Keywords
traumatic brain injuries - vascular system injuries - X-ray computed tomography
Palavras-chave
lesões encefálicas traumáticas - lesões do sistema vascular - tomografia computadorizada por raios X
Introduction
Cerebrovascular injuries associated with cranioencephalic trauma can be classified according to their location, which can be extracranial or intracranial.[1] They are caused by external forces to the skull or neck and their contents,[2] compromising the microvasculature, the great vessels, or both.[3] The present study focuses on intracranial lesions, whose diagnosis is challenging. The most frequently occurring intracranial vascular lesions are pseudoaneurysms,[4]
[5] venous sinus thrombosis, arterial dissections, and arteriovenous fistulas, the most frequent of which is cavernous carotid fistula.
A frequent challenge for the neurosurgeon when treating a patient with cranioencephalic trauma is to determine whether the patient has a vascular lesion, when to suspect it, and what studies to request.[6] There are recommendations or criteria for screening in the literature, such as those of the Japanese guidelines for the management of cranioencephalic trauma.[2] According to some authors, the true positive rates of these screening methods are only between 31.3% and 42%.[2] Therefore, complementary clinical and radiological criteria should be established to more accurately determine who will benefit from additional studies or interventions to clarify the presence of a vascular lesion.[7] Such efforts are of utmost importance because, despite their low prevalence of approximately 1%, these pathologies are accompanied, in a large extent, by hemorrhagic and ischemic complications in up to 50% of the cases if the pathology involves the internal carotid artery, and in ∼ 25% of the cases if the vertebral arteries are involved.[8]
In this context, the present study aimed to characterize the radiological findings on simple computed tomography (CT) scans of the skull that are most frequently associated with intracranial vascular lesions in patients with cranioencephalic trauma.
Materials and Methods
A cross-sectional, retrospective study was conducted. Through convenience sampling, we included patients with cranioencephalic trauma and with suspected vascular injury who were admitted to the Hospital Universitario del Valle between June 2016 and June 2019 and underwent simple CT and digital subtraction angiography. The present study was conducted in accordance with the Declaration of Helsinki, and adhered to the Good Clinical Practice guidelines.
A descriptive analysis was performed. For the continuous variables, measures of central tendency were estimated, and, for the categorical variables, frequencies and percentages were calculated. The distribution of the variables was verified by histograms and box and whisker plots.
A univariate analysis was performed using logistic regression to establish the association between the TC and angiography findings. In addition, a multivariate analysis was performed using logistic regression with stepwise selection and an inclusion probability of 0.1. Statistical analyses were performed with the Stata (StataCorp LLC, College Station, TX, United States) software, version 3.
Results
A total of 138 subjects were identified from the search of the trauma registry databases of the Hospital Universitario del Valle and the angiography service of the institution Hospital Universitario del Valle. A total of 138 subjects were included, and 82% were men (n = 112) with an average age of 32.1 ± 16.04 years. The most frequently reported trauma mechanism was gunshot wound, reported in 59% of the cases (n = 81), followed by blunt trauma, reported in 36% of the cases (n = 49), and sharps injury, reported in 5% of the cases (n = 7).
The results of the logistic regression showed an association between temporal fractures and the presence of lesions on angiography (odds ratio [OR]: 2.1; 95% confidence interval [95%CI]: 0.98 to 4.53), and no association regarding the location of bleeding or fractures. Concerning the Glasgow coma scale ([Fig. 1]), 66 patients (48.18%) presented a Glasgow of 15 on admission, and 23 patients had a Glasgow of 8 or lower on admission (16.79%); on the other hand, 48 patients were admitted with a Glasgow between 9 and 14 (35.1%).
Fig. 1 Distribution of Glasgow scores upon admission.
A regression analysis was performed to evaluate the association with venous sinus injury, and a significant relationship with fractures of the temporal base and of the sphenoid bone ([Table 1]) was found. No statistically significant association was found with pseudoaneurysm. In the multivariate analysis to evaluate the association with the incidence of vascular dissection, a relationship was found with temporal base fractures (OR: 4.6; 95%CI: 1.4 to 14.9), but not with sphenoid hemosinus.
