Apparent diffusion coefficient values on MRI as biomarkers for pre operative prediction of prognostic factors in endometrial cancer

Background/objectives Magnetic resonance imaging (MRI) is essential for preoperative assessment of endometrial carcinoma. Diffusion-weighted imaging (DWI)-derived apparent diffusion coefficient (ADC) values are being explored as quantitative, non-invasive biomarkers of tumor biology and aggressiveness. This study assessed the performance and interobserver reproducibility of ADC measurements for predicting prognostic factors. Methods This study was conducted from June 2023 to June 2025 and included histologically confirmed endometrial carcinoma patients who underwent preoperative pelvic MRI. Patients with 3 T MRI outside the institute or unfit for primary surgery were excluded. Two radiologists independently measured focused-region and whole-lesion ADC values. Interobserver agreement was assessed using paired t -tests and Bland–Altman analysis. Associations with histopathological prognostic factors were assessed using Pearson's coefficient. MRI and histopathological staging concordance was measured using weighted Cohen's kappa. ROC and ANOVA assessed diagnostic performance. Results Fifty-one patients were included, with endometrioid adenocarcinoma being the predominant histological subtype (92.2%). The mean ADC was 747.13 × 10 −6 mm 2 /s. MRI staging showed excellent agreement with histopathology (weighted κ = 0.866). Focused-ROI ADC demonstrated no significant interobserver variability, whereas whole-lesion ADC showed significant interobserver variation. Lower ADC values were associated with serous histology and deep myometrial invasion. ADC significantly differentiated tumor grade, myometrial invasion, and histological subtype. ROC analysis demonstrated good diagnostic performance of focused ADC for tumor grade and pelvic lymph-node involvement. Conclusion Focused ROI-based ADC assessment is practical, reproducible, contrast-free, and may improve risk stratification, surgical planning, and individualized management, especially where molecular testing is limited.