5013 Background: This phase II clinical trial evaluated the efficacy and safety of the combination of bevacizumab, a VEGF inhibitor, and pemetrexed, a multi-targeted antifolate agent inhibiting thymidylate synthase, for recurrent or persistent epithelial ovarian, fallopian tube (FT) or primary peritoneal (PP) cancer. Methods: Patients with measurable recurrent/persistent epithelial ovarian, FT or PP cancer after at least one prior platinum- and taxane-containing regimen were eligible. Patients might have received < 2 prior cytotoxic chemotherapy regimens but no prior bevacizumab. Pemetrexed 500 mg/m2 IV and bevacizumab 15 mg/kg IV were given every 3 weeks until progression, unacceptable adverse effects, or patient/physician choice. 32 patients were needed to have 90% power to detect the primary endpoint of 6-month PFS ≥ 40% with a two-sided 0.05 significance. Secondary endpoints included toxicity and response by RECIST and CA-125 criteria. Results: Thirty-four patients received a median of 7 treatment cycles (range, 2-26). Twenty-eight patients (82%, 95% CI: 66-92) were platinum-sensitive. Median follow-up was 17.1 months (range, 2.7-31.2). The 6-month PFS was 58.2% (95%CI: 40-73) for all patients and 50% (19-81) for platinum-resistant patients. Objective response rates by RECIST criteria included (%; 95%CI): 0 CR, 14 PR (41%; 25-59), 18 SD (53%; 35-70) and 2 PD (6%; 1-20). Of 27 patients evaluable by CA-125, levels declined ≥50% in 17 (62%; 44-79), and ≥75% in 8 (30%; 16-49). 12-month OS was 88% (95%CI: 71-95) and median PFS was 7.8 months (95%CI: 4.7-10.7). Of the 34 patients, grade 3-4 hematologic toxicities occurred in 53% (neutropenia 50%, leukopenia 26%, thrombocytopenia 12%, anemia 9%). Other grade 3-4 toxicities included metabolic (29%), constitutional (18%), pain (18%) and gastrointestinal (15%). No bowel perforations occurred. Conclusions: Combined bevacizumab/pemetrexed is well tolerated and highly active for the treatment of recurrent ovarian cancer. The dose and schedule tested here warrant further investigation in phase III trials.