Validation of CA-125 concentration nadir within the normal range following primary treatment as a predictor of survival for epithelial ovarian cancer (EOC)

Abstract

16007 Background: Recent studies suggest that the CA-125 nadir within the normal range after surgery and chemotherapy treatment is a predictor of survival (Crawford, ASCO 2004; Crawford, Ann Oncol 2005) and relapse (Markman, J Clin Oncol 2006). In order to validate these previous findings, we have conducted a retrospective analysis of patients (pts) treated in our institution for EOC. Methods: Between March 1, 1997, and October 30, 2005, all pts treated for EOC at Vall d'Hebron University Hospital were identified from the tumor registry database and screened retrospectively for their standard prognostic factors (age at diagnosis (=65 vs. >65), stage (III-IV vs. IC-II), and suboptimal vs. optimal cytorreduction). Inclusion criteria: an elevated CA-125 at time of diagnosis (>35 U/mL); primary treatment (PT) that consisted in surgery and intravenous carboplatin/paclitaxel for a maximum of 6–9 cycles; complete clinical and radiological response to initial treatment with normalization of CA-125 (=35 U/mL); and disease status at the time of last follow-up. Standard Kaplan-Meier methods were used to plot the progression-free survival (PFS) of members of each of the nadir groups. The relative contribution of the different potential correlates of prognosis was assessed by the Cox proportional hazards method. Results: 123 pts were identified: 64 Group A (=10 U/mL), 42 Group B (11–20 U/mL), 17 Group C (21–35 U/mL). Median age: 56. Stage IC 25%, II 13%, III 52%, IV 10%. Median follow-up 39.2 months (m). Median PFS was 69.7 m, 27.7 m, and 15.8 m for A, B and C, respectively (p< .0001, log-rank test). The Cox model showed a highly-significant impact on PFS in relation to CA-125 nadir levels, residual tumor after surgery and stage. Hazard ratios (HR) for PFS (95% CI) of B vs. A, C vs. B, and C vs. A were 1.98 (p= .034), 2.35 (p= .02), and 4.67 (p< .001), respectively. HR for PFS (95% CI) of suboptimal vs. optimal cytorreduction and stage III-IV vs. IC-II were 1.84 (p= .058) and 3.2 (p= .002), respectively. Conclusions: The CA-125 nadir in the normal range following PT for EOC is a reproducible predictor of PFS in stage IC-IV. Prospective studies of maintenance-consolidation therapies or different approaches in selected pts based on CA-125 nadir seem warranted. No significant financial relationships to disclose.