Intravenous aflibercept for treatment of recurrent symptomatic malignant ascites in patients with advanced ovarian cancer: a phase 2, randomised, double-blind, placebo-controlled study

Background Targeting of VEGF is a potential therapeutic option in patients with malignant ovarian ascites. We present the final results of a multicentre study of the efficacy and safety of aflibercept, a potent inhibitor of both VEGF and placental growth factor, in the treatment of malignant ascites. Methods In this double-blind, placebo-controlled, parallel-group, phase 2 study, patients with advanced chemoresistant ovarian cancer and recurrent symptomatic malignant ascites were randomly assigned (1: 1) via an interactive voice response system to either intravenous aflibercept (4 mg/kg every 2 weeks) or placebo, stratified by interval of time (<= 2 weeks vs >2 weeks) between the two most recent paracenteses before randomisation. Patients participated in the double-blind period (during which patients, investigators, and sponsor personnel were masked to treatment assignment) until they had a repeat paracentesis and for at least 60 days, and could also participate in an optional open-label period during which all patients received aflibercept. The primary efficacy endpoint was time to repeat paracentesis based on response during the double-blind period alone, and was analysed in the intention-to-treat population with censoring of patients who did not have a repeat paracentesis as of the last day of the double-blind period. Safety analyses included both double-blind and open-label periods. This study is registered at ClinicalTrials.gov, number NCT00327444. Findings 55 patients with a median of four (range two to 11) previous lines of chemotherapy were randomly assigned to receive placebo (n=26) or aflibercept (n=29). Mean time to repeat paracentesis was significantly longer with aflibercept than with placebo (55.1 [SE 7.3] vs 23.3 [7.7] days; difference 31.8 days, 95% CI 10.6-53.1; p=0.0019). In the aflibercept group, two patients did not need a repeat paracentesis during 6 months of double-blind treatment. The most common grade 3 or 4 treatment-emergent adverse events were dyspnoea (six [20%] aflibercept vs two [8%] placebo), fatigue or asthenia (four [13%] vs 11 [44%]), and dehydration (three [10%] vs three [12%]). The frequency of fatal gastrointestinal events was higher with aflibercept (three intestinal perforations) than with placebo (one intestinal fistula leading to sepsis). Interpretation This study shows the effectiveness of VEGF blockade in the reduction of malignant ascites, but confirms the significant clinical risk of fatal bowel perforation in this population of patients with very advanced cancer. VEGF blockade should be used with caution in advanced ovarian cancer with abdominal carcinomatosis, and the benefit-risk balance should be thoroughly discussed for each patient.