Role of routine preoperative lymphoscintigraphy in sentinel node biopsy for breast cancer

Sentinel node biopsy (SNB) is rapidly emerging as the preferred technique for nodal staging in breast cancer. When radioactive colloid is used, a preoperative lymphoscintiscan is obtained to ease sentinel lymph node (SN) identification. This study evaluates whether preoperative lymphoscintigraphy adds diagnostic accuracy to offset the additional time and cost required. 823 breast cancer patients underwent SNB based on lymphoscintigraphy, intraoperative γ probe detection, and blue dye mapping using 99 mTc-nanocolloid and Patent Blue V injected peritumourally. The SNB was followed by standard axillary treatment at the same operation. Preoperative lymphoscintigraphy was performed around 3 h after the radioisotope injection. Preoperative lymphoscintigraphy revealed SNs in 593 (72%) of the 823 patients imaged. SN visualisation on lymphoscintigraphy was less successful in large tumours and tumours involving the upper outer quadrant of the breast (P=0.046, P<0.001, respectively). Lymphoscintigraphy showed internal mammary sentinel nodes in 9% (62/707) patients. The SN was identified intraoperatively in 98% (581) patients who had SN visualised on preoperative lymphoscintigraphy, with a false-negative rate of 7%. In patients who did not have SN visualised on preoperative lymphoscintigraphy, the SN was identified at operation in 90% (204) patients, with a false-negative rate of 7%. The SN identification rate was significantly higher in patients with SN visualised on preoperative lymphoscintigraphy (P<0.001). SN identification rate intraoperatively using the γ probe was significantly higher in the SN visualised group compared with the SN non-visualised group (95% vs. 68%; χ2 (1 degrees of freedom (df)) P<0.001). There was no statistically significant difference in the false-negative rate and the operative time between the two groups. A mean of 2.3 (standard deviation (SD) 1.3) SNs per patient were removed in patients with SN visualised on preoperative lymphoscintigraphy compared with 1.8 (SD 1.2) in patients with no SN visualised on lymphoscintigraphy (P