Advancing Immunotherapy Impact in Bladder Cancer

Immunotherapy agents may be taking the field of urothelial cancer by storm, according to Arlene O. Siefker-Radtke, M.D., but the therapies are still only responding in 15 percent to 25 percent of patients.

“Clearly, immune checkpoint inhibition is here to stay,” says Siefker-Radtke, associate professor, Department of Genitourinary Medical Oncology, Division of Cancer Medicine, The University of Texas MD Anderson Cancer Center, clinical co-leader, Bladder SPORE Executive Committee. “We can stimulate the immune system and we are benefiting some of our patients who were previously incurable, yielding genuine durable partial and complete responses with these drugs. Unfortunately, the fraction is still low.”

Siefker-Radtke lectured on the dramatic shift in bladder cancer treatment during the 2017 OncLive^® State of the Science Summit on Genitourinary Malignancies. In an interview during the meeting, she shed light on the exciting advances in the bladder cancer field, why researchers should move past PD-L1 as a biomarker and what combinations can be expected to move through the pipeline next.

Urothelial cancer treatment has drastically changed in less than one year. What did you highlight in your talk?

Siefker-Radtke: It truly is an exciting time to be in the field of urothelial cancer, since we’ve had our first approval for treatment in over 30 years. To finally have a new drug during my career is quite an exciting time. My presentation focused on the immune treatment and the immune landscape for the treatment of patients with urothelial cancers.

Many people don’t realize that urothelial cancer was the first tumor to have an associated immune response. We saw that with the development of Bacillus Calmette-Guerin (BCG) over 50 years ago, where they actually saw responses with urothelial tumors in patients getting BCG vaccinations. That is what led to the development of BCG as an intravesical therapy for treatment of urothelial cancer.

Over the years, we’ve tried to build on the immune response that we’ve seen in some patients. We even did a trial of interferon alpha given systemically with chemotherapy, which, while it worked and we saw activity, it was actually more toxic than MVAC chemotherapy. Again, this was a poor immune-modulating drug with interferon alpha given systemically.

However, more recently, PD-1/PD-L1 inhibitors are taking the field by storm. There are multiple companies, multiple drugs and my current prediction is that all of them will be approved because they’re showing definite activity for urothelial tumors.

The first was Tecentriq (atezolizumab), which was approved on May 18, 2016. For those in the field, [that date] is the equivalent of knowing where you were when Lady Diana passed away or when the Twin Towers fell. For me, it’s knowing when the first drug was approved. It was Tecentriq in the second-line treatment of patients who received prior cisplatin-based chemotherapy. Response rates were on the order of 15 percent with a median survival of 10 months. What is truly exciting about these drugs is the durability of responses that we are seeing with them.

We are seeing this with other drugs, as well. The PD-1 inhibitor Opdivo (nivolumab) was recently approved for a similar setting — cisplatin-failures and second-line treatment of urothelial cancer. Response rates appear similar; you could argue that the 20 percent to 25 percent is better than 15 percent. Patient selection can certainly be contributing to that. The median survival is again very similar to what we saw with the Tecentriq data.

What all of these drugs are showing are durable responses. I personally have patients on immune-checkpoint inhibitors who have been on therapy with durable partial, and even complete, remissions now for over two and a half years.

When we see the durable long-term benefit in these patients, it does leave one to wonder: are we going to see a cure with checkpoint inhibitors? It is too early to tell. We see a measure of five years in most cases. We are seeing this durable benefit that’s giving everyone hope that we’re going to cure a previously incurable group of patients with bladder cancer.

In your lecture, you suggested it's time we throw PD-L1 as a biomarker by the wayside. Why?

PD-L1 hasn’t been a good marker for patients with urothelial cancer. There have been hints in some studies that the response rate might be improved when you look at the PD-L1 expression on the immune infiltrate in addition to PD-L1 expression on the tumor cell.

But, the truth is, the response rate — even in PD-L1—negative tumors, is better than single-agent taxane. Those responses can also be durable. PD-L1 is a very dynamic biomarker; it can be upregulated in patients who have had chemotherapy, BCG given intravesically or treatment with a PD-L1 inhibitor. That might be the reason why the data have been all over the board. We haven’t seen any definitive statements that you should not consider checkpoint inhibition when PD-L1 is low, because they still respond and may still benefit from treatment.

We also see a lot of variability in the antibodies that have been developed; some may be better than others. We have seen different cutoff points; again, we see responses in those tumors that are completely negative. At the moment, I can’t recommend PD-L1 expression as a needed biomarker because patients who don’t have it may still have benefit, may still respond and may even have that long-term durable response that buys them substantially more time.

Are there any ongoing biomarker studies to move this question along?

There aren’t any conclusive markers. Some of the work that our group has done has been looking at gene expression profiling and trying to characterize bladder cancers on the basis of their gene profile. Gene expression profiling helps us predict underlying biology; we've seen in data that there is a group of tumors that have more basal features. They have markers of stemness, very high proliferation rates and they tend to do very poorly when left untreated. They have the worst overall prognosis.

There is a group of luminal tumors that appear to be intermediate markers of proliferation. They may benefit somewhat from chemotherapy. We also saw FGF receptor expression and FGFR3 mutations enriched in that group of patients.

There is another group with p53 in luminal 2 tumors, where we see more markers of stroma and the lowest proliferation rates in that group. If you look at the survival, the basal tumors do the worst if left untreated. The p53 in the luminal tumors tend to do the best. That probably reflects the proliferation rate of each of these subtypes.

Intriguingly, when we look at the impact of chemotherapy and patients who could receive cisplatin-based chemotherapy, we see these markers of poor prognosis change into more predictive markers. The basal subtype predicts for response to systemic chemotherapy.

We are gaining a sense that bladder cancer is not one disease. There are multiple different types of bladder cancer, and subtyping can start predicting biology and potential benefit from treatment.

We are also seeing that with checkpoint inhibitors. The Tecentriq data suggest that luminal 2 tumors responded better to them, while the luminal 1 tumors responded very poorly to checkpoint inhibitors. The Opdivo data suggest that basal tumors responded better to these agents. But they both showed that luminal 1 tumors had a very poor response to checkpoint inhibitors.

What is intriguing about the biology is if you look at the luminal 1 tumors, this group of tumors appears very immune quiescence. They lack an immune infiltrate. My hope is, with using gene expression profiling, we'll be able to start predicting which tumors respond and which ones benefit from specific therapies. We can use this as a framework that we can build upon to understand the biology underlying the different types of urothelial cancer.

Are combinations next?

Combinations are clearly coming down the road. We saw some data where treatment with Opdivo and Yervoy (ipilimumab) had an improved response rate compared with Opdivo alone. The risk is that there is a higher risk of having an immune-mediated toxicity.

Intriguingly, I wonder if some of those patients with tumors with the immune toxicity might also be the ones who benefit. I have patients who have durable remissions despite having an immune-mediated toxicity and having to stop immunotherapy with the combination of Opdivo plus Yervoy. Clearly, that combination is showing promise. There are other combinations being tested in other trials.

At the moment, we're all hopeful that some of these trials will be positive but, clearly, we have to wait for the data to show us if it definitely is.

What do you think is the main message for community oncologists who sat in your presentation?

Response rates are between 15 percent and 25 percent. There is a group of patients who have their disease slow down or don’t progress as rapidly who also benefit. However, the fraction with the durable responses is still clearly less than 5 percent to 10 percent. We still have a lot of work to go, building the better combination, the better sequence and understanding the biology of these tumors to hopefully lead to greater fractions of patients achieving those durable partial and complete responses that will change the field of urothelial cancer.