Real-World Data for CAR T-Cell Therapy for R/R DLBCL

Matthew J. Frigault, MD: There’s been a lot of talk about real-world data, how those play a role in decision-making for specific products, and whether these products are holding up to the pivotal studies. A lot of these pivotal studies had strict eligibility criteria.

We’re excited to have data that follow up the ZUMA-1 and pivotal JULIET studies. We look forward to having updated data in the real-world setting to follow. Let’s see how they stand up to the TRANSCEND data following FDA approval. I look forward to those data, which should be forthcoming following FDA approval.

What we’re really looking for is, how do these products act in the real world, given that nearly anyone with a hybrid B-cell lymphoma is now eligible for a CAR [chimeric antigen receptor] T-cell therapy.

We’re seeing in the real-world setting that largely, these products are holding up to what we saw in pivotal studies. We’re saying real world, but what we mean is retrospective cohorts that we’re analyzing following approval. There have been a couple of key abstracts that we can discuss that get at that point. The first abstract was from ASH [American Society of Hematology Annual Meeting] this past year by Karl Kilgore. It looked at Medicare patients receiving chimeric antigen receptors. The main point here is looking at whether, because these patients tend to be older, they can tolerate the treatment. We can see that even though they had multiple comorbidities, these patients actually can be treated successfully and safely.

When you look at the period of time leading up to CAR T-cell therapy vs the period of time after CAR T cell therapy, there doesn’t seem to be significant treatment-related mortality for CAR T, despite patients being considered “older.” That’s a very promising and exciting thing to be able to say. However, Lindsey Fitzgerald also published another abstract at ASCO [American Society of Clinical Oncology Annual Meeting] looking at patients over age 75. They were looking at what types of factors predict inferior outcomes.

Once you start going over the age of 65, they were able to show that overall survival decreases, although comparable PFS [progression-free survival] was seen with use of things like axicabtagene. That’s probably because of the increased toxicity and increased need for steroids for neurotoxicity. This is a retrospective cohort of patients, so we’re somewhat limited by selection bias in all these abstracts that we’re looking at. Higher-grade B-cell lymphomas, older age, high-risk features, and poor performance status going in were typically predictive of poorer responses.

But I’ll go back to the same point. There were patients who did well. There were patients who survived, and a majority of patients did well and will likely be cured from these therapies. When we say that they have inferior outcomes, we’re comparing the median progression-free survival of 6 months with almost 100% mortality. When we’re looking at this, we have to put things in context.

There are also some other published retrospective series. Both are published in the Journal of Clinical Oncology. Those were by Loretta Nastoupil from The University of Texas MD Anderson Cancer Center and Caron Jacobson. Both were multicenter, retrospective cohorts analyzing axicabtagene in the real-world setting. We could see that overall, patients did well. All patients appeared to have similar responses, and the durability of responses was holding.

But interestingly, only half the patients in both cohorts actually met eligibility criteria for the initial ZUMA-1 cohort. This suggests that even though patients were sicker and older, had more comorbidities, and would not have been treated on the clinical trial, they did gain clinical benefit from the CAR T cells. Both of these studies were telling. Because there are 2 independent, retrospective, multicenter analyses with hundreds of patients, that largely validated the axicabtagene ciloleucel data from the pivotal ZUMA-1 study. There are also other abstracts that were key. Peter Riedell presented additional data at ASH 2019 looking at comparisons between axicabtagene and tisagenlecleucel prescribing habits across different centers.

Not to compare efficacy directly, but we were able to show that in different independent retrospective series looking at both products, the response rates, safety, and toxicity were largely the same. If anything, for tisagenlecleucel, the safety profile was improved compared with the pivotal study, given that they were using the updated ASTCT [American Society for Transplantation and Cellular Therapy] grading criteria. There are additional data that were updated at ASH 2019, with the most recent by Samantha Jaglowski, et al. It’s actually the CIBMTR [Center for International Blood and Marrow Transplant Research] registry data looking retrospectively, using the FDA-mandated registry as part of the REMS [Risk Evaluation and Mitigation Strategy] program. That is part of the mandated 15-year follow-up with the FDA. We have data for poor patient outcomes, toxicities, and adverse events. That is using the Center for International Bone Marrow Transplant Research Database, or CIBMTR, registry that we used for transplant.

In this cohort, they looked at patients who received tisagenlecleucel. They had a significant number of patients, although it was approved later. Therefore, we only had about 70 patients, at least with significant follow-up to look at. However, we’re seeing that overall response rates were comparable to pivotal studies. CR [complete response] rates were comparable to pivotal studies. Most important, the question of viability was addressed. Following the approval of tisagenlecleucel, the FDA requires a certain testing to be performed on the products. This is for all products.

However, 1 of the criteria is a viability component. Cells have to be greater than 80% viable. That doesn’t mean you don’t have an appropriate CAR T-cell dose. It’s just an independent variable. For some products, it hasn’t been an issue with manufacturing, but it was an issue with tisagenlecleucel. The viability cutoff of 80% at product launch wasn’t being met for a subset of patients. They were able to show in cohorts of patients, who had viability greater or less than 80%, that there were no differences in CR rates, PR [partial response] rates, or overall response rates.

As we start moving forward, this type of release testing should probably be reevaluated by the FDA.Ultimately, you can have a low viable percentage of cell population but have an equal CAR T dose. We have to understand that the real need that matters is the viable CAR-positive dose of cells that are going in. As long as we reach that, we’re seeing clinical benefit.

Transcript Edited for Clarity