Neurosurg Rev (2010) 33:359–366
DOI 10.1007/s10143-010-0251-z
ORIGINAL ARTICLE
D-dimer as a predictor of progressive hemorrhagic injury
in patients with traumatic brain injury: analysis of 194 cases
Heng-Li Tian & Hao Chen & Bing-Shan Wu & He-Li Cao &
Tao Xu & Jin Hu & Gan Wang & Wen-Wei Gao &
Zai-Kai Lin & Shi-Wen Chen
Received: 16 October 2008 / Revised: 8 October 2009 / Accepted: 2 January 2010 / Published online: 27 March 2010
# Springer-Verlag 2010
Abstract This study sought to describe and evaluate any
relationship between D-dimer values and progressive hemor-
rhagic injury (PHI) after traumatic brain injury (TBI). In
patients with TBI, plasma D-dimer was measured while a
computed tomography (CT) scan was conducted as soon as
the patient was admitted to the emergency department. A
series of other clinical and laboratory parameters were also
measured and recorded. A logistic multiple regression
analysis was used to identify risk factors for PHI. A cohort
of 194 patients with TBI was evaluated in this clinical study.
Eighty-one (41.8%) patients suffered PHI as determined by a
second CT scan. The plasma D-dimer level was higher in
patients who demonstrated PHI compared with those who did
not (P<0.001. Using a receiver–operator characteristic curve
to predict the possibility by measuring the D-dimer level, a
value of 5.00 mg/L was considered the cutoff point, with a
sensitivity of 72.8% and a specificity of 78.8%. Eight-four
patients had D-dimer levels higher than the cut point value
(5.0 mg/L); PHI was seen in 71.4% of these patients and in
19.1% of the other patients (P<0.01). Factors with P<0.2 on
bivariate analysis were included in a stepwise logistic
regression analysis to identify independent risk factors for
H.-L. Tian : H. Chen : B.-S. Wu (*) : H.-L. Cao : T. Xu : J. Hu :
G. Wang : W.-W. Gao : Z.-K. Lin
Department of Neurosurgery, Shanghai 6th People Hospital,
Shanghai Jiaotong University,
No. 600, Yishan Road, Xuhui District,
Shanghai, China 200233
e-mail: wubingshan@gmail.com
S.-W. Chen
Department of Radiology, Shanghai 6th People Hospital,
Shanghai Jiaotong University,
No. 600, Yishan Road, Xuhui District,
Shanghai, China
TBI coagulopathy. Logistic regression analysis showed that
the D-dimer value was a predictor of PHI, and the odds ratio
(OR) was 1.341 with per milligram per liter (P=0.020). The
stepwise logistic regression also identified that time from
injury to the first CT shorter than 2 h (OR=2.118, P=0.047),
PLT counts lesser than 100×109/L (OR=7.853, P=0.018),
and Fg lower than 2.0 g/L (OR=3.001, P=0.012) were risk
factors for the development of PHI. When D-dimer values
were dichotomized at 5 mg/L, time from injury to the first
CT scan was no longer a risk factor statistically while the OR
value of D-dimer to the occurrence of PHI elevated to 11.850
(P<0.001). The level of plasma D-dimer after TBI can be a
useful prognostic factor for PHI and should be considered in
the clinical management of patients in combination with
neuroimaging and other data.
Keywords Coagulation . D-dimer . Fibrinolysis .
Progressive hemorrhagic injury . Traumatic brain injury
Introduction
Traumatic brain injury (TBI) poses an important public
health problem. It is a major cause of morbidity and
mortality, while survivors often suffer cognitive or behav-
ioral disorders [20, 26]. Multiple reports have demonstrated
that abnormalities in blood coagulation commonly devel-
oped, which led to bleeding tendency in head trauma
patients [27]. It is clear that the concomitant of coagulation
disorders (however defined) to brain injury contributes
greatly to this morbidity and mortality [4, 25]. Numerous
literatures have shown coagulation abnormality to be an
important factor in various progressive brain injury includ-
ing ischemia, hemorrhage, and/or brain swelling, as well as
enlargement of previously small lesions [15, 16, 22, 24].
360
Bleeding disorders can be classified according to which of
the three hemostatic processes are involved: primary hemo-
stasis (platelet disorders); coagulation (clotting factor disor-
ders); and fibrinolysis (production of inhibitors). Some
disorders involve more than one process. Coagulation test
may also focus on determining whether the clinical problem
involves primary hemostasis (decreased platelet count, etc.),
coagulation [prolonged prothrombin time (PT) and activated
partial thromboplastin time (APTT), etc.], fibrinolysis (in-
creased D-dimer, etc.), or a combination of the three.
Computed tomography (CT) plays an important role in
the management of head-injured patients. The use of
repeated CT scans has made it possible to trace radiograph-
ic variances in brain injuries [5, 9, 24]. Regarding the
enlargement of previously seen hematomas, if the following
CT scan shows deterioration due to new hematomas or an
increase in hematomas present at the time of admission, a
diagnosis of progressive hemorrhagic injury (PHI) can be
made [21, 25].
Moreover, D-dimer, a coagulation end product, has been
shown to increase in peripheral blood in CNS diseases such
as subarachnoid hemorrhage [8], ischemic stroke [2], and
trauma [6], reflecting a systemic activation of coagulation.
In addition, an increased D-dimer production has been
related to poor clinical outcome in these conditions. The
hemorrhagic diathesis resulting from bleeding disorders
after head injury may be involved in PHI. The purpose of
this study was to evaluate any correlation between the
plasma level of D-dimer, a breakdown product in fibrino-
lysis, and the incidence of PHI.
Materials and methods
Patient population
This study was approved by the research ethical committee
of the 6th Affiliated Hospital of Shanghai Jiaotong
University. The files of patients with head injuries who
were admitted between January 1, 2006, and June 30, 2007,
were reviewed. During that year, more than 600 patients
with head injury were admitted to our hospital. Patients
delivered within 12 h whose highest abbreviated injury
score (AIS) was 3 or less (other than head injury) were
considered to be isolated TBI cases and were included.
Patients with known coagulation disorders, such as deep
venous thrombosis or pulmonary embolism, and those on
anticoagulant therapies that could result in coagulopathies
were excluded. Those who died in the emergency depart-
ment before a CT scan was performed, those who
deteriorated and died before a second CT scan was
performed, and those with incurable conditions were also
excluded as having incomplete clinical data. Accordingly,
Neurosurg Rev (2010) 33:359–366
we retrospectively reviewed the hospital records of the
remaining 194 patients (153 men, 41 women), all of whom
had at least two CT scans within 24 h of admission.
Patient management
As soon as the patients arrived at the emergency department,
treatments were given in accordance with the routine guide-
lines for the management of brain injury [3]. Following
clinical and CT evaluation within 4 h of admission,
intraventricular catheters were placed in some patients
presenting with a Glasgow Coma Scale (GCS) of 8 or below
to permit continuous monitoring of intracranial pressure
(ICP), as well as drainage of cerebrospinal fluid when the
ICP exceeded 20 mmHg. Demographic data, injury time,
mechanism of injury, systolic blood pressure, and GCS score
were documented when the patients arrived at the emergency
room. AIS scores were determined according to the 1990
revision of the Abbreviated Injury Scale [1].
All patients underwent a brain CT scan shortly after a
careful neurological evaluation and initial fluid resuscitation.
The second CT scan was routinely obtained within 24 h of
admission, as well as when indicated by clinical deterioration
or a rise in intracranial pressure. All CT scans were reported by
the staff of the radiology department without reference to
clinical or laboratory parameters. The volume of the lesion
was calculated following the formula A×B×C×0.5, where A
and B represent the largest perpendicular diameters through
the hyperdense area on CT scan, and C represents the
thickness of the lesion (the number of 8-mm slices containing
hemorrhage multiply to 0.8) [10]. PHI was defined as the
appearance of new lesion(s) or a conspicuous increase in the
size of hemorrhagic lesion(s), i.e., a 25% increase or more
versus the first postinjury CT scan [17, 21].
Neurological outcome was also recorded using Glasgow
Outcome Scale at 6 months postinjury.
Laboratory analysis
Blood samples were drawn on admission from patients'
antecubital vein by means of venous puncture, collected
into tubes containing different anticoagulant for coagulation
test or blood routine test. Sodium citrate at a ratio of 1
volume to 9 volumes of blood was served as anticoagulant
for coagulation parameters. The plasma D-dimer level was
evaluated with NycoCard® Reader II (Axis-Shield PoC,
Oslo, Norway), using a gold-antibody conjugate in an
immunofiltration test principle which digitally displayed the
concentration in milligrams per liter. D-dimer reference
value was 0–0.30 mg/L, as recommended by the manufac-
turer. Before blood was drawn, no blood product was used
and no operative procedures were performed. Platelet (PLT)
counts and coagulation tests, including PT (reference value,
Neurosurg Rev (2010) 33:359–366 361

