Abstract A027: Leveraging genomic copy-number complexity in uveal melanoma to drive novel therapeutic development

Abstract

Uveal Melanoma (UM) is a rare tumor characterized by activating mutations in GNAQ or GNA11, followed by a secondary mutation in BAP1, SF3B1 or EIF1AX. Notably, chromosome 3 copy-number loss, enriched among BAP1-mutant patients, is associated with an increased likelihood of metastasis. While chromosome 3 loss remains a dominant prognostic marker, patient tumor genomic analyses reveal a substantially more complex landscape, including recurrent arm level alterations across multiple chromosomes such as 1p loss, 8q amplification and 16q loss, that collectively define increasingly aggressive disease states ranging from low risk to ultra-high-risk disease. The purpose of this study is to elucidate how coordinated CN evolution drives transcriptional states, functional dependencies, and exploitable therapeutic vulnerabilities in UM. We integrated genomic data from patient tumors and UM cell line models selected to match patient-defined CN states, enabling inference of transcriptional and functional consequences associated with disease progression. Cytoband-anchored RNA profiling was performed across genomically aligned models to link regional CN evolution with transcriptional output. Functional interrogation included genome-wide and custom library CRISPR loss-of-function screens and drug sensitivity profiling. Cytoband-resolved transcriptional profiling of genomically matched UM models revealed coordinated activation of programs associated with invasion and metastatic fitness that tracked with cumulative CN burden rather than chromosome 3 loss alone. These transcriptional states emerged progressively as CN complexity increased, supporting a genomic evolution continuum underlying UM progression. Integration of CRISPR screening within this framework identified a heightened dependency on CBP and p300 in models corresponding to advanced patient CN states, implicating chromatin-dependent transcriptional control as a core driver of high-risk disease. To address the commonly faced issues of toxicity induced by traditional CBP/p300 inhibitors, we engineered a B7-H3 anchored degrader antibody conjugate (DAC) designed to selectively eliminate the transcriptional coactivators CBP and p300 in B7-H3 expressing cells. Our B7-H3 anchored DAC demonstrated potent, selective activity in vitro, highlighting significant response in both low and high-risk UM models. These findings demonstrate that UM progression reflects an integrated CN-transcriptional ecosystem extending beyond chromosome 3 loss, with cytoband-level genomic evolution shaping malignant fitness and therapeutic response. Anchoring therapeutic strategies within this complex genomic context enables rational targeting of advanced UM. Johnathon Rose, Sanjana Srinivasan, Amy Cooper, Jieqing Chen, Pijus Mandal, Guolin Ma, Chiu Yi Liu, Parth Shah, Rosalba Minelli, Joseph Daniele, Michael Peoples, Jason Gay, Khalida Wani, Anastasia Lopez, Ningping Feng, Christopher Vellano, Joseph Marszalek, Giulio Draetta, Alessandro Carugo, Scott Woodman, Alexander Lazar, Andrew Futreal, Giannicola Genovese, Chunhua Shi, Michael Soth, Faika Mseeh, Timothy Heffernan, Piergiorgio Pettazzoni, Virginia Giuliani. Leveraging genomic copy-number complexity in uveal melanoma to drive novel therapeutic development [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Breaking Barriers in the Fight against Rare Cancers; 2026 Jul 18-20; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2026;86(14_Suppl):Abstract nr A027.