Atezolizumab Approaches FDA Approval in Bladder Cancer, But Biomarkers Remain Unclear

Matthew Galsky, MD

The PD-L1 inhibitor atezolizumab showed promising response and durable activity in patients with metastatic urothelial carcinoma, particularly in those with increased levels of PD-L1 expression, according to results of the phase II IMvigor study recently published in The Lancet.

Based on the IMvigor data the FDA has granted atezolizumab priority review as a treatment for patients with locally advanced or metastatic urothelial carcinoma who progressed during or after platinum-based chemotherapy in the metastatic setting, or whose disease worsened within 12 months of receiving platinum-based chemotherapy before or after surgery.

Study author Matthew Galsky, MD, says the IMvigor findings represent a “watershed moment” for the treatment of metastatic urothelial carcinoma.

“We really have not had a lot of good treatments for patients in this disease state, ever,” says Galsky, who is an associate professor of Medicine, Hematology and Medical Oncology and assistant professor of Urology at Mount Sinai Hospital. “Once patients progress on standard first-line chemotherapy, we really have very little to offer them. Atezolizumab could potentially be the first drug approved in this space.”

In the study, IMvigor 310 patients received atezolizumab. PD-L1 expression was determined by infiltrating immune cells (ICs) in the tumor microenvironment and was defined by the percentage of PD-L1—positive immune cells. Patients were identified as IC0 (<1%), IC1 (≥1% but <5%), and IC2/3 (≥5%).

The primary analysis showed that, compared with a historical control, overall response rate (ORR) of 10%, treatment with atezolizumab resulted in a significantly improved RECIST v1.1 objective ORR for each immune cell group, with a 27% ORR in the IC2/3 group (95% CI, 19-37; P <.0001), 18% ORR in the IC 1/2/3 group (95% CI, 13-24; P = .0004) and 15% ORR in all patients (95% CI, 11-20; P = .0058).

With longer follow-up data by independent review, objective response rates were 26% (95% CI, 18-36) in the IC2/3 group, 18% (95% CI, 13-24) in the IC1/2/3 group, and 15% (95% CI, 11-19) overall in all 310 patients. At a median follow-up of 11.7 months (95% CI, 11.4-12.2), ongoing responses were recorded in 38 (84%) of 45 responders.

Regarding grade 3/4 treatment-related adverse events, fatigue was the most common, occurring in 50 (16%) of 310 treated patients. Grade 3/4 immune-mediated adverse events occurred in 15 (5%) of patients, with pneumonitis, increased aspartate aminotransferase, increased alanine aminotransferase, rash, and dyspnea being the most common. No treatment-related deaths occurred during the study.

In addition to PD-L1 expression being linked to response, an exploratory analysis also showed The Cancer Genome Atlas (TCGA) subtypes and mutation load to be independently predictive for response to atezolizumab. This study was the first to investigate TCGA subtypes with response to immune checkpoint in metastatic urothelial carcinoma.

However, the role of both PD-L1 and TCGA subtyping remains unclear in this patient population, says Galsky.

OncLive: What are the most significant findings from the IMvigor study that oncologists should be aware of?

In an interview with OncLive, Galsky explains the challenges of utilizing molecular subtyping and PD-L1 as biomarkers. He also discusses the impact a potential approval of atezolizumab could have in metastatic bladder cancer, possible combinations that could increase ORR, and next steps in understanding the IMvigor study.Galsky: For patients with metastatic urothelial cancer that has progressed, despite first-line chemotherapy, we really haven’t had any standard therapies for the past several decades. In this study, a subset of patients responded to PD-L1 blockade, and those patients tended to have durable responses. That is the major take-home message of this trial.

What role could subtyping have in predicting response to atezolizumab?

However, there is an ORR of 15% in this study, which means that, unfortunately, the majority of patients will not respond to this treatment, so there is much more work to be done. But this is still very significant because of the lack of really anything working well in this patient population.In the past few years, it has been recognized that bladder cancer, at the transcriptome level and at the molecular level, is not one disease but several different diseases. A few independent groups developed classification schemes based on gene expression profiling of bladder cancer specimens, and those overlap somewhat with what TCGA has found.

We have these different subtypes, and there have been some recent efforts to unify these subtypes so that everyone is talking about a standard classification system. These subtypes we know differ in terms of expression of some genes that are important in the pathogenesis of cancer.