Table 1 Clinical data for 194 patients with traumatic brain injury continuous variables, as median with quartile range (25th,
Clinical data 75th) for the lack of a normal distribution and as percentages
for categorical variables. PHI was dichotomized as positive or
Number of patients 194 negative. And, the patients were divided into two groups,
Mean age, years 43.9±15.4 accordingly. The demographic and clinical characteristics
Gender (M/F) 153/41 between groups were evaluated using bivariate analysis. For
Median GCS score at admission (range) 10 (3–15) continuous variables, t tests were used to compare means for
Median time from injury to the first CT scan, h 1.9 (1.0–3.0) normal data, Mann–Whitney U tests were used to compare
Median time between the first and second CT scan, h 9.0 (5.0–14.4) non-normal distributed data. The Pearson chi-square test was
Cause of injury (number of cases) used for categorical variables; Fisher's exact test and
Motor vehicle accident (%) 131 (67.5) continuity correction were used where appropriate. A
Fall (%) 36 (18.6) backward stepwise logistic regression was applied to
Assault (%) 18 (9.3) determine the predictors for PHI. All variables in the
Other (%) 9 (4.6) bivariate analysis with two-sided P values of ≤0.20 were
candidates for inclusion in the logistic regression model to
assess which variables were associated with PHI. The
relationship between PHI and D-dimer was investigated with
11–14 s), APTT (reference value, 28–40 s), and fibrinogen the final multivariate logistic model. Statistical significance
(Fg; reference range, 2.0–4.0 mg/L) were measured for all was assumed for P values of less than 0.05. We used a
patients, using the routine laboratory assay at the Central receiver–operator characteristic curve (ROC) to assess the
Clinical Chemistry Laboratory of Shanghai 6th Hospital. accuracy of predictions of PHI from the D-dimer values.
The international normalized ratio (INR; reference range, Using the ROC curve, a D-dimer cutoff value and its
0.8–1.2) was also reported and was used as a surrogate for confidence in prognosis could be estimated based on the area
PT to account for institutional variability in the measure- under the ROC curve (AUC).
ment of this marker. PLT count below 100×109/L was
defined as thrombocytopenia. A prolongation of 3 s and 8 s
to the reference value of PT and APTT was regarded as Results
abnormality, respectively. An INR greater than 1.4 was also
considered as PT abnormal. In this study, a cohort of 194 patients (153 men and 41
women; mean age, 43.9±15.4 years) was studied. The GCS
Statistical analysis score ranged from 3 to 15, with median score of 10. The
median time from injury to the first CT scan and the median
Data were analyzed using the Excel (Microsoft Corp., Red- time between the first and second CT scan were 1.9 (1.0,
mond, WA) and SPSS (SPSS, Inc., Chicago, IL) software. 3.0 h) and 9.0 (5.0, 14.4 h), respectively. Their clinical
Values were reported as mean±standard deviation for normal characteristics are summarized in Table 1.

