Purpose: To evaluate the utility of dynamic contrast enhancement (DCE) MRI for distinguishing among benign, borderline and invasive epithelial ovarian tumors.
Materials and methods: We analyzed preoperative MRI studies of 37 patients with ovarian epithelial tumors (10 benign, 11 borderline, and 16 invasive). A DCE-MRI sequence was acquired and regions of interest (ROIs) were drawn in the ovarian tumors and adjacent myometrium. A total of three patterns of enhancement were defined. Dynamic data were parameterized using mathematical models that included the enhancement amplitude (EA), the time of half rising (THR), and the maximal slope (MS). Using myometrium as the internal reference, ratios of EA (EAr), THR (THRr), MS (MSr), and initial area under the curve for 60 seconds after injection (IAUC(60) ratio) were determined.
Results: Morphological criteria such as septa, papillary projection, solid portion, and T2-weighted MR signal intensity of solid tissue were useful for discriminating invasive from noninvasive ovarian tumors (P = 0.01, P = 0.02, P = 0.002, and P < 0.0001 respectively) but not for discriminating benign from borderline tumors. Curve type 3 was specific for invasive ovarian tumors. EAr, MSr, and IAUC(60) ratio were higher for invasive than for benign (P < 0.0001) and borderline tumors (P = 0.005, P = 0.002, and P = 0.001, respectively). The IAU(60) ratio was the most relevant factor for discriminating benign from borderline and invasive tumors. MSr and IAU(60) ratio could be combined to generate a decision tree with 81% accuracy.
Conclusion: DCE-MRI is a useful tool for characterizing epithelial ovarian tumors.
(c) 2008 Wiley-Liss, Inc.