Table 1
Logistic regression for venous sinus injury
|
Variable
|
Odds ratio
|
95% confidence interval
|
p-value
|
|
Temporal skull base fracture
|
4.3
|
1.6–11.8
|
0.004
|
|
Sphenoid skull base fracture
|
0.09
|
0.01–0.72
|
0.023
|
Discussion
The present study evaluated different markers on simple cranial CT scans to establish their association with the presence of vascular lesions, and found that temporal base fractures are associated with the diagnosis of arterial dissection and venous sinus lesions. Traumatic pseudoaneurysms are lesions in which all the layers of the vessel are compromised; they are classified as false aneurysms. Pseudoaneurysms generate the formation of a hematoma, which is transformed into a material rich in collagen and presents the appearance of an aneurysmal sac on neuroimaging studies. In contrast, in a true traumatic aneurysm, the adventitia of the vessel is preserved.[9] The clinical manifestations of aneurysms and traumatic pseudoaneurysms might include intracranial bleeding or persisting epistaxis, decreased visual acuity, and headache. There are specific radiological findings on the simple cranial CT scan that enable the suspicion of a vascular lesion, including the presence of projectile fragments near the base of the skull, the presence of a hematoma near the site of entry of the penetrating trauma,[10] and intracranial hemorrhage. The treatment of this condition is usually transcranial or endovascular surgical management.[9]
Regarding arteriovenous fistulas, the most common location is between the internal carotid artery (ICA) and the cavernous sinus.[11] The pathophysiology is based on the transmission of energy and torsional forces to the cavernous ICA at its point of dural fixation to the clinoid process. The compromise is usually unilateral, and the symptoms and signs are generated by the compromise of the venous drainage of the orbit and the increase in intraorbital pressure, which causes proptosis, limitations to extraocular movements, diplopia, chemosis, orbital murmur, and scleral injection.[9] An important and rare complication of fistulas is intracranial hemorrhage secondary to cortical venous hypertension, which is an emergency, and management is usually interventional.[12] Carotid-cavernous fistulas can be classified as direct or indirect; the latter occurs when there is substitution of the dural branches of the external or internal carotid.[11] The Barrow classification establishes that fistulas can be classified as: type A, when they are direct and have high flow without supply from the external carotid; type B, when they have low flow from the meningeal-feeding branches of the internal carotid; type C, when they have low flow and are fed exclusively by the internal carotid; and type D, when they have low flow but are fed by branches of the internal and external carotid.[10]
Vascular injuries secondary to head trauma are associated with complications with a very high percentage of morbidity and mortality, and present a diagnostic challenge because they can be asymptomatic; therefore, it is necessary to include different radiological criteria in the diagnostic process. Among the criteria that help determine the need to perform a vascular study is the Denver scale for closed trauma.[13] It is widely known that the “gold standard” for ruling out a traumatic vascular injury is cerebral angiography with digital subtraction, which has greater diagnostic capacity in the case of pathologies such as aneurysms and pseudoaneurysms.[14] However, the current recommendation is to perform CT angiography with a minimum of 16 channels, a procedure that has decreased the rate of associated infarcts from 15.2% to 3.8%.[13] With this measure, the time between diagnosis and the start of treatment has been shortened, since cranial CT scan is a technique that can be rapidly performed and has great value for determining vascular lesions.[15] This radiodiagnostic method can be implemented as a predictor of intracerebral vascular lesions.[13]
Conclusions
The limitations of the present study lie in its retrospective nature and its limited power due to the sample size. Prospective studies with larger samples could identify additional predictors. Given the results obtained from the present study, it can be concluded that the presence of temporal and sphenoid skull base fracture can predict the finding of venous sinus injury. In addition, fractures of the base of the temporal skull are associated with arterial dissections.