Table 2 Clinical factors of PHI

and non-PHI patients: Clinical factors PHI P value
continuous parameter
Positive, n=81 Negative, n=113

Normal distributiona
Age, years 48.5±14.7 40.6±15.1 0.000
Fg, g/L 2.33±0.62 2.60±0.65 0.004
Non-normal distributionb
Time from injury to 1st CT, h 1.5 (1.0–2.0) 2.4 (1.2–3.6) 0.000
Time from 1st to 2nd CT, h 8.8 (4.7–14.2) 9.0 (5.5–14.7) 0.326
Systolic blood pressure, mmHg 126 (115–140) 124 (117–132) 0.134
PLT, 109/L 154 (131–201) 190 (157–236) 0.000
PT, s 13.0 (12.0–14.1) 12.4 (11.8–13.1) 0.003
PTINR 1.04 (0.98–1.12) 1.00 (0.94–1.08) 0.008
a
t test, mean±SD APTT, s 26.7 (23.8–34.2) 25.7 (22.3–28.4) 0.001
b
Mann–Whitney U test, quartile D-dimer, mg/L 6.60 (4.30–10.75) 2.10 (1.20–4.20) 0.000
(25th, 75th)
362 Neurosurg Rev (2010) 33:359–366

Table 3 Clinical factors of PHI

and non-PHI patients: Clinical factors PHI P value
categorical parameter
Positive, n=81 Negative, n=113

Gender (male) 41.2 (63) 58.8 (90) 0.753

Time from injury to 1st CT (<2 h) 61.7 (50) 41.6 (47) 0.006
Admission GCS score 0.005
3–8 59.3 (48) 36.3 (41)
9–12 19.8 (16) 26.5 (30)
13–15 21.0 (17) 37.2 (42)
Hypotensiona 3.7 (3) 0.9 (1) 0.395
Thrombocytopenia 13.6 (11) 1.8 (2) 0.003
PT abnormal 7.4 (6) 0 (0) 0.012
APTT abnormal 4.9 (4) 0 (0) 0.061
a
Hypotension was defined as Fg abnormal 33.8 (27) 14.2 (16) 0.001
systolic blood pressure lower than D-dimer (>5 mg/L) 74.1 (60) 21.2 (24) 0.000
90 mmHg

