Read the Full Video Transcript

Charles Ryan: Hello from ASCO 2019. I'm delighted today to be joined by Dr. Jonathan Rosenberg, who is Chief of the Genitourinary Oncology Service at Memorial Sloan Kettering Cancer Center. Thank you for joining us, Jonathan. Today we're going to talk about your Phase III trial in bladder cancer, CALGB 90601. This is a study that looked at bevacizumab plus standard chemotherapy. You're presenting it here. A big important study. Tell us about the results.

Jonathan Rosenberg: So, this trial enrolled 506 patients. They were randomly assigned to either gemcitabine, cisplatin, bevacizumab, or gemcitabine, cisplatin, and placebo. And this study was based on phase two data from two single-arm trials, one with gemcitabine, cisplatin, and bevacizumab, another with gemcitabine and cisplatin and I'm sorry gemcitabine, carboplatin, and bevacizumab. And they both showed better than expected outcomes compared to what we might expect with standard gemcitabine and cisplatin chemotherapy or gemcitabine carboplatin chemotherapy.

In fact, both clinical trials showed a 4-5 month improvement in overall survival compared to historical controls. And so this provided the rationale to move forward. Also, angiogenesis is known to be a negative prognostic indicator in bladder cancer and that the addition of anti-angiogenic agents to chemotherapy and preclinical models actually leads to improved tumor regression and improved outcomes in animal models. So the randomized phase three trial was started in 2009, accrued over five years and accrued 500 patients in that time period.

Patients were treated with the combination of chemotherapy plus bevacizumab or placebo for six cycles and then if they had an ongoing response or stable disease, they were able to continue on bevacizumab or placebo. The trial was designed to look at overall survival as the primary endpoint and progression-free survival greater than or equal to grade three adverse events and objective response rate were secondary endpoints.

Unfortunately, the trial was negative. The hazard ratio for overall survival was 0.87 and the P value was not significant. It was designed to detect a hazard ratio of 0.74 and we obviously, unfortunately, missed the mark. Interestingly enough, progression-free survival with the three-drug regimen, gemcitabine, cisplatin, and bevacizumab was actually better longer than gemcitabine, cisplatin, and placebo with a hazard ratio of 0.75 and the significant P value, but the delta, the difference between the two arms was only 1.1 months. And so not necessarily a clinically significant difference in the setting of a secondary endpoint in a first-line clinical trial.

Charles Ryan: What's striking is from the time that you designed that study and you looked at that initial phase two data and executed the phase three trials you have, a lot has changed in bladder cancer. So the natural question is if PFS is positive, even if it's a minor amount, was subsequent therapy the issue that may have affected the survival endpoint? Do we know that yet?

Jonathan Rosenberg: So we don't have complete data on that yet and we're still looking at that question, but my sense is that very few patients actually received immune checkpoint inhibitors, which would be the major class of agents that came into use after 2014. And so because of the time of the study that when it went on, only a small fraction of patients actually went on to receive immunotherapy and so it's unlikely that that really led to these results.

We also saw that you know, the combination was what you'd expect with adding bevacizumab to chemotherapy. You had some more cardiac events maybe and least severe events. You had some more Cytopenias, particularly thrombocytopenia. You had hypertension and proteinuria, but it didn't seem to be a toxicity problem. The regimen overall, in general, was well tolerated.

Charles Ryan: I see.

Jonathan Rosenberg: So you know this, this study was designed in an era when there was a lot of nihilism about completing randomized phase two bladder cancer trials.

Charles Ryan: Absolutely.

Jonathan Rosenberg: There were history and a legacy of multiple failed clinical trials in advanced bladder cancer. And so the US Intergroup Mechanism, the cooperative groups really managed to pull it together and accrue this trial. And while the trial was negative, it was just doing that in itself is an enormous achievement. And I think it may have changed the mindset of people and suddenly when we had interesting new drugs, in addition to what we've been testing, allowed people to realize that, well, we can actually accrue these clinical trials.

