Risk factors and early diagnosis of cerebral venous sinus occlusion secondary to traumatic brain injury
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.170067
Source of Support: None, Conflict of Interest: None
Objective: This study was designed to investigate the risk factors of, and the strategy for early diagnosis of cerebral venous sinus occlusion (CVSO) secondary to traumatic brain injury.
Keywords: Cerebral venous sinus occlusion; risk factor; skull fracture; traumatic brain injury
A traumatic brain injury (TBI) can trigger venous sinus thrombosis., Factors such as a hematoma or a fracture can compress a venous sinus and lead to venous sinus occlusion., A thrombosis within leading to occlusion of the venous sinuses is referred to as cerebral venous sinus occlusion (CVSO). CVSO is divided into the compression type, thrombosis type and mixed type. Local venous sinus occlusion can influence the blood flow in the upstream or downstream venous sinuses leading to further occlusion of these sinuses. Clinical symptoms of most of the patients having a CVSO are mild. However, a minority of patients with CVSO present with malignant intracranial hypertension, and some even have hemorrhagic infarction, cerebral hemorrhage or edema, which are difficult entities to deal with. Another factor in traumatic brain injury that may increase the chances of CVSO is the use of anti-fibrinolytic drugs. These agents can inhibit the plasminogen activator induced fibrinolysis, thus, playing a role in hemostasis. As a result, antifibrinolytic medications have been used for treating intracerebral hemorrhage. A previous study has shown that blood viscosity increases after treating patients suffering from intracerebral hemorrhage with antifibrinolytic drugs, thus leading to blood stasis and an increased risk of cerebral ischemia., Wang et al., considered that improper use of hemostatic drugs is closely linked to CVSO. However, there are different opinions regarding the association of hemostatic agents with CVSO. It is reported that the adverse reactions caused by antifibrinolytic drugs and thrombin are mainly mild and deep vein thrombosis (DVT) and severe adverse events are seldom observed. A recent study showed that after a single dose of tranexamic acid at an early stage in patients with a severe TBI, the mortality was significantly lower, and no obvious complications were observed.
Cerebral venous sinus occlusion can lead to abnormal intracranial hemodynamics and results in severe neurological deficits, thus affecting the prognosis of the patients.,, Detection of CVSO as early as possible is helpful in starting an early treatment and can improve the prognosis of the patients. However, the symptoms of CVSO are often obscured by the symptoms associated with TBI, often making it difficult to achieve an early diagnosis and to institute an early treatment. This study attempts to analyze the risk factors precipitating a CVSO secondary to TBI, and also to explore the strategies for its early diagnosis.
A total of 240 patients having a TBI were admitted to our department from August 2011 to May 2013. Among the 240 cases, 197 cases were male and 43 cases were female. Their age ranged from 16 to 60 years, with the mean age being 39 years. The inclusion criteria included: Patients with moderate or severe closed TBI; patients in whom treatment was initiated in our hospital within 72 h after sustaining the injury; and, patients who were healthy without prior history of lower extremity deep vein thrombosis (DVT) or pulmonary embolism. The exclusion criteria were: Patients who were unsuitable for venous angiography; or, patients who were in a state of shock in the early period after sustaining the TBI; and, women who were pregnant at the time of sustaining TBI.
The patients underwent a plain computerized tomography (CT) scan after the TBI; and, also a CT venogram (CTV) or magnetic resonance venogram (MRV) as soon as possible once their condition stabilized. If the pathology could not be diagnosed even after the patient had undergone these investigations, a digital subtraction angiography (DSA) was also performed. The diagnostic criteria for CVSO were the presence of a discontinuous flow in the venous sinus or an obvious filling defect revealed by the cerebral computed tomographic venous angiography (CTV) or three-dimensional contrast-enhanced MRV (3D CE-MRV). Patients with coexisting congenital venous sinus dysplasias were carefully identified and excluded.
A plain computed tomography (CT) scan was carried out using the Philips Brilliance iCT Image Gallery (Philips International B.V., Amsterdam, The Netherlands). The scan was performed from the calvarium to the skull base using the following parameters: Slice thickness of 1.0 mm, voltage 120 KV, current 150 mAs, window width 100 HU and window level 50 HU. The presence of a local skull fracture crossing the sinus, natural cranial sutural diastasis and epidural hematoma were evaluated on axial CT images.
