This retrospective cohort study was conducted at Kaohsiung Chang Gung Memorial Hospital and included patients diagnosed with EC between November 2021 and December 2024. Eligible patients were those with a confirmed diagnosis of EC who underwent primary staging surgery incorporating SLN mapping, followed by post-operative pathological ultrastaging. Cases were excluded if they did not undergo primary staging surgery or if they underwent primary staging surgery without SLN assessment and mapping. The complete study workflow is illustrated in Fig. 1. Demographic and clinicopathological data, including age, body mass index (BMI), tumor size, stage, histological subtype, and mismatch repair (MMR) status, were extracted from electronic medical records for analysis. This study was approved by the Institutional Review Board of Chang Gung Memorial Hospital, Taiwan (approval number: 202500410B0) and patient confidentiality was strictly maintained, and informed consent was waived due to the retrospective nature of the study.

Statistical analysis was performed to compare the factors influencing bilateral SLNs success mapping. Continuous variables were compared using the Mann-Whitney U test and presented as medians with ranges. Categorical variables were analyzed using either the chi-squared test or Fisher’s exact test, depending on the expected frequencies, and were expressed as frequencies and percentages. All statistical analyses were performed using SPSS software for Windows, version 26 (IBM, Armonk, NY, USA). A p-value < 0.05 was considered statistically significant.

A total of 488 patients were diagnosed with EC between November 2021 to December 2024, of which 131 uterine-confined cases underwent primary staging surgery with intraoperative SLN assessment and mapping. As shown in Table 1A, most patients were diagnosed at an early clinical stage, with 76.3% (n = 100) classified as FIGO stage IA and 22.9% (n = 30) as stage IB with only one case as stage II. Postoperative pathologic evaluation confirmed stage I disease in 88.5% of cases (n = 116), while the remaining 11.4% (n = 15) were diagnosed with more advanced stages. Notably, two cases were initially staged as IA but were later found to have stage IVB disease postoperatively—one due to omental metastasis and the other due to lung metastasis. Histopathological analysis revealed endometrioid adenocarcinoma as the predominant subtype (90.8%, n = 119), with serous, clear cell, carcinosarcoma, and mixed histology accounting for the remaining cases. Low-grade histology was identified in 85.5% (n = 112) of patients. Medium tumor size was 25 mm (range from 1-108 mm). Most tumors demonstrated positive estrogen receptor (ER) and progesterone receptor (PR) expression, with ER positivity in 87.0% (n = 114) and PR positivity in 75.6% (n = 99) of cases. Molecular profiling revealed MMR deficiency in 26.7% of tumors (n = 35), while 61.1% (n = 80) demonstrated MMR proficiency.

Among the study population, eight cases (6.1%) were pathologically identified with positive lymph nodes (Table 3). Two cases (25%) were classified as pathologic stage IB with isolated tumor cells (ITCs) in observed SLNs, and the remaining 6 cases (75%) were classified as stage IIIC1 due to micrometastasis or macrometastasis of SLNs. Most tumors (n = 7, 87.5%) were low-grade endometrioid carcinoma, except for one case of serous carcinoma. Median tumor size was 41mm (range from 18-108mm), with most cases (n = 6, 75%) exhibited deep myometrial invasion (myoinvasion ≧ 50%). Figure 3 shows a representative case of ITCs identified within a SLN, confirmed by positive AE1/AE3 immunohistochemical staining.

