PubMed2003
OBJECTIVE: To improve the survival of women with stage IV epithelial ovarian cancer the Southwest Oncology Group conducted a phase II trial examining two novel treatment strategies: (a) modification of the chemotherapy regimen (carboplatin/paclitaxel) based on evidence of response during treatment and (b) "maintenance therapy," with the monthly administration of single agent paclitaxel (maximum 2 years). METHODS: Individuals with histologically documented stage IV ovarian, fallopian tube, or peritoneal cancers initially received a maximum of 3 courses of single agent carboplatin (AUC 6), with the decision to administer a subsequent course based on specific response criteria (i.e., declines in serum CA-125 or shrinkage of measurable mass lesions). This was followed by single agent paclitaxel (150 mg/m(2)), delivered every 2 weeks for a maximum of 6 courses, again based on evidence of continued response to the treatment program. Finally, in the absence of disease progression or unacceptable toxicity, patients received single agent paclitaxel (175 mg/m2) for a maximum of 24 monthly courses. The primary study end point was 2-year overall survival, which was to be compared to the survival of a previously published historical control population [stage IV ovarian cancer patients treated with a "standard" cisplatin/paclitaxel regimen (GOG 111; McGuire WP, N Eng E J Med 1996; 334:1)]. This analysis was employed to determine if this novel strategy was worthy of further clinical investigation in a controlled clinical trial. RESULTS: There were 53 patients entered this multicenter study, of whom 50 were evaluable for toxicity and 51 for survival. There were 2 treatment-related deaths. Grades 2 and 3 neutrotoxicity were observed in 18 and 4% of patients, respectively. There were no major treatment-related protocol violations. The overall 2-year survival was 48% (95% CI, 34-62%), compared to 60% for the historical control group. CONCLUSIONS: While this trial demonstrated the feasibility of delivering a treatment regimen requiring frequent alternations in therapy based on predetermined changes in objective parameters of response, further exploration of this specific management approach does not appear warranted in stage IV ovarian cancer, based on the survival outcome compared to a reasonably similar historical control population. As this trial demonstrates, the prospective determination of clinical end points in a phase II study (e.g., 2-year survival) that will lead to continued evaluation of a particular investigative strategy in a large-scale randomized trial can be an effective method to reduce the bias and uncertainty often associated with this complex decision process.