Intravoxel Incoherent Motion Imaging in Differentiation Borderline From Malignant Ovarian Epithelial Tumors: Correlation With Histological Cell Proliferation and Vessel Characteristics

Background The differentiation of borderline from malignant ovarian epithelial tumors (OETs) is difficult based on morphologic characteristics. Diffusion and perfusion information from intravoxel incoherent motion (IVIM) diffusion‐weighted imaging (DWI) might be useful for this distinction. Purpose To investigate the potential of IVIM‐DWI in discriminating borderline from malignant OETs by correlating with cell proliferation and microvessel density (MVD). Study Type Prospective. Subjects Sixty‐six patients with OETs (22 borderline, BOETs; 44 malignant, MOETs) underwent IVIM‐DWI before surgery. Field Strength 3.0T IVIM‐DWI sequence with 12 b ‐values (0–1000 sec/mm 2 ). Assessment Apparent diffusion coefficient (ADC) and IVIM‐DWI parameters (diffusion coefficient [ D ], microvascular volume fraction [ f ], and pseudodiffusion coefficient [ D* ]) were measured. Cell proliferation and MVD was determined by immunohistochemical staining of Ki‐67 and CD‐31, respectively. Statistical Tests Mann–Whitney U ‐test; two‐sample t ‐test; intraclass correlation coefficient; Bland–Altman analysis; receiver operating characteristics (ROC) curves; and Spearman correlation. Results ADC and D in BOETs was significantly higher than those in MOETs ( P < 0.001), while f in BOETs was significantly lower than those in MOETs ( P < 0.001). No significant difference was found in D* between borderline and malignancies ( P = 0.324). In the differential diagnosis of borderline from malignant OETs; D demonstrated the highest area under the curve (AUC) of 0.951, while ADC and f had a lower AUC of 0.921 and 0.847, respectively. The ADC and D was negatively correlated with cell proliferation ( r = –0.682, r = –0.694, respectively, P < 0.001), while f was positively correlated with MVD of the OETs ( r = 0.558, P < 0.001). Data Conclusion IVIM‐DWI may be a useful tool for differentiating BOETs from MOETs. D and f could be image biomarkers to reflect histopathological characteristics of cell proliferation and MVD in OETs. Level of Evidence: 1 Technical Efficacy Stage: 2 J. Magn. Reson. Imaging 2020;51:928–935.