The efficacy of MIS, particularly when combined with ICG fluorescence imaging, has been demonstrated in SLN mapping for EC[12, 13]. MIS was utilized in 87.8% of our cohort, reinforcing its well-established benefits in reducing morbidity while maintaining oncologic outcomes. Our study achieved a high SLN success mapping rate, with an overall success rate of 95.9% and a bilateral success rate of 74.8%. Compared to the FIRES trial, which reported an overall mapping success rate of 86% and a bilateral success rate of 52%, our findings suggest improved SLN mapping efficiency with the incorporation of standardized techniques[14]. Compared with patent blue, the use of ICG significantly improved bilateral SLN mapping success (27.3% vs. 79.2%, p < 0.001), reinforcing its superiority as the preferred tracer for SLN detection. These findings are consistent with the FILM trial, which reported that ICG with near-infrared fluorescence imaging identified more SLNs than isosulfan blue dye in uterine cancers[10]. Eriksson et al. also demonstrated that ICG improves SLN detection, even in obese patients, compared to blue dye[15]. Additionally, our study confirmed that MIS was associated with significantly higher SLN detection rates than open surgery (79.1% vs. 43.8%, p = 0.002), consistent with findings from the LAP2 and SENTI-ENDO trials[13, 16]. However, within the subgroup of patients who received ICG as the tracer agent, no statistically significant difference in success rates was observed between MIS and laparotomy (79.6% vs. 71.4%, p = 0.635). This finding suggests that while MIS generally enhances mapping success, the use of ICG as a tracer may mitigate differences in outcomes between surgical approaches, potentially due to the superior efficacy of ICG in facilitating lymphatic mapping regardless of surgical technique. Taken together, these findings strongly support the routine use of ICG to optimize SLN detection and ensure accurate staging

The SAGE multicenter study by Cianci et al. identified advanced age and higher BMI as independent risk factors for bilateral SLN mapping failure in EC, attributing these associations to age-related lymphatic capillary rarefaction and the technical challenges posed by obesity[17]. In contrast, our retrospective cohort analysis found that neither age nor BMI was significantly associated with bilateral SLN mapping success—both in the overall cohort (p = 0.994 and p = 0.186, respectively) and in the ICG subgroup (p = 0.630 and p = 0.498, respectively). This discrepancy may be explained by differences in sample size and age distribution, as our study population was smaller and had a younger median age (56 years) compared to the SAGE cohort (64 years). Furthermore, the predominant use of ICG as the tracer and adherence to standardized mapping protocols in our study may have minimized the impact of age and BMI on mapping outcomes. Other contributing factors may include high surgeon experience and a lower proportion of patients with extreme obesity, which could have enhanced mapping success in our cohort. Cianci et al. also identified LVSI as a significant negative predictor of successful SLN mapping, suggesting that the presence of LVSI may disrupt effective tracer migration through lymphatic channels, thereby reducing mapping success[17]. However, in our cohort, LVSI was not significantly associated with SLN detection rates. This discrepancy may reflect differences in methodology, such as the predominant use of ICG as a tracer in our study, or variations in patient populations, including a higher proportion of early-stage and low-grade tumors, which could diminish the impact of LVSI on mapping outcomes. Additionally, our definition of LVSI positivity was limited to cases of substantial LVSI, excluding focal involvement, which may differ from the criteria used by Cianci et al. and thus contribute to the observed differences.

Regarding SLN detection rates, we observed an overall detection rate of 95.9%, with detection rates of 94.5% in the right hemipelvis and 90.2% in the left hemipelvis. These results are consistent with the SENTI-ENDO trial, which reported SLN detection rates of 89% (overall), 77% (right), and 76% (left)[16]. Recent studies have demonstrated consistently high SLN detection rates in EC. Fierro et al. reported an overall detection rate of 96.7% using ICG, with 97.3% in laparoscopic surgery and 95.5% in robotic surgery, supporting the reliability of SLN mapping across different minimally invasive surgical approaches[18]. These findings are consistent with meta-analyses by Lin et al. showing bilateral detection rates of 82–86% for both laparoscopic and robotic approaches when ICG is used as the tracer[19]. Additionally, SLNs were most frequently identified in the external iliac, obturator, and internal iliac regions, whereas para-aortic SLN detection was rare, aligning with previous literature. Recent data from studies that performed SLN mapping with reference lymphadenectomy has shown the risk of isolated para-aortic metastasis to be as low as 0.8-1%[12, 14]. These findings further support that the lower paracervical pathway is the predominant drainage route, with the majority of mapped SLNs located in the pelvic region, while para-aortic SLN detection remains limited (~ 5–10%)[16]. The localization of SLNs in the pelvic basin further supports SLN mapping as a reliable alternative to full lymphadenectomy, particularly for low- and intermediate-risk EC patients.

