Intoduction: To review different sentinel lymph node biopsy (SLNB) techniques in patients with thyroid neoplasm. We also compared the detection rates of the different detection methods in these patients. Material and methods: The Medline database from 1998 until December 2010 was searched for the following terms: thyroid cancer, thyroid neoplasm, and sentinel lymph node. Studies in which sentinel lymph nodes were detected by the blue dye technique and/or by a radiotracer in patients with suspected thyroid cancer were analysed.
Results: Twenty five studies were included in the meta-analysis. Based on the technique used for sentinel lymph node (SLN) detection, the included studies were divided into three groups. Group 1 consisted of studies in which only the blue dye technique was used to detect SLNs. Group 2 was made up of studies in which the radioisotope technique was used. Studies in which both techniques were used were grouped into Group 3. There were 18 studies in which the blue dye technique was used to detect SLNs (Group 1), four studies in which only the radioisotope technique was used to detect SLNs (Group 2), and only two studies where both techniques were used (Group 3). Among 891 patients from Group 1, SLN was found in 740 (83.1%) patients. Detection rates in these studies were very different and varied from 0% to 95.5%. Among 160 patients from Group 2, SLN was detected in 158 (98.8%). In the third group of patients, in which both methods were performed, SLN was found in 48 (98%) of 49 patients. Detection rates in those studies were very high (100% and 97.8%).
Conclusions: The analysis proved that SLNB is, technically, fairly easy to perform. However, nodal metastases are of debatable prognostic value in thyroid cancer, so the clinical value of SLNB remains to be proven. It seems reasonable to perform further, prospective studies on larger groups of patients, in which both techniques would be used. They should compare the efficiency of SLNB with elective or selective central lymphadenectomy in reducing local recurrence rates.