To prospectively assess the diagnostic accuracy of a pelvic sentinel lymph node (SLN) algorithm in high-risk endometrial cancer (HREC).

Consecutive women with presumed FIGO stage I-II HREC underwent robotic surgery at two academic centres by five accredited surgeons. An anatomically based algorithm was adhered to, following cervical injection of indocyanine green (ICG), with reinjection of tracer in case of non-display of predefined lymphatic pathways. After removal of SLNs, a pelvic and infrarenal para-aortic lymphadenectomy was performed. Primary end-point was sensitivity of the SLN-ICG algorithm. Secondary end-points were sensitivity of the overall SLN algorithm (including macroscopically suspect nodes as SLNs), SLN mapping rates and morbidity of the SLN procedure.

Two hundred fifty-seven women were analysed; 54 had pelvic lymph node metastases (LNMs), and 52 of those were correctly identified by the SLN-ICG algorithm. In two women (one with false-negative ICG-SLNs and one non-mapped woman), the pelvic LNMs were identified by the overall SLN algorithm. The SLN-ICG algorithm had a sensitivity of 98% (95% confidence interval [CI] 89-100) and a negative predictive value of 99.5% (95% CI 97-100). The sensitivity of the overall SLN algorithm was 100% (95% CI 92-100) and the negative predictive value was 100% (95% CI 98-100). The bilateral mapping rate was 95%. Two women (1%) had isolated para-aortic metastases. No adverse events occurred during the SLN procedure.

With a complete sensitivity to detect pelvic LNMs, the described pelvic SLN algorithm can, in the hands of experienced surgeons, exclude overall nodal involvement in 99% and thereby safely replace a full lymphadenectomy in HREC.