Of the 194 patients, 81 (41.8%) demonstrated PHI on the
second CT scan. Bivariate analysis was performed to
identify risk factors for PHI, and these findings are shown
in Tables 2 and 3. Patients with PHI were older, had a
shorter time from injury to the first CT scan, a lower GCS
score, a lower PLT count, a prolonged PT and APTT, and a
lower Fg value at admission (Table 2). In accordance with
the definition mentioned above, patients with hypotension,
thrombocytopenia, and abnormal PT, APTT, and Fg were
more likely to have PHI than patients who were not
(Table 3).
D-dimer value was tested; the maximum and minimum
values were 0.20 and 22.30 mg/L, respectively; the median
value was 3.75 mg/L, and quartile range was 1.60 and
7.15 mg/L. The plasma D-dimer level was higher in
patients who demonstrated PHI compared with those who
did not (P<0.001, see Table 2). Patients who suffered PHI
had worse outcome (Table 4, χ2 =75.003, P<0.001).
The ROC curve is a plot of the sensitivity of a test versus
its false-positive rate for all possible cutoff points. Using
the ROC and AUC, we found a reliable operating point for
the D-dimer level to be 5.00 mg/L, with a sensitivity of
72.8% and a specificity of 78.8%; the AUC was 0.816 [P<
0.001, 95% confidence interval 0.756–0.876; Fig. 1]. There
were 84 patients who had a D-dimer level higher than
5.00 mg/L; PHI was seen in 71.4% of these patients and in
19.1% of the other patients (P<0.01).
The results of the binary logistic regression are summarized
in Table 5. D-dimer was a predictor of PHI, and the odds
ratio (OR) was 1.341 with per milligram per liter (P=0.020,
Table 5). The stepwise logistic regression also identified that
time from injury to the first CT shorter than 2 h (OR=2.118,
P=0.047), PLT counts lesser than 100×109/L (OR=7.853,
P=0.018), and Fg lower than 2.0 g/L (OR=3.001, P=0.012)
were risk factors for the development of PHI. When D-dimer
values were dichotomized at 5 mg/L, time from injury to the
first CT scan was no longer a risk factor statistically while
the OR value of D-dimer to the occurrence of PHI elevated
to 11.850(P<0.001, Table 6).
Discussion
Our study demonstrated that D-dimer is a predictor for the
clinical occurrence of PHI in patients with TBI. Time from
injury to the first CT, thrombocytopenia, and plasma Fg value
also showed a prognostic value in PHI. The development of
coagulation abnormalities following head injury has been
recognized [6, 8, 25], consisting of the process leading to the
conversion of fibrinogen to fibrin, an insoluble polymer,
which forms a blood clot with trapped cellular elements. In
the presence of cross-linked fibrin, D-dimer, a breakdown
product of fibrin in the plasma, is produced by the activation
of thrombin and the plasmin hybrid effect. With up-

Table 4 Relationship of
Glasgow outcome scale to PHI GOS PHI P value

Positive, n=81 Negative, n=113

Unfavorable outcome (GOS=1, 2, 3) 64.2 (29) 6.2 (7) 0.000

Favorable outcome (GOS=4, 5) 35.8 (52) 93.8 (106)
Neurosurg Rev (2010) 33:359–366 363

ROC Curve Table 6 Multivariate logistic regression to identify independent risk

1.0 factors for PHI—D-dimer value dichotomized at 5 mg/L

Factors OR value 95% CI P value

0.8 Thrombocytopenia 4.247 1.095–37.826 0.039

Fg abnormal 3.209 1.318–7.811 0.010
D-dimer (>5 mg/L) 11.850 5.672–24.758 0.000
0.6
Sensitivity

