Objectives: Recent data suggest that differences in CA125 levels within the normal range may predict progression-free survival (PFS), but limited information is available regarding the value of these differences in predicting overall survival (OS) in patients with epithelial ovarian cancer. The objective of this study was to determine whether CA125 is an independent predictor of OS in patients with surgically defined disease status at the end of primary therapy prior to intraperitoneal (IP) consolidation chemotherapy. A secondary objective was to assess the relationship of CA125 level to PFS.
Methods: Using data from a retrospective cohort of 433 patients who received intraperitoneal (IP) therapy following primary treatment for ovarian cancer between 1984 and 1998, we identified 241 patients with a complete clinical response and CA125 data at the time of second-look surgery prior to IP chemotherapy. Patient demographics and updated follow-up status were abstracted from medical records. Kaplan-Meier survival curves were compared using the log-rank test, and Cox regression models were used for multivariate analysis.
Results: The majority of patients had advanced stage III or IV disease (n=201, 83%) and high-grade histology (n=163, 68%). Taxane was used as part of primary platinum-based therapy in 56% (n=134) of patients, and subsequent IP chemotherapy was platinum-based in 85% (n=206). When considered as a continuous variable, CA125 was a predictor of OS (P=0.029). Using the median CA125 level in our study group as a cut-off, OS was increased in patients with CA125 < or =12 U/ml (median 5.8 years) compared with >12 (3.7 years) (P=0.0027). CA125 level was an independent predictor of OS (HR: 1.410; 95% CI, 1.044, 1.904, P=0.0248) in a multivariate model that included stage (P=0.0166), grade (P=0.0001), and findings at second-look surgery (P=0.0003). CA125 level was also a predictor of clinical PFS (radiographic or CA125 elevation criteria alone) in a subset of 161 patients as a continuous variable (P=0.0036), and when divided at the median (< or = or >12; median 2.8 years vs. 1.7 years; P=0.0017).
Conclusions: In our study population, CA125 level at the end of primary therapy was a predictor of OS and PFS when considered as a continuous variable, or when divided at the median (< or = or >12 U/ml). Further prospective study is required to optimize clinically significant cut-off values within the normal range of CA125 levels for both OS and PFS endpoints.