Diagnostic performance of diffusion tensor imaging in endometrial carcinoma: correlation with the 2023 FIGO staging

Abstract Background The updated 2023 FIGO staging for endometrial cancer (EC) integrates histopathologic and molecular variables to optimize risk stratification. Preoperative Magnetic resonance imaging (MRI) is the key imaging tool for evaluating key prognostic factors. We aimed to evaluate the performance of diffusion-tensor imaging (DTI) in the preoperative evaluation of EC, in correlation with the revised 2023 FIGO staging system, and to compare it with diffusion-weighted imaging (DWI). Methods A prospective analysis was performed on 43 patients with pathology-proven EC who underwent pelvic MRI and DTI. Apparent diffusion coefficient (ADC), fractional anisotropy (FA), and mean diffusivity (MD) values were calculated by two individual radiologists and compared with the final postoperative histopathological data, considering the 2023 FIGO staging. Results There was good interobserver reliability for ADC measurements (ICC = 0.873, 95% CI: 0.778–0.929; p < 0.001) and FA measurements (ICC = 0.777, 95% CI: 0.623–0.873; p < 0.001) with moderate reliability for MD (ICC = 0.633, 95% CI: 0.414–0.783; p < 0.001). Aggressive EC demonstrated significantly lower values for ADC, MD, and FA when compared to non-aggressive types ( p < 0.001). FA exhibited the highest diagnostic performance, achieving near-perfect discrimination, with an AUC of 0.999 at an optimal cutoff of ≤ 0.24. Besides, there was a strong negative correlation between ADC, MD, and FA values and the 2023 FIGO stage. Conclusions Aggressive EC subtypes showed significantly lower ADC, FA, and MD values than non-aggressive ECs. FA revealed the highest diagnostic performance with near-perfect discrimination. There was a significant inverse correlation between FA and MD parameters and the updated FIGO 2023 stage.