Nagelkerke R square=0.452

0.4 aiding in the diagnosis of disseminated intravascular

0.2
0.0
0.0 0.2 0.4 0.6 0.8 1.0
1 - Specificity
Fig. 1 Receiver–operator characteristic curve (ROC) to predict the
possibility of PHI by measuring the level of D-dimer. The value of
5.00 mg/L was considered the cutoff point, with the sensitivity and
specificity of 72.8% and 78.8%, respectively
regulation of fibrinolysis, D-dimer was elevated at admission
as a response to enhanced coagulation activity. Correlated
with other parameters of coagulation and fibrinolysis, D-
dimer reflected approvingly the overall up-regulation of the
hemostatic system at admission [12].
If the D-dimer level is elevated, the coagulation disorder
appears; this is considered a poor prognostic factor. High
D-dimer levels reflect the severity of brain injury [11, 12].
Systemic activation of coagulation leads to widespread
intravascular deposition of fibrin and depletion of PLTs and
coagulation factors. So, we found in our study that the PLT
count and Fg level degraded, and thrombocytopenia and
low Fg level could to be predictors of PHI. This finding is
in line with previous work [7].
An increased D-dimer level has been related to poor
clinical outcome in conditions of subarachnoid hemorrhage,
ischemic stroke, and trauma [2, 6, 8]. Moreover, D-dimer
level reflects plasmin activity not only in blood but also
during tissue degradation and remodeling. In addition to
coagulation (DIC), it is also of clinical use when a
suspicion of deep vein thrombosis or pulmonary embolism
(PE) exists; it is promising as an exclusion test for PE if the
results are negative, while positive results are quite
nonspecific [18, 29]. Nevertheless, whether D-dimer levels
correlate with the incidence of posttraumatic cerebral
infarction, which often develops in patients with moderate
or severe traumatic brain injury, has not been previously
reported [28].
Routine D-dimer laboratory testing seems a suitable and
useful clinical tool to assess the severity of head injury.
Monitoring of D-dimer level may be particularly valuable
to estimate prognosis in patients who are transferred early
after injury and who are not conveniently given only
clinical condition and radiological variables. Sometimes,
patients with mild head injury who are discharged and told
to watch for further symptoms at home may experience
unpredictable neurological deterioration. Therefore, the
blood D-dimer level can provide useful prognostic infor-
mation for physicians concerning whether to transfer
patients to a facilitated hospital.
Although we found D-dimer level associated with hema-
toma growth, the mechanisms that explain the association
between D-dimer and PHI are unclear. D-Dimer, as a marker
of fibrin turnover, might reflect an impaired coagulation and
fibrinolysis pathways. In TBI patients, this abnormal function
may lead to greater intracranial hematoma volume and early
hemorrhagic growth, as we observed. However, D-dimer is
one of the markers of hemostatic function which are acute-
phase reactants. Hence, it is possible that elevated D-dimer
levels in patients with progressing hemorrhage are simply a
marker of a more severe lesion, as part of a reactant
inflammatory process. We may not be able to exclude the

