Stephan Meckel
I chose this topic because the neuroimaging diagnosis of cerebral venous thrombosis (CVT) may sometimes be hard to make for patients presenting at a chronic stage of thrombosis or in the presence of associated anatomic variants of the dural venous sinuses. In particular, chronic thrombosis of dural venous sinus may resemble a true chameleon on standard MR and MRV neuroimaging, because organized thrombus enhances mimicking a patent venous sinus on long-acquisition contrast-enhanced MRV sequences1-3 and TOF MRV may disclose ambiguous findings related to venous sinus asymmetry or retrograde flow phenomena. As our group has vast experience with so-called dynamic, contrast-enhanced MRA sequences (a.k.a. time-resolved MRA or 4D-MRA) in many different diagnostic applications,4-6 we decided to apply this technique in patients with CVT. As these dynamic MRA sequences may only provide lower spatial resolutions, we decided to combine the faster dynamic sequence with a high-resolution static postcontrast T1-weighted 3D sequence (a so-called combo-4D MRV) and study the accuracy of these combined techniques in the diagnosis of CVT. Our study included a sensitivity analysis during the different time-stages of the disease (acute, subacute, and chronic CVT).
The combo-4D MRV technique showed the highest diagnostic accuracy for CVT affecting the dural sinuses, as compared with T2-weighted, GRE, and TOF MRV sequences. We were surprised to find a “specific” sign of chronic dural sinus thrombosis—the delayed dynamic contrast-uptake of the chronically occluded sinus in side-by-side comparison with the regularly appearing enhancement of normal patent sinuses. When studying thrombosis in cortical veins, GRE—a sequence that has (at least in our practice) primarily been replaced by susceptibility-weighted imaging—achieved the highest sensitivity.
The results of our study have changed our practice. We have added the 4D-MRA sequence to our standard contrast-enhanced protocol in the follow-up of patients after CVT, as dural sinus occlusion may persist, or CVT may only partially recanalize despite prompt therapy in a smaller proportion of patients with CVT. Moreover, dural arteriovenous fistula (DAVF) may develop as a secondary complication following CVT,7 elusively mimicking persistent sinus occlusion on standard MR neuroimaging. According to smaller case series, as well as my personal long-standing experience as an interventional neuroradiologist, there is also a small proportion of rather aggressive DAVFs, which may oftentimes occur in association with chronically occluded venous sinuses and, thus, secondary venous outflow obstruction and cerebral venous hypertension. One example is the isolated sinus DAVF with exclusive cortical venous drainage.7,8 In such cases, the preceding CVT was often clinically silent or undiagnosed during the acute stage, and patients primarily presented with symptoms of elevated intracranial venous hypertension or related intracranial hemorrhage, with both DAVF and chronically thrombosed sinuses. For these 2 subsets of patients with CVT, the diagnosis may be rather complicated on standard MR and MRV sequences.