Comparison of tracer application methods for sentinel lymph node detection in open surgery patients with endometrial cancer: a retrospective cohort study.

Following the acquisition of institutional ethical approval (EK-VP-21-0-2023), we retrospectively gathered medical data from patients with endometrial cancer who underwent laparotomy at the University Hospital Kralovske Vinohrady between January 2020 and March 2024. Radiolabeled colloid technetium-99m ( 99m Tc) in an amount of 2 ml with a dose of 40 MBq (1,08 mCi) combined with methylene blue dye (ProveDye 0.5% 2 ml, Provepharm, Marseille, France) were utilized as tracers and were applied to all patients. Three application techniques were employed: (1) intracervical—deep injection (1–2 cm) into the cervical stroma at the 3, 6, 9, and 12 o’clock positions of the uterine cervix using a 25-gauge hypodermic needle at the beginning of the operation; (2) subserosal—two injections, each 1–3 mm deep, performed on both the anterior and posterior surfaces of the uterus at the beginning of the laparotomy after opening the abdominal cavity, followed by an 8-minute wait before beginning the retroperitoneal preparation; and (3) intrafundal— tracer application into the fundus of the myometrium. In this technique, the sheath of a long spinal needle was shortened (Picture S1 , Supplementary material) to allow the needle tip to protrude by 3–5 mm and the tracer was trasncervically inserted into the fundus myometrium (Picture S2 , Supplementary material) of the uterus at the beginning of the operation. The shortened sheath was then introduced into the uterine cavity up to the fundus, and the needle was inserted into the sheath to deliver the tracer to a depth of 3–5 mm within the myometrium at the fundal region, ensuring that the uterine wall was not perforated. The I + S group, which received intracervical and subserosal application, comprised 147 women who subsequently underwent surgery between January 2020 and December 2021. In contrast, the I + I group, which received intracervical and intrafundal application, included 101 patients who underwent surgery between January 2022 and March 2024. Patients were included consecutively as they were indicated for surgery at the Department of Obstetrics and Gynecology, 3rd Medical Faculty, Charles University in Prague, and University Hospital Kralovske Vinohrady. A lower midline laparotomy was performed for all patients. Upon gaining access to the retroperitoneum, careful examination of the pelvic area surrounding the major vessels, including the presacral region, was conducted to identify any marked lymph nodes bilaterally. Following this, a handheld gamma probe (Neoprobe, Devicor Medical Products, Cincinnati, Ohio, USA) was utilized to scan for SLNs. In cases where SLN detection was unsuccessful, side-specific systematic lymphadenectomy was performed in the pelvis, mainly in patients with high-risk (HR) histology, contingent on the patient’s clinical condition and the state of the surgical field. Enlarged LNs were always removed. If enlarged LNs were not present and systematic LND was not performed, sampling LNs was performed from the external iliac vein and supraobturator regions. Subsequently, total abdominal hysterectomy (TAH) and bilateral salpingo-oophorectomy (BSO) were carried out. All retrieved lymph nodes were fixed in a 10% buffered formalin solution, embedded in paraffin wax, and subjected to hematoxylin and eosin staining. All SLNs underwent histopathological ultrastaging, including immunostaining. The surgical procedures were performed by a team of five experienced surgeons. Statistical analyses were conducted using R version 4.2.2 and Rstudio 2024.09.0 [ 7 ]. Missing data (< 10% per variable) were imputed via multiple imputation by chained equations (m = 5, randomforest method; mice R package, seed = 42). Continuous data were presented as means and standard deviations or as medians with 25th and 75th percentiles (interquartile range, IQR), depending on the distribution’s normality. Binary and categorical data were expressed as counts and percentages. The Wilcoxon rank sum test was employed to compare continuous variables, while categorical parameters were analysed using either the Pearson chi-square test or Fisher’s exact test, as appropriate. Univariate logistic regression models (glm, binomial family) estimated odds ratios (ORs) and 95% confidence intervals (CIs) for each of three binary outcomes (overall SLN failure, rightside failure, leftside failure) in both raw and imputed datasets. Multivariate predictor selection was performed separately for each outcome on the imputed data using LASSO penalized logistic regression (glmnet R package) with 10fold cross-validation; variables with nonzero coefficients at the minimum λ (λ_min) were refit in standard logistic regression models to yield final adjusted ORs and 95% CIs [ 8 , 9 ]. The cut-off values for continuous variables were determined using receiver operating characteristic (ROC) curve analysis, employing the Youden index (sensitivity + specificity − 1) as the optimization criterion for threshold determination [ 10 ]. All tests were two-sided, with a significance level set at p  < 0.05.

