Objective: Recent additions of novel chemotherapeutics, such as pegylated liposomal doxorubicin (PLD) and topotecan (TPT), have provided clinicians with multiple options for treating recurrent ovarian cancer. Evaluating treatment response in patients without radiographic or physically measurable disease is problematic, thereby CA-125 values may be the only available objective criteria. It has been advocated that several cycles of novel agents are required prior to an observed CA-125 response. In this study, we sought to gain insight into response patterns regarding CA-125 in responders vs. non-responders and to determine whether specific "cut-off" values could help predict ultimate clinical response.
Methods: Patients with recurrent ovarian cancer who received either single agent PLD, TPT, or both were included. CA-125 levels were evaluated prior to initiation of chemotherapy and thereafter for each additional cycle. The Rustin criteria were utilized to evaluate CA-125 response.
Results: Fifty-four of 120 patients were judged to be responders. When comparing responders to non-responders, as expected, the majority of responders demonstrated a decrease after each of the first 4 cycles. However, nearly 50% of responders who received PLD demonstrated an increase in CA-125 after cycle 1. There were no responders who demonstrated two successive rises in CA-125.
Conclusion: The majority of patients with recurrent ovarian cancer who will ultimately manifest a CA-125 response to novel agents, such as TPT or PLD, will demonstrate a decrease following each cycle. An initial increase in CA-125 should not mandate discontinuation of current therapy, but a successive rise over two or more cycles reliably predicts that a treatment response ultimately is unlikely.