Background: While gemcitabine plus cisplatin has demonstrated significant antitumor activity as 1^st line therapy of BC, there is no effective treatment after failure of gemcitabine-based therapy. REG is an oral multi-kinase inhibitor that targets angiogenesis, oncogenesis and cancer proliferation/metastasis. We evaluated the efficacy of REG in BC. Methods: Patients (pts) with histologically proven BC who progressed on at least one line of systemic therapy received REG 160 mg daily 21 days on 7 days off, in 28 day cycles. The primary endpoint was 6-month (mo) overall survival (OS) and the secondary endpoints were median OS, progression free survival (PFS) and response rates (RR). Pre and post-treatment plasma were collected for cytokine evaluation. Results: A total of 39 pts received at least 1 dose of REG; 32 pts were evaluable for efficacy. Median age was 62 (range: 27-88) years and the primary sites of tumor were intrahepatic cholangiocarcinoma (68.8%), extrahepatic (18.8%), and gallbladder (12.5%). Pts were considered evaluable for efficacy if patients received more than 1 cycle of REG. For 32 evaluable pts, 6 mo OS was 52% with median PFS of 2.8 mo (95% CI: 1.1-4.5) and median OS of 7.9 mo (95% CI: 0-18.7). Median PFS and OS of the pts (n=20) failed 1 line of therapy were 3.7 mo (95% CI: 3.2-4.1) and 13.8 mo (95% CI: 1.8-25.8), respectively. Median PFS and OS of the pts (n=12) failed 2 lines were 1.8 mo (95% CI: 1.63-1.97) and 4.5 mo (95% CI: 2.6-6.3), respectively. RR was 9.4% (2 PR and 1 unconfirmed PR) and DCR was 62.5%. Total 71.8% of grade 3/4 adverse events (AE) were observed, and the most common AE were fatigue (56.4%) and hypertension (53.8%). Dose modification was required in 49% of the pts. Among the 23 cytokines analyzed, elevated baseline VEGF-A was associated with good prognosis (HR 0.62, p=0.01). Elevated baseline TIMP-1 (HR 1.79, p=0.04) and IL-6 (HR 1.33, p=0.05) were associated with poor prognosis. REG treatment decreased BMP-9, GP130, VEGF-R2 and VEGF-R3 and increased IL-6, PIGF, TIMP-1, VCAM-1 and VEGF-A significantly. Conclusions: The primary endpoint was met in this study. VEGF-A may be further evaluated as a predictive biomarker for REG in BC. Further randomized trials are warranted to confirm the efficacy and the correlative data. Clinical trial information: NCT02115542