Search

Objective: To identify predictive clinico-pathologic factors for concurrent endometrial carcinoma (EC) among patients with endometrial intraepithelial neoplasia (EIN) using machine learning., Methods: a retrospective analysis of 160 patients with a biopsy proven EIN. We analyzed the performance of multiple machine learning models (n = 48) with different parameters to predict the diagnosis of postoperative EC. The prediction variables included: parity, gestations, sampling method, endometrial thickness, age, body mass index, diabetes, hypertension, serum CA-125, preoperative histology and preoperative hormonal therapy. Python 'sklearn' library was used to train and test the models. The model performance was evaluated by sensitivity, specificity, PPV, NPV and AUC. Five iterations of internal cross-validation were performed, and the mean values were used to compare between the models., Results: Of the 160 women with a preoperative diagnosis of EIN, 37.5% (60) had a post-op diagnosis of EC. In univariable analysis, there were no significant predictors of EIN. For the five best machine learning models, all the models had a high specificity (71%-88%) and a low sensitivity (23%-51%). Logistic regression model had the highest specificity 88%, XG Boost had the highest sensitivity 51%, and the highest positive predictive value 62% and negative predictive value 73%. The highest area under the curve was achieved by the random forest model 0.646., Conclusions: Even using the most elaborate AI algorithms, it is not possible currently to predict concurrent EC in women with a preoperative diagnosis of EIN. As women with EIN have a high risk of concurrent EC, there may be a value of surgical staging including sentinel lymph node evaluation, to more precisely direct adjuvant treatment in the event EC is identified on final pathology., (© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).)

Objective: To determine the prevalence of underlying high-intermediate (high-IM) and high-risk endometrial cancer (EC) in patients with preoperative diagnosis of Endometrial intraepithelial neoplasia (EIN) and to assess the impact of the information retrieved from the sentinel lymph node (SLN) on adjuvant therapy., Methods: Retrospective cohort study of women undergoing hysterectomy, optional bilateral salpingo-oophorectomy (BSO) and lymph nodes assessment for EIN between December 2007 and August 2021., Results: One hundred and sixty two (162) eligible patients were included, of whom 101 (62.3%) had a final diagnosis of EIN, while 61 (37.7%) were ultimately diagnosed with carcinoma. Out of 15 patients with high-IM to high-risk disease (9.25% of all EIN), 12 had grade 2-3 EC including 8 with >50% myometrial invasion, 2 with serous subtype, 1 with cervical invasion and 2 with pelvic lymph nodes involvement. Of the 3 patients with grade 1 EC, one patient had disease involving the adnexa and 2 patients had tumor invading >50% of the myometrium and with lymphovascular space invasion (LVSI). Ten patients received vaginal brachytherapy after surgery, 3 patients with extrauterine spread were treated with systemic chemotherapy followed by vaginal brachytherapy and pelvic external-beam radiotherapy and 2 patients with early-stage serous carcinoma received chemotherapy followed by vaginal brachytherapy., Conclusions: Information from SLN, even when negative, can be helpful in the management of patients with EC after preoperative EIN, as some patients are found to have high-IM to high-risk disease on final pathology. These patients would require either re-staging surgery or adjuvant external beam radiotherapy, both could be avoided by proper staging., (Copyright © 2022 Elsevier Inc. All rights reserved.)

Objective: To evaluate if the prognostic value of lymphovascular space invasion (LVSI) is different in endometrial cancer patients with negative lymph nodes following sentinel lymph node (SLN) mapping or lymph node dissection (LND) as staging procedure., Material and Methods: A retrospective study of 510 patients diagnosed with endometrial carcinoma in our institution between 2007 and 2014. We excluded patients that were diagnosed with positive nodes (Stage IIIc). We compared patients' characteristics and survival outcomes as function of their LVSI status (positive LVSI vs negative LVSI subgroups) in each cohort separately., Results: 413 patients met the inclusion criteria, out of whom 239 underwent SLN and 174 patients underwent LND only. In the SLN group, life table analysis showed 5-year OS and PFS of 80% and 72% in patients with LVSI compared to 96%, and 93% without LVSI. Same trend was observed among patients with LND with 5-year OS and PFS of 74% and 64% in patients with LVSI compared to 97%, and 90% without LVSI. On multivariable analysis, adjusted for age, FIGO stage, grade and maximal tumor size, the favorable survival of negative LVSI remained only in the LND cohort (SLN cohort: HR 1.2, CI [0.3-4.0], P = 0.8 and HR 1.7, CI [0.7-4.3], p = 0.2 for OS and PFS, respectively; LND cohort: HR 3.1, CI [1.4-6.5], p < 0.001 and HR 2.5, CI [1.2-4.9], p = 0.01 for OS and PFS, respectively)., Conclusions: The prognostic value of LVSI disappears when patients undergo staging with SLN and are found to have negative nodes in contrast to those who have undergone LND. Future studies should confirm our observation on patients with negative sentinel nodes, and plan on tailoring adjuvant treatment to this specific subgroup., Competing Interests: Declaration of Competing Interest The authors report no conflict of interest., (Copyright © 2021 Elsevier Inc. All rights reserved.)

