Objectives: The aims of this study were to develop magnetic resonance (MR)-guided freehand radiofrequency ablation (RFA) using a near-real-time interactive MR platform in an open 1.0-T MR scanner and to determine the feasibility and safety of this new approach in the clinical setting.
Methods: The study was performed using an open 1.0-T MR system and a low-pass filter to prevent interaction between the RFA generator and the scanner. Artifact size of the radiofrequency needle was measured in 2 perpendicular views (transversal [tra] and coronal [cor]) in vitro and in the tra orientation in vivo for diagnostic (T1 high resolution isotropic volume excitation [THRIVE]/T2 turbo spin-echo [TSE]) and near-real-time (T1 fast-field-echo [FFE]) imaging. A liver-specific contrast medium (gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid) was administered 20 minutes before the intervention to enhance lesion visibility. Visibility was rated and compared for both interventional and diagnostic imaging sequences using a 10-point grading scale. Intervention time and complications were recorded.
Results: The mean diameter of needle artifact size for interventional T1 FFE was 17.4 ± 0.7 mm (tra) and 17.1 ± 1.1 mm (cor) in vitro and 15.2 ± 1.5 mm (tra) in vivo. Artifact size for diagnostic imaging was 12.5 ± 1.8 mm (tra) and 11.2 ± 1.4 mm (cor) in vitro and 10.5 ± 1.7 mm in vivo using THRIVE and 8.1 ± 2.4 mm (tra) and 10.8 ± 1.8 mm (cor) in vitro and 9.7 ± 2.0 mm (tra) in vivo using T2 TSE. A total of 57 patients with liver malignancies (mean tumor size, 17 ± 7 mm) underwent freehand MR-guided RFA. In all patients, the ablative procedure was technically successful. Lesion visibility of the diagnostic T2 TSE sequence (4 ± 2) was significantly decreased compared with both the diagnostic (THRIVE, 7 ± 2) and interventional (T1 FFE, 8 ± 1) T1-weighted sequences. Mean time to position the applicator was 7.5 ± 2 minutes. Procedure times ranged from 30 to 60 minutes. The mean in-room time was 57 ± 22 minutes. No major complications were recorded.
Conclusions: Magnetic resonance-guided freehand RFA using a near-real-time interactive MR platform in an open 1.0-T MR scanner is feasible, safe, and applicable in clinical routine. The administration of a hepatocyte-specific contrast agent enhances lesion visualization and therefore improves targeting. Without the need for additional sophisticated devices, this new approach simplifies and shortens the RFA procedure compared with previously published methods.