967 End-of-study results of a prospective longitudinal study on sentinel node detection in endometrial cancer by dual fundic and cervical ICG injection

Introduction/Background* The aim of this study is to present the final results of the study of dual cervical and fundal injection with ICG in endometrial cancer for sentinel lymph node detection. The hypothesis of this study is that fundal injection does not obviate the infundibulo-pelvic drainage pathway and improves the detection of the aortic sentinel lymph node, making it possible to add aortic mapping to the excellent pelvic mapping achieved by cervical injection. Methodology This is a 6-year (June 2014-December 2020) prospective, observational cohort study with 332 patients evaluated. Inclusion criteria were early endometrial cancer, and low/intermediate or high risk criteria of the ESMO/ESGO/ESTRO 2014. Sentinel lymph node (SLN) biopsy with dual cervical and fundal indocyanine green injection was performed in all cases, associating aorto-caval lymphadenectomy up to the left renal vein in intermediate and high risk cases. Result(s)* Detection rate (DR) was 94%. Pelvic DR, Bilateral Pelvic and Aortic Dr was 91.3%, 70.5% and 67.2%. The DR of all the anatomical areas (left & right pelvic & aortic) was 53.6%. Sentinel node detection improved overall detection by 3%. There were a total of 16.9% positive nodes for macrometastases or low-volume disease, 1 out of 4 isolated lymph node metastases corresponds to the aortic area. Sensitivity was 98.3% (95% CI 91-99.7), specificity 100% (95% CI 98.5-100), NPV 99.6% (95% CI 97.8-99.9) and PPV 100% (95% CI 93.8-100). Conclusion* Dual fundic and cervical injection is a feasible technique, which provides additional mapping of the aortic area and allows improved sentinel node detection rates in endometrial cancer, allowing detection of otherwise missed aortic lymph node involvement, which corresponds to a quarter of stage IIIC, and thus a greater number of cases of lymph node involvement that are not limited to the pelvic area.