000215 PREDICTION OF CERVICAL INVOLVEMENT IN ENDOME-TRIAL CANCER BY DIFFERENT PREOPERATIVE EVALUATION METHODS: CLINICAL EXAMINATION, D&C HYSTEROSCOPY

Objective: Assess if preoperative Ca125 levels can predict stage & likelihood of debulking in ovarian cancer. Methods: Ca125 assessed in 591 patients with ovarian cancers managed in Sheffield. Results: (Table 1) Ca125 was raised (.35IU) in: 85% of all ovarian cancers; 89% of epithelial ovarian cancers (EOC); 100% of primary peritoneal tumours (PPT); 60% of borderline tumours. Ca125 levels increased from FIGO stage I to IV and were significantly higher in stage III/IV compared to stage I, and in patients with residual disease compared to those with no residual disease post-operatively (p = 0.02 Independent sample t-test). No significant difference in Ca125 was seen between optimally or sub-optimally debulked patients. Receiver operator curves (ROC) curves using Ca125 as a diagnostic predictor, show good separation of stage I from stage II/ III/IV disease and of patients with no residual disease from those with residual disease, with AUC (area under curve) values of 0.690.82. If Ca125 cut-off levels for predicting residual disease and stage II/III/IV disease are selected as 284IU & 205IU respectively, sensitivity of 71% and 74%, and specificity of 70% and 74% respectively are achieved. Conclusions: Pre-operative Ca125 levels can be used as a predictor of patients likely to have residual disease and FIGO stage greater than stage I disease, with sensitivities and specificities of 70-74%. Ca125 cannot however accurately predict exact stage or likelihood of optimal versus sub-optimal debulking.