Uveal melanoma (UM) is the most common intraocular melanoma, derived from transformed melanocytes of the uvea. Although treatment of primary UM is usually successful, there is a high risk (up to 50%) of liver metastasis with negligible long-term survival. There are currently no reproducible patient-derived animal models that faithfully recapitulate the latter stages of metastatic dissemination of UM, hindering the discovery of curative treatments. To overcome this problem and to accelerate the development of new metastatic UM treatments, we developed a patient-derived zebrafish xenograft (zf-PDX) model, using spheroid cultures generated from metastatic and primary UM tissues. Engrafted UM cells derived from these spheroid cultures give rise to metastatic lesions and recapitulate the molecular features of UMs and their potential drug sensitivity. Importantly, harnessing this versatile model, we reveal a high sensitivity of circulating UM cells to ferroptosis induction in vivo by Erastin and RSL3. Our findings are further corroborated by supportive analysis of patient data implicating ferroptosis as a new, and druggable, target for the treatment of metastatic UM patients, specifically in those with BAP1 loss in the tumor.
