Development of deep learning model predicting immune checkpoint inhibitor outcomes in urothelial bladder carcinoma

Bladder cancer is a common disease affecting hundreds of thousands of people each year around the world. One type, called urothelial bladder cancer, is the most frequently diagnosed. In cases where the cancer has spread or become advanced, many patients are treated with a type of immunotherapy called immune checkpoint inhibitors (ICIs). These medications help the body’s immune system recognize and attack cancer cells. However, not all patients benefit from this treatment, and doctors currently have limited ways to predict who will respond well.
This research focuses on developing a new way to help doctors better predict whether a patient with urothelial bladder cancer will respond to ICI treatment. Specifically, we are studying routine blood tests that patients already receive before and during treatment. These blood tests include simple measurements, such as the number of certain types of white blood cells and the level of a protein called lactate dehydrogenase (LDH). Previous studies suggest that these results may be linked to how well a patient will do on immunotherapy.
We plan to use deep learning, a type of computer-based model that can recognize complex patterns, to analyze large sets of blood test data collected from patients in earlier clinical trials. By doing so, we aim to build a tool that can identify patterns in the blood that are linked to better or worse responses to ICIs.
The goal of this research is to create a reliable, easy-to-use tool that can help doctors make more personalized treatment decisions. If successful, this approach could lead to better outcomes for patients by identifying those most likely to benefit from immunotherapy while helping others avoid treatments that might not work for them.