The serum CA-125 concentration data assists in evaluating CT imaging information when used to differentiate borderline ovarian tumor from malignant epithelial ovarian tumors

Abstract

Objective: We wanted to evaluate the diagnostic value of serum CA-125 concentration, when used in combination with the preoperative contrast-enhanced CT results, to differentiate borderline ovarian tumors (BOTs) from stage I malignant epithelial ovarian tumors (MEOTs).

Materials and methods: Ninety-eight masses (46 BOTs and 52 stage I MEOTs) from 87 consecutive patients (49 with BOTs and 38 with stage I MEOTs) who had undergone preoperative contrast-enhanced computed tomography (CT) and surgical staging were evaluated retrospectively and independently by two radiologists. The preoperative serum CA-125 concentration was measured in all patients. The utility of analyzing serum CA-125 concentration in combination with the CT results was evaluated by receiver operating characteristic (ROC) curve analysis.

Results: An irregular tumor surface and lymphadenopathy were predictive of a MEOT. ROC analysis showed that the combination of CT data and the serum CA-125 level resulted in a higher diagnostic performance than did using the CT alone for differentiating BOTs from MEOTs. The areas under the curves (AUCs) without and with the use of the serum CA-125 level data were 0.67 (95% confidence interval [CI]: 0.57-0.77) and 0.78 (95% CI: 0.68-0.85), respectively, for reader 1 (p = 0.029) and 0.71 (95% CI: 0.61-0.80) and 0.81 (95% CI: 0.72-0.89), respectively, for reader 2 (p = 0.009).

Conclusion: The serum CA-125 concentration is of additional diagnostic value when used in conjunction with the CT imaging results for differentiating BOTs from MEOTs.

Keywords: Borderline ovarian tumor; CA-125; CT; Malignant ovarian tumor.

Fig. 1

Box plots showing CA-125 serum concentrations in patients with stage I malignant epithelial ovarian tumors (MEOTs) and borderline ovarian tumors (BOTs). Mean serum CA-125 levels in patients with malignant epithelial ovarian tumors and borderline ovarian tumors were 421.0 U/ml and 54.7 U/ml, respectively (p = 0.0003).

Fig. 2

Receiver operating characteristic curve analyses showing additional utility of CA-125 concentration data for differentiating between borderline ovarian tumors and stage I malignant epithelial ovarian tumors. Dotted lines = diagnoses performed using both CT images and serum CA-125 concentrations. Solid lines = diagnoses made by CT imaging alone. A. Areas under curve for reader 1 were 0.71 and 0.81 before and after addition of serum CA-125 concentration data, respectively, to diagnoses obtained using CT imaging (p = 0.009). B. Areas under curve for reader 2 were 0.67 and 0.78 before and after addition of serum CA-125 concentration data, respectively, to diagnoses obtained using CT imaging (p = 0.029). ^*Reader 1_pre = diagnoses using CT images only, reader 1_post = diagnoses using CT images and CA-125 concentration data, reader 2_pre = diagnoses by CT images only, reader 2_post = diagnoses by CT images and CA-125 concentrations

Fig. 3

Images from 47-year-old woman with mucinous borderline ovarian cancer. A. Axial postcontrast CT scan of mid-pelvic level showing well-defined mass (arrows) with solid portion (arrowheads), and this was later found to be mucinous fluid by histopathologic examination. All of ascites, peritoneal implantation and pleural effusion were absent on CT images. Serum CA-125 concentration was 7 U/ml. B. Photograph of cut surface of ovarian mass showing mucinous fluid within mass.

Fig. 4

44-year-old woman with FIGO stage I ovarian serous cystadenocarcinoma. A. Axial postcontrast CT scan of mid-pelvic level showing well-defined cystic mass (arrow). No ascites, peritoneal implant or pleural effusion was evident on CT images. Serum CA-125 concentration was 263 U/ml. B. Photograph of cut surface of ovarian mass showing solid and cystic characteristics, with hemorrhagic serous fluid in cystic portion.