Eur Radiol (2006) 16: 2687–2699 DOI 10.1007/s00330-006-0163-z UROGENITAL Marc Bazot Jinane Nassar-Slaba Isabelle Thomassin-Naggara Annie Cortez Serge Uzan Emile Daraï Received: 24 August 2005 Revised: 8 December 2005 Accepted: 12 January 2006 Published online: 18 March 2006 # Springer-Verlag 2006 MR imaging compared with intraoperative frozen-section examination for the diagnosis of adnexal tumors; correlation with final histology Abstract The aim was to compare the accuracy of magnetic resonance im- aging (MRI) and intraoperative con- sultation (IC) for the diagnosis of adnexal masses, with reference to final histology. MRI was performed in 136 women with sonographically indeter- minate adnexal masses. IC included macroscopic and frozen-section ex- amination. Macroscopic examination and MRI determined size, nature, and presence of vegetations or solid por- tions within masses. All masses were characterized as benign or malignant according to previously published MR imaging and histopathologic criteria. Sensitivities, specificities, and predic- tive values for the diagnosis of malig- nancy of MRI and IC were assessed. Histology revealed 168 adnexal masses (99 benign, 23 borderline and 46 invasive). Frozen sections were examined in 151 cases. Among the 151 adnexal masses studied by both MRI and IC, respective sensitivities, specificities, positive and predictive values of both methods for the diag- nosis of malignancy were 89.7% and 84.5%, 91.4% and 100%, 86.7% and 100%, and 93.4% and 91.3%. Sensi- tivities of MR imaging and frozen section for the diagnosis of serous versus mucinous borderline tumors were 33.3% and 93.3%, and 62.5% and 12.5%, respectively. MRI is less accurate than IC for characterizing adnexal masses. However, MRI may increase the relevance of IC for bor- derline mucinous tumors. Keywords Ovarian tumors . Comparative studies . Frozen section examination . Magnetic resonance imaging Introduction Adnexal masses are the leading indication for gynecolog- ical surgery [1]. The main goal of imaging techniques in this setting is to differentiate malignant from benign tumors, and to determine the surgical strategy. Ultrasonography (US) is the main imaging technique used to assess the nature of adnexal masses. Previous studies have reported accuracies ranging from 60% to 95% [2–5]. For adnexal masses that appear complex or possibly malignant on US, magnetic resonance (MR) imaging adds significant further information, with a reported accuracy as high as 91% [4, 6–8]. In this specific setting, some criteria are highly suggestive for benign ovarian lesions (e.g., fatty components, shading on T2-weighted images) [9, 10], whereas vegetations and solid portions within cystic masses have been shown to be the best predictors of malignancy [7, 11]. Using these different criteria, a preliminary diagnosis can be made before surgery. Thus, the procedure can be discussed with the patient and the surgical approach (laparoscopy versus laparotomy) can be determined. Despite the value of preoperative MR imaging, intraop- erative consultation, including macroscopic and frozen- section examinations of vegetations or solid portions, is frequently requested to confirm the malignant nature of adnexal masses before radical surgery, or to rule out malignancy during the course of surgery for presumed ap-hop-paris: Assistance Publique-Hôpitaux de Paris M. Bazot (*) . J. Nassar-Slaba . I. Thomassin-Naggara Department of Radiology, Hôpital Tenon, 4, rue de la Chine, 75020 Paris, France e-mail: marc.bazot@tnn.ap-hop-paris.fr Tel.: +33-1-56016453 Fax: +33-1-56016402 A. Cortez Department of Pathology, Hôpital Tenon, 4, rue de la Chine, 75020 Paris, France S. Uzan . E. Daraï Department of Obstetrics and Gynecology, Hôpital Tenon, 4, rue de la Chine, 75020 Paris, France