Magnetic Resonance Imaging of Internal Jugular Veins in Multiple Sclerosis: Interobserver Agreement and Comparison with Doppler Ultrasound Examination

Abstract

Doppler ultrasound (US) has been widely used to evaluate the cervical venous system of multiple sclerosis patients according to the hypothesis of chronic cerebrospinal venous insufficiency with contradictory results. Venous anatomy and pathology can be examined with less operator-dependent magnetic resonance imaging (MRI). Our aim is to assess the interobserver agreement in measuring internal jugular vein (IJV) cross-sectional area (CSA) in MR images and to explore the agreement between US and MRI in the detection of calibers of ≤0.3 cm2 in the IJV CSA in the prospective study. Thirty-seven multiple sclerosis patients underwent MRI of the cervical venous system. Two independent neuroradiologists measured the CSA of IJV at the mid-thyroid level. Furthermore, the time from contrast enhancement of common carotid arteries to that of each IJV (transit time in seconds) was assessed, and recorded whether IJV or the vertebral plexus visualized first during the contrast passage. US examination had been performed earlier. Interobserver agreement for assessing IJV CSA in MR images was substantial: the measurements differed >0.5cm2 between the examiners in only 5 IJVs (7%), Cohen's kappa 0.79. Transit times from common carotid artery to IJV varied between 5.1 and 14.1sec. Fifteen patients had left-to-right asymmetry in the speed of IJV contrast filling. IJV CSA ≤0.3cm2 was found in 51 IJVs on the basis of US. Ten of these IJVs (19.6%) showed IJV CSA≤0.3cm2 also in MRI. All IJVs defined as CSA ≤ 0.3cm2 in MRI met this caliber criterion also in US. Interobserver agreement at the thyroid level of the IJV was good at MRI measurements. The US defines more IJVs as narrow (CSA≤0.3cm2) than MRI. The US measurements for IJV CSA are not comparable with these methods. The US seems too sensitive in terms of finding venous stenosis.