Study design, performance of VABB and follow-up of patients
This study enrolled 177 women with a median age of 51 (range 32-81) years, (mean ± SD: 52.4 ± 9.6), who underwent VABB for non-palpable mammographic lesions (BI-RADS 4). Exclusion criteria were: current use of aspirin or non-steroid anti-inflammatory drugs, oral anticoagulants, vitamin E, herbal supplements that may contribute to bleeding and history of bleeding disorders.
VABB was performed on the digital prone table (Mammotest, Fischer Imaging, Denver, CO, USA) using 11-gauge Mammotome vacuum probes, under local anesthesia, according to standardized protocol; 24-96 cores were obtained, according to the results of a double blind study . All procedures were performed by the same surgeon, in the same unit. For local anesthesia in VABB, the two-step approach was adopted: 5 cm3 1% lidocaine without epinephrine (superficial), and 10 cm3 1% lidocaine with epinephrine (deep) were administered . After VABB, mammograms were obtained from the women's breast, as well as from the specimens excised, confirming the successful sampling of the suspicious lesion.
The volume of blood suctioned into the device's bin was registered at the end of the procedure in all cases; importantly, no rinsing saline fluid has been pushed into the VABB probe, as this could have interfered with the measured blood aspiration volume. Hematoma formation was noted after the examination of all traditional mammogram projections (CC, MLO) independently by two radiologists (GG and DK). Compression bandages were applied so as to prevent hematoma. In all cases, the occurrence of clinically significant (i.e. subsequently organized, diameter larger than 3 cm) hematomas within the subsequent four weeks was followed up by ultrasound and clinical examination of the breast.
Concerning the histopathological evaluation, the type of the lesion was noted, i.e. benign lesion, precursor lesion (atypical ductal hyperplasia or lobular neoplasia) [5, 6], carcinoma (in situ or invasive).
For the evaluation of the associations between the volume of blood suctioned/hematoma and the duration of the procedure, age, menopausal status, and histopathological diagnosis, non-parametric tests, i.e. Mann-Whitney-Wilcoxon test for independent samples (MWW) or Spearman's rank correlation coefficient, were implemented. This choice has been made given that non-parametric tests are appropriate for non-normal distributions and that they are not heavily influenced by outliers, since the tests allocate ranks to the observations
Focusing on the blood suctioned-hematoma association, logistic regression was performed. Hematoma formation, treated as a dichotomous variable (0: no, 1: yes), was set as dependent variable, whereas the volume of blood suctioned was set as independent variable. Higher-order associations (second- and third-order were evaluated).
Statistical analysis was performed with STATA 8.0 statistical software (Stata Corporation, College Station, TX, USA).
Informed consent was obtained by all participants in this study. This study has been approved by the local Ethics Committee, in compliance to the Helsinki Declaration.