2050 Background: Temozolomide has shown modest efficacy in the treatment of recurrent brain metastases. We designed a regimen combining temozolomide with vinorelbine, a lipophilic agent that crosses the blood-brain barrier, trying to improve efficacy. Methods: This is a phase II trial with 28-day cycles using temozolomide (150 mg/m2, days 1–7 and 15–21) and vinorelbine on days 1 and 8. We previously reported a phase I trial that established an MTD of 30 mg/m2 of vinorelbine in this combination, but the dose was decreased to 25 mg/m2 in this phase II trial. The phase II component was planned as a two-stage clinical trial. Since two or more responses were observed after the 20 initial patients, 15 more assessable patients were required. This design had a 91% probability to detect a true response rate of 20% or more. The primary endpoint was objective radiographic response. Secondary endpoints include OS, PFS and toxicity. Patients = 18 years old with KPS = 60, adequate organ function and progressive or recurrent brain metastases were eligible. Results: Thirty-six patients (13 men, 23 women) with a median age of 56 years (range, 38–76) and median KPS of 80 were enrolled. The primary tumor sites were lung (n=19), breast (n=11), colon (n=2), bladder (n=1), endometrium (n=1), head/neck (n=1) and kidney (n=1). Prior therapies included whole-brain radiation therapy (81%), chemotherapy (97%), radiosurgery (42%) and brain metastasis resection (47%). Objective radiographic response was 7% (1 CR and 1 minor response); 4 patients had SD and 23 PD. Three patients withdrew consent and did not undergo follow-up scans, 2 patients did not receive the planned treatment and 2 patients recently began treatment and have not been assessed. The median follow-up was 12.3 weeks and 72% of patients have died. Median PFS and OS were 8.3 weeks and 5 months, respectively. Grade 3/4 toxicities were mainly hematological and 3 patients were removed from the study due to myelosuppression. Conclusions: In this heavily pretreated population of patients with brain metastases, adding vinorelbine to temozolomide does not seem to improve response rates as compared to temozolomide alone. Single-agent temozolomide also has a more favorable toxicity profile. [Table: see text]