CA-125 KINETICS - A COST-EFFECTIVE CLINICAL TOOL TO EVALUATE CLINICAL-TRIAL OUTCOMES IN THE 1990S

OBJECTIVE: Our purpose was to determine the importance of the rate of decline of CA 125 relative to conventional prognosticators of ovarian cancer survival to develop a cost effective management algorithm that supports clinical trial research. STUDY DESIGN: By use of a retrospect ive chart review the slope of the CA 125 exponential regression curve was calculated for 126 women undergoing combination chemotherapy for e pithelial ovarian cancer. Univariate and multivariate survival analyse s evaluated conventional parameters including age, grade, stage, histo logic features, time to initial chemotherapy, dose and treatment inten sity, number of cycles to normal CA 125 levels, the intercept from the regression equation, and the slope of the exponential curve. RESULTS: The ideal CA 125 regression rate was calculated at 7.6 days (95% conf idence interval 5.9 to 10.7). Univariate analysis determined slope of the CA 125 exponential regression curve (p = 0.0003), number of cycles to normal CA 125 levels (p = 0.0001), residual disease (p = 0.0006), and platinum treatment intensity (p = 0.0001) as the most important pr edictors of survival. Cox proportional-hazard regression analysis iden tified slope of the CA 125 exponential regression curve and number of cycles to normal CA 125 levels as the most significant factors for act uarial survival, replacing such conventional parameters as patient age , stage, grade, chemotherapy intensity, and residual disease. None of the factors investigated predicted treatment outcome for patients with out residual disease. Multiple linear regression analysis of the slope of the CA 125 exponential regression curve identified intercept of th e regression equation, stage, age, and time to initial chemotherapy as important determinants of the slope. CONCLUSION: The slope of the CA 125 exponential regression curve is the single most important prognost icator of actuarial survival for the patient with a CA 125-positive ov arian carcinoma. Treatment algorithms based on this slope may be helpf ul in developing novel cost-effective clinical trials.