Differentiation between malignant and benign ovarian tumors by magnetic resonance imaging / 中国医学科学杂志(英文版)

ABSTRACT

<p><b>OBJECTIVE</b>To determine the magnetic resonance (MR) imaging findings of an ovarian mass which are most predictive of malignancy and assess the value of intravenous gadolinium administration in the characterization of an ovarian mass.</p><p><b>METHODS</b>Totally 74 consecutive patients with a clinically or sonographically indeterminate adnexal mass underwent MR imaging, of whom 59 had subsequent surgical resection of 70 adnexal masses. These 59 patients formed the study population. MR imaging studies were prospectively and independently reviewed by a senior and a junior radiologist. The senior radiologist also reevaluated the studies in a blind fashion after a minimum 6 months interval. The sensitivity, specificity, positive predictive value, and negative predictive value of contrast-enhanced and unenhanced MR imaging were evaluated.</p><p><b>RESULTS</b>The most predictive MR imaging findings for malignancy were presence of vegetations in a cystic lesion and presence of necrosis in a solid lesion. The odds ratio was even higher when the ancillary finding of peritoneal metastasis or ascites was present. Contrast media contributed significantly to lesion characterization. Total 70 ovarian masses were detected by contrast-enhanced MR imaging including 37 malignant ovarian masses and 33 benign ovarian masses with 87% (61/70) accuracy, 86% (32/37) sensitivity, 88% (29/33) specificity, 89% (32/36) positive predictive value, and 85% (29/34) negative predictive value, whereas 70 ovarian masses were detected by unenhanced MR imaging with 74% (52/70) accuracy, 73% (27/37) sensitivity, 76% (25/33) specificity, 77% (27/35) positive predictive value, and 71% (25/35) negative predictive value. There were significant differences in accuracy (P < 0.01), sensitivity (P < 0.01), specificity (P < 0.01) between contrast-enhanced and unenhanced MR imaging.</p><p><b>CONCLUSION</b>Contrast-enhanced MR imaging is highly accurate in detection and characterization of complex adnexal masses.</p>