Sentinel node study in early breast cancer

Since October 1997 60 patients with early breast cancer (T <3 cm) were stud ied. All patients underwent lymphoscintigraphy with two types of colloid: t he first (17 pts) with a particle size <1000 nm; the second (43 pts) with a particle size <80 nm. The standard procedure consists of injection, on the day before surgery, of 70 Msg of the smaller nanocolloid in 0.4 cc saline divided over four sites, around the lesion or subdermally around the surgic al scar. We utilize a low-energy, high-resolution LFOV camera for scintigra phy and a probe specific for the sentinel node during surgery. In 56/60 pat ients (93.3%) lymphoscintigraphy showed the sentinel node (SN). In two case s the SN was not detected presumably because of lymphatic interruption by a n old surgical scar; in the other two cases the sites of injection were too close to the SN, thus masking it. In five cases (9%) the SN was not visual ized with the surgical probe but in two of these drainage to the internal m ammary chain was observed. The apparently lower sensitivity of intraoperati ve localization was due to the extra-axillary lymphatic drainage or to the vicinity of the SN to the primary lesion. The SN proved to be metastatic in 12 cases. No false-negative SNs were found. In five cases (10%) the radiol abeled lymph node was the only node containing tumor cells (micrometastases ): this result depends on the combined use of hematoxylin-eosin and rapid c ytokeratin staining, The application of blue dye was useful for easier iden tification of the SN but did not allow detection of more SNs. Our prelimina ry results are extremely encouraging. Considering that at the early stages of breast cancer the likelihood of lymph node metastases is low (20% in our series) and no false negative were reported in this study, we conclude tha t with SN biopsy axillary lymph node dissection can be avoided, making surg ery less aggressive but maintaining accuracy.