Uveal melanoma (UM) is the most common primary intraocular cancer in adults and remains difficult to treat once it spreads. Although primary tumors can often be controlled, up to 50% of patients eventually develop metastases, most often in the liver. Survival in metastatic disease remains poor.
One reason progress has been slow is the lack of experimental models that accurately reproduce both the genetic changes and the immune environment seen in human UM. In a new study published in Cancer Research, Xu and colleagues describe a genetically engineered mouse model that recreates the step-by-step genetic changes involved in human UM while maintaining an intact immune system.
Human UM usually begins with activating mutations in GNAQ or GNA11. These mutations are common in benign choroidal nevi and represent an early event in tumor development. However, additional genetic alterations are required for malignant transformation. Two of the most important are: loss of BAP1, a tumor-suppressor gene associated with high-risk disease; gain of chromosome 8q, which often increases activity of the cancer-promoting gene MYC.
To reproduce this sequence, the researchers engineered mice carrying a conditional GNAQQ209L mutation, a common oncogenic variant. When this mutation was activated specifically in the uveal tract, the mice developed choroidal nevi, closely resembling early human disease.
Removing Bap1 increased the number of nevi but rarely led to melanoma. However, when activated MYC was added to the model, the mice developed fully penetrant intraocular melanomas, with a median survival of about three months.
This stepwise progression mirrors what is seen in human patients: GNAQ mutation as the initiating event; BAP1 loss linked to high-risk disease; Chromosome 8q/MYC gain driving aggressive tumor growth.
Unlike many experimental tumor models, which rely on transplanting tumor cells into immunodeficient mice, this system preserves a functioning immune system. Because immune interactions remain intact, the model may be particularly useful for studying immunotherapies, an area where clinical progress in UM has so far been limited.
The researchers also identified two main types of tumor cells within the melanomas: a melanocytic state, which resembles differentiated pigment-producing cells; a more primitive neural crest-like state.
Analyses suggested that tumor cells can transition from the melanocytic state to the neural crest-like state as the tumor progresses.
The neural crest-like cells showed gene expression patterns similar to high-risk Class 2 human UM and were associated with greater genetic instability, including copy number changes comparable to chromosome 8q gains and chromosome 3 loss seen in patients.
These findings suggest that UM progression involves not only accumulating genetic mutations but also changes in tumor cell identity, which may contribute to metastasis and immune evasion.
By combining physiologic oncogene expression, stepwise genetic changes, an intact immune system, and metastatic behavior, this model provides a powerful new tool for studying uveal melanoma. It offers researchers a platform to investigate key drivers of tumor progression, explore tumor–immune interactions, and test potential targeted or immune-based therapies.
For a field that has long lacked reliable experimental systems, the work represents an important step toward translating laboratory discoveries into improved treatments for patients with uveal melanoma.
