A comparison of dye versus fluorescence methods for sentinel lymph node mapping in endometrial cancer

Objective: Sentinel nodes (SNs) have been observed in several reports from Japan and overseas in cases with endometrial cancer; however, no consensus has been reached regarding the types of tracers or the method of their injection. A combination of the radioisotope (RI) and dye method is considered to be desirable. We assessed SN mapping using either dye or near-infrared fluorescence imaging to clarify a suitable method in cases of endometrial cancer. Methods: Patients were enrolled from 92 patients diagnosed with endometrial cancer and having no extrauterine metastasis by the preoperative imaging between 2009 and 2014 at our institution. To identify the SNs, we performed 3 methods using either dye or fluorescence solutions in conjunction with a RI method. In the dye method, we injected indocyanine green in the uterine subserosa, visually identifying SNs as stained green. In the fluorescence method, a dilute indocyanine green solution (0.5 mg, fluorescence A or 0.25 mg, fluorescence B, each per 10 mL of solvent) was injected and the SN identified by the HyperEye Medical System. Results: The SN detection rates were 100%, 100%, and 96% using dye and fluorescence A or B solution, respectively. Pelvic SNs were detected by the 3 methods in 98%, 100%, and 96% of cases and para-aortic SNs in 65%, 88%, and 74%, respectively. Fluorescence A solution was somewhat better than dye in detecting para-aortic SNs, although not significantly so (P = 0.07). The sensitivity and negative predictive values for detecting SNs with metastases with the dye method were 92% and 98% compared with 100% and 100%, respectively, for both fluorescence solutions. Conclusions: Although both dye and fluorescence methods performed well, no method perfectly identified para-aortic SNs. The concomitant use of the RI method is required to detect para-aortic SNs.