Objective: With this retrospective, multi-centric study, the authors are showing the technique of Vacuum assisted biopsies under ultrasound guidance and comparing it with the other widely used diagnostic techniques. Material and method. Six hundred and fifty biopsies were performed between May 2000 and December 2004, on 644 patients, in 3 centres, following a unique protocol. Lesions were categorized, using the classification from Stavros, between "probably benign", "indeterminate", "probably malignant" and "malignant" Histology was validated only after review of the clinical and radiological data, as well as surgical data when available. All benign cases were included in an on-going follow-up protocol.
Results: We have identified 471 benign lesions and 179 malignant lesions. The mean size of the lesions was 9 mm. Three cancers were diagnosed in the cases of "probably benign lesions" and in the cases of "probably malignant lesions" 18 (27%) were inflammatory disorders. In 5 cases vacuum biopsy underestimated the pathology with regard to surgery: 2 cases of atypical duct hyperplasia (HCA) were in situ ductal carcinoma (DCIS) at surgery and 3 cases of DCIS were infiltrative ductal carcinoma (DCI) at surgery. With this technique we have avoided surgery for 71% of all women who presented an "indeterminate" or "probably malignant" condition. Specificity is excellent with no cancer detected so far among the patients with benign findings, under follow-up.
Conclusion: Ultrasound guided Vacuum assisted biopsy is a fairly recent minimally invasive technique, with short learning curve. The ability to collect a larger volume of tissue overcomes the targeting issues on small lesions and avoids underestimation of heterogeneous and larger abnormalities and some specific at-risk lesions such as papilloma. This technique thus appears indicated in such cases because it overcomes some of the limitations of core needle biopsy and should be considered as an alternative to surgical biopsy.