Computed tomographic venogram (CTV) was performed with the same scanner and the same scanning parameters as the plain CT scan. The scanning parameters were as follows: Slice thickness 1 mm, reconstruction thickness 0.45 mm, pitch 1, voltage 120 KV and current 400 mAs. The contrast agent (Bayer HealthCare China, Guangzhou, China; 300 mgI/ml) was injected intravenously with a high-pressure syringe. The amount of the contrast agent was 1.0–1.5 ml/Kg, the injection rate was 5 ml/s and the scanning delay time was 36 s.
Siemens Trio Tim 3.0T superconducting magnetic resonance scanner (Siemens Medical Solutions, Erlangen, Germany) was used for MR imaging. The imaging method of three dimensional fast low angle (FLASH) sequence scanning (TR/TE: 2.6/1.1ms; flip angle: 20°; bandwidth: 930 Hz/pixel; Voxel: 1.2 mm × 1.1 mm × 1.2 mm; TA: 30 s) was taken for 3D CE-MRV.
Cerebral venous sinus occlusion severity scoring
The major intracranial dural sinuses were divided into 16 parts according to the method adopted for analyzing the intracranial venous sinuses by Sun et al. The superior sagittal sinus was divided into three sections, including the frontal 1/3, the central 1/3, and the posterior 1/3. The torcula was considered as 1 section. Each side of the transverse sinus was divided into the medial 1/2 and the lateral 1/2, with 4 sections of bilateral transverse sinuses in total. Each side of the sigmoid sinus was divided into a vertical segment and a horizontal segment. The straight sinus was considered as 1 section, the inferior sagittal sinus as 1 section, the jugular foramen including the jugular bulb and the proximal internal jugular vein as 1 section, and the great cerebral vein as 1 section. Each involved venous sinus was given 1 point, and the number of involved venous sinuses in a patient decided his/her severity score.
Criteria for the classification of cerebral venous sinus occlusion
Fujii et al., classified CVSO into the "compression" and "thrombosis-occlusion" types according to the condition that was responsible for the venous sinus occlusion based on the CTV findings. In this study, while the subgroups of the patients being studied were based on the classification of Fujii et al., the classification was modified after considering the findings of the entire range of the plain CT scan, CTV and MRV imaging data. CVSO in our study was divided into three types: Type I consisted of the "thrombosis occlusion" type, in which venous sinus occlusion was caused by venous sinus thrombosis; Type II consisted of the "compression" type, in which the epidural hematoma crossing a sinus, or a skull fracture crossing a sinus compressed it and led to venous sinus occlusion; and, Type III was the mixed type, in which venous sinus occlusion was caused by both thrombosis and compression within the venous sinus.
Based on the previous reports ,, and clinical observations, seven risk factors were included in this study for further statistical analysis. These seven risk factors included a cranial fracture or a skull sutural separation across the sinus (that included an intracranial venous sinus or the jugular bulb); an epidural hematoma directly compressing the venous sinus; the Glasgow Coma Scale (GCS) score at admission; the presence of transtentorial herniation; intravenous use of hemostatic drugs; past history of venous thromboembolism, including lower extremity venous thrombosis as well as pulmonary embolism; and, the type of TBI sustained. Based on the duration of coma as well as the neurological and vital signs, the risk factors were divided into the mild, medium and severe type. These seven risk factors were analyzed based on whether or not there was an associated intracranial venous sinus occlusion in the patients.
Statistical analyses were performed by using the SPSS (Statistical Package for the Social Science19.0) for Windows Software (SPSS Inc., Chicago, IL, USA). The risk factors associated with CVSO were analysed using binary logistic regression analysis. Risk factors for each type of CVSO were further evaluated using multinomial logistic regression analysis. A P value < 0.05 was considered as statistically significant.
