Introduction Women with epithelial ovarian cancer (EOC) are conventionally treated with primary cytoreductive surgery. For those with a low probability of optimal primary surgical debulking, an alternative management option is primary chemotherapy with delayed primary surgery. Selection criteria are required to identify women who may benefit from this approach.

Patients and methods Patient age, presence of ascites, preoperative serum CA-125 level, surgical procedures performed, postoperative residual disease, FIGO stage, and histology data were collected on 97 women with preoperative clinical evidence of advanced EOC. Univariate and multivariate analysis was performed to identify which preoperative factors predict disease that will be suboptimally debulked. Receiver–operator curves were constructed for CA-125 level as a predictor for residual disease.

Results The best predictor of disease suboptimally cytoreduced was serum CA-125 level (OR = 22.76, 95% CI = 7.13–72.69). Other predictive factors included age over 60 years (OR = 3.16, 95% CI = 1.04–9.56) and clinical evidence of ascites (OR = 3.30, 95% CI = 1.03–10.62). The optimal cut-off for serum CA-125 as a predictor of suboptimal debulking was 586 IU (sensitivity 80.0%, specificity 88.5%, PPV 85.7%).

Conclusion Serum CA-125 level is a reliable component of the preoperative assessment of women with EOC.