Damon Runyon Researchers

The treatment of metastatic melanoma has been transformed over the past decade with the development of (1) targeted therapies that target a very common gene mutation (BRAF mutations in 50-60% of tumors) in melanoma and (2) two different types of immune therapies that induce the immune system to attack the cancer (CTLA-4 and PD-1 inhibition). However, not all patients respond to either targeted or immune therapy, and there is evidence suggesting that patients who quickly develop resistance on the initial therapy (whether targeted or immune) have worse outcomes (e.g. rapid resistance) when switched to other therapy approaches. Thus, being able to predict which patients will respond to targeted and immune therapies is critically important to personalize therapy and improve patient outcomes. Dr. Liu proposes to address this issue by analyzing large cohorts of melanoma patients treated with targeted and immunotherapy with deep genetic sequencing and molecular characterization of their tumors, developing algorithms to identify molecular features that predict differential response or resistance to therapy. Importantly, he will apply approaches from machine learning to develop and validate a predictive model integrating genetic and clinical tumor characteristics to predict response to therapy in individual patients. The validated molecular features in the model will shed light on the biological mechanisms that underpin response and resistance to therapy, identifying novel therapeutic targets and potential synergistic therapy combinations. Taken together, the success of this proposal will have important biological and clinical implications, informing future drug and combination therapy development as well as impacting clinical care.