At Northwestern Medicine in Chicago, researchers have cracked open one of cancer immunotherapy's biggest puzzles: why the same treatment saves some patients' lives while leaving others without relief. Scientists led by Dr. Joshua Meeks discovered that hidden in a patient's own immune cells lies a predictor of whether bladder cancer treatment will work—a finding published in the Journal of Clinical Investigation that could reshape how doctors personalize care for high-risk bladder cancer.
Bacillus Calmette-Guérin, or BCG, has been the standard immunotherapy for high-risk non-muscle-invasive bladder cancer for years. It works by injecting a weakened bacterial strain into the bladder to provoke an immune response and attack tumors. Yet despite decades of clinical use, doctors have remained puzzled about why some patients experience dramatic improvement while others see their cancer persist unchanged. Understanding this variation matters enormously: bladder cancer strikes tens of thousands of Americans annually, and better treatment predictions could mean the difference between cure and progression for vulnerable patients.
To solve this mystery, Meeks' team used single-cell RNA sequencing—a cutting-edge technology that allows scientists to examine gene expression in individual immune cells—to compare patients who responded well to BCG with those who did not. They analyzed immune cells collected directly from the bladder and from patients' bloodstreams, looking for clues written in the genetic fingerprints of the immune system itself.
What they found was striking. Patients who responded well to BCG showed a marked increase in a specific type of immune cell called Th17-like Th1 cells. These cells are highly inflammatory, producing powerful signaling molecules called cytokines that help the immune system mount an effective attack on tumors. But in patients who failed to respond, the immune picture looked entirely different. Their tumors were filled with exhausted CD8+ T-cells—immune soldiers worn out and less effective at killing cancer—and populated with regulatory T-cells that actively suppress immune activity. Together, these factors create an immunosuppressive environment where cancer thrives despite treatment.
The key to success, Meeks explained, lay in how different immune cells communicate. "The interaction of the bladder macrophages with the T-cells is what's determining the response to the therapy," he said. In responders, myeloid cells encouraged inflammatory, tumor-fighting behavior. In nonresponders, those same cells promoted immune suppression instead. It's a stark reminder that the same player can have opposite effects depending on the overall game being played.
Using machine learning, the team then identified specific biomarkers—certain cytokines associated with Th17-like Th1 cells—that predicted treatment success with remarkable accuracy. They validated these biomarkers in a separate group of patients, strengthening confidence in their predictive power.
While the work is not yet ready for routine clinical practice, it represents a crucial step toward a more intelligent form of cancer medicine. As Meeks noted, "Our previous work provided details about the tumors, and now we've found that the patient's immune system determines treatment response." His laboratory is already planning the next phase: studying whether doctors can deliberately manipulate a patient's immune response to improve treatment success. For bladder cancer patients, that possibility represents genuine hope.