BCAN's Funded Research Awards

Meera Chappidi, MD, MPH

Acting Instructor and Fellow in Urologic Oncology

Institution:
University of California, San Francisco

Research:

Optimization and external validation of a histopathology-based, machine learning biomarker to predict neoadjuvant chemotherapy response in muscle‑invasive bladder cancer

Summary:

What the Study Is About 

When bladder cancer grows into the muscle of the bladder, it becomes very hard to treat. For these patients, chemotherapy before surgery to remove the bladder—called neoadjuvant chemotherapy, or NAC—can sometimes improve survival. But not all patients benefit. In some, the tumor disappears completely, while in others, the chemotherapy does little or even allows the cancer to get worse. 

Right now, doctors cannot reliably predict which patients will respond to NAC. This means some patients go through a difficult treatment with serious side effects without gaining any benefit, while others may miss out on the best possible care for their type of cancer. 

How the Study Will Be Done 

This study is working to build a new tool that can predict which patients with muscle-invasive bladder cancer (MIBC) will respond to NAC. Researchers will use artificial intelligence (AI) to carefully study tumor images that are already collected during routine diagnosis. Early results suggest that this AI approach can accurately predict whether NAC will work. 

The team plans to improve the tool by adding more details, such as medical report findings and genetic information from the tumor. To make sure the tool is trustworthy, they will test it on patient samples from several different medical centers. 

Why This Study Matters 

If this tool is successful, it could transform how doctors decide on treatment for patients with MIBC. Patients unlikely to benefit from chemotherapy could avoid the side effects of a treatment that won’t help them, while others might be spared from major surgery if their cancer is more likely to respond well. 

Because the tool uses information already gathered during diagnosis, it could be put into practice quickly. This would make personalized treatment decisions easier and more widely available, giving patients the best chance at both longer survival and better quality of life. 

Citations:

None Reported as of August 2025

Additional Research:

None Reported as of August 2025

Project Collaborators:

NA

Project Status:
Active