Purpose
To evaluate the uniformity and reproducibility of thermal lesion ablation on selected breast tumours ≤ 2 cm quantifying the volume of tissue destruction induced by employing radiofrequency ablation performing contrast-enhanced 3T MRI of ablation necrosis before surgical breast resection

Method And Materials
A cooled-tip electrode was realized to induce wide thermal lesions in swine in vivo, and in resected mammary glands from human surgical mastectomy. 24 patients with biopsy proven non-palpable invasive breast carcinomas underwent sonographically guided RFTA under local anesthesia. Surgical excision was performed one week later. All patients had 3T breast MRI scans before and after thermal ablation. Histopathological analysis of H&E and NADH-stained specimens was performed to verify the completeness of tumour coagulation necrosis and the marginal clearance. Skin or thoracic wall damage and patient reports of pain and procedural acceptability were also recorded.

Results
In resected mammary glands the mean diameter of the radiofrequency-induced lesion was 3.7 ± 0.4 cm corresponding volume of 26.67 ± 9.59 cm3. In vivo, the mean diameter was 3 ± 0.4 cm corresponding volume of 11.16 ± 3.65 cm3. A pre-RFTA MRI scan showed enhancing tumors in all the 24 patients. A post-RFA MRI scan revealed no residual lesion enhancement in 22 of 24 patients. Two patients had residual enhancement consistent with residual tumour that was confirmed histologically. Histopathological examination revealed in all the remainig ablated lesions the complete coagulation necrosis of the tumor tissue and the typical signs of complete or coagulation necrosis along the free margins of the thermoablated areas. The procedure was well tolerated from all patients, without skin or thoracic wall damages.

Conclusion
US-guided percutaneous ablation of small selected invasive breast carcinoma is feasible and reliable. Dynamic contrast-enhaced 3T MRI is an ideal method to qualify and quantify procedure effectiveness.