The aim of this study was to evaluate the feasibility of multi–detector row computed tomography (MDCT) in the differentiation between borderline ovarian tumors and ovarian cancer on the basis of tumor morphology and specific features of tumor vascularity in correlation with the results at pathology.
A triphasic MDCT protocol was used for the analysis of tumor vascularity. The following features were taken into account: (1) The number of vessels in papillary projections, solid-tissue component, and septa (2 vs >2), (2) serpentine and chaotic configuration of vessels, (3) presence of microaneurysms, and (4) presence of arteriovenous microfistulas. Masses with at least 3 of 4 features were considered ovarian cancer (group A) and masses with 2 features or less as borderline tumor (group B). Radiological findings were compared with results of postoperative pathology.
Pathologic vessels were found in all 56 patients. Thirty-two patients were included in group A and 24 in group B. The results of pathology were as follows: in group A: 31 malignant tumors, including 31 ovarian carcinomas and 1 benign cystadenoma; in group B: 22 borderline ovarian tumors, 1 benign cystadenoma, and 1 ovarian cancer.
Morphological evaluation of tumor vascularity in MDCT seems to be an efficient method of differentiating between borderline ovarian tumors and ovarian carcinomas. Because of a small number of cases in the current study, a further research seems justified to confirm our results. The presented MDCT-angiographic criteria showed high sensitivity (97%) and specificity (96%) in differentiation of borderline ovarian tumors and ovarian cancers as compared with pathology. The presented CT-angiographic criteria of malignancy showed an excellent interobserver agreement.
*2nd Department of Radiology and †2nd Department of Obstetrics and Gynecology, Medical University of Warsaw, Warsaw, Poland.
Address correspondence and reprint requests to Pawel Derlatka, MD, PhD, 2nd Department of Obstetrics and Gynecology, Medical University of Warsaw, Karowa 2, 00-315 Warsaw, Poland. E-mail: firstname.lastname@example.org.
The authors declare no conflicts of interest.
Received April 14, 2013
Accepted August 4, 2013