of primary thrombiphilia, we found that
these disorders varied depending on each study and did not
follow any basis. Thus, we can recognize that not only clinical
characteristics of CVT were diverse but also primary risk factors
were varied.
4.4.2. Factors of acquired thrombophilia
Among 33 female patients with CVT in our study,
percent of female patients using oral contraceptive drugs was
30.3%, percent of pregnant patients was 3.03%, percent of
puerperium patients was 24.2%, no case was being treated by
formone replacement therapy. When compared to control group,
we found that the risk of CVT in patients using oral contraceptive
drugs was high with OR = 4.05 (KTC 95%: 0.88-25.1); we could
not calculate OR for pregnant, puerperium and hormon

replacement therapy due to limitation of sample size.
20
According to the study of Tran Thanh Tung, among 21
female patients with deep vein thrombosis, the percent of patients
related to administration of oral contraceptive drugs was 13.8%,
percent of abortions was 10.3%, no case was recognized as
relating to pregnancy and hormone replacement therapy.
According to the study of Dentali, this was a meta-study of
previous studies, found that among 263 female patients with
CVT, percent of patients using oral contraceptive drugs was
58.9%. When compared to control group, OR = 5.59 (KTC 95%:
3.95 to 7.91). It might be due to in this study the author mainly
assessed risk factors of contraceptive drug administration then he
did not refer to the risk of pregnancy and puerperium on CVT.
Overall, on female subjects, all studies showed that
administration of oral contraceptive drugs and puerperium were
related to CVT. Especially in 2 studies of Martinelle and Bruijin,
the percents and OR relating to administration of oral
contraceptive drugs were very high, this might be due to female
subjects in these study aged from 18 to 54 then the possibility of
using contraceptive drugs were very high.
4.5. CHARACTERISTICS OF D-DIMER
4.5.1. Comparing mean D-dimer and testing time
In our study, mean concentration of D-dimer of
experimental group was higher than control group (1890.92 vs
146.19) and the difference was statistically significant (p <
21
0.000). The mean testing time of patients after onset was 10.24 ±
9.67.
When compared results of 3 studies of: Kosinski,

Crassard , Ghaffarpour. We found that mean concentration of D-
dimer of the experimental group in our study was lower than
result of the study of Kosinski (1890.9 vs 2052), but higher than
results of Crassard (1890.9 vs 1521) and Ghaffarpour (1890.9 vs
1380), while the mean concentration of D-dimer of the control
group in our study was lower than results of Kosinski (146.19 vs
375) and Ghaffarpour (146.19 vs 388). We also recognized that
the standard deviations of experimental groups were very high in
studies of Kosinski (1286), Crassard (938), and Ghaffarpour
(920).
Overall, the difference in D-dimer concentration of our
study to studies of other authors might be due to the following
reasons: First, because the nature of D-dimer concentration is to
indirectly evaluate the fibrin breakdown of thrombosis, while
clinical characteristics of thrombosis are varied then D-dimer
concentration also varying greatly; second, CVT is a rare disease
then sample sizes in these studies were not large enough leading
to representative for patients was not high; third, D-dimer
concentration may change depending on patient's age, testing
time, accompanied diseases and used techniques.
4.5.2. Identify the cut off of D-dimer test
22
In our study, as showed in table 3.35, the cut off of D-
dimer concentration (µg/L) ≥ 256; ≥ 280; ≥ 302; ≥ 424; ≥ 502; ≥
604. We recognized that: at the cut off of D-dimer concentration
≥ 302 µg/L the ROC curve value was best as compared to other
cut off. Then, at this cut off, the sensitivity, specificity, and area
under ROC curve were optimal (91.33%, 89.47% and 0.904%).
This result showed that our study was equivalent to other studies.
According to the studies of Kosinski, Crassard,

Ghaffarpour, Gouda, they chose the cut off ≥ 500 µg/L, with the
exception of the study of Tnislaz chose a low cut off ≥ 500 µg/L.
At this cut off ≥ 500 µg/L , the study of Kosinski would provide
the highest sensitivity and specificity (97.1% and 91.2%),
followed by the study of Tanislav with 90% sensitivity and 90%
specificity and area under ROC curve was 0.921. The study of
Gouda obtained sensitivity and specificity were 85.7% and
85.5%, respectively. The study of Ghaffarpour obtained low
sensitivity (83.1%) as compared to other authors in table 4.43 and
specificity was 95.2% respectively.
In summary, due to the epidemiological, clinical and
imaging characteristics of CVT are very varied and the D-dimer
concentration fluctuates greatly because it depends on many
factors as mentioned then it's hard to avoid the difference of
diagnostic cut off between studies.
CONCLUSIONS
23