Table 5 Multivariate logistic

regression to identify indepen- Factors OR value 95% CI P value
dent risk factors for PHI—
continuous D-dimer value Time from injury to 1st CT (<2 h) 2.118 1.012–4.433 0.047
Thrombocytopenia 7.853 1.418–43.495 0.018
Fg abnormal 3.001 1.270–7.092 0.012
D-dimer, mg/L 1.341 1.206–1.492 0.020
Nagelkerke R square=0.426
364
possibility of an acute-phase reaction in cases of more severe
brain injury and greater hemorrhage.
The finding that severe patients had a higher D-dimer level
is interesting. Proinflammatory states in critically ill hospital-
ized patients lead to elevated D-dimer levels via cytokine
activation of the coagulation cascade [30]. When the blood is
clotting, thrombin and other intermediate products are
generated. Some of these products, and in particular
thrombin, can activate the inflammatory cascade [13]. D-
dimer as an end product of both coagulation and fibrinolysis
may trigger some deleterious actions directly. D-dimer itself
may stimulate monocyte synthesis and release of proinflam-
matory cytokines such as interleukin-6 [19], which has been
related to both development of edema and hematoma
enlargement. Thus, activated inflammation and activated
blood coagulation may be related and both contribute to the
occurrence of PHI. Additionally, severe head-injured patients
with high D-dimer concentration may suffer organ failure
resulted from microthrombosis. Liver failure may induce the
reduction of blood coagulation factor generation and severe
bleeding may result [14]. This agrees with our findings.
Some 63.7% severe patients underwent PHI while the
percentage in mildly injured patients was only 21.9%.
Delayed and progressive posttraumatic hemorrhage has
long been linked to abnormal clotting onset time. Our study
found that all of the patients with abnormal PT or APTT
suffered PHI, while the non-PHI group had no patient with
a prolonged PT or APTT. Engstrom et al. established that
thrombocytopenia was a risk factor for PHI in TBI, and
those patients with PHI had a more affected coagulation
system than patients without. Moreover, although they
found no significant difference between the groups in PLT
count at admission, they found significant differences in
terms of decreased PLT counts, particularly those below
150×106/L [7]. We found the same in our study, and in our
cohort of patients, significant differences in PLT counts
were observed at admission. Patients with severe depletion
of PLTs and coagulation factors tend to be subject to diffuse
and ongoing bleeding or to be at high risk for bleeding.
Antifibrinolytic treatment is effective in patients with
bleeding, but the use of these agents in patients with high
levels of D-dimer, which may indicate a state of DIC, is not
recommended. Because the deposition of fibrin in this
disorder appears to be due in part to insufficient fibrinoly-
sis, further inhibition of the fibrinolytic system would not
seem to be appropriate. Anticoagulants, however, have not
yet been tested in any controlled clinical trial; the relatively
high risk of bleeding associated with the use of these
compounds may be a limiting factor.
The experience with recombinant activated factor VII
(rFVIIa) has increased; studies show that the administration
of rFVIIa is safe and cost-effective [23, 31]. Although the
early use of rFVIIa in head-injured patients can rapidly
Neurosurg Rev (2010) 33:359–366
reverse coagulopathy and reduce the occurrence of enlarge-
ment of contusions, the requirement of further operation, and
adverse outcome, the indication of using rFVIIa remains
controversial [23, 31]. There were no patients who received
rFVIIa in our study; so, we cannot assess the effects of using
rFVIIa in preventing PHI. By the same token, heparin, which
may produce a reverse effect where bleeding is the main
clinical problem, was not given in this study either.
The use of a serial CT scan in patients with head injuries
can disclose new or progressive abnormalities compared
with the scan performed at admission. However, most
studies do not recommend routine early follow-up CT
examinations in patients with head injuries in the absence
of other clinical changes [5, 9]. Our study determined a
high risk of PHI when the plasma D-dimer level was above
5.00 mg/L; such a value alone may be an appropriate
reason for a follow-up CT scan.
The main limitation of this study was its retrospective
nature. This might have introduced a significant bias in
patient selection and data collection. PHI was diagnosed
according to the second CT scan comparing with the first
one. Many mildly injured patients had been excluded for
the lack of further clinical data including second CT scan,
so we could only calculate detection rate rather than
incidence. Additionally, not all the patients received
intraventricular catheters, so data on ICP was incomplete
and excluded. Therefore, prospective controlled studies are
needed.
Conclusions
Our study demonstrated that the occurrence of PHI can be
predicted within hours after injury based on many clinical and
laboratory findings. A high level of D-dimer indicates
fibrinolysis, and a compensatory reaction to hypercoagulation
at the time of admission is particularly likely to presage the
development of PHI. Patients with high levels of D-dimer and
other high-risk findings should be further examined for causes
of PHI and to enable rapid intervention.
References
1. Association for the Advancement of Automotive Medicine (1990)
The Abbreviated Injury Scale, 1990 Revision. Association for the
Advancement of Automotive Medicine, Des Plaines, IL
2. Barber M, Langhorne P, Rumley A, Lowe GD, Stott DJ (2004)
Hemostatic function and progressing ischemic stroke: D-dimer
predicts early clinical progression. Stroke 35:1421–1425
3. Bullock R, Chesnut RM, Clifton G, Ghajar J, Marion DW, Narayan
RK (1996) Guidelines for the management of severe head injury.
Brain Trauma Foundation. Eur J Emerg Med 3:109–127
4. Carrick MM, Tyroch AH, Youens CA, Handley T (2005)
Subsequent development of thrombocytopenia and coagulopathy
Neurosurg Rev (2010) 33:359–366
in moderate and severe head injury: support for serial laboratory
examination. J Trauma 58:725–730
5. Chao A, Pearl J, Perdue P, Wang D, Bridgemann A, Kennedy S et
al (2001) Utility of routine serial computed tomography for blunt
intracranial injury. J Trauma Inj Infect Crit Care 51:870–876
6. Delgado P, Alvarez-Sabin J, Abilleira S, Santamarina E, Purroy F,
Arenillas JF et al (2006) Plasma D-dimer predicts poor outcome
after acute intracerebral hemorrhage. Neurology 67:94–98
7. Engstrom M, Romner B, Schalén W, Reinstrup P (2005)
Thrombocytopenia predicts progressive hemorrhage after head
trauma. J Neurotrauma 22:291–296
8. Juvela S, Siironen J (2006) D-dimer as an independent predictor
for poor outcome after aneurysmal subarachnoid hemorrhage.
Stroke 7:1451–1456
9. Kaups KL, Davis JW, Parks SN (2004) Routinely repeated
computed tomography after blunt head trauma: does it benefit
patients? J Trauma Inj Infect Crit Care 56:475–481
10. Khotari RU, Brott T, Broderick JP, Barsan WG, Sauerbeck LR,
Zuccarello M et al (1996) The ABCs of measuring intracerebral
hemorrhage volume. Stroke 27:1304–1305
11. Kuo JR, Chou TJ, Chio CC (2004) Coagulopathy as a parameter
to predict the outcome in head injury patients—analysis of 61
cases. J Clin Neurosci 11:710–714
12. Kuo JR, Lin KC, Luc L, Lin HJ, Wang CC, Chang CH (2007)
Correlation of a high D-dimer level with poor outcome in
traumatic intracranial hemorrhage. Eur J Neurol 14:1073–1078
13. Lee K, Kawai N, Kim S, Sagher O, Hoff JT (1997) Mechanisms
of edema formation after intracerebral hemorrhage: effects on
thrombin on cerebral blood flow, blood-brain barrier permeability,
and cell survival in a rat model. J Neurosurg 86:272–278
14. Levi M, Ten CH (1999) Disseminated intravascular coagulation.
N Engl J Med 341:586–592
15. Lobato RD, Sarabia R, Cordobes F (1988) Posttraumatic cerebral
hemispheric swelling. Analysis of 55 cases studied with comput-
erized tomography. J Neurosurg 68:417–423
16. Mirvis SE, Wolf AL, Numaguchi Y (1990) Posttraumatic cerebral
infarction diagnosed by CT: prevalence, origin, and outcome. Am
J Roentgenol 154:1293–1298
17. Oertel M, Kelly DF, Mcarthur D, Boscardin WJ, Glenn TC, Lee
JH et al (2002) Progressive hemorrhage after head trauma:
predictors and consequences of the evolving injury. J Neurosurg
96:109–116
18. Rathbun SW, Whitsett TL, Vesely SK, Raskob GE (2004) Clinical
utility of D-dimer in patients with suspected pulmonary embolism
and nondiagnostic lung scans or negative CT findings. Chest
125:851–855
19. Robson SC, Shepard EG, Kirsch RE (1994) Fibrin degradation
products D-dimer induces the synthesis and release of biologically
active IL-1B, IL-6 and plasminogen activator inhibitors from
monocytes in vitro. Br J Haematol 86:322–326
20. Rutland-Brown W, Langlois JA, Thomas KE, Xi YL (2006)
Incidence of traumatic brain injury in the United States. J Head
Trauma Rehabil 21:544–548
21. Sanus GZ, Taner T, Ilker A, Sabri A, Mustafa U (2004) Evolving
traumatic brain lesions: predictors and results of ninety-eight
head-injured patients. Neurosurg Q 14:97–104
22. Servadei F, Nanni A, Nasi MT, Zappi D, Vergoni G, Giuliani G et al
(1995) Evolving brain lesions in the first 12 hours after head injury:
analysis of 37 comatose patients. Neurosurgery 37:899–907
23. Stein DM, Dutton RP, Kramer ME, Scalea TM (2009) Reversal of
coagulopathy in critically ill patients with traumatic brain injury:
recombinant factor VIIa is more cost-effective than plasma. J
Trauma 66:63–72, discussion 73-5
24. Stein SC, Spettell C, Young G, Ross SE (1993) Delayed and
progressive brain injury in closed-head trauma: radiological
demonstration. Neurosurgery 32:25–31
365
25. Stein SC, Young GS, Talucci RC, Greenbaum BH, Ross SE
(1992) Delayed brain injury after head trauma: significance of
coagulopathy. Neurosurgery 30:160–165
26. Tagliaferri F, Compagnone C, Korsic M, Servadei F, Kraus J
(2006) A systematic review of brain injury epidemiology in
Europe. Acta Neurochir 148:255–267
27. Talving P, Benfield R, Hadjizacharia P, Inaba K, Chan LS,
Demetriades D et al (2008) Coagulopathy in severe traumatic
brain injury: a prospective study. J Trauma 66:55–61, discussion
61-62
28. Tian HL, Geng Z, Cui YH, Hu J, Xu T, Cao HL et al (2008) Risk
factors for posttraumatic cerebral infarction in patients with
moderate or severe head trauma. Neurosurg Rev 31:431–437
29. Wells PS, Anderson DR, Rodger M, Forgie M, Kearon C, Dreyer
J et al (2003) Evaluation of D-dimer in the diagnosis of suspected
deep-vein thrombosis. N Engl J Med 349:1227–1235
30. Williams MT, Aravindan N, Wallace MJ, Riedel BJ, Shaw AD
(2003) Venous thromboembolism in the intensive care unit. Crit
Care Clin 19:185–207
31. Zaaroor M, Soustiel JF, Brenner B, Bar-Lavie Y, Martinowitz U,
Levi L (2008) Administration off label of recombinant factor-VIIa
(rFVIIa) to patients with blunt or penetrating brain injury without
coagulopathy. Acta Neurochir (Wien) 150:663–668, Epub 2008
May 12
Comments
Jorge Humberto Tapia-Pérez, Magdeburg, Germany
In this study, the authors tried to explain the role of D-dimer as
predictor for progression of hemorrhagic injury (PHI) in trauma
patients. The sufficient number of patients could have been shown
more detail; nevertheless, with the given data, many interesting
conclusions are obtained. The patients with PHI displayed more
frequent abnormalities in coagulation parameters; based on these
observations, it could be expressed that a kind of coagulopathy is
associated. Pathophysiologic considerations allow us, assuming that
inflammation and endothelial injury are contributing. Recently, a
study from Canada showed that coagulation abnormalities detected in
routine laboratory tests in the first hours are predictors of hemorrhagic
progression [1]. This study provides support to this observation. The
causal relationship between coagulopathy and progression will require
further studies.
The relevance as therapeutic target has been described by Narayan
et al. [2] in their study about factor VIIa. A consideration is that the
delay or limitation of PHI could not redound in a clinical
improvement, despite the association of bad outcome described by
Tian et al. In this way, targeting traumatic coagulopathy could be
just an additional therapy, if the future studies do not demonstrate an
increased risk of thrombotic events.
The first regression model presented showed the time to first CT scan
as risk factor for PHI, but it should not necessarily be considered as that.
An early CT scan could not detect lesions because some of them do not
develop yet. It is probably an evidence of a natural PHI after trauma,
which is linked to coagulation disorders; further studies are needed.
The authors' conclusion provides us an easy clinical tool.
Reasonably, in patients with severe head injury, admitted early, and
with abnormalities in coagulation (especially DD) can the perfor-
mance of a CT scan be very useful. Evidently, the therapeutic decision
must be taken in context of each case.
References
1. Allard CB, Scarpelini S, Rhind SG, Baker AJ, Shek PN, Tien H,
Fernando M, Tremblay L, Morrison LJ, Pinto R, Rizoli SB. Abnormal
366 Neurosurg Rev (2010) 33:359–366