The presence of CVSO secondary to TBI was evaluated using the CT, CTV and MRV images. Among the 240 TBI patients, 40 patients were diagnosed as having a CVSO. Of these, twenty-four cases were diagnosed using a MRV and the other 16 cases using a CTV. The subgroup analysis revealed that patients with Type I CVSO, Type II CVSO and Type III CVSO were 9, 18 and 13, respectively. The average age of the patients was 36 years. The range of duration between the occurrence of trauma and the imaging diagnosis of CVSO was 6 h–7 days, with the mean diagnosis time of 3.9 days. In 20 patients, the Glasgow Coma Scale (GCS) score was between 13-15; in 9 patients, it was between 9-12, and, in 11 patients, it was <9. Twenty-one patients had recurrent and severe headache. However, the findings on the cranial CT scan could not provide a valid explanation for the headache. There were 3 patients with mild headache and in 4 patients, the headache was accompanied by vomiting and blurred vision. Two patients suffered from sudden onset unconsciousness with bilateral limb paralysis. Recurrent seizures were found in 5 patients on the first day after the onset of trauma. The other patients had no specific symptoms. These results indicate that in CVSO, no characteristic symptomatology may be present.
In order to diagnose and assess the extent of cranio-cerebral injury, a CT, CTV and/or MRV were carried out as soon as possible after the patients was admitted. Among the 240 patients, 152 underwent a CTV, 77 underwent a MRV, 1 patient underwent both a MRV and a DSA. On the axial plain CT scan images, 10 patients showed a high-density area within the sinus [Figure 1]. Among them, 9 cases showed the hyperdensity within the vertical section of unilateral sigmoid sinus and one case showed it in a unilateral transverse sinus. A skull fracture crossing the sinus was found in 34 cases, with 3 cases of depressed fracture across the superior sagittal sinus, and the remaining 31 cases of linear skull fracture crossing the involved venous sinus [Figure 1]. There were 18 cases of epidural hematoma across the venous sinus [Figure 2], with the average thickness of the epidural hematoma being 7.5 mm. Among the 18 cases, 16 cases were associated with transverse sinus and 2 cases associated with superior sagittal sinus involvement. There were three cases where an epidural hematoma across the transverse sinus was associated with the simultaneous presence of an intracranial pneumatosis. The cranial CT scan showed an intracerebral hemorrhage with infarction in 4 cases. These patients showed a superior sagittal sinus thrombosis and the hemorrhagic infarct was in bilateral parietal lobes [Figure 3].
There were 16 cases examined with a CTV that included 3 cases having Type I CVSO [Figure 1]. These 3 cases of Type I CVSO showed a crossover between the involved sinus and a linear skull fracture. These cases included 1 patient having thrombosis affecting the downstream internal jugular vein. Six patients had Type II CVSO [Figure 2]. Among them, 4 cases did not show enhancement of the spindle-shaped low-density zone inside the skull, representing the venous sinuses on the original serial axial scans. However, the peripheral venous sinuses filled well in these 4 cases. The other 2 of the 6 Type II CVSO cases showed a sinus filling defect on the part compressed by the fractured bone fragment. Seven patients had Type III CVSO [Figure 4]. In 1 patient, thrombosis formed on the downstream segment of the compressed superior sagittal sinus. In another patient, thrombosis formed on the upstream segment of the compressed superior sagittal sinus, and in the other 5 cases, thrombosis occurred in the internal jugular vein or the downstream segment of the compressed transverse -sigmoid sinus.
Eight Type I CVSO patients were diagnosed by MRV, with the interval time of 3–7 days between the occurrence of injury and the time of diagnosis. The thrombosis was present as a hyper- to isointense T1 and T2 signal [Figure 1]. There were 12 cases of Type II CVSO. Among them, 8 cases showed a spindle-like abnormal signal on axial and sagittal planes [Figure 4]; and, the remaining 4 cases showed sinus compression by a low signal bone fracture. Two cases of superior sagittal sinus CVSO had evidence of an established collateral venous circulation between the frontal vein and the lateral fissure vein [Figure 3]. Four cases had Type III CVSO. In 1 case, no fracture or epidural hematoma could be found in the involved sinus. However, thrombosis had formed in the superior sagittal sinus in the upstream segment of the venous sinus circulation [Figure 5]. Compression by an epidural hematoma was found in all the remaining 3 cases compressing on the involved venous sinus. These results suggest that a CT is helpful in the early screening of CVSO while both the CTV as well as MRV are useful for establishing an unequivocal diagnosis of CVSO.