In some studies, there has been a link with these subtypes and response to standard chemotherapy, but this is really the first study to explore whether or not these subtypes are relevant, with regards to response to immune checkpoint blockade.

Interestingly, what this study shows is that there is enrichment in responses in certain subtypes. Those subtypes are ones that are also enriched for immune-related genes.

Do you see molecular subtyping becoming common practice in bladder cancer?

Not only do we see this enrichment of response that could potentially have critical implications, in terms of identifying patients who are most likely to respond, but it also establishes some biological proof of concept because the enriched responses are occurring in these sets of patients whose tumors are enriched to immune-related genes.It is a work in progress for several reasons. We have a few different groups that have developed these subtypes, which are largely overlapping but not totally. There needs to be some standardization and harmonization before subtyping is ready for prime time.

The second challenge—and this is really a challenge that we’ve seen with a lot of biomarkers in the era of immune checkpoint blockade&mdash;is, although there is an enrichment of response within a certain TCGA subtypes, there are patients in every TCGA subtype who have responded to treatment.

Is PD-L1 expression an effective biomarker in this space?

There are also many patients in every subtype who didn’t respond to treatment. It is going to be hard to use this type of information currently to determine whether or not a patient should get a treatment like this. We would most certainly deny some patients treatment who might respond and don’t otherwise have good treatment options.The way that this study looked at PD-L1 expression is by identifying PD-L1 expression in tumor infiltrating cells—immune cells that are in the immune microenvironment. That is sort of different than what other companies are arguing in their trials. They are, oftentimes, looking at PD-L1 expression in tumor cells themselves rather than the immune cells.

Using those definitions, it is clear from this study that patients with a higher level of PD-L1 expression have a higher likelihood of responding to treatment. Based on an independent review, the response rate was 26% of patients in the highest expression group, compared with a 15% response rate in all comers. We can clearly enrich the percentage of patients who are going to respond using a biomarker like this.

However, the problem is similar to what I outlined regarding subtypes. Twenty-six percent of patients respond, so there are a lot of patients who don’t respond even with this biomarker. Moreover, there are patients with no biomarker expression who respond and who have good responses.

Is there potential to eventually combine atezolizumab with other agents to attempt to increase response?

What are the next steps planned for the IMvigor study?

Therefore, it is very difficult to imagine the clinical utility that this will have in the immediate term, given the lack of other good treatment options for bladder cancer. That being said, we will see if this drug is approved, whether there is an indication on the label to restrict the treatment to certain subsets, based on this data. Most people in the field are of the opinion that, because the response rate in patients with no expression of PD-L1, which is about 10%, is as good as anything that we have for patients in this setting, it is hard to justify using that biomarker to select patients.Absolutely. There are many rational combinations that are being considered. There is emerging data, as well as previous data, suggesting the potential of combinations of atezolizumab with standard chemotherapy, radiation, targeted therapies, other immune therapies, vaccines, and other immune checkpoint inhibitors. Across the board, there are a large number of combinations that are worthy of investigation, many of which have already been initiated.There were 2 cohorts of patients enrolled. The cohort of patients that we recently reported consists of patients who have already received chemotherapy for metastatic disease and, unfortunately, had their cancers progress despite that chemotherapy.

There is a second cohort which also represent a major unmet need. Those are patients with metastatic urothelial cancer who are ineligible to receive cisplatin-based chemotherapy for a variety of reasons, including poor kidney function, which is not uncommon in this patient population.

What is on the horizon for immunotherapy in bladder cancer?

There are really no standard treatment approaches in such patients, although we do use some default chemotherapy regimens. In that cohort of patients that has not been reported yet, that data will be forth coming. It will represent a first-line treatment for those patients, which will give us insight into the potential of using atezolizumab earlier.It will become the next pillar for standard treatment that we have for this disease, alongside standard chemotherapy, radiation, and surgery. The two main areas that need to be explored are these types of drugs in other disease states of bladder cancer—adjuvant, neoadjuvant, and first-line settings. All of those studies are currently ongoing.

The other area is to figure out how we can improve this response rate by developing rational combination strategies. Those are also being actively explored.

Rosenberg JE, Hoffman-Censits J, Powles T, et al. Atezolizumab in patients with locally advanced and metastatic urothelial carcinoma who have progressed following treatment with platinum-based chemotherapy: a single-arm, multicentre, phase 2 trial [published online March 4, 2016]. Lancet. doi:10.1016/S0140-6736(16)00561-4.