The Latest in Endometrial Cancer: Liquid Profiling, LVSI, and More

Photo Credit: Tippapatt

Recent research into endometrial cancer has explored minimally invasive versus open surgical approaches, liquid profiling, recurrence prediction, and more.

This year, a wealth of research into endometrial cancer (EC) has been published, from AI models to DNA sequencing. This roundup offers highlights from recently published studies and reviews.

Using Tech to Predict Recurrence

A deep learning model called HECTOR improved the prediction of distant recurrence in patients with EC, according to a study published in Nature Medicine.

• Developed to address the need for better prognostic tools, HECTOR analyzes tumor images and clinicopathological data to estimate distant recurrence risk. The model was trained on data from 2,751 patients with stages 1 to 3 EC, using hematoxylin and eosin-stained hysterectomy specimens, molecular data, and clinical outcomes.
• HECTOR performed well in validation, achieving C-indices of 0.789, 0.828, and 0.815 in one internal and two external test sets. It also successfully stratified patients into low-, intermediate-, and high-risk groups, predicting 10-year distant recurrence-free survival rates of 97%, 77.7%, and 58.1%, respectively.
• Per the findings, HECTOR outperformed the current gold standard of pathological and molecular analysis in predicting distant recurrence and was predictive of adjuvant chemotherapy benefit in the PORTEC-3 trial.
• The authors noted that while HECTOR shows promise in improving individualized prognostication and reducing costs, further validation in diverse prospective trials is needed.

Comparing Early EC Surgical Approaches

Minimally invasive surgery, using robotic-assisted or laparoscopic techniques, appears safer and more effective than open surgery for early-stage EC, according to a review published in Cancers.

• Current guidelines recommend minimally invasive surgery for early-stage EC due to its lower postoperative morbidity and comparable oncologic outcomes. However, existing trials may not fully capture the risks and benefits of different surgical approaches.
• The authors conducted a systematic review and network meta-analysis, including 99 studies with 181,716 women and 14 outcomes.
• Laparoscopic and robotic-assisted surgeries resulted in less blood loss, shorter hospital stays, and fewer complications than open surgery, though they took longer to perform. Robot-assisted surgery showed advantages in reducing intraoperative complications and improving disease-free survival.
• The authors concluded that minimally invasive surgery offers significant benefits for early EC treatment despite longer operation times.

Predicting Lymphovascular Space Invasion

The authors of a prospective study published in Radiology evaluated using MR elastography to noninvasively predict lymphovascular space invasion (LVSI) prior to surgery.

• Identifying LVSI typically requires post-hysterectomy pathological examination, but MR elastography, which assesses tumor stiffness, could offer a predictive alternative.
• The study included 101 patients, with MR elastography and MRI performed to measure tumor stiffness and develop clinical-radiologic models.
• The LVSI-positive group had significantly higher tumor stiffness, cancer antigen 125 levels, and tumor volumes compared with the LVSI-negative group.
• A combined model incorporating clinical-radiologic and MR elastography data predicted LVSI positivity with an area under the curve (AUC) of 0.93, outperforming the clinical-radiologic model alone (AUC, 0.77). The combined model also demonstrated improved specificity (89%) and positive predictive value (77%).
• The findings suggest that MR elastography and clinical data may provide a more accurate, noninvasive method for predicting LVSI in patients with EC.

Personalizing Treatment in Advanced EC

Cell-free DNA (cfDNA) sequencing was valuable for classifying metastatic or recurrent EC into molecular subtypes for targeted therapies, according to a study published in Cancer.

• In a cohort of 61 patients, the researchers successfully conducted cfDNA analysis in 89% of cases, with molecular classifications 87.5% accurate compared with tissue biopsies.
• The authors found actionable genetic alterations in 65% of cases, and 16% of patients received targeted therapies, achieving a 56% response rate and a median progression-free survival of 7.7 months.
• According to the study, cfDNA profiling is a valuable tool for guiding treatment decisions in metastatic endometrial cancer, though some discrepancies exist between liquid and tissue biopsies. The findings demonstrated the feasibility of using liquid biopsy to personalize treatment in advanced EC, with contributive results in over 88% of cases.