Sequential Angiogenic Blockade for the Treatment of Recurrent Mullerian Malignancies

OBJECTIVES: Primary * Assess the efficacy of a sequential antiangiogenic blockade regimen of bevacizumab then cyclophosphamide with bevacizumab at disease progression in patients with recurrent ovarian cancer as measured by the proportion of patients who remain on study at three months (4 cycles) * Assess the safety profile with respect to gastrointestinal perforations Secondary * Assess other toxicity/ safety profile of this metronomic antiangiogenic approach * Assess preliminary response rate and proportion of patients on study at 6 months * Assess progression-free survival, time to progression and overall survival Correlative * Determine if biological correlates of angiogenesis are altered by the addition of sequential therapy * Determine if changes in biological markers are correlated with clinical state of cancer * Determine whether biomarkers of angiogenesis can predict and measure response * Determine whether oncogenic mutations predict response * Determine whether hypertension and urinary biomarkers such as varying levels of albuminuria predict response to bevacizumab * Determine patient risk factors for hypertension and proteinuria as toxicities of bevacizumab STATISTICAL DESIGN: This study used a two-stage design to evaluate safety and efficacy of sequential antiangiogenic blockade with a regimen of bevacizumab then cyclophosphamide added at disease progression. Safety was measured by the incidence of grade 3-5 gastrointestinal perforation (GIP) during the first 3 months of therapy and efficacy by completion of at least 3 months of therapy. The sample size was determined based on efficacy with the null and alternative therapy completion rate of 50% and 80%, respectively. If 6 or more patients enrolled in the stage one cohort (n=9 patients) completed at least 3 months of therapy then accrual would proceed to stage two (n=11 patients) if there were fewer than 2 cases of grade 3-5 GIP. There was 0.75 probability of stopping the trial at stage one if the true therapy completion rate was 50%. If 13 or fewer patients remained on therapy for at least 3 months by the end of stage two, this regimen would be deemed ineffective. The power to reject the null hypothesis with a one-sided binomial test was 87% assuming 5% significance.