摘要
Background: Despite excellent per-lesion performance for peritoneal staging, the additional clinical value of diffusion-weighted magnetic resonance imaging (DWI/MRI) compared to computed tomography (CT) remains to be established in ovarian cancer. Our purpose was to evaluate whole body (WB)-DWI/MRI for diagnosis, staging and operability assessment of patients suspected for ovarian cancer compared to CT. Methods: One hundred and sixty-one patients suspected for ovarian carcinoma underwent 3 TWB-DWI/MRI and contrast-enhanced CT. WB-DWI/MRI and CT were compared for confirmation of the malignant nature and primary origin of the ovarian mass, Federation Internationale de Gynecologie et d Obstetrique (FIGO) staging and prediction of incomplete resection using institutional operability criteria. Interobserver agreement between two readers was determined for WB-DWI/MRI and CT. Results: WB-DWI/MRI showed a significantly higher accuracy than CT (93 versus 82%, p = 0.001) to confirm the malignant nature of the ovarian mass and correctly identified 26 of 32 (81%) cancers of non-ovarian origin compared to 10/32 (31%) for CT (p < 0.001). WB-DWI/MRI assigned more ovarian carcinoma patients to the correct FIGO stage (82/94, 87%) compared with CT (33/94, 35%). For prediction of incomplete resection, WB-DWI/MRI showed significantly higher sensitivity (94 versus 66%), specificity (97.7 versus 77.3%) and accuracy (95.7 versus 71.3%) compared to CT (p < 0.001). Interobserver agreement was almost perfect (kappa = 0.90) for WB-DWI/MRI and moderate (kappa = 0.52) for CT for prediction of incomplete resection. Conclusions: WB-DWI/MRI was superior to CT for primary tumour characterisation, staging and prediction of incomplete resection in patients suspected for ovarian cancer.
- 出版日期2017-9