Avascular stereoelectroencephalography planning: comparison between MRA and T1-weighted MRI with double contrast

OBJECTIVE: The risk of hemorrhage during stereoelectroencephalography (SEEG) is low but not negligible. The planning of avascular trajectories together with the accuracy of the implantation technique plays an important role for the reduction of this risk. The aim of this study was to compare vessel visualization using dedicated MR arteriography-venography (MRAV) versus frequently used contrast-enhanced T1-weighted MRI for the planning of avascular SEEG trajectories. METHODS: Among 93 patients with drug-resistant epilepsy, 100 SEEG schemes from 100 consecutive SEEG procedures (86 patients with single SEEG and 7 patients with 2 SEEG explorations) with 1525 electrodes were included in this study. Every SEEG scheme was a result of a multidisciplinary discussion and aimed to test a hypothesis for the localization of the epileptogenic zone based on the results of previous noninvasive investigations. All patients had dedicated MRI for SEEG planning including the following 3D sequences/techniques: T1-weighted, FLAIR, T2-weighted, MRAV, and T1-weighted with a double dose of contrast (T1+2C). Avascular planning was based on the individual SEEG scheme with 10-22 trajectories (mean 15 trajectories). A distance of 2.5 mm from the trajectory to the closest vessel on the pial surface was considered as an obligatory safety margin. All vessels closer than this safety margin on the pial surface were considered dangerous. Two neurosurgeons planned independently all the 100 SEEG schemes on T1+2C or MRAV. The same neurosurgeons performed cross-checking with the alternative vascular sequence while looking for dangerous vessels (i.e., T1+2C was checked with MRAV and MRAV was checked with T1+2C). Finally, the rate of detection of dangerous vessels on T1+2C after planning on MRAV and the rate of detection of dangerous vessels on MRAV after planning on T1+2C were calculated and compared. RESULTS: MRAV visualized a dangerous vessel at the pial entry point in 96 of 100 SEEG explorations that were first planned on T1+2C. The number of dangerous vessels found on MRAV after T1+2C planning ranged from 0 to 5, most frequently 4 dangerous vessels per planning. Overall, 291 of 1525 trajectories (19.1%) were found in 100 SEEG procedures in which MRAV visualized a vessel in the safety zone after SEEG was initially planned on T1+2C. In contrast, there was no vessel visualized on T1+2C in the safety zone when the SEEG was initially planned on MRAV. CONCLUSIONS: These findings indicate that MRAV allowed better vessel visualization than T1+2C during SEEG planning.