coagulation tests are associated with progression of traumatic
intracranial hemorrhage. J Trauma 2009;67:959–967.
2. Narayan RK, Maas AI, Marshall LF, Servadei F, Skolnick BE,
Tillinger MN; rFVIIa Traumatic ICH Study Group. Recombinant
factor VIIA in traumatic intracerebral hemorrhage: results of a dose-
escalation clinical trial. Neurosurgery 2008 ;62:776–778.
Ignacio J. Previgliano, Buenos Aires, Argentina
This paper by Tian et al. is very interesting in addressing the issue
of bleeding disorders following head injury.
Although based on retrospective data, their finding that D-dimer at
a cutoff point of 5 mg/l with a sensitivity of 72.8% and a specificity of
78.8% is an important clue for further investigation.
The authors gave substantial bibliographic support for the use of
D-dimer as a prognostic toll of progressive hemorrhagic injury in
patients with traumatic brain injury. Most of their findings showed that
patients with D-dimer above 5 mg/l were older and had a severe head
injury according to the initial GCS, so they underwent early to
diagnostic studies. They also pointed out in the paper that D-dimer is
augmented in many acute neurological disorders mainly to the
activation of the coagulation process due to the liberation of tissue
factor.
Nevertheless, some important points should be highlighted:
(a) In the logistic regression model, only patients with determi-
nations above the cutoff point were included.
(b) It was not possible to identify the exact coagulation problem in
most of the study population.
(c) Regarding this dilution, coagulopathy and disseminated
intravascular coagulation (DIC) diagnosis should be ruled out for the
reason that a different therapeutic approach is needed.
The International Society on Thrombosis and Haemostasis devel-
oped a simple scoring system for the diagnosis of overt DIC (see
Table 1). A score of 5 or greater indicates overt DIC, whereas a score
of less than 5 does not rule out DIC, but may indicate non-overt DIC.
Studies have demonstrated the DIC score to be 93% sensitive and
98% specific for DIC.

