Transcript:Daniel P. Petrylak, MD: As for similarities, the data seem to be similar in this situation. We’ve got activity of these agents. What about moving these agents up front?

Arjun V. Balar, MD: Right. So, the natural and tantalizing question after you demonstrate durable responses and complete responses in the second-line setting is, is there a role for immunotherapy in the frontline setting? And, Dr. Bajorin, you touched on the cisplatin-ineligible patient population, where we use a standard of care of gemcitabine and carboplatin and we see median survival of 9 months, no long-term survival. The curve from EORTC-30986 is a very depressing curve. So, that’s the context in which we are dealing with immunotherapy in the frontline setting. And I was privileged to lead 2 large studies of PD-1 and PD-L1 blockade in the first-line setting, and this is IMvigor-210 cohort 1 study and the KEYNOTE-052 study.

I’ll first touch on cohort 1 of IMvigor-210, and this focused on roughly 119 patients. They were treated with atezolizumab, 1200 mg every 3 weeks, until RECIST progression, which is a little bit different from the second-line cohort. So, RECIST-defined progression is what allowed patients to come off trial. And 119 patients who were considered cisplatin-ineligible in the first-line setting, here, they used some of the consensus criteria that had been developed. Patients who have ECOG poor performance status of 2, impaired kidney function, and GFR of less than 60, peripheral neuropathy or hearing loss that is grade 2 or worse, these are some of the criteria that we’ve developed over time to define the cisplatin-ineligible patient population. And what this study showed was roughly about 23% to 24% response rate in the frontline setting—the initial analysis was actually at 24% and the subsequent analysis that was ultimately published was 23%. We saw a very loose numerical correlation in terms of PD-L1 expression and response where the highest level of PD-L1 expression had a response rate of 28%. Whereas, the patients with lowest level of PD-L1 expression had a response rate of 21%. Again, these are numerical, but there’s clearly overlap in confidence intervals, so there’s not much we can say about this.

But what also was striking, very similar to what we saw in the second-line cohort, is that there were complete responses, roughly 6%, and they were durable. The median duration of response has not yet been reached. And, at a median follow-up of over 17 months, the median survival now is 15.9 months. Which, if you put that in the context of what we typically see with this patient population, this is quite striking. Again, you can’t make comparative claims. This is a single arm study, but that is what has led a lot of people to be excited about this therapy. What’s also key here is the safety profile. Similar to what we saw in second-line studies, the grade 3 and 4 adverse event rate is 16%, immune related toxicities are manageable with corticosteroids. In fact, we did not see any treatment-related deaths, and all side effects were managed with steroids and did not require any additional agents. So, when you think about this therapy and you put it in the context of patients who were receiving previously with carboplatin-based therapy with a 21% rate of treatment discontinuation from toxicity, this is a complete sea change in terms of how we manage these patients.

Now KEYNOTE-052—a similarly designed trial, much larger study—ultimately enrolled 370 patients. What I presented at ESMO was the analysis of the first 100 patients who were treated. They received pembrolizumab at a 200-mg flat dose every 3 weeks; similar patient population, almost identical in some respects. In the first 100 patients, we saw a 24% response rate and about 6% or 7% complete response rate. Median duration of response had not yet been reached, and there was a 15% grade 3 and 4 adverse event rate. So, it was very similar to what we saw in the frontline atezolizumab study. In that particular study, we actually had an ongoing process whereby which we developed a biomarker. To Betsy’s point earlier, this was a discovery set, the first 100 patients. The approach that Merck has taken with PD-L1 expression in that particular study was to focus both on immune infiltrating cells as well as the tumor cells, and look at PD-L1 expression on both. And that’s an interesting take because it really does have a very composite and overall look at the immune system rather than just focusing on one cell type.

And then, in this discovery set, the 10% PD-L1 expression—which they defined as a combined positive score—had the strongest level of enrichment for response as well, so minimizing the false negative rate. And that was a response rate in the high 30s. What we found is that in the validation cohort, which was presented at the ASCO GU symposium, is that the biomarker actually got better in the remaining 270 patient population, which is something that’s very unique about this particular trial. We did not see a major change in terms of the safety events. The grade 3 and 4 adverse event rate was similar, and median duration of response has not yet been reached. So, what we take together from both of these studies is that frontline immunotherapy is clearly active, well tolerated in the cisplatin-ineligible patient population. The key question we have to ask now is, does this represent a new standard of care in the cisplatin-ineligible patient population? I think both these studies do make a compelling argument toward that fact.

Daniel P. Petrylak, MD: And there are randomized trials now. There are several that are going on that are looking at frontline immune therapy, either single agent or in combination. Perhaps you could describe some of those.

Arjun V. Balar, MD: Right. So, there are several key studies. KEYNOTE-361 is a randomized phase III trial comparing pembrolizumab versus chemotherapy alone. In this particular setting, they’re using platinum-based chemotherapy, including cisplatin- and carboplatin-based therapy versus pembrolizumab in combination with platinum-based chemotherapy. And this is a study that’s actively ongoing and recruiting patients. Similarly, IMvigor-211, which is a frontline chemotherapy trial of atezolizumab alone versus atezolizumab in combination with carboplatin-based chemotherapy versus chemotherapy alone, I think has been amended. And so, in that particular study, both studies will really aim to compare single-agent immunotherapy, single-agent or combination chemotherapy versus combination chemotherapy with immunotherapy. And the results of both these trials are eagerly anticipated.

Daniel P. Petrylak, MD: And there are other trials, too. There’s the DANUBE study that’s looking at chemotherapy versus checkpoint plus CTLA4 or PD-L1 plus CTLA4. So, other trials are looking at this particular question.

Transcript Edited for Clarity