Charles Ryan: Right. And you and I work closely together in the cooperative group system and we're both big advocates for it. And you know, my perspective is, even though it's a negative trial, we're still going to learn a lot about bladder cancer from this study. I'm going to ask you about that in a second. But the other thing is just to congratulate you and to really praise the cooperative group system, which is able to ask the questions that industry's not asking.

And you know, we didn't know what the answer was going to be when we started this study many, many years ago. So instead of calling this a failure, we say, "Well that question, the answer's no. But we know it now." And so where do you go from here in terms of what you're going to tell us about bladder cancer moving forward?

Jonathan Rosenberg: Well I think I just want to make one quick point that the phase two data looked really promising and so it really shows the importance of confirmatory phase three trials that phase two studies are not really able to look at overall survival. And that's why we need good phase three clinical trials. And so if we had changed practice based on phase two data, we would have been exposing patients to an agent that is not adding to their care in general.

Charles Ryan: With some cost and some risk.

Jonathan Rosenberg: Exactly. So just highlights the importance of doing this and as we get larger single-arm phase two trials where we're getting accelerated approval of drugs based on those data, the confirmatory trials really are important to understand what the contribution of those, of the actual treatment, is to the patient benefit.

Because we've known that the pretreatment prognostic factors for bladder cancer are actually very important in determining the outcome of a phase two trial. You can have a combination or a drug that looks spectacular if you have patients with lymph node only metastases and you can have a combination that looks terrible for patients who have liver metastases and they may be the exact same or it may be that the one that looked worse is actually better, but you'll never know unless you actually do the trial.

Charles Ryan: Right, right.

Jonathan Rosenberg: So I think it's very important that we did this. We are going to learn a lot from this trial actually, and we do have a DOD grant focusing on untangling the biology of bladder cancer in the metastatic setting. We are looking at DNA mutations. We're looking at RNA expression profiles and expression subtypes, immune subtypes, angiogenesis signatures, as well as extreme responders, as well as really developing a catalog of chemosensitivity mutations. There's ample data out there nowadays suggesting that there are at least some bladder cancers that have exuberant responses to cisplatin-based chemotherapy.

We're developing data that's actually allowing us to identify those patients and in fact, there are clinical trials that are accruing both nationally and regionally to answer that question and allow us to see whether or not we can identify an exquisitely sensitive population to chemotherapy. And so hopefully we'll learn information about that. It also will serve as an important control group when we analyze the data from the randomized phase three trials that are adding immunotherapy to chemotherapy in bladder cancer or immunotherapy alone.

Because we'll have an analogous data set in a chemotherapy only population of about 250 patients on the control arm that we can then use as a historical comparator, another dataset of chemotherapy-treated patients for us to understand the biology.

Charles Ryan: Sure, so here's a great example where an NIH sponsored trial, although negative for its primary endpoint, may end up helping bladder cancer patients in the future because of the prognostics work that you'll be able to do, the biomarker work that you'll be able to do. Really important work. This is hard. This takes a long time. It requires a lot of patience and a lot of patients and your work is really appreciated.

And I think, yeah, as I'm listening to you talk, I'm also thinking for all the excitement about immunotherapy and other targeted therapies in bladder cancer, chemotherapy is not going away. So we need to continue to focus on how we can make it better.

Jonathan Rosenberg: Right. There are a small proportion of patients who are cured with chemotherapy, with this disease, about 10%. And we see that in the data. There is a plateau in the curve and actually given the historical lack of ability to crew studies, we minimize the sample size, which means we had to maximize the number of events. And in fact, there's a group of patients that never dies of bladder cancer on these trials.

And so that kind of matched up and that's why the trial took so long to revamp because there is a tail on the curve with chemotherapy.

Charles Ryan: That good problem of people doing well.

Jonathan Rosenberg: Right. A small percentage albeit.

Charles Ryan: Right.

Jonathan Rosenberg: We need to raise that bar.

Charles Ryan: Yeah. Yeah. Well, congratulations again and thank you so much for joining me today.

Jonathan Rosenberg: Thank you. Very good. Thanks.