A total of 248 women with histologically confirmed endometrial cancer staged I or II based on ultrasound criteria, who were indicated for open TAH and BSO with SLN detection, were included in the study. According to the final histopathological evaluation, the tumors were classified following FIGO 2009 staging system as stage IA in 105 patients (42.9%), stage IB in 69 (27.8%), stage II in 46 (18.5%), stage IIIA in 9 (3.6%), stage IIIB in 4 (1.6%), and stage IIIC1 in 15 (6.0%) patients (Table  1 ). The cohort was divided into two groups: the I + S group consisted of 147 patients (59.3%) who underwent intracervical and subserosal application of tracers, while the I + I group comprised 101 women (40.7%) who received intracervical and intrafundal application. No statistically significant differences were observed between the two groups regarding age, body mass index (BMI), or other clinical and histopathological characteristics of the carcinomas (Table  1 ). Table 1 Clinicopathologic characteristics of patients Variable All patients, n  = 248 n (%) Intracervical + intrafundal, n  = 101 n (%) Intracervical + subserosal, n  = 147 n (%) p -value Age (years) in mean (interquartile range) 67.1 (40, 86.6) 68 (60, 76) 68 (62, 74) 0.7 Body mass index (kg/m 2 ) in mean (interquartile range) 31.6 (18.3, 54.4) 32 (28, 36) 31 (26, 35) 0.12 Pregnancy in mean (median) 2.2 (2) 2.26 (2) 2.10 (2) 0.7 Delivery in mean (median) 1.6 (2) 1.55 (2) 1.57 (2) 0.8 Smoking 37 (14.9%) 14 (13.9%) 23 (15.6%) 0.7 Conisation history 18 (7.3%) 4 (4.0%) 14 (9.5%) 0.1 Cesarean section history 21 (8.5%) 8 (7.9%) 13 (8.8%) 0.8 Pelvic surgery history 36 (14.5%) 13 (12.9%) 23 (15.6%) 0.5 Myomatosis 100 (40.3%) 37 (36.6%) 63 (42.3%) 0.6 Adenomyosis 29 (11.7%) 12 (11.9%) 17 (11.6%) 0.8 Cervix infiltration 48 (19.4%) 19 (18.8%) 29 (19.7%) 0.9 LVSI 91 (36.7%) 35 (34.7%) 56 (38.1%) 0.6 HR ( high risk) histology 71 (28.6%) 25 (24.8%) 46 (31.3%) 0.2 Operation lenght (min) mean (median) 90.7 (90) 92 (90) 89 (90) 0.3 Estimated blood loss (ml) mean (median) 210 (200) 220 (200) 195 (200) 0.14 Stage (FIGO 2009)  IA 105 (42.3%) 42 (42.4%) 63.0 (43.2%) 0.7  IB 69 (27.8%) 27 (26.7%) 42 (28.6%)  II 46 (18.5%) 18 (18.2%) 28 (19.2%)  IIIA 9 (3.6%) 3 (3.0%) 6 (4.1%)  IIIB 4 (1.6%) 2 (2.0%) 2 (1.4%)  IIIC1 15 (6.0%) 9 (9.1%) 6 (4.1%) Histology  endometrioid 203 (81.9%) 84 (83.2%) 116 (78.9%) 0.2  serous cancer 36 (14.5%) 10 (9.9%) 26 (17.7%)  clear cell 4 (1.6%) 3.0 (3.0%) 1 (0.7%)  carcinosarcoma 7 (2.8%) 3.0 (3.0%) 4.0 (2.7%)  adenosarcoma 2 (0.8%) 1 (1.0%) 1 (0.7%) Procedure on SLNs  SLN biopsy only 54 (21.8%) 26 (25.7%) 28 (19.0%) 0.5  SLN + sampling 114 (46.0%) 44 (43.6%) 70 (47.6%)  SLN + PLN 80 (32.3%) 31 (30.7%) 49 (33.3%) LVSI - lymph-vascular space invasion; ml – milliliter; SLN -sentinel lymph node; PLN – pelvic lymphadenectomy Clinicopathologic characteristics of patients Age (years) in mean (interquartile range) Operation lenght (min) mean (median) Estimated blood loss (ml) mean (median) LVSI - lymph-vascular space invasion; ml – milliliter; SLN -sentinel lymph node; PLN – pelvic lymphadenectomy In 54 patients (21.8%), only SLN detection was performed alongside TAH and BSO. In 114 patients (46.0%), pelvic lymph node sampling (LNS) was conducted following SLN detection, and in an additional 80 patients (32.3%), pelvic systematic lymphadenectomy (PSL) was performed (Table  1 ). Among patients with only SLN detection, an average of 3.6 lymph nodes (95% CI: 2.0-5.2) was retrieved, while those undergoing LNS had a mean of 6.6 lymph nodes (95% CI: 5.5–7.7), and patients subjected to PSL had an average of 15.1 lymph nodes (95% CI: 13.7–16.4) removed. Successful simultaneous detection of SLNs on both sides was achieved in 39.9% of patients (99/248) across the entire cohort (Table  2 ). In the I + S group (intracervical + subserosal application), SLNs were successfully detected