Assessment and treatment planning of lateral intracranial dural arteriovenous fistulas in 3 T MRI and DSA: A detailed analysis under consideration of time-resolved imaging of contrast kinetics (TRICKS) and ce-MRA sequences.

OBJECTIVES: The current gold standard in the assessment of lateral intracranial dural arteriovenous fistulas (LDAVF) is digital subtraction angiography (DSA). However, magnetic resonance imaging (MRI) is a non-invasive emerging tool for the evaluation of such lesions. The aim of our study was to compare the DSA to our 3 T MR-imaging protocol including a highly spatial resolved (ce-MRA) and a temporal resolved ("time-resolved imaging of contrast kinetics", TRICKS) contrast-enhanced MR angiography to evaluate if solely DSA can remain the gold-standard imaging modality for the treatment planning of LDAVF. METHODS: We retrospectively reviewed matched pairs of DSA and 3 T MRI examinations of 24 patients with LDAVF (03/2008-04/2014) by the same list of relevant criteria for an endovascular LDAVF treatment planning. In particular, we determined intermodality agreement for the Cognard classification, the identifeication of arterial feeders, and the detailed assessment of each venous drainage pattern. RESULTS: Intermodality agreement for the Cognard classification was excellent (ĸ = 1.0). Whereas MRI failed in identifying small arterial feeders, it was superior to the DSA in the assessment of the sinus and the venous drainage pattern. CONCLUSIONS: The combination of MRI and DSA is the new gold standard in LDAVF treatment planning. KEY POINTS: • DSA is superior to the MRI in detecting LDAVF arterial feeders. • MRI excellently evaluates the venous side of an LDAVF. • MRI can replace DSA in initial diagnosis and monitoring of LDAVF. • MRI and DSA combined are the new gold standard in LDAVF treatment planning.

Sensitivity and reproducibility of fast-FLAIR, FSE, and TGSE sequences for the MRI assessment of brain lesion load in multiple sclerosis: a preliminary study.

Fast fluid-attenuated inversion recovery (fast-FLAIR), fast spin echo (FSE), and turbo-gradient spin echo (TGSE), new pulse sequences for magnetic resonance imaging (MRI), are able to display multiple sclerosis (MS) lesions more conspicuously (fast-FLAIR) and with shorter imaging times (FSE, TGSE) than is conventional spin-echo MRI. In this study, we scanned 7 MS patients, using fast-FLAIR (18 axial brain slices), FSE (27 slices), and TGSE (9 slices) sequences in the same session, to compare the brain MRI lesion loads detected by these different sequences and the intraobserver reproducibility of these measurements. On the subset of slices (n = 9) covered by all three measurements, the mean lesion load was 7.577 mm3 on fast-FLAIR, 5.248 mm3 on FSE, and 3.080 mm3 on TGSE (p = 0.006) sequences. The mean intraobserver coefficients of variation were 2.92% for fast-FLAIR, 2.86% for FSE, and 4.31% for TGSE (not significant). These findings demonstrate that both fast-FLAIR and FSE sequences may be potentially useful for serial MRI studies for monitoring clinical trials, while TGSE might be useful for speeding diagnostic MRI in MS patients. Longitudinal, clinically correlated studies using these new MRI sequences are needed to confirm these preliminary data.