Clinicopathologic analysis of sentinel lymph node mapping in early breast cancer

Axillary nodal status is the most significant prognosticator for predicting survival and guiding adjuvant therapy in breast cancer patients. Sentinel lymph node biopsy (SLNB) represents a minimally invasive procedure with low morbidity for staging axillary nodal status. In this article we review and report our experiences in patients with early breast cancer who underwent SLNB at the Revlon/UCLA Breast Center. Between September 1998 and May 2000, a total 83 SLNBs were performed in 81 patients with proven breast cancer and negative axillary examination who elected to have SLNB as the first step of nodal staging. Two patients had bilateral breast cancer. SLNB was localized by using both 99Tc sulfur colloid (83 cases) and isosulfan blue dye (75 cases). Data of these patients were prospectively collected and analyzed. The clinical and pathologic characteristics of women with positive and negative sentinel lymph nodes (SLNs) were compared to identify features predictive of SLN metastasis. Of the 83 cases, the SLN was successfully localized in 82 (98.8%). Sixty-three percent of patients had SLNs found in level I only, 18.3% in both level I and II, and 4.9% in level II alone. The vast majority (84.3%) of these cases had T1 breast cancer with an average size of 1.55 cm for the entire series. Twenty-three patients (28%) had positive SLNs, with an average of 1.5 positive SLNs per patient. Fifteen had metastases detected by hematoxylin and eosin staining and 8 had micrometastases detected by immunohistochemistry (IHC) using anticytokeratin antibodies. Ten of the former group agreed to and 2 of the latter group opted for full axillary lymph node dissection (ALND). An average of 17.5 lymph nodes were removed from each ALND procedure. Additional metastases or micrometastases were found in seven patients (in a total of 28 lymph nodes). Three patients with completely negative SLNs experienced additional axillary lymph node removal due to their election of free flap reconstruction. None had metastases detected in these lymph nodes. The absence of estrogen and progesterone receptors (ER/PR) by IHC (p = 0.036) and the presence of lymphatic/vascular invasion (LVI) (p = 0.002) predicted positive SLNs in patients with early breast cancer in a univariate analysis; in a multivariate analysis only LVI was predictive (p = 0.0125). Histologic type, nuclear grade, tumor differentiation, HER-2/neu and p53 status, S-phase fraction, and DNA ploidy did not predict SLN status. Immediate postoperative complications were uncommon and delayed complications completely absent. Because of the high detection rate, accurate staging, and minimal morbidity, SLNB should be offered as a choice to women with small breast cancers and clinically negative nodes. Because positive LVI and negative ER/PR status are highly predictive of pathologically positive SLNs in small breast cancers, women whose cancers meet these criteria should be advised preoperatively about their risk of having a positive SLN and may benefit from intraoperative assessment (frozen section and/or touch preparation) of their SLNs.