on both sides of the pelvis in 38.1% (56/147) of patients, while in the I + I group (intracervical + intrafundal application), this was observed in 42.6% (43/101). SLNs were identified in 32.7% (81/248) of patients on only one side of the pelvis, with a detection rate of 31.3% (46/147) in the I + S group and 34.7% (35/101) in the I + I group. Additionally, no SLNs were detected in 27.4% (68/248) of subjects, comprising 30.6% (45/147) from the I + S group and 22.8% (23/101) from the I + I group (Table  2 ). Although the success rate of SLN detection was higher in the I + I group (intracervical + intrafundal application) and on the right side of the pelvis regardless of the detection method, these differences were not statistically significant (Table  2 ). SLNs were detected on the right side of the pelvis in 58.9% (146/248) of women in the overall cohort, with rates of 55.1% (81/147) in the I + S group and 64.4% (65/101) in the I + I group. On the left side, SLNs were identified in 53.6% (133/248) of women in the total cohort, 52.4% (77/147) in the I + S group, and 55.4% (56/101) in the I + I group, with no statistically significant differences observed (Table  2 ). The locations of detected sentinel lymph nodes are detailed in Table  3 . Table 2 SLN detection rate SLN detection rate All patients, n  = 248 n (%) Intracervical + intrafundal application, n  = 101 n (%) Intracervical + subserosal application, n  = 147 n (%) p Detected on both sides 99 (39.9%) 43 (42.6%) 56 (38.1%) 0.4 Detected on right side 146 (58.9%) 65 (64.4%) 81 (55.1%) 0.15 Detected on left side 133 (53.6%) 56 (55.4%) 77 (52.4%) 0.6 Detected only on one side 81 (32.7%) 35 (34.7%) 46 (31.3%) 0.4 Not detected on any side 68 (27.4%) 23 (22.8%) 45 (30.6%) SLN detection rate Table 3 SLN detection sites Right side, n  = 147 p Left side, n  = 135 p Intracervical + intrafundal Intracervical + subserosal Intracervical + intrafundal Intracervical + subserosal external iliac artery 1 (1.5%) 4 (4.9%) 0.14 1 (1.7%) 2 (2.6%) 0.8 external iliac vein 37 (56.1%) 35 (43.2%) 34 (58.6%) 46 (59.7%) common iliac vessels 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) iliac bifurcation 1 (1.5%) 0 (0.0%) 1 (1.7%) 3 (3.9%) supraobturator 26 (39.4%) 42 (51.9%) 21 (36.2%) 26 (33.8%) presacral 1 (1.5%) 0 (0.0%) 1 (1.7%) 0 (0.0%) SLN detection sites In the analysis of risk factors associated with overall SLN detection failure on both sides, the only statistically significant factor identified was age (Table  4 ). Specifically, an age exceeding 66.3 years was recognized as a critical risk factor for successful detection. Other factors assessed, including BMI, histological type and grade, presence of leiomyomas or adenomyosis, history of pelvic surgery (including cesarean section), endometriosis, total tumor volume, and depth of invasion into the myometrium, did not demonstrate a statistically significant impact on detection success. In the examination of detection failure on either side of the pelvis, additional significant risk factors were identified: depth of tumor invasion greater than 19% of myometrial thickness ( p  = 0.041) on the right side, history of pelvic surgery ( p  = 0.023), and total tumor volume exceeding 153,822 mm³ ( p  = 0.003) on the left side (Table  4 ). Table 4 Risk factors for SLN detection failure Risk factor SLN on both sides SLN on right side SLN on left side Not detected on either side n  = 68 n (%) Detected at least on one side n  = 180 n (%) p Not detected n  = 102 n (%) Detected n  = 146 n (%) p Not detected n  = 115 n (%) Detected n  = 133 n (%) p Age > 70 36 (52.9%) 70 (38.9%) 0.046 55 (53.9%) 51 (34.9%) 0.003 58 (50.4%) 48 (36.1%) 0.023 Age > 66.3 49 (72.1%) 88 (48.9%) 0.001 73 (71.6%) 64 (43.8%)  30.7 43 (63.2%) 90 (50.0%) 0.062 62 (60.8%) 71 (48.6%) 0.059 66 (57.4%) 67 (50.4%) 0.3 Pelvic surgery 12 (17.6%) 24 (13.3%) 0.4 16 (15.7%) 20 (13.7%) 0.7 23 (20.0%) 13 (9.8%) 0.023 Cesarean section 4 (5.9%) 17 (9.4%) 0.4 8 (7.8%) 13 (8.9%) 0.8 9 (7.8%) 12 (9.0%) 0.7 Adenomyosis 7 (10.3%) 22 (12.2%) 0.7 9 (8.8%) 20 (13.7%) 0.2 12 (10.4%) 17 (12.8%) 0.6 Myomatosis 25 (36.8%) 75 (41.7%) 0.5 39 (38.2%) 61 (41.8%) 0.6 48 (41.7%) 52 (39.1%) 0.7 Conisation history 4 (5.9%) 14 (15.8%) 0.8 8 (7.8%) 10 (6.8%) 0.8 8 (7.0%) 10 (7.5%) 0.9 HR histological type 21 (30.9%) 49 (27.4%) 0.6 29 (28.7%) 41 (28.1%) 0.9 29 (25.2%) 41 (31.1%) 0.3 LVSI 26 (38.2%) 65 (36.1%) 0.8 39 (38.2%) 52 (35.6%) 0.7 41 (35.7%) 50 (37.6%) 0.8 Invasion depth > 19% 55 (82.1%) 126 (70.0%) 0.056 81 (80.2%) 100.0 (68.5%) 0.041 88 (77.2%) 93 (69.9%) 0.2 Cervix infiltration 15 (22.1%) 33 (18.3%) 0.5 23 (22.5%) 25 (17.1%) 0.3 26 (22.6%) 22 (16.5%) 0.2 Tumor volume > 153,822 (mm 3 ) 22 (38.6%) 36 (25.0%) 0.055 27 (31.8%) 31 (26.7%) 0.4 36 (39.1%) 22 (20.2%) 0.003 Risk factors for SLN detection failure The multivariate model for bilateral SLN detection failure included the following predictors: age > 66.3 years ( p  = 0.002), BMI > 30.7 ( p  = 0.063), invasion depth > 19% ( p  = 0.12), and tumor volume > 153,822 mm³ ( p  = 0.3). Among these, age > 66.3 years emerged as the strongest predictor (aOR = 0.371, 95% CI: 0.196–0.679), indicating that older patients had significantly lower odds of successful SLN detection (Table  5 ). Figure S1 presents the regularization path illustrating the coefficients of variables selected by LASSO regression for predicting sentinel lymph node detection failure on both sides, while Figure S2 shows the model performance for SLN detection failure (Supplementary materials). Table 5 Univariate and multivariate logistic regression for risk factors of SLN detection both sides failure Risk factor Not detected on either side n  = 68 n (%) Detected at least on one side n  = 180 Univariate logistic regression Multivariate logistic regression aOR (95%CI) p aOR (95%CI) p Age > 66.3 49 (72.1%) 88 (48.9%) 0.371 (0.199–0.670) 0.001 0.371 (0.196–0.679) 0.002 BMI > 30.7 43 (63.2%) 90 (50.0%) 0.581 (0.325–1.025) 0.062 0.568 (0.310–1.025) 0.063 Pelvic surgery 12 (17.6%) 24 (13.3%) 0.718 (0.342–1.574) 0.4 Cesarean section 4 (5.9%) 17 (9.4%) 1.669 (0.591–5.967) 0.4 Adenomyosis 7 (10.3%) 22 (12.2%) 1.213 (0.515-3.200) 0.7 Myomatosis 25 (36.8%) 75 (41.7%) 1.229 (0.695–2.204) 0.5 Conisation history 4 (5.9%) 14 (15.8%) 1.349 (0.464–4.897) 0.8 HR histological type 21 (30.9%) 49 (27.4%) 1.162 (0.624–2.120) 0.6 LVSI 26 (38.2%) 65 (36.1%) 0.913 (0.515–1.636) 0.8 Invasion depth > 19% 55 (82.1%) 126 (70.0%) 0.500 (0.239–0.980) 0.052 0.560 (0.260–1.139) 0.12 Cervix infiltration 15 (22.1%) 33 (18.3%) 0.793 (0.405–1.609) 0.5 Tumor volume > 153,822 (mm 3 ) 22 (38.6%) 36 (25.0%) 0.556 (0.306–1.019) 0.055 0.714 (0.379–1.361) 0.3 p = p-value, CI = Confidence Interval, aOR = Adjusted Odds Ratio Univariate and multivariate logistic regression for risk factors of SLN detection both sides failure p = p-value, CI = Confidence Interval, aOR = Adjusted Odds Ratio For right-sided SLN detection failure, significant independent predictors included age > 66.3 years ( p   30.7 ( p  = 0.024), and invasion depth > 19% ( p  = 0.032), with HR histological type demonstrating borderline significance ( p  = 0.079). For left-sided SLN detection failure, significant predictors included age > 66.3 years ( p  = 0.003), history of pelvic surgery ( p  = 0.032), lymphovascular space invasion (LVSI) ( p  = 0.037), and tumor volume > 153,822 mm³ ( p  = 0.002). Notably, age > 66.3 years was consistently associated with SLN detection failure on both sides, whereas other risk factors exhibited side-specific patterns, suggesting distinct mechanisms influencing SLN detection failure on each side (Table  6 ). Table 6 Univariate and multivariate logistic regression for risk factors of SLN detection failure on each side Risk factor Right side Left