Selective Internal Radiotherapy (SIRT) and Chemosaturation Percutaneous Hepatic Perfusion (CS-PHP) for Metastasized Uveal Melanoma: A Retrospective Comparative Study

SIMPLE SUMMARY: Uveal melanoma usually shows a liver-dominant metastasis spread. Even in the light of multiple treatment options, including liver-targeted therapies, patients die early from this disease. Two promising treatment options are transarterial radioembolization (SIRT) and chemosaturation-percutaneous hepatic perfusion (CS-PHP), and in this study, we retrospectively compared their efficiency on 62 patients (SIRT = 34, CS-PHP = 28) receiving multiple treatment cycles. By a standardized evaluation of the tumor response using RECIST 1.1, we saw a disease control rate of 18% for SIRT and in 30% for CS-PHP. The median of progression-free survival was 127.5 days for SIRT and 408.5 days for CS-PHP, and advanced analysis showed this to be not significant. The median overall survival after treatment was 300.5 days for SIRT and 516 days for CS-PHP, and advanced analysis showed this to be significant. We conclude that for these patients, CS-PHP might therefore be preferable. ABSTRACT: Even with liver-targeted therapies, uveal melanoma with hepatic metastasis remains a challenge. The aim of this study was to compare the outcome of patients treated with either SIRT or CS-PHP. We included 62 patients with hepatic metastasized uveal melanoma (n = 34 with SIRT, receiving 41 cycles; n = 28 with CS-PHP, receiving 56 cycles) that received their treatments between 12/2013 and 02/2020 at a single center. We evaluated their response according to the RECIST 1.1, as well as progression-free survival (PFS) and overall survival (OS), after the initiation of the first cycle of the liver-directed treatment using Cox regression, adjusted via propensity score analysis for confounders, including the amount of hepatic involvement. The disease control rate was 18% for SIRT and 30% for CS-PHP. The median (range) of PFS was 127.5 (19–1912) days for SIRT and 408.5 (3–1809) days for CS-PHP; adjusted Cox regression showed no significant difference (p = 0.090). The median (range) of OS was 300.5 (19–1912) days for SIRT and 516 (5–1836) days for CS-PHP; adjusted Cox regression showed a significant difference (p = 0.006). In our patient cohort, patients treated with CS-PHP showed a significantly longer OS than patients treated with SIRT. CS-PHP might therefore be preferable for patients with liver-dominant metastatic uveal melanoma.