Magnetic resonance imaging in the differential diagnosis of ovarian masses: Capabilities of quantitative multiparametric evaluation

Objective. To estimate the capabilities of comprehensive magnetic resonance imaging (MRI) in the differential diagnosis of ovarian tumors. Subjects and methods. In 2011 to 2015, a total of 256patients with 289 ovarian masses underwent small pelvic and abdominal MRI (if necessary) to clarify the pattern and extent of the tumor process. MRI was performed on a 1.5 Tl scanner. The MRI protocol was to obtain T2-weighted images (WI) in three projections, STIR, T1-WI, DWI with b-factors 0, 1000м/мм2, to calculate the values of the diffusion coefficient, and to make the diffusion maps Dynamic 3D FatSat. Postprocessing involved an analysis of signal intensity-time curves in a given area of interest (8-45 pixels). MRI findings were compared with intraoperative tumor histological structural data or verified during a follow-up for at least 6 months. Results. The pattern of detected abnormalities was as follows: true ovarian tumors (71%), endometriomas (16%), cysts (11%), and tubo-ovarian abscesses (2%). Among the true tumors, there were serous epithelial tumors (51%), mucinous epithelial tumors (26%), endometrioid cysts (2%), dermoid cysts (6%), clear-cell carcinomas (2%), granulosa cell tumors (6%), fibromas (4%), Brenner tumors (1%), and metastatic tumors (3%). The degree distribution of the tumors identified was the following: benign tumors (49%), borderline tumors (12%) (occurring only in a group of epithelial tumors), and malignant tumors (39%). FIGO staging of the borderline ovarian tumors classified as Stages IA (66%) and IC (34%). That of the malignant tumors classified as Stages IA (7.3%), IIA (17%), IIB(12.2%), IIC(17%), IIIB (21.9%), IIIC(14.6%), and IV(9.7%). Quantitative estimation of the parameters of perfusion images showed that the amplitude of contrast agent accumulation was significantly higher in malignant tumors (167% (115.2-212.5%)) than in benign tumors (61.2% (41.2-99.0%)) (P < 0.001) and borderline ones (85.7% (58.3-138.2%)), (P < 0.01); the signal intensity semi-elevation period was significantly longer in benign tumors (35.1 sec (30.8-42.5sec)) than in borderline tumors (27.9sec (23.5-29.8 sec) (P < 0.05), and malignant ones (23.1 sec (20.5-30.9 sec)) (P = 0.01). The largest curvature (inflection) of the curve (%/sec) amounted to 1.78 (1.0-2.6); 2.86 (2.01-3.95), and 6.1 (4.19-9.46) for benign, borderline, and malignant tumors, respectively, and it was significantly higher in invasive carcinomas (P < 0.01). Malignant tumors have significantly lower mean apparent diffusion coefficients (ADC) than benign tumors (1.012±0.18 and 1.54±0.25 mm2/sec x 10-3, respectively); the value intervals did not intersect. The threshold ADC value of malignant ovarian tumors was lower than 1.139 mm2/sec x 10-3. The information values of advanced MRI techniques were an accuracy of 92.1%, a sensitivity of 93.6%, and a specificity of 91.2%. Conclusion. The incorporation of MRI with quantitative evaluation of perfusion parameters and diffusion-weighted images in a comprehensive examination algorithm allows differentiation of the degree of malignancy of ovarian tumors with a high degree of accuracy, by determining the opportunities for optimizing management tactics for patients.

ovarian cancer, magnetic resonance imaging, diffusion-weighted images, neoadjuvant chemotherapy, optimal cytoreduction