side Univariate logistic regression Multivariate logistic regression Univariate logistic regression Multivariate logistic regression aOR (95%CI) p aOR (95%CI) p aOR (95%CI) p aOR (95%CI) p Age > 66.3 0.310 (0.179–0.528) < 0.001 0.325 (0.184–0.566)  30.7 0.611 (0.364–1.017) 0.060 0.533 (0.305–0.917) 0.024 0.754 (0.455–1.244) 0.3 0.795 (0.460–1.371) 0.4 Pelvic surgery 0.853 (0.419–1.759) 0.7 0.433 (0.203–0.889) 0.025 0.428 (0.192–0.916) 0.032 Cesarean section 1.148 (0.465–3.003) 0.8 1.168 (0.476–2.965) 0.7 Adenomyosis 1.640 (0.733–3.939) 0.2 1.258 (0.578–2.818) 0.6 Myomatosis 1.159 (0.692–1.951) 0.6 0.896 (0.538–1.491) 0.7 Conisation history 0.864 (0.329–2.340) 0.8 1.087 (0.414–2.943) 0.9 HR histological type 1.067 (0.608–1.861) 0.8 0.538 (0.266–1.072) 0.079 0.731 (0.416.1.272) 0.3 0.770 (0.407–1.444) 0.4 LVSI 0.519 (0.273–0.978) 0.043 1.087 (0.648–1.830) 0.8 2.087 (1.057–4.230) 0.037 Invasion depth > 19% 0.894 (0.530–1.512) 0.7 0.494 (0.255–0.928) 0.032 0.679 (0.380–1.199) 0.2 0.734 (0.377–1.417) 0.4 Cervix infiltration 0.710 (0.376–1.342) 0.3 0.678 (0.358–1.276) 0.2 0.675 (0.334–1.349) 0.3 Tumor volume > 153,822 (mm 3 ) 0.875 (0.499–1.541) 0.6 0.407 (0.228–0.716) 0.002 0.346 (0.176–0.662) 0.002 p = p-value, CI = Confidence Interval, aOR = Adjusted Odds Ratio Univariate and multivariate logistic regression for risk factors of SLN detection failure on each side p = p-value, CI = Confidence Interval, aOR = Adjusted Odds Ratio

Endometrial carcinoma (EC) is a common malignancy in women globally with a stable incidence. According to the latest global cancer statistics, the total number of new cases in 2022 was estimated at 420,242 with 97,704 associated deaths, reflecting a cumulative lifetime risk of 1.01% for developing corpus uteri cancer by the age of 74 [ 1 ]. In 2018, 121,578 women in Europe were diagnosed with endometrial cancer [ 2 ] and the EUROCARE-5 study reported a five-year survival rate of 76% for European women [ 3 ]. Due to the presence of early symptomatology, approximately 75% of patients are diagnosed at an early stage. The primary treatment modality is surgical intervention, with only a subset of patients subsequently receiving adjuvant therapy. The standard surgical approach involves total hysterectomy along with surgical-pathological staging of the lymphatic nodes. According to the guidelines established by the National Comprehensive Cancer Network (NCCN) [ 4 ], sentinel lymphatic node (SLN) biopsy is considered an acceptable alternative to systematic lymphadenectomy for lymph node staging in patients with stage I and II disease, as endorsed by the joint guidelines of the European Society of Gynaecological Oncology, European Society for Radiotherapy and Oncology, and European Society of Pathology (ESGO/ESTRO/ESP) [ 5 ] and the Society of Gynecological Oncology (SGO) [ 6 ]. We conducted a retrospective, non-randomized study involving patients with endometrial cancer who underwent open surgical procedures due to comorbidities. Our investigation compared the success rates of sentinel lymph node (SLN) detection between two groups: one receiving tracer application via both intracervical and intrafundal methods, and the other via intracervical and subserosal methods. The risk of failure in SLN detection was evaluated based on the application technique and additional patient characteristics.

Our analysis did not reveal significant differences in SLN mapping success between groups receiving intracervical + intrafundal and intracervical + subserosal tracer applications among patients with endometrial cancer treated via open surgery. Overall, older age emerged as the most critical risk factor for SLN detection failure. Cervical injection of the tracer is straightforward and appears to achieve a sufficient detection rate in endometrial cancer patients when compared to intrafundal and/or subserosal applications.

