Narrative review of sentinel lymph node biopsy in breast cancer: a technique in constant evolution with still numerous unresolved questions

The aim of this narrative review was to provide an update on the use of sentinel lymph node biopsy (SLNB) for breast carcinoma (BC). Relevant studies published between 01/01/1994 and 15/08/2020 assessing the accuracy and the usefulness of SLNB were reviewed. SLNB was first used in 1977 for penile cancers. However, it took 17 years to enter in clinical practice for BC. The first procedures were based on two methods of non-specific marking of LN vmacrophages using a radioisotope (99mTc) and a blue dye (BD, Isosulfan, Patent or Methylene). To overcome side effects of radioisotopes (radiation exposure) and BD (allergic reactions), innovative tracers such as indocyanine green (ICG), superparamagnetic iron oxide (SPIO), and microbubbles have been explored. The SLN intraoperative examination is no longer performed, due to its low impact on the rate of reoperation and high time and cost of surgery. Likewise, immunohistochemistry, which can lead to an unnecessary ALND in some cases of occult metastases, is no more recommended. Except cases with metastasized LN, all contraindications aim to avoid situations where the risk of false negative would be too high (notably T3-T4 or multicentric tumors). The current indications for invasive BC are T0-T1-T2 N0 or N1 (after an accurate LN evaluation with ultrasound and/or cytology or core biopsy) and for DCIS treated by mastectomy or presenting as a palpable mass. After SLNB, axillary recurrence rates are generally below 2% after a follow up of 8-10 years, comparable to those observed after ALN. Likewise, when the SLN contains less than 2 metastases, axillary recurrence rates remain low even when ALN is omitted. In case of more than 2 metastatic SLN or capsular effraction, ALND is still indicated. For most teams, SLNB can be performed in clinically node-negative patients receiving neoadjuvant systemic therapy. The results of the literature consistently show that SLNB is extremely reliable in selected BC, as long as it is performed with a rigorous technique by teams having undergone multidisciplinary training and having gained the necessary experience