Background: Sentinel lymph node (SLN) mapping has emerged as a promising solution to the ongoing debate regarding lymphadenectomy in the initial surgical management of endometrial cancer. Currently, little is known about its possible impact on location of disease recurrence compared to systematic lymphadenectomy., Methods: In this retrospective study, 472 consecutive patients with endometrial cancer who underwent either SLN mapping (SLN cohort, n=275) or systematic lymphadenectomy (LND cohort, n=197) from sequential, non-overlapping historical time points were compared. Clinical characteristics were extracted from a prospectively gathered electronic database. Both overall and pelvic sidewall recurrence free survival (RFS) were evaluated at 48-month post-operative follow-up., Results: No significant difference in overall RFS could be identified between the cohorts at 48months (HR 0.74, 95% CI 0.43-1.28, p=0.29). However, the SLN cohort had improved pelvic sidewall RFS compared to the LND cohort (HR 0.32, 95% CI 0.14-0.74, p=0.007). The pelvic sidewall recurrences accounted for 30% of recurrences in the SLN cohort (8 out of 26 recurrences) compared to 71.4% in the LND cohort (20 out of 28 recurrences)., Conclusions: SLN mapping may enable more efficient detection of the LNs at greatest risk of metastasis and help to guide adjuvant therapy, which in turn seems to decrease the risk of pelvic sidewall recurrences., (Copyright © 2017 Elsevier Inc. All rights reserved.)

Background and Aims: With the debate over extent of lymphadenectomy in endometrial cancer, sentinel lymph node (SLN) mapping may provide a focused approach to evaluate the most relevant lymph nodes (LN) while minimizing the complications. We evaluated SLN mapping using filtered technetium(99), indocyanine green (ICG), and blue dye., Methods: Prospective evaluation of 100 patients who underwent SLN mapping by using submucosal and deep stromal cervical injections of technetium(99), ICG, and blue dye as part of the staging for endometrial cancer., Results: 286 SLNs were mapped (2.9 per patient) in 92% of patients. The bilateral detection rate was 76%. ICG had a significantly higher SLN detection rate than blue dye in both overall (87% vs 71%, respectively; p=0.005) and bilateral (65% vs 43%, respectively; p=0.002) detection, but similar SLN detection rates compared to technetium(99) in both overall (87% vs 88%, respectively; p=0.83) and bilateral (65% vs 71%, respectively; p=0.36) detection. In eight cases, the SLN was in the para-aortic area and in 14 cases in the pre-sacral, hypogastric vein, or parametrial area. In nine cases, the SLN was positive for metastasis, and in seven cases the SLN was the only positive node. One SLN was falsely negative. No complications or anaphylactic reactions occurred., Conclusion: Intra-operative SLN mapping using cervical injection is feasible in patients with endometrial cancer and yields adequate detection rates. It allows mapping of SLNs in areas (pre-sacral, hypogastric vein, parametrial) not routinely sampled. Given the poorer performance of blue dye, surgeons may omit its use if a combination of ICG and technetium(99) is used., (Copyright © 2015 Elsevier Inc. All rights reserved.)

Objective: The objective of this study is to evaluate the detection rate and diagnostic accuracy of sentinel lymph node (SLN) mapping using intra-operative cervical injection of filtered 99mTc-sulfur colloid (99mTc-SC) and patent blue in patients with endometrial cancer., Methods: Prospective evaluation of the first 100 endometrial cancer patients undergoing SLN mapping using cervical injection of patent blue combined with filtered 99mTc-SC in the operating room was done. Patients underwent robotic-assisted lymphatic mapping with frozen section, hysterectomy, BSO, and completion bilateral lymphadenectomy (including para-aortic nodes in grade 2 and 3 tumors)., Results: At least one SLN was detected in 92% of patients; in 66 of these (72%) bilateral SLN were detected, and in 15 cases the SLN was in the para-aortic area. Eleven percent of all patients had lymph node metastases, and 4 of which had pre-operative grade 1 tumor. The SLN was the only positive node in 44% of the cases with positive nodes. Sensitivity was 89% with 1 false negative result, yielding a negative predictive value of 99% (95% CI 93-100). Specificity was 100% (95% CI 94-100), and positive predictive value was 100% (95% CI 60-100). No complications or anaphylactic reactions were noted., Conclusions: Intra-operative SLN biopsy, using cervical injection of patent blue and filtered 99mTc-SC in endometrial cancer patients is feasible and yields adequate detection rates., (Copyright © 2012 Elsevier Inc. All rights reserved.)