We compared two groups of patients with endometrial cancer who utilized different tracer application techniques for sentinel lymph node marking. All patients underwent total hysterectomy and bilateral salpingo-oophorectomy with SLN detection via an open midline incision laparotomy, necessitated by their medical conditions. In some cases, lymph node sampling or systematic pelvic lymphadenectomy was performed; however, paraaortic lymphadenectomy was not indicated due to the patients’ comorbidities. The primary objective of surgical management for EC is to excise the primary tumor and to identify prognostic factors that determine the necessity of adjuvant therapy. Laparoscopic or robotic hysterectomy is a viable option for most patients when medically appropriate [ 4 – 6 , 11 – 14 ]. Randomized trials have shown that minimally invasive surgery (MIS) for EC staging results in lower rates of perioperative and postoperative complications compared to laparotomy, without adversely affecting oncologic outcomes [ 15 – 17 ]. However, many patients with EC are often obese and have multiple comorbidities that may contraindicate minimally invasive procedures. For those in whom MIS is not feasible, laparotomy through a midline incision remains the appropriate surgical approach [ 18 , 19 ]. There remain numerous medical facilities worldwide where only open surgical access is utilized, and this study may provide valuable insights for these regions [ 20 – 22 ]. According to the joint European guidelines from ESGO/ESTRO/ESP [ 5 ], as well as NCCN guidelines from the United States [ 4 ], Germany [ 23 ], and the United Kingdom [ 24 ], a negative SLN biopsy is accepted as confirmation of pN0 status for uterine-confined EC, provided the procedure adheres to the guidelines’ principles [ 25 ]. A hospital-based study involving over 45,000 patients in the United States reported an increase in SLN mapping from 1.8% in 2012 to 25.3% in 2018 [ 26 ], a trend that was not linked to an increase in cancer-specific mortality [ 27 ]. The same change in operational approach is also noted in Europe [ 13 ]. An additional benefit of SLN identification is the potential for histopathological ultrastaging, including immunostaining. When pN0 status is unequivocally confirmed, adjuvant radiotherapy is typically avoided, even in cases where it would have previously been indicated. This approach significantly reduces patient morbidity, including the risk of radiation-related fractures [ 28 ]. There are four potential methods for tracer application in SLN detection in EC: (1) intracervical (1–2 cm into the cervix); (2) superficial (1–3 mm into the cervix); (3) subserosal (1–3 mm deep into the uterine fundus); and (4) intrafundal (1–2 cm into the uterine body). Some studies have reported a statistically significant increase in successful SLN detection following cervical injection compared to fundal injection [ 29 – 33 ], while others found no statistically significant differences between the methods [ 34 – 36 ]. A recent meta-analysis published in 2024, encompassing 62 eligible studies, revealed an overall false negative rate of SLN detection of 4% (95% CI: 3–5) [ 37 ]. Notably, the false negative rate was significantly lower for cervical injections compared to uterine injections (4% vs. 10%, p  = 0.024). In contrast to previous studies, we administered the tracer intracervically to all patients and compared the efficacy of subserosal and intrafundal injection methods. Our study demonstrated a slightly higher success rate in SLN detection for the group receiving intracervical and intrafundal injections, although these results were not statistically significant (Table  2 ). The detection rate in our study was relatively low. However, a meta-analysis of 55 studies ( n  = 4915) reported bilateral detection rates ranging from 6 to 88%, with a pooled average of 50% (95% CI, 44–56) [ 38 ], which is comparable to our results. Another systematic review of 17 studies involving more than 30 patients found bilateral detection rates from 19 to 80% [ 39 ]. The overall detection rate of SLNs following cervical injection varied from 62 to 100%, while the rates for corporeal injection ranged from 73 to 95% [ 39 ]. A prospective trial indicated a one-sided detection rate of 76.8% (53/69) after hysteroscopic ICG injection [ 33 ], which aligns closely with our findings of a detection rate of 77.2% (23/101) following cervical + intrafundal application (Table  2 ). According to the Cochrane database, the detection rates for SLNs using blue dye alone were 77.8% (95% CI: 70.0-85.6), with Technetium-99 m alone at 80.9% (95% CI: 63.9–97.9), a combination of Technetium-99 m and blue dye at 86.3% (95% CI: 80.7–91.9), and with ICG at 92.5% (95% CI: 81.8–97.1) [ 40 ]. In our study, we successfully detected at least one SLN in 72.6% (180/248) of all patients (Table  2 ), a finding that closely resembles the outcomes reported in the Cochrane database for blue dye usage. In conclusion, contemporary practice guidelines advocate for both superficial (1–3 mm) and deep (1–2 cm) cervical injections of indocyanine green (ICG), which effectively delivers the dye to the primary layers of the cervical and uterine lymphatic channels [ 4 – 6 , 11 , 38 , 41 ]. Although ICG demonstrates a superior detection rate for SLNs and most current