To evaluate the ability of preoperative serum CA-125 level to predict the outcome of primary cytoreductive surgery in patients with epithelial ovarian carcinoma.

We performed a retrospective chart review of 85 consecutive patients with epithelial ovarian carcinoma. All patients had preoperative serum CA-125 levels measured. We used a receiver operating characteristics curve (ROC) to determine the CA-125 level with the maximal power in predicting the outcome of primary cytoreductive surgery.

The median CA-125 level was 890.9 U/mL for all patients. Preoperative CA-125 level had significant correlations with histology, tumor grade, stage, and the presence of ascites (p<0.05). Also, preoperative CA-125 level showed significant difference between patients with suboptimal cytoreduction and those with optimal cytoreduction (2584.9 U/mL vs. 524.8 U/mL, p<0.05). Using the ROC, we found that preoperative CA-125 level of 1050 U/mL had the most powerful ability in predicting the outcome of primary cytoreductive surgery, but a poor negative predictive value (sensitivity 66.7%, specificity 64.0%, PPV 81.6%, NPV 44.4%). Optimal cytoreductive surgery was achieved in 81.6% (40/49) among patients with CA-125 <1050 U/mL, but 55.6% (20/36) among those with CA-125=1050 U/mL (p<0.05).

We think that preoperative CA-125 level may be used for selection of candidates for neoadjuvant chemotherapy before primary cytoreductive surgery. But preoperative CA-125 level was a weak negative predictor of primary optimal cytoreductive surgery. Thus, preoperative CA-125 level could not be a primary predictor of the outcome of primary cytoreductive surgery and should be considered in the context of other preoperative features.