Role of multiparametric magnetic resonance imaging in the differentiation of benign and malignant uterine myometrial lesions

Abstract Background Differentiating uterine leiomyosarcomas from atypical and degenerated forms of benign leiomyomas remains of critical clinical and radiological challenge, often resulting in mismanagement. Several magnetic resonance imaging (MRI) descriptors, sequences, and algorithmic approaches have been proposed to improve this distinction. The aim of this study was to assess the diagnostic accuracy of various MRI descriptors in the differentiation, with the objective of achieving high accuracy and specificity enhancing diagnostic confidence. Methods This prospective cross-sectional analytic study included 35 female patients with uterine myometrial neoplastic masses. MRI key sequences previously proposed in the literature, along with specific MRI descriptors—namely ill-defined margins, high T1 signal (hemorrhage), intermediate T2 signal, enhancing low T2 signal, high T2 signal (degeneration), enhancing high T2 signal, marked enhancement, and mean diffusion restriction lower (R-DWI < Endo) and greater than endometrium (R-DWI > Endo)—were evaluated for their sensitivity, specificity, and diagnostic accuracy in predicting sarcoma. Results A well-defined mass, enhancing low T2 signal, and higher apparent diffusion coefficient (ADC) values were significantly more common in benign lesions than in malignant ones ( P < 0.05). Conversely, features such as ill definition, marked enhancement, enhancing high T2 signal, T1 hyperintensity suggestive of hemorrhagic necrosis, and R-DWI > Endo, along with other malignant MRI characteristics (infiltration, lymph nodes, and peritoneal deposits), were significantly associated with sarcomas ( P < 0.05). When combining the criteria of ill-defined margins, enhancing high T2 signal soft tissue, marked enhancement, and R-DWI > Endo, considering three or more of these features as positive for malignancy, the diagnostic performance improved markedly, demonstrating a higher sensitivity (90.9%), specificity (95.8%), and overall diagnostic accuracy (94.3%) ( P < 0.00001). Conclusions Diagnostic criteria based on ill-defined mass, high T2 solid tissues with post-contrast/marked enhancement, R-DWI > Endo and low ADC, with or without the presence of peritoneal deposits and lymphadenopathy, help distinguish malignant sarcomas from degenerated/atypical leiomyomas.