guidelines favor ICG, it is not universally available in all medical centers worldwide. Thus, we aimed to investigate whether SLN detection could be enhanced by employing various application techniques with standard tracers. Radiocolloid 99m Tc is a widely utilized radioactive isotope in gynecologic oncologic surgery for SLN mapping, particularly in patients with clinically early-stage vulvar cancer [ 42 ]. However, this technique exposes both patients and healthcare professionals to radiation. Although radioactive tracers can be prepared up to 24 h before surgery, employing ultrashort protocols is advantageous in minimizing unnecessary radiation exposure for healthcare providers [ 43 ]. Many countries enforce stringent regulations governing the handling, storage, and disposal of radioactive materials. Nevertheless, hospitals equipped with nuclear medicine departments have established standardized protocols to ensure the efficient and safe implementation of this technique. According to a recent questionnaire study, which involved a total of 302 participants of the European Society of Gynecologic Oncology (ESGO) and the International Gynecologic Cancer Society (IGCS), 77.5% of respondents use ICG alone to detect SLN in endometrial carcinoma and on the other hand 16% never use ICG [ 44 ]. However, most participants came from developed countries (e.g. 56.3% of participants from Europe) and centers from developing countries were minimally represented (e.g. only 0.7% from Africa). This shows that ICG is not available in all oncogynecological facilities in developed countries and is minimally available in developing countries, where standard tracers are routinely used. None of our patients underwent para-aortic lymphadenectomy, even in cases of (HR) histology. In our study, both techniques primarily localized SLNs on the medial aspect of the external iliac vessels, ventral to the hypogastric region, and in the upper part of the obturator region (Table  3 ), consistent with common SLN detection sites [ 4 ]. Recent studies that have conducted SLN mapping in conjunction with systematic lymphadenectomy indicated that the risk of isolated para-aortic metastasis is only 0.8–1% [ 45 – 47 ], with some sources reporting rates as high as 2% [ 4 ]. This finding underscores the reliability of SLN mapping when adhering to the SLN algorithm, supporting the omission of routine para-aortic lymphadenectomy, even in patients with HR histology [ 11 , 48 ]. A meta-analysis of nine prospective studies involving 429 patients with high-grade EC demonstrated that SLN mapping accurately identified 80 out of 87 patients with positive lymph nodes [ 49 ]. The observed false-negative rate was 8%, comparable to rates found in low-grade endometrial cancer. Recent retrospective studies have shown similar prognoses for patients undergoing complete lymphadenectomy compared to those undergoing SLN biopsy alone among patients with HR histology [ 50 , 51 ]. A study comparing 326 patients with HR endometrial cancer who underwent SLN mapping with another 326 patients who received pelvic lymphadenectomy found no significant difference in three-year disease-specific survival probability (88.2% vs. 82.7%, p  = 0.056). Notably, patients who underwent SLN mapping had a higher rate of positive pelvic lymph nodes (18% vs. 14%, p  = 0.04) and exhibited better overall survival ( p  = 0.047) [ 52 ]. Additionally, another meta-analysis of 17 original studies involving 1,322 women with HR endometrial cancers further corroborated SLN detection as a valuable technique for diagnosing lymph node involvement, offering a viable alternative to conventional lymphadenectomy [ 53 ]. Age emerged as the only significant factor associated with bilateral SLN detection failure in our study (Table  4 ). Interestingly, we observed distinct risk factors for SLN detection failure on each side of the pelvis. The depth of myometrial invasion ( p  = 0.041) significantly affected SLN mapping failure on the right side, whereas tumor size ( p  = 0.003) and a history of pelvic surgery ( p  = 0.023) were significant factors on the left side (Table  4 ). Although several trials have examined risk factors for SLN detection failure, none have specifically analyzed discrepancies in detection rates between the sides of the pelvis. These findings are unlikely to be attributed to embryological factors, as the uterus develops from the fusion of the right and left Müllerian ducts (paramesonephric ducts). While fusion abnormalities may lead to developmental anomalies such as uterus duplex, uterus bicornis, or uterus arcuatus, they do not account for potential differences in lymphatic drainage. While the fusion of the two uterine components could theoretically contribute to variability in vascular and lymphatic drainage, the observed results should not yield statistically significant differences. Instead, this variability may be explained by the differing topographical arrangements of the superficial and deep lymphatic trunks on the right and left sides, with the left side potentially exhibiting a distinct course due to the presence of the rectosigmoid [ 54 ]. A significant influence on side-to-side detection success may be related to the anatomical positioning of the iliac veins, where compression of the left common iliac vein by the right