Purpose: To compare the accuracy of magnetic resonance-guided focused ultrasound (MRgFUS) with MR-guided needle-wire placement (MRgNW) for the preoperative localization of nonpalpable breast lesions.
Materials and methods: In this experimental ex vivo study, 15 turkey breasts were used. In each breast phantom an artificial nonpalpable "tumor" was created by injecting an aqueous gel containing gadolinium. MRgFUS (n = 7) was performed with the ExAblate 2000 system (InSightec). With MRgFUS the ablated tissue changes in color and increases in stiffness. A rim of palpable and visible ablations was created around the tumor to localize the tumor and facilitate excision. MRgNW (n = 8) was performed by MR-guided placement of an MR-compatible needle-wire centrally in the tumor. After surgical excision of the tumor, MR images were used to evaluate tumor-free margins (negative/positive), minimum tumor-free margin (mm), and excised tissue volume (cm(3)).
Results: With MRgFUS localization no positive margins were found after excision (0%). With MRgNW two excision specimens (25%) had positive margins (P = 0.48). Mean minimum tumor-free margin (+/-SD) with MRgFUS was significantly larger (5.5 +/- 2.4 mm) than with MRgNW (0.9 +/- 1.4 mm) (P < 0.001). Mean volume +/- SD of excised tissue did not differ between MRgFUS and MRgNW localization, ie, 44.0 +/- 9.4 cm(3) and 39.5 +/- 10.7 cm(3) (P = 0.3).
Conclusion: The results of this experimental ex vivo study indicate that MRgFUS can potentially be used to localize nonpalpable breast lesions in vivo.
(c) 2009 Wiley-Liss, Inc.