Comparing Real-World Evidence With Data From Clinical Trials

Grzegorz S. Nowakowski, MD: There’s a new wave of real-world data from closely monitoring patients for a number of covariates. Some of this analysis was done in the setting of a tafasitamab-lenalidomide study. There was the RE-MIND study, which compared it with single-agent lenalidomide, and then RE-MIND2, which looked at frequently used regimens in the relapsed/refractory setting. What’s your take on studies like this? I’ll start with Matt.

Matthew Lunning, DO, FACP: It definitely gave us the opportunity to discuss this analysis. It seemed as if they got traction from the agency. One major point of contention with this analysis is that it’s real-world data against clinical trial patients. If the analysis had been done in real-world tafasitamab and lenalidomide patients who were then matched to the other populations, then it would hold more water. But it certainly brought to the table the discussion about tafasitamab-lenalidomide and its place. It allowed some regimens to go under retirement. Single-agent lenalidomide is a good example. But I don’t think it’s going to retire CAR [chimeric antigen receptor] T-cell therapy.

Caron Jacobson, MD: Some of these analyses started with CAR T cells because you had single-institution studies of CD19 CARs that were doing a phenomenal job of getting patients into remission, and it became unethical to think about a randomized trial. You had to set a benchmark. With a study like the L-MIND with tafasitamab-lenalidomide, I’m not sure it couldn’t have been a randomized study. They aren’t necessarily the same stakes. I don’t know that we should be adapting what worked to establish studies of CAR T cells into other pharmaceutical studies.

Grzegorz S. Nowakowski, MD: Do you have any other comments on that?

Leo Gordon, MD: In certain situations, the real-world look is important. CAR T is important. I was surprised to see that the data looked fairly similar to the clinical trial data. I thought the restrictions on CAR T on the trials were fairly strict, and I thought people would say, “Now that this is available, I’m going to give it to everybody.” But people are being pretty careful about it in the real world.

Frederick Locke, MD: First, I have a comment about SCHOLAR-1, which was an establishment of the baseline that Caron was mentioning. I agree, you don’t necessarily need that for all trials to establish. We didn’t know how a refractory patient population would do, so establishing that was important for CAR T. Now we have randomized trials, so we can look to those data.

In the real-world setting for CAR T, Caron and I published 2 complementary papers in JCO [Journal of Clinical Oncology] showing real-world data from consortiums that we helped put together. This was specifically for axicabtagene ciloleucel, which we and the centers had the most experience with. The results showed that the PFS [progression-free survival] and OS [overall survival] curves look almost superimposable on the ZUMA-1 trial that led to its approval. We’re being careful, but the majority of those patients wouldn’t have been eligible for ZUMA-1 because of some comorbidity or factor. The curves look the same.

It raises the question of why they’re doing the same if they have more comorbidities. My interpretation is that the ZUMA-1 trial allowed for refractory patients, patients who were getting active therapy and had progressive or at best stable disease or relapsed after an autologous transplant. In the real-world setting, we’re getting patients who responded to second-line therapy and then relapsed 4 months later. If you stratify by refractory disease vs relapsed disease, you can see that those with refractory disease don’t do as well. That’s why we’re getting those relapsed patients and why those curves look so similar.

Caron Jacobson, MD: In our series, when we did the PFS curves for patients who were eligible for ZUMA-1 vs those who were ineligible, the group that was eligible did better than the patients on ZUMA-1. It’s probably because of what Fred is saying.

Matthew Lunning, DO, FACP: Because we’re getting real, let’s talk about the curves. You mentioned it. Is there a tail or is it a tale? This is an important dichotomizer, because the real issue is what I call the brain-to-vein time, which is when you want to do a CAR T-cell therapy to when you can do an apheresis. Those data aren’t in the clinical trials or the real-world experience. It shows the evolution of commercial cellular therapy. How quickly can you get a person for whom you want to do CAR T-cell therapy into the machine and get those T cells pulled out?