Laboratory Platelet count

coagulation tests D-dimer and FDPs
Fibrinogen
PT and aPTT

Scoring Platelet count: >100 = 0 points, <100 = 1 point, <50 = 2 points

Elevated fibrin marker: No elevation = 0 points, moderate increase =
2 points, strong increase = 3 points
Prolonged PT: <3 sec = 0 points, >3 <6 = 1 point, >6 = 2 points
Fibrinogen level: >1 g/L = 0 points, <1 = 1 point

Calculate score Greater than or equal to 5 = compatible with overt DIC, repeat
scoring daily
Less than 5 suggestive of non-overt DIC

Table 1. DIC scoring system. Modify from Taylor Jr FB, Thromb
Haemost 2001;86:1327–1330.
I think this is a good tool for the neurosurgeon to confront acute
bleeding disorders in head injury patients.
As stated above, this paper is a good starting point for further
research in such important issue as the prognosis of delayed
hematomas.
Susanne Mink, Zurich, Switzerland
It still remains challenging to find outcome predictors for
progression of acute hemorrhagic injury (PHI) of traumatic brain-
injured patients. In this comprehensive article, the authors summarize
their experience with D-dimer levels in nearly 200 patients with
traumatic brain injury. The results and conclusion of the study are very
interesting and may be one part of the facts for clinical judgement and
therapeutic decisions in the acute state of injury. A relevant limitation
of the study represents the retrospective character. However, more
detailed and additional information should be addressed in further
prospective studies. Thereby, subgroup analysis and enhanced CT
time management would be of interest in these cases. Early CT scans
can miss lesions which are not developed at the time of examination.
This leads to the necessity for definition of exact time intervals of D-
dimer plasma levels and further coagulation parameters. Sawamura et
al. (1) described the parameters of disseminated intravascular
coagulation in general traumatic patients in more detail, considering,
additionally, ratios to be more specific and significant in outcome
prediction. Generally, contributing factors in PHI are endothelial
damage and associated inflammation which seek still for causal
relationship between coagulopathy and the PHI.
Reference
1. Sawamura A, Hayakawa M, Gando S, et al. Disseminated
intravascular coagulation with a fibrinolytic phenotype at an early
phase of trauma predicts mortality. Thromb Res 2009;124:608–613.