David Maloney, MD, PhD: We’re going to turn our attention to use of CAR [chimeric antigen receptor] T-cell therapy in patients with high-risk lymphomas. We actually have 2 commercial agents approved in this setting. Fred, do you want to take us through some of the emerging data with axicabtagene ciloleucel?
Frederick Locke, MD: Axicabtagene ciloleucel is a CD28 costimulatory domain-based CD19 CAR T-cell therapy. It was tested in the ZUMA-1 trial in adult patients with refractory diffuse large B-cell lymphoma. Before we talk about the results of that trial and the durability of results, it’s important to look at the historical data in patients who were not treated with CAR T-cell therapy. The SCHOLAR-1 trial was a large patient-level meta-analysis of patients who had prior lines of chemotherapy and had not responded. There were patients who, at best, had progressive disease or stable disease to their last line of systemic therapy. These patients were followed in 2 large clinical trials and 2 separate large databases, and those patients who didn’t respond to their last line of therapy had a 26% chance of responding to the next line of therapy and a 7% chance of having a complete response to their next line of therapy.
That sets the stage for the historical data. The ZUMA-1 trial took those same patients and gave them axicabtagene ciloleucel. The overall response rate was over 80%, and the complete response rate was over 50%, again comparing very favorably with historical treatment for these patients. Here, at ASH [the American Society of Hematology 2019 Annual Meeting], we now have even longer-term follow-up of the ZUMA-1 clinical trial with a median of 3-year follow-up in all patients. What we can see at a 39.1-month median follow-up is that the median overall survival has now been reached at 25.8 months. That’s pretty impressive for patients, compared with the SCHOLAR-1 trial, where the median overall survival was 6.3 months.
David Maloney, MD, PhD: One of the criticisms of that analysis is that the patients described in SCHOLAR-1 did not have to jump through any hoops of eligibility criteria. In the ZUMA-1 trial, they did. This could obviously bias the population slightly. But I think some of the data we’re seeing now in the real world can relax some of our concerns about that.
Frederick Locke, MD: Yeah, so several different data sets in the real world with standard-of-care use of axicabtagene ciloleucel—which is, of course, now an FDA-approved product—were presented here at ASH. Dr Mike Jain presented an oral abstract looking at a large cohort of patients across 17 academic medical centers that provided axicabtagene ciloleucel for large B-cell lymphoma patients. With close to a year follow-up, the durable response rates compare very favorably in these patients. In fact, on an intent-to-treat basis, at least intent defined as a patient who is apheresed for manufacturer of CAR T-cell therapy, the results are still quite impressive and compare very favorably with the SCHOLAR-1 data set.
David Maloney, MD, PhD: Caron, I know you have a [International Lymphoma Epidemiology] Consortium study as well. Would you like to comment?
Caron Jacobson, MD: Yeah. I think we’re seeing the same thing. We’re actually seeing that the number of patients who were apheresed and treated in the real world are actually comparable with what we saw in ZUMA-1, and it amounts to a very small drop-off. That’s why the intent-to-treat analysis looks so good. It is so comparable with what was seen on the ZUMA-1 study. But in our series, about 60% of patients would not have been eligible for ZUMA-1 for a variety of reasons, partly because they received bridging therapy, which was not allowed on ZUMA-1. But at least 40% of them were ineligible because of organ dysfunction, impairment, and other parameters. When you put that patient into a combined pool, the group actually does similarly well to what we saw in ZUMA-1.
David Maloney, MD, PhD: You bring up a really controversial area. I want to ask Fred about this. Bridging therapy was not allowed in ZUMA-1, but it is used in all the other clinical trials.
Frederick Locke, MD: I want to add 1 thing and then talk about the bridging. When we looked at eligibility for ZUMA-1, we defined it as at the time of apheresis. About 40% to 45% of the patients in our data set, and close to 300 patients total, would not have been eligible for ZUMA-1. The 1 thing that seems to come out of that analysis that predicts for worse outcomes in those patients is a poor ECOG performance status. That’s 1 area where, outside clinical trials, patients do seem to do worse. It’s not that they can’t obtain remissions and durable remissions, but treating patients with an ECOG of 3 or 4, or maybe even 2, is something that really has to be considered carefully.
Caron Jacobson, MD: And we saw this. We saw the same thing. That was the only thing that came out, in terms of being predictive for response rate. We didn’t look at that variable with duration of response.
Frederick Locke, MD: And then we looked at bridging therapy. On the ZUMA-1 trial, patients were apheresed. Then they had to wait for receipt of their CAR T cells. That turnaround time was, on average, 17 days, so [it was] relatively rapid. But in the real world, in the standard-of-care setting, patients are receiving bridging therapy—whether corticosteroids, radiation, combination chemotherapy, or more molecularly targeted agents. We did an analysis in that 298-patient data set in the Consortium and found that on univariate analysis, patients who had received bridging therapy had worse progression-free survival and overall survival. And that held on a multivariable analysis as well.
However, when we looked deeper, when we did a matched propensity score analysis, we found that progression-free survival was no longer different when we matched for characteristics in the bridged patients versus nonbridged patients. That’s likely because the bridged patients were the worst patients. They had worse performance status. They had all the features that would suggest that they had more rapidly progressive disease. And they were the ones who were selected for bridging chemotherapy, because their physician probably felt that they couldn’t make it to the 17 days. That said, on the propensity score matched analysis, the overall survival was worse in bridged patients. And there is a question about whether bridging should be given. In my practice, I try to avoid giving bridging therapy unless I feel the patient can’t make it that 17 days, or 3 weeks.
David Maloney, MD, PhD: What about in your practice?
Jason Westin, MD: I agree with Fred. We give bridging therapy when it’s needed. However, we don’t give it just to hold somebody to receive therapy if they don’t have an urgent need for bridging. The controversy about bridging, and I agree with all the points, is it’s either a surrogate for bad lymphoma or it may be doing something to their microenvironment that could potentially be harmful to the CAR T cells. We don’t yet know. There are no data to suggest why.
Transcript Edited for Clarity