The feasibility of ex vivo gene therapy as an alternative to liver transplantation for the treatment of liver metabolic diseases needs to be analyzed in large animal models. This approach requires appropriate gene transfer vectors and effective hepatocyte engraftment. Lentiviral vectors have the ability to transduce nondividing differentiated cells, such as hepatocytes, and portal vein occlusion increases hepatocyte engraftment. We investigated whether reversible portal vein embolization combined with ex vivo lentivirus-mediated gene transfer is an effective approach for successful hepatocyte engraftment in nonhuman primates and whether the transgene remains expressed in the long term in transplanted hepatocytes in situ. Simian hepatocytes were isolated after left lobe resection, and the left and right anterior portal branches of animals were embolized with absorbable material. Isolated hepatocytes were labeled with Hoechst dye or transduced in suspension with lentiviruses expressing green fluorescent protein under the control of the human apolipoprotein A-II promoter and transplanted via the inferior mesenteric vein. The whole procedure was well tolerated. The embolized liver was revascularized within 2 weeks. The volume of nonembolized liver increased from 38.7% +/- 0.8% before embolization to 55.9% +/- 1% after embolization and hepatocytes significantly proliferated (10.5% +/- 0.4% on day 3 after embolization). Liver repopulation after transplantation with Hoechst-labeled hepatocytes was 7.4% +/- 1.2%. Liver repopulation was 2.1% +/- 0.2% with transduced hepatocytes, a proportion similar to that obtained with Hoechst-labeled cells, given that the mean transduction efficacy of simian hepatocyte population was 34%. Transgene expression persisted at 16 weeks after transplantation. Conclusion: We have developed a new approach to improve hepatocyte engraftment and to express a transgene in the long term in nonhuman primates. This strategy could be suitable for clinical applications. (HEPATOLOGY 2009.).