Assessing the Diagnostic Accuracy of 3D Transvaginal Ultrasonography for Endometrial Cancer Staging: A Comparative Analysis of Different Methods

Background: Endometrial carcinoma (EC) is among the most prevalent malignancies in the female population and represents the only gynecological cancer currently characterized by a rising trend in both incidence and mortality, ranking as the third most frequent carcinoma among women aged 50 to 69 in Italy. Recently, the understanding of the disease has shifted from the traditional Bokhman dualistic model to a complex molecular framework (TCGA), identifying four distinct prognostic subgroups. The main therapeutic strategy involves a tailored surgical approach, typically incorporating total extrafascial hysterectomy, bilateral salpingo-oophorectomy, and sentinel lymph node (SLN) mapping, which has largely replaced systematic lymphadenectomy for early-stage disease. Following the surgical procedure, an integrated anatomopathological and molecular staging is conducted, in accordance with the landmark 2023 FIGO staging system and the updated 2025 ESGO-ESTRO-ESP guidelines. To ensure the most appropriate surgical management and accurately evaluate critical prognostic parameters, such as the depth of myometrial invasion, cervical involvement, and lymph node status, dedicated preoperative imaging is mandatory. While pelvic Magnetic Resonance Imaging (MRI) remains a cornerstone and the gold standard modality for local staging, its routine application can be limited by costs and patient-related contraindications. Consequently, three-dimensional transvaginal ultrasound (3D-TVUS) has emerged as a pivotal and highly accessible diagnostic tool. By allowing multiplanar reconstruction and objective volumetric assessments, current evidence demonstrates that 3D-TVUS is a dependable imaging modality, offering a diagnostic test accuracy comparable to MRI for the preoperative staging of deep myometrial and cervical stromal invasion in patients with endometrial cancer. Aim: To assess the suitability of three-dimensional transvaginal ultrasound (3D-TVUS) as a valid support or alternative to Magnetic Resonance Imaging (MRI) during the preoperative assessment of endometrial cancer (EC). Specifically, we aim to evaluate and compare their diagnostic accuracy in detecting deep myometrial invasion (≥ 50%) as a primary endpoint, alongside cervical and vaginal involvement as secondary endpoints. Materials and method: A prospective, double-blind, non-profit observational study was conducted on patients with a histopathological diagnosis of EC referred to the Gynecologic Oncology Clinic of the University of Padua. Patients underwent 3D-TVUS and pelvic MRI to assess myometrial (≥ 50%), cervical, and vaginal invasion. For each patient, clinical data (e.g., date of birth, parity, BMI, menopausal status), histological data from diagnostic biopsies and surgical specimens (histotype and grade), and radiological parameters (infiltration of the myometrium, cervix, vagina, and the presence of pathological lymph nodes) were collected. The diagnostic performances of 3D-TVUS and MRI were evaluated and compared using the final surgical histopathological examination as the reference standard. Results: 3D-TVUS showed a higher overall agreement with final histology compared to MRI (Cohen's Kappa 0.29 vs 0.25). For deep myometrial invasion, 3D-TVUS demonstrated slightly superior sensitivity (75.7% vs 71.4%) and NPV (78.1% vs 74.4%), with comparable overall accuracy (67.4% vs 66.7%). Regarding cervical invasion, MRI was more sensitive (58.3% vs 41.7%), but 3D-TVUS maintained an excellent specificity (98.7%), making it an optimal method for excluding cervical involvement. Conclusions: 3D-TVUS is a highly reliable diagnostic tool, comparable to MRI for the preoperative staging of endometrial cancer. Given its rapid accessibility and lower costs, 3D-TVUS is proposed as a first-line imaging modality for immediate and personalized risk stratification, reserving MRI for doubtful cases.