Scoring assessment of the severity of cerebral venous sinus occlusion
To evaluate the severity of CVSO, a CVSO severity scoring was performed. Among the 40 patients of CVSO, 21 patients scored 1 point, 13 patients scored 2 points, 1 patient scored 3 points, 4 patients scored 4 points and 1 patient scored 5 points. The mean degree of venous sinus involvement was assessed at 1.8 points. Two patients had transverse sinus involvement without a high density present in them, and there was no skull fracture or epidural hematoma crossing the venous sinus. These results showed that most of the patients had only 1 venous sinus involvement; however, some of them had multiple venous sinuses involved.
Results of logistic regression analysis results
To identify the risk factors associated with CVSO, logistic regression analysis was performed. As shown in [Table 1], single factor logistic regression analysis indicated that the significant risk factor was skull fracture together with epidural hematoma crossing the venous sinus. The multivariate logistic regression analysis results, as shown in [Table 2], demonstrated that skull fracture crossing the venous sinus (odds ratio [OR] = 8.026; 95% confidence interval [CI]: 3.107–20.734) and epidural hematoma crossing the venous sinus (OR = 8.026; 95% CI 3.107–20.734) were significant risk factors associated with CVSO. Skull fracture crossing a venous sinus more significantly correlated with CVSO than an epidural hematoma crossing a venous sinus. As described in [Table 3], multinomial logistic regression analysis results showed that female gender was risk factor (OR = 10.306; 95% CI: 1.715–61.943) for Type I CVSO, an epidural hematoma crossing the venous sinus was a risk factor (OR = 5.653; 95% CI: 1.767–18.084) for Type II CVSO, while skull fracture crossing the venous sinus, epidural hematoma crossing the venous sinus, and a history of previous venous thrombosis were risk factors for Type III CVSO. These results indicate that the three types of CVSO have their separate and independent risk factors.
Traumatic brain injury may lead to CVSO. In this study, patients with CVSO (n = 40) accounted for 16.7% of the all the TBI patients (n = 240) encountered in the series, a figure that is slightly higher than that seen in the previous reports., This might be related to the higher proportion of patients with a skull fracture present in this study. In this study, it was also found that CVSO occurred even when there was no evidence of a skull fracture or an epidural hematoma crossing the venous sinus. However, the incidence of CVSO not associated with a skull fracture or an epidural hematoma was extremely low, which was consistent with the previous report by Delgado et al. The latter study found that all patients with CVSO were associated with a skull fracture crossing the venous sinus and that in the patients suffering from TBI without a skull fracture crossing the venous sinus, the incidence of CVSO was very low. We propose that the occurrence of CVSO in patients who do not show any evidence of a skull fracture or an epidural hematoma crossing a venous sinus might be due to several associated causes. First, trauma may tear the emissary vein apart and cause thrombosis inside the emissary vein. The thrombus may extend to the venous sinus, thus inducing sinus thrombosis. Second, thrombosis or occlusion in the upstream or downstream venous sinus may affect the hemodynamics of other sinuses, thus, resulting in the formation of venous sinus thrombosis in distant parts of that sinus. Third, there may be a preexisting venous sinus thrombosis before the injury. However, the possibility of this condition being present is indeed very small. Fourth, a minor skull fracture that could not be displayed by the imaging technology might have damaged the endothelial cells of the venous sinus. Finally, craniocerebral trauma may just be a predisposing factor for the precipitation of CVSO. In addition, there may also be other, hereto unknown, risk factors responsible for the development of CVSO.
A skull fracture and an epidural hematoma crossing a venous sinus were associated risk factors for the development of CVSO secondary to TBI. A skull fracture crossing a venous sinus was prone to injuring the sinus wall, thus promoting venous thrombosis formation. The venous sinus wall lacks flexibility. Even a small amount of epidural hematoma across a sinus may compress the venous sinus and thus influence the dynamics of venous flow. Both the presence of venous sinus wall damage and the presence of external compression could participate the formation of venous sinus thrombosis or occlusion.