In my opinion, in our Medicare territory, the fastest brain-to-vein time is Medicare with a supplement. But if you have a third-party payer, it’s been very hard to negotiate single-case agreements. That’s a 2-way mirror, your institutions negotiating that with a third-party payer. But that’s the true denominator of CAR T. That’s the true access issue. Until we start counting that period of time, which requires institutions to show their cards of the brain-to-vein time, I’m not sure we have the true denominator, the true impact that cellular therapy is bringing us in the real world.

Leo Gordon, MD: We also have that for clinical trials. There’s always delay getting somebody through the process of a clinical trial.

Matthew Lunning, DO, FACP: But the biggest Achilles’ heel is the concept called off the shelf. Whether it’s a cellular therapy, a bispecific, or lenalidomide-tafasitamab, that’s off the shelf. All these things are in play. They’re all semantics, but they’re topics that we should be discussing in a forum like this.

Frederick Locke, MD: Brain-to-vein is an interesting concept. We’ve thought a lot about this at our center. The problem is: how do you define that? When does brain-to-vein start? At our center, we have a consult process for CAR T, so we know it’s intentional that this is a patient who we think could be eligible and bring them forward. But other centers may not do that. You’re seeing different kinds of patients in your clinic. When do you decide to do CAR T? How do you record that? We try to work with centers in Europe to look at their brain-to-vein time because some of their outcomes look different from ours. We say, “Maybe this is the issue.” Even there, it’s hard to capture when brain-to-vein starts. They have central reviews across the country or regionally.

Matthew Lunning, DO, FACP: Everybody’s brain is different. That’s true. But institutionally, you can set your own brain-to-vein time, and the importance of this time period is maybe not understanding your own evolution. We all think we should be shortening that brain-to-vein time. If you do it at a single center and you know you’re shortening it, then you’re doing justice to your patients by comparing Moffitt [Cancer Center] with the University of Nebraska or Moffitt with the entire country of France, which may be different. But if we aren’t having this conversation, then we aren’t being impactful to the whole patient population. Do the PFS curves that we’re reporting in the clinical trials and real-world experience have true trails on the curves? I think they do, but this should be an intent-to-CAR discussion rather than “I got apheresed, and I got infused.”

Frederick Locke, MD: We’ve had this in the transplant world forever.

Matthew Lunning, DO, FACP: Intent to CAR, intent to transplant.

Frederick Locke, MD: How do you do that? I agree that it’s something we have to think about.

Caron Jacobson, MD: There are also differences among the different manufacturers and products in terms of what allows you to have the shortest brain-to-vein time, because it isn’t just the insurance carriers that delay access to manufacturing slots.

Matthew Lunning, DO, FACP: You’re sitting there ready to go and you have an apheresis slot that’s 2 weeks away.

Leo Gordon, MD: Or a month away.

Matthew Lunning, DO, FACP: There are plenty of factors and variables that go beyond the institution or the payer, but we need to have those explained to the patient. Because right now, the patient feels this time. They’re feeling the vein-to-vein time and the brain-to-vein time. If we’re going to get true access into the community, we need other providers’ brains, the presynaptic brain-to-vein time between a community oncologist to a cellular therapy doctor or a lymphoma doctor to realize and make that synapse that they can be part of the shortening of this period.

Leo Gordon, MD: That’s where these kinds of forums help. You’re going to see people who know that this may be a second-line therapy, and referrals will come faster. At least part of this raises the question of what we do in the future. You mentioned off the shelf. Where are allogeneic CARs going to go? Maybe a bigger question is: in the next 10 years, are we going to still be sending cells to manufacturers across the country? Or are we going to be doing this at our own sites?

Caron Jacobson, MD: Or are we going to be injecting something into the patient, and CARs are going to manufactured in vivo?

Frederick Locke, MD: Exactly.

Transcript edited for clarity.