Several methods were employed for slicing lymph nodes during ultrastaging in EC, each with distinct protocols, advantages, and limitations. Commonly used macroscopic slicing techniques include longitudinal slicing section and bread-loaf slicing section[8]. Longitudinal sectioning, on the other hand, slices the lymph node along its longest axis, which is technically simpler, faster to perform, and preserves nodal architecture[20]. Bread-loaf slicing involves cutting the lymph node into multiple thin sections, typically 2–3 mm thick, perpendicular to the long axis. This approach ensures comprehensive sampling of the entire node, increasing the likelihood of detecting low-volume metastases, including micrometastases and ITCs[20, 21]. In our study, bread-loaf slicing was adopted as the ultrastaging protocol for SLN assessment, following the methodology of the FIRES trial[14]. Specifically, SLNs larger than 8 mm were sliced perpendicular to the long axis into 2–4 mm sections, with subsequent histological and immunohistochemical evaluation. This protocol was chosen based on robust evidence from a systematic review by Burg et al., which demonstrated that bread-loaf slicing achieved a significantly higher SLN metastasis detection rate (53%) compared to longitudinal slicing (33%, p < 0.05)[9]. The enhanced sensitivity of bread-loaf slicing is attributed to its cross-sectional approach, which increases the probability of detecting metastatic deposits that might be missed with longitudinal slicing. Furthermore, the review found that increasing the number of IHC slides did not significantly improve detection rates, highlighting that the slicing method itself is the most influential factor in ultrastaging sensitivity. Despite the absence of standardized international guidelines for SLN ultrastaging, our findings align with previous studies supporting the bread-loaf slicing technique as an optimal approach. Given its higher diagnostic accuracy, our study further supports the need for standardizing SLN ultrastaging with the bread-loaf technique to optimize metastatic detection and improve staging accuracy.

According to Katsoulakis et al., an analysis using the SEER registry evaluated the incidence of lymph node metastasis in clinical stage I, low-grade, endometrioid EC and reported a 3.5% risk of pelvic lymph node metastasis based on systematic pelvic lymphadenectomy[22]. Similarly, a recent study reported by Praiss et al. using pelvic lymphadenectomy for nodal assessment found a comparable incidence of 4.0% LN metastasis in low-grade, stage I ECs, further confirming the low but notable risk of nodal involvement[6]. In our study, we employed a different methodology, utilizing intraoperative ICG for SLN mapping and postoperative ultrastaging with the bread-loaf slicing method, yielding a comparable lymph node metastasis detection rate of 4.0%. This suggests that our combined approach of ICG dye for mapping and bread-loaf slicing for ultrastaging may detect micrometastases missed by conventional lymphadenectomy, highlighting the diagnostic advantage of this method while reducing surgical morbidity in early-stage, low-risk EC patients. As of this manuscript submission, none of the patients diagnosed with pathological stage I disease—whether through successful SLN mapping or conventional lymph node dissection following mapping failure—have experienced regional or distant recurrences. This suggests a favorable prognosis for patients treated at an early stage. We will maintain ongoing surveillance to ensure early detection and timely intervention if needed.

This study has several important limitations that should be acknowledged. First, its retrospective design inherently introduces selection bias and limits the ability to control for confounding variables that may influence SLN detection rates and metastasis identification. Second, as a single-institution study, the findings may not be fully generalizable to other clinical settings with different patient populations and levels of surgical expertise. Lastly, the study lacked a direct comparison group using longitudinal slicing within the same patient population, which would have provided stronger evidence for the superiority of the bread-loaf technique. Despite these limitations, this study offers valuable insights into optimizing lymph node metastasis detection in uterine-confined EC. Addressing these methodological issues in future prospective, multi-institutional research could enhance the reliability and clinical applicability of the findings.