A past history of venous sinus thrombosis of the lower limbs suggests that the patient might be having a preexisting congenital hypercoagulable state. The congenital hypercoagulable state is an important factor for the occurrence of venous sinus thrombosis, and is especially a risk factor for Type III CVSO. In this study, we did not screen for congenital thrombophilia. However, it is important that during the treatment of CVSO, the history of any previous episode of venous thromboembolism should not be overlooked.
Previous studies ,, have shown that the incidence of venous sinus thrombosis is higher in female patients. The majority of female patients with venous sinus thrombosis received previous hormone therapy. A history of oral contraceptive intake was a major factor that was responsible for the induction of thrombosis. This study found that women had a predisposition to developing Type I CVSO. It is also interesting to note that none of these patients were on hormone therapy at the time of onset of the CVSO.
It is reported that the injudicious use of hemostatic medication is closely related to the development of CVSO. However, on multinomial logistic regression analysis, we found that intravenous use of hemostatic medication was not a risk factor for the occurrence of CVSO. This may be because only a small amount of hemostatic agents were used in the early hours following the sustenance of trauma, which, could not lead to an increased risk of developing CVSO. Although there was no evidence that hemostatic drugs increased the risk of developing CVSO in patients with TBI in this study, the medications should be used with discretion. A large dose and its usage for prolonged periods should be avoided to prevent the occurrence of DVT and other adverse effects.
According to Delgado et al., in all their 195 patients exhibiting a traumatic dural venous sinus thrombosis, the thrombosis occurred within 24 h after trauma. This study clearly indicates that CVSO had a propensity to develop in the early hours after sustaining a TBI. This was consistent with previous reports,, providing a valid reason for the early screening of CVSO.
A thin-layer CT could eliminate the influence of partial volume effect on the imaging quality and enhance the rate of discovery of a skull fracture and a small epidural hematoma. We, therefore, recommend the use of a thin-layer plain CT scan for the early detection of CVSO. The sensitivity of CTV in diagnosing a CVSO is approximately 95%, and performance of this imaging in patients suffering from TBI is also relatively convenient. The CTV images, when combined with the original CT image, could not only help in ascertaining the factor responsible for the development of the venous sinus occlusion but also directly displayed the venous sinus thrombosis and the secondary brain change. We recommend the use of CTV in those patients who do not cooperate for the conduction of a MRV. The most commonly used techniques of MRV are the time of flight (TOF-MRV) imaging and the CE-MRV. A contrast agent is not needed in the TOF-MRV, but the imaging quality of the TOF-MRV is poor, and there is a certain proportion of false positives. TOF-MRV can be used in patients who are allergic to contrast agents. In patients, who are able to fully cooperate and having no history of allergy, a CE-MRV is preferred as the imaging modality to diagnose CVSO. DSA is rarely used in the diagnosis of post-traumatic venous sinus thrombosis. It is only used when the noninvasive examination results are equivocal or when an interventional thrombolytic therapy is to be simultaneously admininstered.
The clinical symptoms of traumatic CVSO vary with the severity of the disease and lack specificity. Hence during the treatment process, the clinical symptoms of the patients should be closely observed. Following the occurrence of craniocerebral trauma, in patients with a coexisting intracranial hypertension as well as inexplicable headache or epilepsy, a traumatic CVSO needs to be considered. The diagnosis of traumatic CVSO should direct ones attention towards the risk factors. A careful evaluation using a thin-layer skull CT image should be emphasized, to detect a fracture or a small epidural hematoma, that may easily be missed. MRV is the preferred method for the diagnosis of traumatic CVSO; however, it is not feasible in patients who do not cooperate or are seriously ill; moreoever, the CTV examination may help in making the diagnosis safely and quickly. In patients with skull fractures traversing across the sinus or having the presence of an epidural hematoma related to a sinus, a CTV or MRV evaluation should be carried out as early as possible to make an early diagnosis of the presence of a CVSO so that early treatment may be instituted to prevent further secondary damage to the brain.
This work was supported by Key Project of Fujian Province Science and Technology Plans (No. 2010Y0043).
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2], [Table 3]