14519 Background: The treatment of unresectable liver confined metastatic disease from colorectal cancer is a challenging issue. Although locoregional treatments such as HAI claim the advantage of delivering higher doses of anticancer agents directly into the affected organ, the evidence for benefit in terms of overall survival (OS) is conflicting. We aimed to quantitatively summarize the results so far described in randomized trials comparing HAI to systemic chemotherapy. Methods: To date, ten randomized controlled trials (RCT) have been published, for a total of 1,277 patients enrolled. Seven RCT (n=1,098) enrolling at least 50 patients per arm (the minimum sample size required to achieve a type II error lower than 20%) were considered for OS meta-analysis, which was based on the inverse of variance method. For tumor response rates, hazard ratios (HR) and their 95% confidence intervals (CI) were obtained from raw data; for OS, HR and CI were obtained from reported Cox model survival analyses or were calculated from Kaplan-Meier survival curves according to the Parmar method. Results: HAI regimens were based on floxuridine (FUDR) in nine out of ten RCT, while in one case 5-fluorouracil (5FU) + leucovorin (LV) were used. Systemic chemotherapy consisted of FUDR, 5FU, 5FU + LV, or a miscellany of 5FU and best supportive care in three, one, four and two studies, respectively. Tumor response and median OS were better in the HAI arm in eight and three RCT, respectively. At meta-analysis, median tumor response rate was 38.84% and 17.30% for HAI and systemic chemotherapy, respectively (HR: 2.24, CI: 1.80- 2.81; P < 0.0001); on average, median OS was 16.04 and 12.64 months, respectively (HR: 0.83, CI: 0.58–1.19; P value 0.30). Conclusions: Although HAI regimens are followed by significantly higher tumor response rates, no OS advantage can be demonstrated for this locoregional treatment as compared to systemic chemotherapy. Since HAI has been so far compared to systemic chemotherapy regimens not containing modern and more effective antineoplastic agents (e.g. oxaliplatin, irinotecan), these findings do not support the clinical or investigational use of FUDR-based HAI. No significant financial relationships to disclose.