Preoperative marking of musculoskeletal tumors guided by magnetic resonance imaging

BACKGROUND

The purpose of this study was to evaluate a new stereotactic method for preoperative coil-marking of musculoskeletal tumors with use of interventional magnetic resonance imaging.

METHODS

Nine patients with a soft-tissue or bone-marrow tumor were referred to our department for preoperative marking of the extent of the lesion. In one patient, two lesions were marked. Guidance for the punctures and the delivery of the coils was provided by an open low-field magnetic resonance imaging system with horizontal access. After imaging of the extent of the lesion, magnetic resonance imaging-compatible titanium coils were placed with use of nearly real-time or step-by-step magnetic resonance imaging control. The coils were placed up to seventy-two hours before the surgery. The inclusion of the tumor borders within the area of the excision was examined with cross-sectional histological analysis of surgical specimens.

RESULTS

The tumor-marking intervention was successfully performed with the guidance of magnetic resonance imaging only in all patients. Preoperatively, nineteen coils were used to mark the ten lesions in the nine patients. All of the coils were easily located with intraoperative fluoroscopy. No coil migrated between the time of the percutaneous marking and the surgery. Histological examination of the resection borders revealed no residual tumor cells. No complications were observed, and, after a mean of twenty-three months of follow-up, no tumor had recurred.

CONCLUSIONS

Preoperative coil-marking guided by magnetic resonance imaging for exact delineation of a musculoskeletal tumor is technically feasible and can readily demonstrate the full extent of the tumor. Use of magnetic resonance fluoroscopy reduces the time needed for the intervention. We recommend the coil-marking technique.

Magnetic Resonance Imaging Characteristics of Six Radiofrequency Electrodes in a Phantom Study

PURPOSE

To evaluate and compare visibility and artifacts in magnetic resonance (MR) compatible radiofrequency (RF) electrodes for MR-guided RF ablation.

MATERIAL AND METHODS

Six different MR compatible electrodes for RF ablation including two internally cooled single needles, one internally cooled cluster needle, two expandable needles and one perfused needle were tested in a phantom study at 0.2 Tesla and at 1.5 Tesla field strength. Fluoroscopic, T1- and T2-weighted fast spin echo (FSE) and gradient echo (GE) sequences, which are usually used for MR-guided interventions, were evaluated. Qualitative and quantitative evaluations were performed. Length, width, noise, tip artifacts, global artifacts and global visualization of the RF electrodes that showed all sequences at different angles.

RESULTS

Qualitative analysis showed that electrodes were well visualized at all angles and sequences and on both MR imagers. Quantitative analysis showed that artifact-induced widening of the shaft was increased in all electrodes by: a). use of fluoroscopic sequences, GE sequences, and fat saturation, b). increasing the angle between the needle and main magnetic field, and c). high field strength (1.5 T). Expandable needles produced fewer tip artifacts but broader signal voids along the shaft compared to nonexpandable needles. Cluster electrodes produced less widening than the other electrodes.

CONCLUSION

Visibility and artifacts in all six MR compatible RF electrodes are satisfactory and these electrodes could be used for MR-guided radiofrequency ablation procedures.

MR-guided biopsy of musculoskeletal lesions in a low-field system

Thirty magnetic resonance (MR)-guided biopsies were obtained from 20 skeletal and 10 soft-tissue lesions in 31 patients using an open 0.2 T MR system equipped with interventional accessories. The results from aspiration (N = 3), core biopsy (N = 15), and transcortical trephine biopsy (N = 12) were evaluated for accuracy and clinical efficacy. Specimens were successfully obtained from 29 patients. Results were clinically effective in 23 patients, rated definitive in 16, nonconclusive in 9, and unspecific in 2 patients. A false diagnosis due to sampling error occurred in 2 patients, and biopsy sampling was impossible in one case. The best diagnostic yield was achieved from nontranscortical biopsies of osteolytic or soft-tissue masses. Results from transcortical biopsies were less specific due to the predominance of benign lesions. MR fluoroscopy for needle guidance was applied in 13 patients. Complete needle placement inside the magnet could be performed in 16 patients. MR-guided biopsy using an open low-field MR imager is feasible and clinically effective and will become a valuable tool in the management of musculoskeletal lesions. J. Magn. Reson. Imaging 2001;13:761-768.