17th World Congress on Ultrasound in Obstetrics and Gynecology Oral communication abstracts OC198: Table Prevalence Technique Sensitivity Specificity PPV NPV %(n) (%) (%) (%) (%) Endometrioma 30.5 (276) B-mode 84 98 96 93 Cystic teratoma 12.6 (114) B-mode 88 100 97 98 Serous cyst 16.8 (152) B-mode 80 96 79 96 Ovarian cancer premenopausal group 7.5 (51) B-mode 100 89 42 100 Ovarian cancer premenopausal group Power doppler 96 94 57 100 Ovarian cancer postmenopausal group 42 (94) B-mode 99 74 73 99 Ovarian cancer postmenopausal group Power doppler 98 82 80 98 diagnosis of an ovarian mass and select masses for power Doppler evaluation. Methods: Between 1996 and 2006, 905 masses in 831 patients were referred to the Department of Obstetrics and Gynecology, University of Cagliari, Italy, and underwent ultrasound evaluation before surgery. Sixty-hundred and eighty masses were in premenopausal women and 225 in postmenopausal patients. The histological type of each mass was predicted on the basis of the B-mode typical benign findings as in case of endometrioma (‘ground glass’ endocystic pattern with a clear demarcation from the ovarian parenchyma), cystic teratoma (echogenic pattern with or without acoustic shadow) and serous cyst (anechoic unilocular or bilocular cyst with a thin regular wall without endocystic vegetations). Any cystic mass containing excrescences, thick septations, multiple irregular septations or a solid mass in which the echo architecture was not suggestive of benign histology was categorized as malignant. All masses suspected to be malignant were submitted to power Doppler evaluation to identify the location of flow. A mass was suspected to be malignant at power Doppler examination when central location of flow was found. Results: The sensitivity, specificity, positive (PPV) and negative (NPV) predictive values of B-mode and power Doppler for each evaluated histological type (divided into two groups, premenopausal and postmenopausal, for the diagnosis of ovarian cancer) are reported in the table. Conclusions: The use of simple ultrasonographic findings has high specificity for the prediction of several types of benign cyst with a good sensitivity. Besides the different prevalence of ovarian cancer in premenopausal and postmenopausal patients, the use of power Doppler seems to improve the specificity of B-mode in the diagnosis of ovarian cancer in both groups of patients. OC199 The frequency of typical ultrasound morphological features in borderline ovarian tumors J. Yazbek 1 , L. Valentin 2 , T. K. Holland 1 , A. Testa 3 , D. Timmerman 4 , D. Jurkovic 1 1 Early Pregnancy and Gynaecology Assessment Unit, King’s College Hospital, London, United Kingdom, 2 Department of Obstetrics and Gynecology, University Hospital Malm ¨ o, Sweden, 3 Istituto di Clinica Ostetrica e Ginecologica, Universit ` a Cattolica del Sacro Cuore, Italy, 4 University Hospitals Leuven, Belgium Objectives: To assess how frequently typical morphological features are present in different subtypes of borderline ovarian tumors (BOT) on gray-scale ultrasound scan. Methods: Ultrasound data of women with a histological diagnosis of BOT who underwent a preoperative ultrasound examination were collected from nine different centers in three European countries who all took part in the IOTA collaboration. The ultrasound description of the tumors, including locularity, presence of papillary projections and fluid content, was reviewed and compared to the known typical morphological features of the different subtypes of BOTs; a cyst with multiple papillary projections was considered typical of serous BOT (sBOT) and endocervical mucinous BOT (mBOT), and a cyst with a ‘honeycomb nodule’ and thick echogenic fluid content typical of gastrointestinal mucinous BOT (GI mBOT). Results: The data of 153 women were analyzed; 31/153(20%) BOTs were not given a subtype on histology and were excluded from the final data analysis. Overall, 74/122 (61%) BOTs with specified histological subtypes had typical ultrasound features on gray-scale ultrasound. Typical morphological features on gray-scale ultrasound were present in: 32/39 (82%) GI mBOTs, 38/74 (51%) typical sBOTs, and 4/9 (44%) endocervical mBOTs. Some 15/122 (12%) BOTs were unilocular cysts with no papillary projections and had a median tumor volume of 603 (range, 55–5561) mL; of these, six (40%) were GI mBOTs, six (47%) were sBOT, and two (13%) endocervical type mBOTs. Six of 74 (8%) sBOTs and 2/9 (22%) endocervical mBOTs were multilocular cysts with no papillary projections; 17/74 (23%) sBOT and 1/9 (12%) endocervical mBOTs were multilocular cysts with at least one papillary projection. Conclusions: Some 61% of BOTs have typical ultrasound features on gray-scale ultrasound, which facilitates a correct preoperative diagnosis. The remaining tumors with non-typical presentation resemble benign ovarian cystadenomas. OC200 Ultrasound-based triage for surgical management of adnexal masses in asymptomatic women J. L. Alcazar , P. Royo, A. Ruiz-Zambrana, J. A. M´ ınguez, M. Jurado, G. L ´ opez-Garc´ ıa Department of Obstetrics and Gynecology. University of Navarra, Spain Objectives: The aim of the present study was to evaluate prospectively an ultrasound-based scoring system as a method for triaging asymptomatic women presenting with an adnexal mass for surgical treatment. Methods: From June 2003 to December 2006, 292 adnexal masses in 265 asymptomatic women (mean age 40 years, SD 12.6 years) undergoing elective surgical treatment at our institution were included in this prospective study. Patients were evaluated by transvaginal power Doppler ultrasound prior to surgery. Patients were classified as low risk or high risk for malignancy according to our ultrasound-based scoring system. Patients at low risk for malignancy were scheduled for laparoscopy and patients at high risk for malignancy were scheduled for laparotomy. Some patients at high risk were scheduled for advanced oncological laparoscopic surgery. Patients with a low risk but tumor size > 10 cm or clinical suspicion of pelvic adhesions were scheduled for laparotomy. Results: Some 205 masses were considered as low risk and treated by laparoscopy (203 tumors were benign and two malignant); 64 masses were considered as high risk (43 masses were scheduled for primary laparotomy and 21 masses were scheduled for advanced oncologic laparoscopy; 54 tumors were malignant and 10 benign). Twenty-three tumors were considered as low risk but scheduled for primary laparotomy. In this group, 21 tumors were benign and two were malignant. Sensitivity, specificity, PPV, NPV and accuracy were 93%, 96%, 85%, 98% and 95%, respectively. Conclusions: Ultrasound-based triage of asymptomatic women diagnosed as having a persistent adnexal mass is effective for determining the most appropriate surgical approach. 428 Ultrasound in Obstetrics & Gynecology 2007; 30: 367–455