Background: Uveal melanoma is the most common primary intraocular malignancy in adults. Despite successful control of the primary tumor, metastatic disease will ultimately develop in approximately 50% of the patients, with the liver being the most common site. The median survival for patients with liver metastases is about 6-12 months, and there are only few systemic treatment options available providing only small survival benefits. The SCANDIUM trial previously demonstrated significantly superior response rate (40% vs 4.5%) and progression free survival (7.4 vs 3.3 months), compared to best alternative care, in patients with liver metastases of uveal melanoma receiving first-line treatment with isolated hepatic perfusion (IHP). Here we present the primary endpoint, overall survival (OS) rate at 24 months. Methods: In this multicenter randomized, controlled, phase III trial, adult patients with a performance status ECOG 0-1, and with previously untreated isolated liver metastasis from uveal melanoma, were randomized 1:1 between 2013 and 2021 to receive a one-time treatment with IHP or best alternative care (control group). No crossover from the control group to the IHP group was allowed. The primary endpoint was OS rate at 24 months, with the hypothesis of a treatment effect leading to a 50% OS rate in the IHP group compared to 20% in the control group. Results: A total of 93 patients were randomized, with three patients in each group being excluded due to either withdrawal of consent or inappropriate enrollment, and a total of 87 patients were assigned to either IHP group (43 patients) or control group (44 patients). In the IHP group, 41 (89%) patients received IHP. In the control group, the first-line of treatment was chemotherapy (49%), immunotherapy (39%) or localized treatment interventions (9%). In the intention-to-treat (ITT) population, the OS rate at 24 months in the IHP group was 46.5% (95% CI, 31.2-60.4%) compared to 29.5% (95% CI, 17.0-43.2%) in the control group (p = 0.12, Fisher’s exact test). The median OS in the IHP group was 21.7 months (95% CI, 19.1-NA months) compared to 17.6 months (95% CI, 13.5-21.4 months) in the control group (p = 0.10, log-rank test), with a hazard ratio of 0.64 (95% CI 0.37-1.10) in favor of the IHP group. Conclusions: In the SCANDIUM trial, patients with metastatic uveal melanoma receiving IHP experienced a significantly improved PFS. At two years, OS was longer in patients receiving IHP, but the difference was not statistically significant. This could in part be attributed to the control group performing better than expected, potentially due to the introduction of immunotherapy during the study period. Prolonged follow-up of the cohorts will further elucidate how IHP affects OS in patients with uveal melanoma. Clinical trial information: NCT01785316.