Response Assessment of Metastatic Uveal Melanoma Treated With Rose Bengal Disodium

Abstract

Background: Rose bengal disodium (PV-10) is a small molecule oncolytic immunotherapy in clinical development for treatment of solid tumors. Upon intralesional injection, it can produce immunogenic cell death and a T-cell mediated immune response against treatment-refractory and immunologically-cold tumors. Methods: PV-10-LC-01 (NCT00986661) is an open-label phase I basket study evaluating safety, tolerability and preliminary efficacy of PV-10 in patients (pts) with solid tumors metastatic to liver. PV-10 is administered percutaneously to 1 or more hepatic tumors 1.0-4.9 cm in diameter with response assessment via contrast-enhanced CT, MRI and/or FDG PET at Day 28 then q12w. Pts with multiple injectable tumors may receive further PV-10 after Day 28. Eligible pts can receive concomitant standard care checkpoint blockade immunotherapy. In a single-center cohort of uveal melanoma pts we compared overall response and progression-free survival (PFS) using 2-dimensional European Association for the Study of Liver (2D-EASL) criteria and criteria that consider lesions outside of the liver (RECIST 1.1, irRC, irRECIST and iRECIST). Results: Eight pts who received at least 1 injection of PV-10 and had baseline and follow-up imaging were assessed: 5 received 2 doses and 3 received 1 dose. There was a difference in overall response by RECIST vs. the other criteria in 4 pts: these pts developed progression on RECIST but remained stable by 2D-EASL, irRC, irRECIST and iRECIST with disparate PFS of 91 days (RECIST) vs. 145 days. Progression by RECIST was based on appearance of new lesions for 3 pts and increase in target lesion size in 1 pt; in 1 pt the injected lesion demonstrated complete response based on 2D-EASL obtained after progression as determined by RECIST. For the remaining pts, overall response of stable disease was concordant between all response criteria; 2D-EASL demonstrated sustained partial response in injected lesions in 2 of these pts. Conclusion: Imaging response assessment of pts treated with PV-10 using RECIST may lead to premature and inaccurate determination of progression. 2D-EASL provides information specific to the behavior of injected lesions. Immunotherapy-centric criteria (irRC, irRECIST and iRECIST) could be a useful alternative to 2D-EASL.