common iliac artery against the lumbar spine could create an obstruction or a site predisposed to blood stasis. Additionally, this anatomical relationship may predispose individuals to venous thromboembolism and the development of May-Thurner syndrome, which, considering the associated lymphatic vessels, may also affect lymphatic flow rates [ 55 ]. Taskin et al. analyzed 357 patients with clinically uterine-confined endometrial cancer, employing cervical application of ICG (64.7%) or blue dye (35.3%) for SLN mapping during MIS (79.3%) and open surgeries (26.1%). The overall and bilateral SLN detection rates were 91.9% and 71.4%, respectively. The mapping rates using ICG and blue dye were comparable ( p  = 0.526). The only observed risk factors for unmapped patients were higher BMI ( p  = 0.037) and, similar to our study, advanced age ( p  = 0.030). Other factors monitored, including stage, histology type, grade, tumor size, myometrial invasion, and presence of LVSI, did not significantly impact SLN mapping failure [ 56 ]. A more recent study of 71 women with endometrial cancer revealed that the likelihood of SLN mapping failure increased by 1.6 times for each 5-unit increment in BMI ( p  = 0.040). BMI was established as an independent risk factor for mapping failure in both univariate (odds ratio [OR] 3.267, p  = 0.045) and multivariate analyses (OR 5.779, p  = 0.020) [ 57 ]. Obesity as a risk factor for SLN detection in EC was also identified in a retrospective ALIEN study, which compared two groups: 172 EC patients who received 2 mL of indocyanine green (ICG) and 180 EC patients who received 4 mL of ICG [ 58 ]. While the overall detection rate differed significantly between the two groups (92.4% vs. 97.8%, respectively; p  = 0.024), the bilateral detection rate did not show a statistically significant difference (84.9% vs. 86.1%, p  = 0.76). However, a higher BMI was associated with lower overall detection rates in univariate analysis ( p  = 0.0006). Specifically, the detection rate significantly declined from 97 to 91% when BMI exceeded 30 ( p  = 0.05) [ 58 ]. In our study, a BMI greater than 30.7 was identified as a potential factor for detection failure, although this did not reach statistical significance (Table  4 ). A meta-analysis indicated that only cervical injection compared with uterine injection ( p  = 0.003) and the use of ICG ( p  = 0.001) were significant factors associated with higher rates of bilateral SLN detection, while other factors such as BMI, histology type, tumor grade, and surgical approach did not demonstrate significant differences in SLN detection [ 38 ]. All of these differing conclusions highlight the complexity, if not the impossibility, of clearly defining the anatomical and histopathological risk factors for SLN failure. Most studies, in line with our findings, confirmed that older age significantly impacts the success of SLN detection. The increased risk of failure in older patients may be attributed to the heterogeneity in lymphatic vessel uptake due to age-related changes in lymphatic vessel walls, surrounding tissues, and musculature. We observed statistically significant effects of myometrial invasion depth ( p  = 0.041) on the right side and tumor volume ( p  = 0.003) on the left side with respect to SLN failure (Table  4 ). Notably, no published studies have specifically examined risk factors for side-specific SLN detection failure. However, the characteristics related to overall detection failure were also analyzed by Sozzi et al., who did not find statistical significance for myometrial infiltration ( p  = 0.58) or tumor size ( p  = 0.57) [ 59 ]. Another potential risk factor, depth of myometrial invasion exceeding 50%, similarly showed no statistically significant impact in a separate study [ 60 ]. Several prospective trials are currently underway, investigating various aspects of SLN detection in EC patients. The ECLAT trial ( NCT03438474 ) aims to determine whether para-aortic nodal assessment, either through SLN biopsy or systematic lymphadenectomy, is necessary and to identify the patient subgroups that may benefit from this approach [ 61 ]. The SELYE trial ( NCT04845828 ) evaluates three-year progression-free survival in patients undergoing laparoscopic or robotic SLN biopsy compared to systematic lymph node dissection [ 62 ]. The ALICE trial ( NCT03366051 ) seeks to generate data on oncological outcomes in patients with HR endometrial cancer who undergo SLN biopsy versus routine lymphadenectomy [ 63 ]. The limitations of this study include its retrospective design, which restricts the strength of the evidence. Additionally, as some procedures were performed recently, the complete oncological outcomes remain unknown, raising questions about the proxy for oncological safety. Conversely, this research was conducted at a single tertiary cancer center, focusing on a large cohort of patients undergoing open surgery for endometrial cancer, thereby minimizing selection bias. The overall detection rates in both groups align with findings from similar studies not utilizing ICG. To our knowledge, this study represents the first comparative analysis of these two tracer application methods in endometrial cancer patients undergoing open abdominal surgery.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4