Multimodal diagnosis of cortical vein thrombosis

Introduction: Thrombosis of the cortical veins (CVT) is a very rare disorder and its clinical and radiological diagnosis is challenging. The aim of our study was to evaluate the sensitivity and the specificity of different imaging modalities, including unenhanced multidetector-row CT (MDCT), multidetector-row CTA (MDCTA), and different MRI sequences in the detection of CVT.

Methods: 10 patients with clinical and radiological confirmed CVT, either isolated (ICVT) or in combination with a sinus thrombosis (=combined cortical venous thrombosis, CCVT) and 10 patients without a venous pathology were included in the study. Analysis was performed independently by three blinded readers. Readers evaluated the different imaging modalities and MR sequences separately (unenhanced MDCT, MDCTA, diffusion-weighted images (DWI), T1-, PD-, T2-, T2*-, and FLAIR weighted sequences, and MR venography). Sensitivity and specificity parameters of the different modalities were calculated.

Results: T2*- and FLAIR weighted sequences showed the highest sensitivity for the detection of CVT: 95.2 and 100%, respectively, followed by MR venography (74.1%) and T1-weighted images (54.2%). The sensitivity of unenhanced MDCT, MDCTA, DWI, and PD-weighted images was under 25%. T2*-weighted sequence showed false positive results in 2 patients, and thus a specificity of 92.6%, while the other modalities and sequences showed a specificity of 100%.

Discussion: Our study demonstrates that MR is superior to MDCT and MDCTA in the diagnosis of CVT and revealed a high sensitivity of T2*- and FLAIR-weighted MR images in the detection of CVT, followed by MR venography and T1-weighted images. Thus, imaging protocols should include those four sequences for the diagnosis of CVT. Care has to be taken concerning the specificity of T2*-weighted images, to avoid false positive results.