Optimizing the Sensitivity of a Pelvic Sentinel Node Algorithm Requires a Hybrid Algorithm Combining Indocyanine...

to investigate the incidence of non-mapped isolated metastatic pelvic lymph nodes at pre-defined anatomical positions.

Between June 2019 and January 2024, women with uterine-confined endometrial cancer (EC) deemed suitable for robotic surgery and the detection of pelvic sentinel nodes (SLNs) were included. An anatomically based, published algorithm utilizing indocyanine green (ICG) as a tracer was adhered to. In women where no ICG mapping occurred in either the proximal obturator and/or the interiliac positions, defined as "typical positions", those nodes were removed and designated as "SLN anatomy". Ultrastaging and immunohistochemistry were applied to all SLNs. The proportion of isolated metastatic "SLN anatomy" was evaluated.

A non-mapping of either the obturator or interiliac area occurred in 180 of the 620 women (29%). In total, 114 women (18.4%) were node-positive and five of these women (4.3%) had isolated metastases in an "SLN anatomy", suggesting a similar lower sensitivity of the ICG-only algorithm.

In an optimized SLN algorithm for endometrial cancer, to avoid undetected nodal metastases in 4.3% of node-positive women, if mapping fails in either the proximal obturator or interiliac area, nodes should be removed from those defined anatomic positions, despite mapping at other positions.