Venous sinus thrombosis. How to avoid underdiagnosis on non-enhanced head CT?

A 15-year-old man with history of acute lymphocytic leukaemia in treatment with Asparaginase and Cytarabine presented to the Emergency Department (ED) with headache that worsened with effort. The patient’s haematocrit was noted to be 28.5 %.

A non-enhanced brain computed tomogram (CT) showed hyperdense venous superior sagittal sinus and cortical veins (Fig. 1). A region of interest (ROI) placed in the superior sagittal sinus measured 60 Hounsfield Units (HU) and the Hounsfield unit-to-haematocrit ratio (H:H ratio) was 2.2. Based on these findings, diagnosis of cerebral venous thrombosis was made.

Cerebral venous thrombosis may have an atypical presentation or even absence of clinical symptoms. The imaging findings are often subtle, and non-enhanced CT has a sensitivity of 64.6%. Underdiagnosis of cerebral venous thrombosis can lead to severe consequences such as haemorrhagic infarction. [1]

A moderate increase in attenuation may indicate cerebral venous thrombosis. To avoid possible false-negative or false-positive imaging findings on unenhanced CT scans, Buyck et al. [2] proposed active attenuation measurement within the dural sinuses, as it can be helpful in the detection of acute cerebral venous thrombosis. A threshold of 62 HU was established to discriminate patients with acute cerebral venous thrombosis from those without. They also suggested routine calculation of the H:H ratio to increase the sensitivity of non-contrast CT of the brain in the diagnosis of cerebral venous thrombosis. Their results showed an optimal threshold of 1.52, with a sensitivity of 95%, specificity of 100%, and accuracy of 97.5%. [2]

For the diagnosis enhanced CT or MRI can also be used. On enhanced CT a filling defect in a sinus, the ‘empty delta sign’, may be seen.

MRI is able to visualise both the clot and the complications. On MRI, the signal intensity of venous thrombi on T1-weighted and T2-weighted sequences varies according to the time between the onset of thrombus formation and the moment of performing the exploration. GRE and SWI sequences may be useful in acute-stage thrombosis, as blood paramagnetic products can be detected. Vascular MRI is currently considered to be the noninvasive test of choice for evaluation of the dural sinus thrombosis. [1, 2]

To treat this condition it is important to stabilise the patient, to prevent herniation and to treat the underlying cause. A recent study has shown that endovascular treatment did not improve clinical outcome in patients with severe cerebral venous thrombosis. [3]

To sum up, when performing a nonenhanced head CT the use of density measurements in the venous sinus and the calculation of the H:H ratio may be helpful in the diagnosis of the patient with cerebral venous thrombosis.

Written informed patient consent for publication has been obtained.