Oncologic safety of sentinel lymph node–based nodal de-escalation in apparent uterine-confined endometrial cancer: a multicenter international cohort study

Background Sentinel lymph node (SLN) mapping has increasingly replaced systematic lymphadenectomy in apparent uterine-confined endometrial cancer (EC). However, concerns persist regarding the risk of recurrence following nodal surgical de-escalation, particularly in patients with aggressive histologic subtypes. We aimed to evaluate the oncologic safety of SLN-based nodal de-escalation by analyzing recurrence patterns and recurrence-free survival in patients with apparent early-stage EC. Methods We conducted a retrospective multi-institutional study including women with apparent uterine-confined EC who underwent primary surgery including SLN mapping, with or without pelvic and/or para-aortic lymphadenectomy. Patients were grouped according to nodal staging strategy: SLN-only, SLN plus pelvic lymphadenectomy (PLND), and SLN plus PLND plus para-aortic lymphadenectomy (PALND). The primary endpoints were progression-free survival (PFS) and patterns of recurrence. Results We included 2123 patients from 15 centers in six countries. SLN-only staging was performed in 1,341 patients (63.2%), SLN+PLND in 483 (22.8%), and SLN+PLND+PALND in 299 (14.1%). With a median follow-up of 44.9 months (IQR 19.1–73.4), 121 recurrences were observed (5.6%). No statistically significant differences in PFS were observed among nodal staging groups. On multivariable Cox analysis, SLN-only staging was not associated with inferior PFS compared with SLN+PLND (HR 1.12, p=0.61), and the addition of PALND did not confer a significant benefit. Endometrioid high-grade histology, non-endometrioid high-risk histotypes, deep myometrial invasion, and lymphovascular space invasion were independently associated with recurrence. Isolated nodal relapse was uncommon (14.9%) and similarly distributed across groups. Exploratory molecular analysis did not show statistically significant differences in PFS across molecular subgroups, although expected survival trends were observed. Conclusions In apparent uterine-confined EC, no significant differences in recurrence or nodal relapse were observed across nodal staging strategies. These findings support the use of SLN mapping as an adequate staging approach within a biology-driven framework, although they should be interpreted in light of the retrospective design.