Insights From: Daniel J. George, MD, Duke University Medical Center
Daniel J. George, MD: Immunotherapy, as exciting as it is in bladder cancer, doesn’t respond in everybody. And the reality is that there are patients who we can probably identify up front and who are less likely to respond. These are typically patients who have more visceral disease: liver metastases, other soft tissue metastases, central nervous system metastases, or bone metastases. These patients tend to have response rates of 10% or less. It’s not that they don’t respond at all, it’s just that they’re much less likely to respond. Whereas, our lung-only and our lymph node patients tend to respond more. We tend to see response rates in the 30% or higher in these groups. Additionally, when we look at patients who have upper tract disease, historically that has been a more difficult patient population to treat. They have less kidney function, their disease tends to spread more aggressively, and yet they have as good or better response rates in many studies to PD-1, PD-L1 inhibitors than even patients with more historical bladder cancers.
In many tumors, PD-L1 expression corresponds with a worse prognosis. If you think about it, it is a mechanism for how tumors can evade the immune system, and bladder cancer is no exception. We see these are difficult tumors to treat. These are tumors that are associated with immune escape, and we don’t have a lot of treatments in bladder cancer to begin with. Adding PD-L1 status to this just makes this tumor that much harder, and yet that has been one of the interesting features of early studies with a PD-1 and PD-L1 inhibitors. And it suggests that this is an important marker. Interestingly though, it hasn’t been the clear predictor it is for response and survival benefit that we see in other cancers.
PD-L1 status is typically done on archival specimens in bladder cancer. Many of these patients don’t have time to get a biopsy or access to tissue. Using archival specimens, we do see PD-L1 status and it does vary in terms of its level of expression and in terms of both the tumor tissue as well as in the surrounding infiltrating lymphocytes. And although there is generally a trend that’s positive in favor of the PD-L1-positive tumors responding better to PD-L1 and PD-1 inhibitors, it has not been something that has been statistically significant. So, in general, this hasn’t panned out as a biomarker for this population. The fact is because this has been a population that has gotten so few treatment options, we’re going to treat everybody regardless of PD-L1 status and give them the best chance. And there are some really durable and remarkable responses in PD-L1-negative tumors.
Other factors that are also relatively either prognostic or predictive for response to checkpoint inhibitors would be things like performance status, prior chemotherapy, patients who have a high mutational burden, and patients who have any signs of microsatellite instability. These have all been associated with higher response rates.
Lastly, probably one of the most common and important association with response to checkpoint inhibitors has been smokers. Historically, these are the patients who have been the most difficult to treat with bladder cancer, both because of their comorbidities as well as the complexity of their disease brought on by the carcinogenesis associated with smoking. And yet it may be just precisely that environmental factor that makes these tumors more responsive to checkpoint inhibitors. And so, for 50% or more of the population of bladder cancer patients who have a significant history of active or prior smoking, some good news to offer them with checkpoint inhibitors.